Changing Societies & Personalities, 2024 Vol. 8, No. 1, pp. 151–172 https://doi.org/10.15826/csp.2024.8.1.269 Received 15 October 2023 Accepted 26 February 2024 Published online 5 April 2024 ARTICLE How the Enlargement Affects European Union Legislative Process Nikolay Yu. Kaveshnikov Moscow State Institute of International Relations (MGIMO University), Moscow, Russia Aleksey O. Domanov Institute of Europe, Russian Academy of Sciences, Moscow, Russia ABSTRACT This article contributes to the ongoing debate regarding the systemic impact of EU enlargements on the duration of the legislative process. Two methods, interrupted time series analysis and survival analysis (the Cox model), are used to show the effects of enlargements, using empirical data comprising EU secondary law directives and regulations. A key distinction of this study from most similar research lies in its focus on distinguishing between legislative and implementing acts, which mitigates the risk of conflating the analysis due to substantial differences in their adoption processes. The methodology and research design help us disentangle the enlargement effects from those of the Treaty reforms and other institutional and structural parameters of the EU decision- making process. The findings reveal a significant acceleration of the legislative process at the moment of the 2004 enlargement, essentially confirming our prior research results. The validity of our conclusions is substantially enhanced by the improvements in modeling techniques. The article also explores potential reasons for the acceleration of the legislative process and concludes that the most likely cause is the refinement of working methods in the Council. KEYWORDS European Union, EU enlargement, legislative process, decision- making, legislative duration, survival analysis, Cox model, interrupted time series analysis © 2024 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov n.kaveshnikov@inno.mgimo.ru, domanov.aleksey@gmail.com https://changing-sp.com/ mailto:n.kaveshnikov@inno.mgimo.ru mailto:domanov.aleksey@gmail.com 152 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov Introduction The 2004 enlargement of the European Union (EU) has given impetus to the research on how the increase in the number of Member States (MSs) complicates and slows down the EU decision-making process. This topic goes beyond mere academic interest, as significant disagreements on politically crucial issues including the 2015 migration crisis, the Nord Stream 2 gas pipeline, and reforms of the Emissions Trading System have captured public attention. Despite this widespread discussion and numerous examples, we still lack a clear understanding of the systematic negative impact of EU enlargements on the duration and efficiency of the decision-making process. Our article aims to contribute to the current body of research on the duration of the EU legislative process by specifically examining the impact of EU enlargements. Using a database and methods that differ from previous studies, we intend to more accurately assess the impact of enlargements, isolating it from the effects of institutional changes. Theoretical Background Several approaches theorize the possible impact of an increase in the number of actors on decision-making process. The most well-known ones are spatial analysis, club theory, voting power approach, and transaction costs theory. Spatial analysis proceeds from the basic assumption that decision-making efficiency depends on distance between the positions of actors and institutional parameters of the decision-making process (Enelow & Hinich, 1984). From this perspective, an increase in the number of EU MSs should generally lead to a more complicated decision-making process and an increased duration of the adoption of legislative acts (Klüver & Sagarzazu, 2013; König, 2007). However, the result and duration of the legislative process are affected not so much by the number of actors (MSs) themselves as by the level of heterogeneity of their preferences. If the preferences of the new MSs fall within the range of preferences of the old MSs, this may not significantly impact the duration of decision-making (Steunenberg, 2002; Tsebelis, 2003). Thus, from the conceptual point, an increase in the number of MSs may not affect the duration of decision-making. In addition, changes in institutional parameters can both accelerate and slow down the legislative process, overlapping with the influence of the increase in the number of actors. The substance of institutional parameters also matters, however. Theoretically, with rigid, super-majoritarian decision-making rules, the growth in the number of actors (and in the heterogeneity of preferences) should lead to a slowdown or ACKNOWLEDGMENTS The research was carried out with the financial support of the Russian Science Foundation in the framework of scientific project No. 23-28- 00096, https://rscf.ru/en/project/23-28-00096, MGIMO University. https://rscf.ru/en/project/23-28-00096 Changing Societies & Personalities, 2024, Vol. 8, No. 1, pp. 151–172 153 even a deadlock of the legislative process, since it becomes more difficult to build a coalition, for which any change of legislation is more profitable than the status quo. Quite the contrary, with soft decision-making rules, the opposite effect is possible, since an increase in the number of actors facilitates the formation of various coalitions in support of changes (Golub, 2007; Toshkov, 2017). The theory of clubs (or the theory of club goods) based on the theory of public goods explains that along with an increase in the number of actors, their willingness to contribute to the production of club goods decreases (Cornes & Sandler, 1996). This phenomenon should be particularly evident in financial matters, especially when it involves contributing to common financial instruments and distributing these shared funds. Thus, an increase in the number of actors theoretically should lead to an increase in disputes and a slowdown in the decision-making process. There is research evidence to some extent confirming this point for the EU (Thomson, 2009; Zimmer et al., 2005). The voting power index method involves the analysis of all possible winning coalitions and the calculation of the power index of each actor based on how often it becomes a key member of the coalition (in other words, its participation is essential for the coalition to gain the necessary number of votes). A commonly used Banzhaf index (Banzhaf, 1965) have been regularly applied in the studies of qualified majority voting (QMV) in the Council of the European Union. According to this approach, EU enlargements should complicate the adoption of legislation (Baldwin & Widgren, 2004) unless it is accompanied by significant institutional changes. The transaction costs theory postulates that an increase in the number of actors should raise transaction costs unless there are changes in institutional parameters that could offset this effect (Buchanan & Tullock, 1965). Theoretical studies of EU institutions’ activities have brought about similar conclusions (König, 2007; Scharpf, 2006). Summing up, the existing theoretical approaches reveal several causal mechanisms that can engender a slowdown of the EU decision-making because of an increase in the number of MSs. However, this forecloses neither the conclusion that such a slowdown has ever occurred nor the extent of it. Furthermore, the majority of theoretical approaches emphasize the significance of institutional parameters of the decision-making process, supposing that changes in these parameters could potentially outweigh the negative effects of enlargement. Despite this, the question regarding the actual impact of the past EU enlargements remains unanswered. Review of Empirical Studies Several empirical studies consistently reveal common trends regarding the impact of various factors on EU decision-making duration. The participation of the European Parliament, particularly through the ordinary legislative procedure, tends to slow down the legislative process, while QMV in the Council accelerates it (Golub, 1999, 2007; Golub & Steunenberg, 2007; Hertz & Leuffen, 2011; Kaveshnikov & Domanov, 2022; Klüver & Sagarzazu, 2013; König, 2007; Rasmussen & Toshkov, 2011; Schulz & König, 2000). The adoption of new legislative documents typically takes longer https://changing-sp.com/ 154 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov than amendments. Additionally, the process tends to be more time-consuming for permanent acts compared to provisional ones (Drüner et al., 2018; Hurka & Haag, 2020; Kaveshnikov & Domanov, 2022; Rasmussen & Toshkov, 2011). The higher is the complexity of the document, the longer is the decision-making process (Hurka & Haag, 2020; Kaveshnikov & Domanov, 2022). However, Rasmussen and Toshkov (2011) did not find any clear dependence. The packaging of proposals typically extends the decision-making timeline (Kirpsza, 2022). Enhanced transparency in the Council has been observed to facilitate decision-making (Hagemann & Franchino, 2016). Several empirical studies relying on the methodology of duration studies have assessed the impact of various EU enlargements on the speed of the legislative process, either directly or indirectly. First, let’s examine studies that investigate the impact of EU enlargements on the speed of the legislative process. On the basis of survival analysis and the Cox model, Golub and Steunenberg concluded that the increase in the EU membership to 9, 10, 12, and even 15 members accelerated the decision-making process (Golub, 2007; Golub & Steunenberg, 2007). However, their research is based on database comprising EU legislative acts between 1968 and 1998, with the period of 1968–1972 (before the 1973 enlargement) used as a reference. It thus makes sense to suggest that the observed acceleration was notable mainly in contrast to the relatively slow decision-making process during those years, which was partly attributed to the policies of France under the Charles de Gaulle government. Klüver and Sagarzazu (2013) investigated the influence of ideological differences between the Commission, the Council, and the European Parliament on the EU legislative process using the same Cox model. Additionally, they concluded that enlargements did not affect the duration of decision-making. The fact that their research used the period from 1979 to 2010 raises questions about whether they adequately considered the influence of various changes in institutional parameters, including Treaty reforms, on the duration of decision-making. The study of Best and Settembri (2008) based on descriptive analysis did not reveal the impact of the 2004 enlargement on such parameters as the number of adopted acts and the average period of adopting acts. However, the number of acts adopted before and after the enlargement is not a reliable indicator, since this approach does not take into account the possible increase of EU competence and the emergence of new sectoral policies. Moreover, the analysis of the average adoption period of acts does not consider many other factors that affect the duration of the legislative process, e.g., the type of acts, legislative procedure, and decision-making procedure in the Council. Two relatively recent papers are entirely devoted to the study of the effect of EU enlargements on the duration of the legislative process. Hertz and Leuffen (2011), having analyzed data for the years 1976–2006 with the Cox model, concluded that the enlargement of the EU to 10, 12 and especially to 15 and 25 MSs slowed down the decision-making process. Although, the results of their study should be regarded with some reservations. The use of such a long time series complicates the evaluation of the impact of the numerous changes in institutional Changing Societies & Personalities, 2024, Vol. 8, No. 1, pp. 151–172 155 parameters that have occurred in the EU over this long period, particularly the Treaties reforms and the evolution of the QMV practice in the Council. The Luxembourg Compromise of 1966, amongst other things, cast a long shadow on EU development. The political veto principle led to the fact that until the mid-1980s, the Council legislated exclusively on the unanimity basis, which inevitably slowed down the decision-making. The new dynamics of the EU development in the late 1980s and 1990s in this context can be misinterpreted as the effect of the 1986 and 1995 enlargements. In their pursuit of more accurate results, Hertz and Leuffen (2011) compared various periods characterized by different numbers of MSs, presenting paradoxical findings. For instance, no significant change in the duration of decision-making was observed when comparing the EU-12 with the EU-10 and the EU-15 with the EU-12. However, a substantial slowdown emerged in the EU-15 compared to the EU-10. Given that the enlargement from 10 to 12 MSs and from 12 to 15 MSs on a standalone basis did not have a significant impact on the duration of decision- making, there is a reason to doubt that the difference between the EU-10 and EU-15 periods is caused by the increase in membership, and not by other factors. Similarly, when scrutinizing the EU-25 compared to the EU-15, Hertz and Leuffen (2011) failed to identify a substantial difference. Yet, they found a significant slowdown when comparing the EU-25 with both the EU-12 and the EU-10, which raises doubts about the interpretation of the results. Toshkov (2017) conducted a thorough study of the impact of the 2004 enlargement on the legislative process through a range of methods, such as (a) the descriptive analysis of the number of acts adopted, (b) the Kaplan–Meier curve, (c) the heterogeneity of preferences of MSs in terms of their initial positions and of voting results using network analysis logic. He concluded that the 2004 enlargement did not have a negative impact on the duration of the decision-making process. The individual methods employed are susceptible to criticism. The analysis of the number of acts before and after the enlargement fails to consider the potential increase in the EU competence and the development of new sectoral policies. The Kaplan–Meier curve cannot take into account factors beyond the EU membership, such as the amendments to the EU Treaties, different legislative procedures, etc. The study of the heterogeneity of the initial positions of MSs relied on the data from the DEU-II project (Thomson et al., 2012), comprising MSs’ positions on 331 issues of 125 legislative acts, constructed through expert interviews. While this database is the most comprehensive of its kind, it may not fully align with the analysis of the EU legislative process as a whole. The analysis of the preferences of Member States’ preferences based on voting results has some acknowledged shortcomings (Toshkov, 2017, p. 186). Nevertheless, when considering all the methods collectively, a tentative conclusion can be made regarding the absence of any “strong and systemic effects” of the 2004 enlargement on the decision-making process (Toshkov, 2017, p. 189). In summary, the existing body of research on the subject yields contradictory results. This inconsistency may be partly attributed to variations in how databases are constructed, that is, whether they include only directives or encompass all acts, as well as the specific periods under examination. Notably, the research does not differentiate https://changing-sp.com/ 156 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov between documents of secondary and tertiary law, an important factor given that the process of adoption of tertiary legislation acts has specific features because of acts’ technical nature. Another notable limitation in most studies is their methodology, which falls short in effectively distinguishing the influence of enlargement from other factors. This includes changes in institutional parameters, notably Treaties reforms and the gradual evolution of decision-making practices. Some researchers openly acknowledge this challenge, in particular Toshkov (2017), who points out that “the precise effect of enlargement remains impossible to disentangle from all other contemporarious institutional, political, and societal developments that affected the EU” (p. 178). Research Methods Our goal is to evaluate the influence of the 1995 and 2004 EU enlargements on the duration of the decision-making process. To this end, we employ two methods, namely interrupted time series (ITS) analysis and survival analysis employing the Cox model. The basic idea of ITS analysis (Morgan & Winship, 2007) is to model the behavior of a dependent variable (in our case, the duration of the decision-making process) before and after the event. The proper construction of the model and the choice of variables, as well as the absence of other relevant change in the survey period, allow us to interpret the discontinuity at the time of the event as the casual effect of this event. As regards the EU legislative process, the ITS method was previously used by Bølstad and Cross (2016) to assess the impact of the Amsterdam, Nice, and Lisbon Treaties. We investigate the effect of each enlargement via linear regression, which was calculated separately for the period before and after the event, as proposed, for example, by Imbens and Lemieux (2008). We chose symmetric windows covering a period of two years before and two years after the event. These windows should be small enough to minimize the influence of other possible factors, and large enough so that the empirical data (the number of acts adopted) ensures acceptable accuracy of calculations. Linear regression is calculated using Equation 1. For each enlargement, a binary variable E (enlargement) is introduced, reflecting whether the document was adopted before (E = 0) or after (E = 1) the enlargement1. The time variable t is centred on the moment of enlargement, i.e., t = 0 when E changes from zero to one. X is a vector of additional variables reflecting fixed effects. In this linear regression equation, coefficient β1 reflects the immediate effect of the enlargement, β2 sets the trend of the duration of the legislative process before the enlargement, and β3 stands for trend change after the enlargement. ln𝑌𝑌𝑌𝑌𝑡𝑡𝑡𝑡 = 𝛽𝛽𝛽𝛽0 + 𝛽𝛽𝛽𝛽1𝐸𝐸𝐸𝐸𝑡𝑡𝑡𝑡 + 𝛽𝛽𝛽𝛽2𝑡𝑡𝑡𝑡 + 𝛽𝛽𝛽𝛽3𝐸𝐸𝐸𝐸𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 + ∑ 𝛽𝛽𝛽𝛽𝑚𝑚𝑚𝑚+3𝑋𝑋𝑋𝑋𝑚𝑚𝑚𝑚𝑡𝑡𝑡𝑡 𝑀𝑀𝑀𝑀 𝑚𝑚𝑚𝑚=1 + 𝜀𝜀𝜀𝜀𝑡𝑡𝑡𝑡 (1) 1 Legislative acts initiated before the enlargement, but adopted after, were coded as adopted after the enlargement. We assume that a significant part of the negotiations on the legislative file was held before the enlargement, and the role of the new MSs during the final stages of legislative process was of low significance. Changing Societies & Personalities, 2024, Vol. 8, No. 1, pp. 151–172 157 To consider the influence of institutional and systemic factors, we incorporated a set of control parameters into our model. These parameters reflect fundamental factors whose impact on the duration of decision-making was confirmed by the numerous studies mentioned above. The model includes the following control parameters: the type of the act (directive or regulation); the legislative procedure (ordinary legislative procedure/cooperation/consultation/assent/the Council acts without participation of the European Parliament2); the decision-making rule in the Council (unanimity or QMV); the novelty of the act (new act or amendments); the provisional/permanent nature of the act; and the complexity of the act. We take into account these control parameters in the form of fixed effects by introducing additional variables. To correctly calculate fixed effects in the model, we should disentangle documents adopted on the same day. To achieve this, we divided each day of the study period into 34 time slots based on the maximum number of documents adopted in a single day during the given period3, which allowed us to implement the model at the level of individual acts. To consider the nonlinear nature of the dependency between variables and not overload the model with hard-to-compute components, link functions are used. Considering the characteristics of our database, we use a logarithmic link function, as recommended by Fox (2015, p. 392) and Hilbe (2011, p. 193). We test our model for first-order auto-regressive auto-correlation using the Breusch–Godfrey test, and for auto-regressive conditional heteroscedasticity using the Engle test. The test results are generally unproblematic4, enabling the implementation of the model at the level of individual acts. The second method we applied was a survival analysis using the semiparametric Cox model for calculations. In survival analysis, the crucial choice is between parametric and semiparametric models. Parametric models are very sensitive to characteristics of baseline hazard distribution, while semiparametric models including the Cox model could be used, on the contrary, even if hazard varies with predictor level. Following the intensive theoretical discussions in the early 2000s, it was concluded that the specific characteristics of empirical data on the EU legislative process and the nonuniform probability distribution make the Cox model more suitable for researching the duration of EU legislation (Golub, 2007, p. 162; Zorn, 2007, p. 568). The Cox model has been actively used in political science since the second half of the 1990s (Box-Steffensmeier & Jones, 2004). Since the early 2010s, the Cox model has been the most frequently used method in the study of the activities of EU institutions in the context of duration studies. Among eleven studies of this kind, eight employed the Cox model (Brandsma & Meijer, 2020; Chalmers, 2014; Drüner et al., 2018; Hertz & Leuffen, 2011; Hurka & Haag, 2020; Kaveshnikov & Domanov, 2 Officially, when the Council acts without the participation of the European Parliament, the procedure is designated as a non-legislative procedure. However, many acts adopted in this manner possess all the legal characteristics of secondary law, which justifies our inclusion of this procedure in the list. 3 34 acts were adopted on April 29, 2004. 4 Signs of autocorrelations are evident in the database of documents around the 1995 enlargement (all acts, models 1 and 2), which aligns with the overall low accuracy of calculations related to the 1995 enlargement (see the Results section) and serves as another reason to exercise caution in drawing conclusions in this segment of the research. https://changing-sp.com/ 158 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov 2022; Klüver & Sagarzazu, 2013; Rasmussen & Toshkov, 2013). Two studies used variations of linear and logistic regression, while one employed the interrupted time series (ITS) method. The Cox model is based on the multiple regression method (Cox, 1972). This semiparametric method predicts the risk of occurrence of an event (hazard risk) for the object under consideration and evaluates the influence of independent variables on this risk. As a result, the model determines the value of the hazard ratio (HR) between the risk indicators in the experimental and control groups. In this study, we define an event as the adoption of a legislative act. A control group is an array of legislative documents adopted before enlargement; the experimental group consists of documents adopted after the enlargement. Risk is the probability that a document that has not been adopted on day N since the beginning of the legislative procedure will be adopted on day N+1. Thus, the hazard ratio shows to what extent the probability of document adoption differs before and after the enlargement. In our study, we established two-year periods both before and after the respective enlargements. This approach serves a dual purpose: firstly, it ensures comparability between the results obtained from the Cox model and the ITS method; secondly, these designated periods help minimize the potential influence of various changes in the institutional parameters of the decision-making process. Calculations were carried out using a multifactorial model to take into account the influence of institutional and systemic factors. The control parameters are similar to those used in the ITS model. The calculation results were checked for the proportional effect assumption by the Grambsch–Therneau test and passed the test as far as concern the enlargement variable. The empirical database of our study covers EU secondary law directives and regulations extracted from the official EU legislation portal EUR-Lex5. Our database principally differs from the data used in most similar studies; it includes only legislative measures (secondary law) but not implementing measures (tertiary law). Implementing measures contain detailed provisions, which make it possible to implement existing legislative decisions in practice. Some types of implementing measures are adopted by the Council6, they are manifold and could muddle up the analysis of decision-making if taken into account together with legislative acts. After the Lisbon Treaty, it became easy to differentiate between secondary and tertiary legal acts, due to the inclusion of terms like “delegated” or “implementing” in the titles of acts of tertiary legislation. However, in the periods we examined, these markers were not used. For empirical data collection, we relied on a crucial distinction, which was that acts of secondary legislation were identified by their legal basis, i.e., particular articles of basic Treaties, while acts of tertiary legislation have a different legal basis, namely acts of secondary law (Lenaerts & van Nuffel, 2005, p. 570). 5 https://eur-lex.europa.eu/homepage.html 6 After the Lisbon Treaty, delegated acts are adopted by the Commission (Article 290 TFEU, see Consolidated version of the Treaty on the Functioning of the European Union, 2012a), while implementing acts are adopted by the Commission according to comitology procedures or by the Council alone (Article 291 TFEU, see Consolidated version of the Treaty on the Functioning of the European Union, 2012b). https://eur-lex.europa.eu/homepage.html Changing Societies & Personalities, 2024, Vol. 8, No. 1, pp. 151–172 159 We decided to exclude decisions from the database since these acts lack a fundamental characteristic of legislation: they do not establish general rules but rather apply legislative rules to individual situations. The number of secondary law decisions is significantly higher than the combined total of directives and regulations. Including all these documents in the database could significantly compromise the results of the analysis. We excluded legal documents directly related to the 1995 and 2004 enlargements (those citing Accession Treaties as their legal base) from the database. These documents are usually rapidly adopted7 because they formalized agreements reached during earlier accession negotiations. The final database thus includes 1,538 legal documents for both study periods (Table 1). It is challenging to distinguish the effects of enlargements from those of the basic treaty reforms because these events occurred closely in time. Without careful handling, the absence of this distinction or the use of improper methods can cast doubt on the validity of conclusions in the studies mentioned earlier. The Maastricht Treaty (MT) took effect within two years before the 1995 enlargement, and the Nice Treaty (NT) within two years before the 2004 enlargement. To tackle this challenge, we created two models for both the Cox and ITS methods. Model 1 uses the entire dataset, where the impact of enlargement is mixed with the potential impact of the Treaty amendments. For Model 2, we refined the dataset by excluding documents based on those articles that were changed respectively by the MT and NT. We omitted such Treaty changes as the introduction of new competencies, change of legislative procedures in specific policy areas (e.g., new policy areas subject to consultation or the ordinary legislative procedure), and shifts from unanimity to majority voting in the Council. To maintain dataset consistency, we excluded such documents adopted both before and after the respective Treaty entered into force. As a result, Model 2 accurately represents the pure impact of the enlargement. Table 1 Description of the Database Four years around 1995 enlargement Four years around 2004 enlargement Total Before enlargement After enlargement Total Before enlargement After enlargement Directive 179 102 77 195 138 57 Regulations 677 383 294 487 314 173 All acts 856 485 371 682 452 230 To enhance the validity of the findings, we employed two methods of analysis. The exclusion of tertiary law documents and decisions allows us to assess the patterns of the legislative process stricto sensu. The chosen methods, control variables, and survey periods help us accurately separate the impact of the enlargements from the 7 Most directives of this kind are adopted within 50–60 days. The average time for the adoption of a typical directive is 658 days in the first research period (1993–1996) and 765 days in the second period (May 1, 2002–April 30, 2006). https://changing-sp.com/ 160 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov influence of the Treaty changes and other institutional parameters. Model 2, excluding relevant documents from the dataset, serves to assess the potential impact of the Treaty changes. The potential impact of the gradual evolution of decision-making practices, such as trilogues and the erosion of consensual practices in the Council (details below), is minimized as these trends develop slowly and may not significantly affect decision-making patterns in the four-year periods around the corresponding enlargements. It should be noted that the short timeframes for analysis require us to rely on relatively small data samples, potentially diminishing the statistical accuracy of the model results. Results and Discussion The results of the modeling are presented in Tables 2–5. Before discussing the impact of the enlargements, it is worth noting that the Cox model confirms well-known patterns related to control parameters8. The participation of the European Parliament, especially the use of the ordinary legislative procedure9, slows down decision-making while the Council’s qualified majority voting accelerates this process. The adoption of new, complex, and permanent acts takes more time than the adoption of amendments, simple and provisional acts. Table 2 Impact of 1995 Enlargement on the Duration of Decision-Making (ITS Model) Variable Model 1 Model 2 (Maastricht Treaty impact excluded) All acts Directives Regulations All acts Directives Regulations Enlargement (Е), β1 0.74 0.97 0.81 0.69 0.85 0.74 Time (t), β2 1.0001 1.0005* 1.0001 1.0004 1.0005 1.0003 Е × t, β3 1.0009 0.9997 1.0005 1.0004 0.9995 1.0003 Constant, β0 250.9*** 508.0*** 162.9*** 170.4*** 517.0*** 138.1*** Dispersion 0.80*** 2.26*** 0.88*** 0.84*** 0.41 0.95*** Fixed effects (directive/regulation) yes – – yes – – Fixed effects (other control parameters a) yes yes yes yes yes yes Number of acts (N) 856 179 677 677 81 596 Period, months 48 48 48 48 48 48 Note. a Control parameters: legislative procedure, unanimity/QMV in the Council, new act/amendments, permanent/provisional act, complexity of the act. * p ≤ .05; ** p ≤ .01; *** p < .005. 8 We have excluded these data from the tables to avoid clutter but they are available in Appendix (Tables A1, A2). 9 In the survey periods, this procedure was officially called the co-decision procedure. Changing Societies & Personalities, 2024, Vol. 8, No. 1, pp. 151–172 161 Table 3 Impact of 1995 Enlargement on the Duration of Decision-Making (the Cox Model) Variable Model 1 Model 2 (Maastricht Treaty impact excluded) All acts Directives Regulations All acts Directives Regulations Hazard ratio (HR) 0.83** 0.76 0.88 1.19* 1.15 1.02 Control parameter directive/regulation yes – – yes – – Other control parameters a yes yes yes yes yes yes Number of acts (N) 856 179 677 677 81 596 Period, months 48 48 48 48 48 48 Note. a Control parameters: legislative procedure, unanimity/QMV in the Council, new act/amendments, permanent/provisional act, complexity of the act. * p ≤ .05; ** p ≤ .01; *** p < .005. Table 4 Impact of 2004 Enlargement on the Duration of Decision-Making (ITS Model) Variable Model 1 Model 2 (Nice Treaty impact excluded) All acts Directives Regulations All acts Directives Regulations Enlargement (E), β1 0.36*** 0.47*** 0.43*** 0.34*** 0.47*** 0.40*** Time (t), β2 0.9998 0.9992* 1.0005 0.9998 0.9992 1.0005 Е × t, β3 1.0019*** 1.0026*** 1.0004 1.002*** 1.0026*** 1.0004 Constant, β0 514.6*** 1111.1*** 275.9*** 521.2*** 1106.9*** 275.3*** Dispersion 0.96*** 0.57 0.93*** 0.94*** 0.57 0.91*** Fixed effects (directive/ regulation) yes – – yes – – Fixed effects (other control parameters a) yes yes yes yes yes yes Number of acts (N) 682 195 487 649 191 458 Period, months 48 48 48 48 48 48 Note. a Control parameters: legislative procedure, unanimity/QMV in the Council, new act/amendments, permanent/provisional act, complexity of the act. * p ≤ .05; ** p ≤ .01; *** p < .005. https://changing-sp.com/ 162 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov Table 5 Impact of 2004 Enlargement on the Duration of Decision-Making (the Cox Model) Variable Model 1 Model 2 (Nice Treaty impact excluded) All acts Directives Regulations All acts Directives Regulations Hazard ratio (HR) 1.28*** 1.26 1.20 1.32*** 1.23 1.14 Control parameter directive/regulation yes – – yes – – Other control parameters a yes yes yes yes yes yes Number of acts (N) 682 195 487 649 191 458 Period, months 48 48 48 48 48 48 Note. a Control parameters: legislative procedure, unanimity/QMV in the Council, new act/amendments, permanent/provisional act, complexity of the act. * p ≤ .05; ** p ≤ .01; *** p < .005. Practically identical results of Models 1 and 2 demonstrate that the impact of the Maastricht and Nice Treaties on the duration of legislative process was negligible, at least within our study period. These results confirm the conclusions of Bølstad and Cross (2016) about the Nice Treaty. Regarding the 1995 enlargement, two methods employed have yielded divergent results. The Cox method did not reveal the impact of the enlargement on decision-making. The HR for all acts is 0.83 (Model 1); it is considered that the factor is insignificant if HR differs from 1.0 by less than 0.2. In Model 2 (database refined of Maastricht Treaty impact), we can see negligible effect of opposite nature (HR = 1.19). However, the accuracy of calculations in both cases is not very high, which casts doubt on any meaningful interpretations. The ITS method at first glance indicates a significant instantaneous effect of enlargement: β1 is equal to 0.74 in Model 1 and 0.69 in Model 2, that means considerable acceleration of decision-making. However, the accuracy of calculations in both Models is below the significance level. Moreover, when analysing directives and regulations separately, the results of both methods do not have the accuracy that would allow us to make any relevant conclusions. Thus, our research does not allow us to make an unambiguous conclusion about the effect of the 1995 enlargement due to the low statistical accuracy of calculation results. Nevertheless, we can cautiously assert that the 1995 enlargement did not have a distinctively negative effect on the duration of the decision-making process. More reliable conclusions can be drawn about the impact of the 2004 enlargement. Paradoxically, both methods indicate that after the enlargement the decision-making process has noticeably accelerated. This acceleration was instantaneous, long- term trends of decision-making durations did not change. The Cox method gives HR = 1.28 in Model 1 and 1.32 in Model 2 (database refined of Nice Treaty impact). The ITS method gives the coefficient β1 = 0.36 in Model 1 and 0.34 in Model 2. Changing Societies & Personalities, 2024, Vol. 8, No. 1, pp. 151–172 163 Similar results that are obtained when analysing directives and regulations separately confirm the systemic nature of the enlargement effect. The calculations have high statistical accuracy both for all acts (both methods) and for directives and regulations separately (ITS method). Our findings challenge the common notion of a slowdown that occurred in decision-making following the 2004 enlargement. The conceptual approaches presented in the first part of the article do not provide a clear explanation for this phenomenon. Nevertheless, several assumptions can be considered. The seemingly obvious explanation that the decision-making process speed up because of the Nice Treaty is not valid. Model 2, which was designed specifically to test this assumption, convincingly refuted it. Other options to explain the acceleration of decision-making process include the influence of trilogues and shifts in consensual practices within the Council. The period in question saw the growing use of the practice of trilogies, informal meetings involving representatives from the Commission, the European Parliament, and the Council to discuss legislative proposals (Brandsma, 2015; Kaveshnikov, 2021b). The initiation of regular trilogues during the third reading of the co-decision procedure in 1995 marked a pivotal moment, proving to be highly effective. Over time, trilogues became a standard practice during the first and second readings as well. This shift is evident in the legislative outcomes during the fifth and sixth EP legislatures (2000– 2009), where approximately 60% of co-decision legislation was adopted at the first reading, around 30% at the second reading, and only 10% required a third reading and a meeting of the conciliation committee (Rasmussen, 2012, p. 743). Empirical studies further confirm the significant role of trilogues in expediting the legislative process (Toshkov & Rasmussen, 2012). The erosion of the consensus culture in the Council could also have a certain impact on the duration of decision-making. Even when the Council de jure can decide by vote, MSs usually continue negotiations and try to reach a consensus (Mattila & Lane, 2001). Thus, between 1994 and 2002, 81% of all decisions in the Council were adopted without a vote (Heisenberg, 2005, p. 66). However, in recent decades, it has become increasingly difficult for Council members to reach a consensus: in the first half of the 2010s, the vote share increased to 35–40% of the number of issues where voting is provided for by the Treaties (Wallace et al., 2015, p. 83). When building our database, we coded decision-making procedures in the Council (voting or unanimity) on the basis of the articles of the Treaties. When the relevant article provided for QMV, we did not check whether the vote was held in practice or the document was adopted by consensus. Thus, the gradual increase in the share of legislative documents adopted by the QMV, unaccounted for in our models, could distort the results and create the illusion of acceleration of the legislative process due to the enlargement. It is important to note that the evolution of both trilogue practices and the diminishing consensus culture in the Council has been a gradual process. It is highly improbable that these factors could produce an immediately noticeable effect, as indicated by the ITS model, specifically through coefficient β1 reflecting the immediate impact of the enlargement. The observation that the enlargement does not influence https://changing-sp.com/ 164 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov the long-term trend (as indicated by coefficient β3 in the ITS method) further supports this assertion. The absence of a noticeable slowdown in the legislative process after the 2004 enlargement can be explained by the fact that the preferences of new MSs lied generally within the range of preferences of old MSs, especially in the first decade after the new MSs joined the EU. Research on Member States’ positions and voting outcomes shows that the conflict dimension between Western and Eastern MSs is usually not significant (Mattila, 2009; Plechanovová, 2011; Thomson, 2009; Toshkov, 2017). This dimension becomes noticeable only in specific policy areas, such as climate policy or migration and asylum policy (Thomson, 2009, p. 767; Toshkov, 2017, p. 188). Sectoral studies strongly demonstrate that a sustained divergence in preferences (West vs. East) is notable in specific policies, including migration and asylum (Geddes, 2018; Potemkina, 2019), redistributive policies, e.g., encompassing structural funds, common agricultural policy, and environmental policy (Veen, 2011), climate policy (Jevnaker & Wettestad, 2017; Kaveshnikov, 2021a), and energy security (Goldthau & Sitter, 2015; Mišík, 2016; Youngs, 2020). It should be noted, however, that these policy areas constitute only a small fraction of EU activities. Additionally, most of these policy divisions gained significance and became politically contentious many years after the 2004 enlargement. Consequently, the divergence in the preferences of MSs on these issues had no significant influence on the overall dynamics of decision- making immediately following the enlargement. The distribution of preferences among new MSs may explain the absence of a slowdown, but it does not account for the notable and immediate acceleration in the legislative process after the 2004 enlargement, as demonstrated in our models. Therefore, it would be safe to assert that this acceleration was primarily due to changes in the Council’s working methods. Just before the accession of the new MSs, for the specific purpose of streamlining decision-making in the enlarged Council, new Rules of Procedure were adopted10, which provided for several measures to improve its effectiveness. In particular, the new Rules of Procedure established that at the stage of preparation of the Council meeting: (a) a file is submitted to Coreper only when there is reasonable prospect of progress; (b) the Presidency shall undertake efficient consultations between meetings and encourage MSs’ delegations to communicate among themselves; (c) MSs’ delegations are recommended to communicate their positions in written form before the meeting; wherever possible, written input shall be submitted jointly by delegations maintaining identical positions. As far as concern the meeting, Rules of Procedure established inter alia that: (a) no item shall be placed on the agenda for information only, such information should instead be transmitted to delegations whenever possible in advance in written form; (b) the Presidency shall limit the maximum length of interventions; (c) like-minded delegations are encouraged to entrust a single spokesperson to express their common 10 The Rules of Procedure were adopted by the Council’s Decision 2004/338 on March 22, 2004 (Council Decision, 2004). Changing Societies & Personalities, 2024, Vol. 8, No. 1, pp. 151–172 165 position on a specific point; (d) delegations shall make concrete drafting proposals, rather than merely expressing their disagreement with a particular point. Despite the technical nature of these procedural changes, they successfully simplified discussions in the Council, playing a pivotal role in accelerating the overall decision-making process in the EU, as our study demonstrates. Conclusions This article contributes to the debate on the systemic impact that EU enlargements may have on the duration of the legislative process. As for the 1995 enlargement, the low reliability of calculations allows us to only state very cautiously that it did not affect negatively (decelerated) the decision-making process. Results of our research demonstrate that a significant instantaneous acceleration of the legislative process took place at the moment of the 2004 enlargement. The absence of a slowdown can be explained by the fact that the preferences of the new MSs were generally within the range of preferences of the old ones. A new axis of contestation (Western vs. Eastern MSs) emerged only in a limited number of EU policy areas and it happened quite some time after the enlargement. Even if there was a temporary deceleration of the legislative process due to this enlargement, it has been more than compensated for by subtle institutional changes. Most likely, the new Rules of Procedure of the Council, which was specifically approved to adapt the institution to the increased number of MSs, significantly enhanced the Council’s efficiency and positively influenced the duration of the legislative process. The increase in membership to EU-25 neither diminished the functionality of the Council nor undermined the generally smooth legislative process within the Union. These conclusions essentially confirm our previous research but are more valid due to a significant refinement in our modeling technique. The inclusion of new control parameters in the ITS method enables us to consider the effects of corresponding institutional parameters on the speed of decision-making. By dividing each day of the study period into 34 time slots, we can more precisely evaluate the impact of control parameters. Models 1 and 2 clearly show that the Maastricht and Nice Treaties had no discernible impact on the speed of the legislative process. We have also developed additional arguments to discuss potential reasons for the acceleration of decision- making after the 2004 enlargement. Our research demonstrates that the relationship between the number of actors (Member States) and decision-making speed in the EU is not straightforward, with institutional parameters playing a significant role. While research often focuses on major changes like fundamental Treaty reforms, we find that minor changes could have a more profound impact. The Union’s institutional system is intricate, with numerous forums, actors, procedures, and entry points. Within this complex structure, many elements can be fine-tuned to influence the efficiency and direction of the policy process. https://changing-sp.com/ 166 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov References Baldwin, R. E., & Widgren, M. (2004). Council voting in the Constitutional Treaty: Devil in the details (CEPS Policy Brief No. 53). Centre for European Policy Studies. https://cdn.ceps.eu/wp-content/uploads/2009/08/1133.pdf Banzhaf, J. F. (1965). Weighted voting doesn’t work: A mathematical analysis. Rutgers Law Review, 19, 317–343. Best, E., & Settembri, P. (2008). Surviving enlargement: How has the Council managed? In E. Best, T. Christiansen, & P. Settembri (Eds.), The institutions of the enlarged European Union. Continuity and change (pp. 34–53). Edward Elgar. https://doi.org/10.4337/9781848443785.00009 Bølstad, J., & Cross, J. P. (2016). Not all treaties are created equal: The effects of treaty changes on legislative efficiency in the EU. Journal of Common Market Studies, 54(4), 793–808. https://doi.org/10.1111/jcms.12349 Box-Steffensmeier, J., & Jones, B. S. (2004). Event history modeling in political science. Cambridge University Press. https://doi.org/10.1017/CBO9780511790874 Brandsma, G. J. (2015). Co-decision after Lisbon: The politics of informal trilogues in European Union lawmaking. European Union Politics, 16(2), 300–319. https://doi.org/10.1177/1465116515584497 Brandsma, G. J., & Meijer, A. (2020). Transparency and the efficiency of multi- actor decision-making processes: An empirical analysis of 244 decisions in the European Union. International Review of Administrative Sciences, 88(3), 626–643. https://doi.org/10.1177/0020852320936750 Buchanan, J. M., & Tullock, G. (1965). The calculus of consent. Logical foundations of constitutional democracy. University of Michigan Press. https://doi.org/10.3998/mpub.7687 Chalmers, A. W. (2014). In over their heads: Public consultation, administrative capacity and legislative duration in the European Union. European Union Politics, 15(4), 595–613. https://doi.org/10.1177/1465116514529849 Consolidated version of the Treaty on the Functioning of the European Union (2012a, October 26). Article 290. EUR-Lex. http://data.europa.eu/eli/treaty/tfeu_2012/ art_290/oj Consolidated version of the Treaty on the Functioning of the European Union (2012b, October 26). Article 291. EUR-Lex. http://data.europa.eu/eli/treaty/tfeu_2012/art_291/oj Cornes, R., & Sandler, T. (1996). The theory of externalities, public goods, and club goods (2nd ed.). Cambridge University Press. https://doi.org/10.1017/ CBO9781139174312 Council decision 2004/338/EC, Euratom. (2004, March 22). EUR-Lex. http://data.europa.eu/eli/proc_rules/2004/338/oj https://cdn.ceps.eu/wp-content/uploads/2009/08/1133.pdf https://doi.org/10.4337/9781848443785.00009 https://doi.org/10.1111/jcms.12349 https://doi.org/10.1017/CBO9780511790874 https://doi.org/10.1177/1465116515584497 https://doi.org/10.1177/0020852320936750 https://doi.org/10.3998/mpub.7687 https://doi.org/10.1177/1465116514529849 http://data.europa.eu/eli/treaty/tfeu_2012/art_290/oj http://data.europa.eu/eli/treaty/tfeu_2012/art_290/oj http://data.europa.eu/eli/treaty/tfeu_2012/art_291/oj https://doi.org/10.1017/CBO9781139174312 https://doi.org/10.1017/CBO9781139174312 http://data.europa.eu/eli/proc_rules/2004/338/oj Changing Societies & Personalities, 2024, Vol. 8, No. 1, pp. 151–172 167 Cox, D. R. (1972). Regression models and life tables. Journal of the Royal Statistical Society, Series B: Statistical Methodology, 34(2), 187–202. https://doi.org/10.1111/j.2517-6161.1972.tb00899.x Drüner, D., Klüver, H., Mastenbroek, E., & Schneider, G. (2018). The core or the winset? Explaining decision-making duration and policy change in the European Union. Comparative European Politics, 16(2), 271–289. https://doi.org/10.1057/cep.2015.26 Enelow, J., & Hinich, M. (1984). The spatial theory of voting: An introduction. Cambridge University Press. Fox, J. (2015). Applied regression analysis and generalized linear models. SAGE. Geddes, A. (2018). The politics of European Union migration governance. Journal of Common Market Studies, 56(S1), 120–130. https://doi.org/10.1111/jcms.12763 Goldthau, A., & Sitter, N. (2015). A liberal actor in a realist world. The European Union Regulatory State and the global political economy of energy. Oxford University Press. Golub, J. (1999). In the shadow of the vote? Decision-making in the European Community. International Organization, 53(4), 733–764. https://doi.org/10.1162/002081899551057 Golub, J. (2007). Survival analysis and European Union decision-making. European Union Politics, 8(2), 155–179. https://doi.org/10.1177/1465116507076428 Golub, J., & Steunenberg, B. (2007). How time affects EU decision-making. European Union Politics, 8(4), 555–566. https://doi.org/10.1177/1465116507082814 Hagemann, S., & Franchino, F. (2016) Transparency vs. efficiency? A study of negotiations in the Council of the European Union. European Union Politics, 17(3), 408–428. https://doi.org/10.1177/1465116515627017 Heisenberg, D. (2005). The institution of “consensus” in the European Union: Formal versus informal decision-making in the Council. European Journal of Political Research, 44(1), 65–90. https://doi.org/10.1111/j.1475-6765.2005.00219.x Hertz, R., & Leuffen, D. (2011). Too big to run? Analysing the impact of enlargement on the speed of EU decision-making. European Union Politics, 12(2), 193–215. https://doi.org/10.1177/1465116511399162 Hilbe, J. (2011). Negative binomial regression (2nd ed.). Cambridge University Press. https://doi.org/10.1017/CBO9780511973420 Hurka, S., & Haag, M. (2020). Policy complexity and legislative duration in the European Union. European Union Politics, 21(1), 87–108. https://doi.org/10.1177/1465116519859431 Imbens, G. W., & Lemieux, T. (2008). Regression discontinuity designs: A guide to practice. Journal of Econometrics, 142(2), 615–635. https://doi.org/10.1016/ j.jeconom.2007.05.001 https://changing-sp.com/ https://doi.org/10.1111/j.2517-6161.1972.tb00899.x https://doi.org/10.1057/cep.2015.26 https://doi.org/10.1111/jcms.12763 https://doi.org/10.1162/002081899551057 https://doi.org/10.1177/1465116507076428 https://doi.org/10.1177/1465116507082814 https://doi.org/10.1177/1465116515627017 https://doi.org/10.1111/j.1475-6765.2005.00219.x https://doi.org/10.1177/1465116511399162 https://doi.org/10.1017/CBO9780511973420 https://doi.org/10.1177/1465116519859431 https://doi.org/10.1016/j.jeconom.2007.05.001 https://doi.org/10.1016/j.jeconom.2007.05.001 168 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov Jevnaker, T., & Wettestad, J. (2017). Ratcheting up carbon trade: The politics of reforming EU emissions trading. Global Environmental Politics, 17(2), 105–124. https://doi.org/10.1162/GLEP_a_00403 Kaveshnikov, N. (2021a). Analiz vliianiia Evropeiskogo parlamenta i Soveta ES na primere reformy sistemy torgovli parnikovymi gazami [Analysis of the influence of the European Parliament and the Council of the EU exemplified by the EU Emissions Trading System reform]. World Economy and International Relations, 65(6), 21–32. https://doi.org/10.20542/0131-2227-2021-65-6-21-32 Kaveshnikov, N. (2021b). Obychnaia zakonodatel’naia protsedura v ES kak primer kooperativnykh praktik [The ordinary legislative procedure in the EU as an example of cooperative practices]. MGIMO Review of International Relations, 14(1), 126–147. https://doi.org/10.24833/2071-8160-2021-1-76-126-147 Kaveshnikov, N., & Domanov, A. (2022). Skorost’ zakonodatel’nogo protsessa v Evropeiskom Soiuze: Kolichestvennyi analiz [Factors behind legislative duration in the European Union: Quantitative analysis]. International Trends, 20(1), 80–108. https://doi.org/10.17994/IT.2022.20.1.68.3 Kirpsza, A. (2022). Time is of the essence: Explaining the duration of European Union lawmaking under the co-decision procedure. International Political Science Review, 43(4), 564–579. https://doi.org/10.1177/01925121211036423 Klüver, H., & Sagarzazu, I. (2013). Ideological congruency and decision-making speed: The effect of partisanship across European Union institutions. European Union Politics, 14(3), 388–407. https://doi.org/10.1177/1465116512472938 König, T. (2007). Divergence or convergence? From ever-growing to ever-slowing European legislative decision-making. European Journal of Political Research, 46(3), 417–444. https://doi.org/10.1111/j.1475-6765.2006.00648.x-i1 Lenaerts, K., & van Nuffel, P. (2005). Constitutional law of the European Union (2nd ed., R. Bray, Ed.). Sweet & Maxwell. Mattila, M. (2009). Roll call analysis of voting in the European Union Council of Ministers after the 2004 enlargement. European Journal of Political Research, 48(6), 840–857. https://doi.org/10.1111/j.1475-6765.2009.01850.x Mattila, M., & Lane, J.-E. (2001). Why unanimity in the Council?: A roll call analysis of Council voting. European Union Politics, 2(1), 31–52. https://doi.org/10.1177/1465116501002001002 Mišík, M. (2016). On the way towards the Energy Union: Position of Austria, the Czech Republic and Slovakia towards external energy security integration. Energy, 111, 68–81. https://doi.org/10.1016/j.energy.2016.05.056 Morgan, S. L., & Winship, C. (2007). Counterfactuals and causal inference: Methods and principles for social research. Cambridge University Press. https://doi.org/10.1162/GLEP_a_00403 https://doi.org/10.20542/0131-2227-2021-65-6-21-32 https://doi.org/10.24833/2071-8160-2021-1-76-126-147 https://doi.org/10.17994/IT.2022.20.1.68.3 https://doi.org/10.1177/01925121211036423 https://doi.org/10.1177/1465116512472938 https://doi.org/10.1111/j.1475-6765.2006.00648.x-i1 https://doi.org/10.1111/j.1475-6765.2009.01850.x https://doi.org/10.1177/1465116501002001002 https://doi.org/10.1016/j.energy.2016.05.056 Changing Societies & Personalities, 2024, Vol. 8, No. 1, pp. 151–172 169 Plechanovová, B. (2011). The EU council enlarged: North-South- East or core-periphery? European Union Politics, 12(1), 87–106. https://doi.org/10.1177/1465116510390720 Potemkina, O. (2019). Posle krizisa: “Novyi start” migratsionnoi politiki ES [After the crisis: “A new start” of the EU migration policy]. Sovremennaya Evropa, 6, 19–29. http://doi.org/10.15211/soveurope620191829 Rasmussen, A. (2012). Twenty years of co-decision since Maastricht: Inter- and intrainstitutional implications. European Integration, 34(7), 735–751. https://doi.org/10.1080/07036337.2012.726012 Rasmussen, A., & Toshkov, D. (2011). The inter-institutional division of power and time allocation in the European Parliament. West European Politics, 34(1), 71–96. https://doi.org/10.1080/01402382.2011.523545 Rasmussen, A., & Toshkov, D. (2013). The effect of stakeholder involvement on legislative duration: Consultation of external actors and legislative duration in the European Union. European Union Politics, 14(3), 366–387. https://doi.org/10.1177/1465116513489777 Scharpf, F. W. (2006). The joint decision trap revisited. Journal of Common Market Studies, 44(4), 845–864. https://doi.org/10.1111/j.1468-5965.2006.00665.x Schulz, H., & König, T. (2000). Institutional reform and decision-making efficiency in the European Union. American Journal of Political Science, 44(4), 653–666. https://doi.org/10.2307/2669273 Steunenberg, B. (2002). An even wider Union: The effects of enlargement on EU decision-making. In B. Steunenberg (Ed.), Widening the European Union: The politics of institutional change and reform (pp. 97–118). Routledge. Thomson, R. (2009). Actor alignments in the European Union before and after enlargement. European Journal of Political Research, 48(6), 756–781. https://doi.org/10.1111/j.1475-6765.2009.00848.x Thomson, R., Arregui, J., Leuffen, D., Costello, R., Cross, J., Hertz, R., & Jensen, T. (2012). A new dataset on decision-making in the European Union before and after the 2004 and 2007 enlargements (DEUII). Journal of European Public Policy, 19(4), 604–622. https://doi.org/10.1080/13501763.2012.662028 Toshkov, D. (2017). The impact of the Eastern enlargement on the decision- making capacity of the European Union. Journal of European Public Policy, 24(2), 177–196. https://doi.org/10.1080/13501763.2016.1264081 Toshkov, D., & Rasmussen, A. (2012). Time to decide: The effect of early agreements on legislative duration in the EU. European Integration online Papers, 16, Article 11. https://doi.org/10.1695/2012011 Tsebelis, G. (2003). Veto players: How political institutions work. Princeton University Press. https://doi.org/10.1515/9781400831456 https://changing-sp.com/ https://doi.org/10.1177/1465116510390720 http://doi.org/10.15211/soveurope620191829 https://doi.org/10.1080/07036337.2012.726012 https://doi.org/10.1080/01402382.2011.523545 https://doi.org/10.1177/1465116513489777 https://doi.org/10.1111/j.1468-5965.2006.00665.x https://doi.org/10.2307/2669273 https://doi.org/10.1111/j.1475-6765.2009.00848.x https://doi.org/10.1080/13501763.2012.662028 https://doi.org/10.1080/13501763.2016.1264081 https://doi.org/10.1695/2012011 https://doi.org/10.1515/9781400831456 170 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov Veen, T. (2011). The dimensionality and nature of conflict in European Union politics: On the characteristics of intergovernmental decision-making. European Union Politics, 12(1), 65–86. https://doi.org/10.1177/1465116510391918 Wallace, H., Pollack, M., & Young, A. (Eds.). (2015). Policy-making in the European Union (7th ed.). Oxford University Press. Youngs, R. (2020). EU foreign policy and energy strategy: Bounded contestation. Journal of European Integration, 42(1), 147–162. https://doi.org/ 10.1080/07036337.2019.1708345 Zimmer, C., Schneider, G., & Dobbins, M. (2005). The contested Council: Conflict dimensions of an intergovernmental EU institution. Political Studies, 53(2), 403–422. https://doi.org/10.1111/j.1467-9248.2005.00535.x Zorn, C. (2007). Temporal change and the process of European Union decision-making. European Union Politics, 8(4), 567–576. https://doi.org/10.1177/1465116507082815 https://doi.org/10.1177/1465116510391918 https://doi.org/10.1080/07036337.2019.1708345 https://doi.org/10.1080/07036337.2019.1708345 https://doi.org/10.1111/j.1467-9248.2005.00535.x https://doi.org/10.1177/1465116507082815 Changing Societies & Personalities, 2024, Vol. 8, No. 1, pp. 151–172 171 Appendix The Cox Models of the Decision-Making Duration Table A1 Impact of 1995 Enlargement on the Duration of Decision-Making (The Cox Model) Parameter Model 1 Model 2 (Maastricht Treaty impact excluded) All acts Directives Regulations All acts Directives Regulations EU membership EU-12 1.20** 1.32 1.14 1.19* 1.15 1.02 EU-15 0.83** 0.76 0.88 0.84* 0.87 0.98 Control parameters Type of the act Directive 0.33*** – – 0.31*** – – Regulation 3.04*** – – 3.19*** – – Legislative procedure Ordinary legislative procedure 0.65** 0.73 0.40* Cooperation 0.48*** 0.99 0.22*** 0.39*** 0.81 0.39*** Consultation 0.69*** 1.38 0.63*** 0.52*** 1.23 2.55*** Council without EP 3.45*** 3.29*** 3.05*** 6.81*** Assent 1.17 1.18 0.98 Decision-making rule in the Council Unanimity 0.46*** 0.62 0.39*** 0.48*** 0.41* 0.45*** QMV 2.16*** 1.60 2.56*** 2.10*** 2.46* 2.23*** Novelty of the act Amendments 1.26*** 1.59** 1.27* 1.36*** 1.97** 1.21* New act 0.79*** 0.62** 0.79* 0.73*** 0.51** 0.83* Temporary scope of the act Permanent 0.37*** 0.18*** 0.42*** 0.42*** 0.27*** 0.40*** Provisional 2.71*** 5.46*** 2.40*** 2.37*** 3.70*** 2.48*** Complexity (length) of the act Short 1.24* 1.38 0.95 1.09 1.97* 1.07 Medium 0.89 0.71 1.08 0.87 0.52* 0.93 Long 0.78 0.97 0.98 1.10 0.95 0.90 Number of acts (N) 856 179 677 677 81 596 Period, months 48 48 48 48 48 48 Note. The main entries are hazard ratios (HR). Empty cells mean that data are not applicable because of absence/small number of documents with such parameter. * p ≤.05, ** p ≤ .01, *** p < .005 https://changing-sp.com/ 172 Nikolay Yu. Kaveshnikov, Aleksey O. Domanov Table A2 Impact of 2004 Enlargement on the Duration of Decision-Making (The Cox Model) Parameter Model 1 Model 2 (Nice Treaty impact excluded) All acts Directives Regulations All acts Directives Regulations EU membership EU-15 0.78*** 0.79 0.84 0.75*** 0.81 0.87 EU-25 1.28*** 1.26 1.20 1.32*** 1.23 1.14 Control parameters Type of the act Directive 0.48*** – – 0.46*** – – Regulation 2.08*** – – 2.19*** – – Legislative procedure Ordinary legislative procedure 0.35*** 0.88 0.29*** 0.34*** 0.89 0.43 Cooperation 2.29 2.04 2.19 2.69 Consultation 1.17 0.95 1.08 1.15 1.13 0.96 Council without EP 4.59*** 9.35*** 4.70*** 4.49*** 2.86 Assent 0.92 0.98 0.98 0.34 Decision-making rule in the Council Unanimity 0.45*** 1.15 0.66** 0.68*** 1.02 0.58*** QMV 2.23*** 0.87 1.51** 1.48*** 0.98 1.72*** Novelty of the act Amendments 1.56*** 1.72*** 1.33** 1.57*** 1.55** 1.38*** New act 0.64*** 0.58*** 0.76** 0.64*** 0.64** 0.73*** Temporary scope of the act Permanent 0.66** 0.43** 0.72 0.63*** 0.41** 0.78 Provisional 1.51** 2.34** 1.41 1.58*** 2.45** 1.27 Complexity (length) of the act Short 1.50*** 1.55** 1.52*** 1.44*** 1.65*** 0.95 Medium 0.77*** 0.68* 0.87 0.79* 0.61** 1.08 Long 0.76*** 0.92 0.60*** 0.69*** 0.73 0.98 Number of acts (N) 682 195 487 649 191 458 Period, months 48 48 48 48 48 48 Note. The main entries are hazard ratios (HR). Empty cells mean that data are not applicable because of absence/small number of documents with such parameter. * p ≤ .05, ** p ≤ .01, *** p < .005.