Abstract
The Lisbon Treaty fundamentally changed EU legislation. It introduced a new hierarchy of acts, new types of acts and new decision-making procedures. However, instead of replacing the pre-Lisbon types of acts and procedures, it added to them. The legislative landscape is therefore now more complex than before. However, to political scientists, this complexity represents new opportunities to study power and influence. But these opportunities have been under-exploited so far. The article explains the complex post-Lisbon situation and uses legislative output from the period 2012–2017 to discuss how much the various types of acts and decision procedures are used in practice. The advantages and drawbacks of the EUR-Lex database are discussed, and a guide to navigate this database is provided. The article concludes by pointing to the potential of using the full complexity of the post-Lisbon legislative landscape and the EUR-Lex database for analysing the politics of the European Union.
Introduction
The Lisbon Treaty brought fundamental changes to the European Union (EU) legislative landscape. It introduced a new hierarchy of acts, new types of acts and new decision-making procedures. However, it added to, rather than replaced, existing acts and procedures. The legislative landscape is therefore arguably more complex than before. However, to political scientists, this complexity represents new opportunities to study power and influence, but these opportunities have been under-exploited so far. The purpose of this article is to make the case for more use of post-Lisbon legislation to study the EU.
For many years, empirical studies of decision-making in the EU had to rely on case studies. For example, Tsebelis’ (1994) famous study of the European Parliament’s conditional agenda-setting power relied on formal–theoretical analysis and a case study of legislation on small-car emission standards. This situation changed with the introduction of the EU’s legislative databases, especially CELEX (today EUR-Lex). Political scientists have made good use of this opportunity. The databases have been used to address a number of important questions, such as the power of the European Parliament under the consultation procedure (Kardasheva, 2009), the internal workings of the Council of Ministers (Häge, 2008), decision-making efficiency in the EU (Golub, 1999) and the relative power of the European Parliament and the Council of Ministers under the co-decision procedure (Hagemann and Høyland, 2010). In addition, the use of the databases has generated a lively debate on the relative merits of different approaches and ways of categorizing the various types of legislation (e.g. Golub, 2007; Häge, 2011; König, 2008).
However, these debates have largely taken place before the Lisbon Treaty. This Treaty brought fundamental changes to EU legislation, but did not make a clean break with the past. So, many pre-Lisbon decision procedures and types of acts live on. For political scientists, this means that there are new opportunities for using EU legislation to study power and influence in the EU. However, so far, debates about the new legislative landscape have mainly taken place among EU legal scholars, not political scientists (Bergström and Ritleng, 2016; Craig and de Búrca, 2015; Hofmann, 2009; Mendes, 2013; Peers and Costa, 2012). In fact, political scientists have hardly discussed it at all (but see Brandsma and Blom-Hansen, 2017; Christiansen and Dobbels, 2013; Kaeding and Stack, 2015).
This article argues for the research opportunities offered by post-Lisbon legislation and the EU’s legislative databases. To illustrate, three specific examples of how the post-Lisbon legislative scene may provide political scientists with more analytic leverage over their research questions can be offered.
First, Junge et al. (2015) argue that when the legislature (Council and Parliament) is divided, the Commission becomes more active because legislative overrule becomes less likely. This relationship is logically dependent on the legislative decision rule for overruling the Commission. However, since the study is based on pre-Lisbon data, there is limited variation in this decision rule. A similar study based on post-Lisbon data would have considerable variation available. The decision rule for the member states to overrule the Commission now varies between blocking minority (implementing acts, variant of examination procedure), super-qualified majority (delegated acts), normal qualified majority (implementing acts, variant of examination procedure), simple majority (implementing acts, variant of examination procedure) and no overrule possible (implementing acts, advisory procedure or no procedure). In addition, there is variation in the involvement of the European Parliament in decisions to overrule (delegated acts versus implementing acts).
Second, Hagemann and Høyland (2010) study bicameral politics in the EU based on pre-Lisbon data. But the study only focuses on one legislative procedure, co-decision. However, the consultation and consent procedures as well as procedures not involving the European Parliament are still empirically relevant in the post-Lisbon era. A complete understanding of bicameral politics would therefore benefit from taking all decision procedures into account.
Third, Heritier and Moury (2011) argue that the Commission and the Council may collude to circumvent the Parliament by shifting regulation from the legislative arena to the executive arena, where rules are made by the Commission acting under comitology procedures. Their study is based on pre-Lisbon data, but actors have more options open for arena shifting after the Lisbon Treaty, which introduced two delegation regimes (the delegated and implementing acts regimes) providing the Council and Parliament with different control positions. A similar study in the post-Lisbon era would therefore have more empirical variation available and thus potentially have more analytic leverage over the arena-shifting argument.
As these examples suggest, researchers can now more effectively address questions about legislative overrule, bicameral politics and delegated rule-making in the EU. These questions are, however, only teasers; more will be discussed in the conclusion. A secondary purpose of the article is to provide some help in navigating both the legislative landscape and the EUR-Lex database.
In the following, I first introduce the Lisbon ideal, the simplification of the EU’s legislation pursued by the Treaty reformers. Second, I evaluate the limited extent to which this ideal has been met in practice. The resulting complex post-Lisbon legislative landscape is then laid out in detail. Third, I turn to the legislative output to discuss how much the various types of acts and decision procedures are used in practice. In this context, I discuss the potentials and limits of the EUR-Lex database and illustrate them with the amount of legislative output from the period 2012–2017. Before concluding, I briefly discuss a type of act that raises a particular challenge for researchers: EU decisions with addressees.
The Lisbon ideal
The Lisbon Treaty introduced a new legal hierarchy. It distinguishes between legislative and non-legislative acts and introduces new types of acts; delegated acts and implementing acts. It is sometimes said to imply a three-tiered hierarchy between legislative acts and the new delegated and implementing acts. However, the exact relationship between delegated and implementing acts is far from clear and actively debated by legal scholars (Bast, 2016; Bradley, 2016; Craig, 2016). This is not the place to settle that question, so I pragmatically illustrate the Lisbon ideal with delegated and implementing acts placed side-by-side. In this pragmatic reading, the Lisbon ideal is a two-tiered hierarchy 1 with legislative acts at the top and non-legislative acts at the bottom (see Table 1).
The Lisbon ideal of EU legislation.
In the following, I lay out the types of acts in the Lisbon ideal in more detail. In the next section, I discuss how incomplete the ideal is as a representation of the EU’s post-Lisbon legislative landscape.
Legislative acts are acts decided by the ordinary legislative procedure or by special legislative procedures. The aim of the Lisbon Treaty was to turn the co-decision procedure into the standard legislative procedure. It was therefore renamed the ordinary legislative procedure, and its use was applied to more areas. However, in a number of instances, other procedures are still used. According to the Treaty, when these procedures are used, acts are to be adopted by the Council with the participation of the European Parliament or by the latter with the participation of the former. Collectively, they are known as special legislative procedures (Treaty on the Functioning of the European Union (TFEU), Article 289). In practice, these procedures are the consultation procedure (e.g. TFEU, Article 118) or the consent procedure (e.g. TFEU, Article 86), both well known from the pre-Lisbon period.
Turning to delegated acts, the first type of non-legislative acts, these are a true innovation of the Lisbon Treaty. The idea of the Treaty makers was to introduce a distinction between political and technical delegation and to install control mechanisms accordingly (Brandsma and Blom-Hansen, 2017: 55–69). Political delegation was to be done by the new delegated acts system. Under this system, rule-making power delegated to the Commission is controlled by both the European Parliament and the Council. When the Commission adopts a delegated act, it does not enter into force until the expiry of a scrutiny period (normally two months) in which both the Parliament and the Council may veto it. More dramatically, the Council and the Parliament may also revoke delegated powers (TFEU, Article 290).
The second type of non-legislative act is the implementing act, which was meant for technical delegation. While delegated acts can only be made by the Commission, the power to issue implementing acts can also ‘in duly justified specific cases’ be delegated to the Council (TFEU, Article 291). When the power to adopt implementing acts is delegated to the Commission, the European Parliament and the Council may decide that the Commission must first consult a committee of member state representatives. This is known as comitology, a control system well known from the pre-Lisbon period. Comitology committees discuss the Commission’s draft implementing acts and vote on the final version. If the vote is unfavourable, the Commission may have to abandon its proposed implementing act.
The Lisbon ideal: An incomplete picture
The Lisbon ideal presented above provides a first map to navigate the post-Lisbon legislative landscape. However, the map is misleading since there are many types of acts that do not fit into its categories. The main reason is that the pre-Lisbon situation continues in a number of respects.
First, the Lisbon Treaty’s distinction between legislative and non-legislative acts is seriously misleading because there are many more types of non-legislative acts than suggested by the ideal. The most dramatic example is perhaps that acts adopted by the European Parliament and the Council through the consultation and consent procedures can be either legislative or non-legislative acts depending on the exact formulation of the relevant Treaty article. An act is only legislative if the expression ‘legislative procedure’ appears in the Treaty article, which functions as the legal base of the act (Bast, 2012; Craig and de Búrca, 2015: 113–114). However, there are many instances of Treaty articles specifying the consultation or consent procedure without stating that this constitutes a special legislative procedure. Acts adopted with such Treaty articles as legal base are non-legislative acts. Examples are provided in Table 2. The distinction between legislative and non-legislative acts may appear legalistic. But it has important real-world consequences (Bast, 2016). For example, only legislative acts can delegate the power to adopt delegated acts; the requirement for the Council of Ministers to meet in public when it negotiates and decides only applies to legislative acts; and the involvement of national parliaments under the subsidiarity principle only applies to legislative acts.
Examples of Treaty bases for legislative and non-legislative acts adopted by the European Parliament and Council.
Note: Italics added.
Second, even after the Lisbon Treaty, there are still a number of Treaty provisions that privilege the Council as legislator and do not provide for involvement of the European Parliament. Examples include TFEU, Article 31 (common customs union), Article 43 (common fisheries policy) and Article 122 (economic policy). In these instances, the Council decides based on a proposal from the Commission. However, this is not a legislative procedure as defined in TFEU, Article 289, so these Council acts are non-legislative acts.
Third, the EU treaties stipulate that other actors than the European Parliament and the Council may adopt acts. The most important such actor is the Commission. 2 Thus, a number of Treaty articles directly empower the Commission to issue acts, for example, TFEU, Article 105 (competition policy) and Article 108 (state aid). Commission acts based on direct Treaty power are non-legislative acts, but do not fit easily into the Lisbon ideal.
Fourth, until the Lisbon Treaty, tertiary acts, i.e. acts based on power delegated in secondary acts, were issued by the Commission acting under comitology procedures, although in some cases by the Council. The ambition of the Lisbon Treaty was to create a new control structure. From now on, tertiary regulation was to be done by the new delegated and implementing acts. However, the Treaty left an alignment issue: what to do with all the provisions in existing secondary acts, which specified that control of the Commission’s delegated powers was to take place by the old comitology procedures? Should they be left in place or should they be aligned to the new delegation regimes? A new 2011 comitology regulation provided a partial solution: pre-Lisbon comitology procedures were automatically changed into the new comitology procedures under the implementing acts regime (Regulation 2011/182, Article 13). However, this held with one exception: the regulatory procedure with scrutiny. The Commission considered this particular pre-Lisbon comitology procedure similar to the new delegated acts regime, so it proposed automatic alignment to the new delegated acts regime. However, this turned out to be controversial, and inter-institutional negotiations dragged on for years and are still not settled at the time of writing. The main issue is that the member states are not ready to give up ex-ante comitology control for the ex-post control mechanisms in the new delegated acts system (Brandsma and Blom-Hansen, 2017: 88–92). Therefore, the regulatory procedure with scrutiny is still used in a number of cases. Acts adopted under this procedure are tertiary non-legislative Commission acts – but neither implementing nor delegated acts. Thus, again, we have a type of act that does not fit easily into the Lisbon ideal.
Finally, in a number of instances, secondary legislation from the pre-Lisbon period empowers the Commission to adopt acts without comitology control. Since there is no comitology procedure involved, there is no alignment issue in relation to the delegation regimes introduced by the Lisbon Treaty. These legal bases are therefore likely to remain in place for a considerable amount of time and only gradually align with the implementing acts regime as the need to revise the basic acts arise. Again, we have a type of act that does not square well with the Lisbon ideal.
We are now ready to provide a more complete picture of the post-Lisbon EU legislative landscape (see Table 3). 3 The main point is that the types of acts and decision-making procedures are more plentiful than ever. This should provide food for thought for political scientists because it represents a fertile empirical testing ground for many theories of power and influence in the EU.
The post-Lisbon EU legislative landscape.
Note: The table excludes some rarely used types of acts, see Note 3.
How much are the EU’s decision-making procedures used?
This section provides an introduction to the relative importance of the EU’s types of acts and decision-making procedures. This is done by overviews of the number of acts adopted under the respective procedures in the six-year period 2012–2017. 4 I focus on binding acts (directives, regulations, decisions) and exclude non-binding acts (recommendations and opinions) and make a gross count of all binding acts. 5
The overviews are based on searches in EUR-Lex, the official legislative database of the EU. The database is run by the EU’s Publications Office. The first version of CELEX (Communitatis Europeae Lex), the predecessor of EUR-Lex, became operational in 1971. It was opened for public access in 1980, but only in French and not with full coverage of EU law. An Internet version was launched in 1997. Work on EUR-Lex began in the late 1990s and was meant to be an integrated service built around CELEX providing access to all EU legal documents (Düro, 2009). After the turn of the millennium, EUR-Lex was merged with CELEX and later with PRE-Lex, a database that made it possible to follow documents through the legislative process. Today, EUR-Lex is therefore the EU’s main legal database. 6 It provides free access to the Official Journal of the European Union, to EU law, to preparatory acts and to the decision-making procedures of acts.
In the following, oversights of the annual number of acts adopted under the EU’s various decision-making procedures are provided. Sometimes it takes a creative approach to identify acts in EUR-Lex. Therefore, this section (and the Online appendix) also serve as an introduction to using the EUR-Lex database. The following tables are built upon the category of acts developed above in Table 3. To keep things manageable, the presentation is divided into three separate tables corresponding to sections I–III in Table 3.
Table 4 shows the number of legislative acts in the period 2012–2017. Identifying legislative acts is relatively straightforward in EUR-Lex (see the Online appendix for guidance). Two things stand out in Table 4. First, the ordinary legislative procedure has indeed become the most common legislative procedure, although it may be less dominant than commonly thought. The special legislative procedures are still used, but less frequently. Second, the EU’s legislative activity took a rather dramatic turn downwards with the Juncker Commission, which took over from the Barroso Commission in late 2014. Thus, there is some evidence to suggest that Juncker delivered on his promise to ‘focus on the “big things”… [and] not present proposals that do not contribute to these priorities’ (European Commission 2014: 2).
Number of legislative acts, 2012–2017.
Source: EUR-Lex.
Note: See the Online appendix for the exact EUR-Lex search commands used to identify the acts.
Turning to non-legislative acts based on Treaty provisions (section II in Table 3), identifying the different types of acts, is more challenging. There is no quick command in the EUR-Lex search function that allows an unambiguous identification. Therefore, some creativity is necessary (see the Online appendix for guidance). Table 5 shows how numerous Treaty-based non-legislative acts are in the period 2012–2017. It may come as a small surprise that this type of act, which is rarely mentioned in the literature, is in fact quite common. However, again, we see a drop in legislative activity in 2015 when the Juncker Commission took over from the Barroso Commission – although less clearly so than in the area of legislative acts.
Number of non-legislative acts based on Treaty provisions, 2012–2017.
Source: EUR-Lex.
Note: See the Online appendix for the exact EUR-Lex search commands used to identify the acts.
Turning finally to acts based on provisions in secondary acts (section III in Table 3) – also known as tertiary acts – delegated and implementing acts are relatively straightforward to identify in EUR-Lex, while both Commission acts adopted under the regulatory procedure with scrutiny and Commission acts based on pre-Lisbon secondary acts and adopted without comitology control are challenging to identify (but see the Online appendix for guidance).
As is evident from Table 6, tertiary acts constitute the most numerous acts in the EU. The first type of tertiary act is the delegated act, the genuinely new type of act introduced by the Lisbon Treaty. As Table 6 shows, delegated acts had a slow start. Until around 2014–2015, they were not used very much. This is mainly because the member states were reluctant to delegate powers to the Commission in this way. They did not like to exchange well-known ex-ante comitology control for the new ex-post veto option. Therefore, they objected whenever the Commission proposed delegation provisions in secondary acts that made use of the new delegated acts system. This controversy quickly turned into one of the most vehement interinstitutional battles in the post-Lisbon period (Brandsma and Blom-Hansen, 2017).
Number of non-legislative acts based on provisions in secondary acts (tertiary regulation), 2012–2017.
Source: EUR-Lex.
Note: See the Online appendix for the exact EUR-Lex search commands used to identify the acts.
The next type of tertiary act is the implementing act. This type of act was also introduced by the Lisbon Treaty, but is well known from the pre-Lisbon period. As is evident from Table 6, implementing acts are by far the most numerous type of acts in the post-Lisbon legislative landscape.
The two final types of tertiary acts are left-overs from the pre-Lisbon period. They do not attract a lot of scholarly attention, but as Table 6 shows, they are not empirically irrelevant.
The most striking result in Table 6 is perhaps just how numerous tertiary acts are. In a purely quantitative sense, EU legislation is all about tertiary legislation. It is also noteworthy that, in contrast to secondary legislation, there is no discernable drop in legislative activity in 2015 when the Juncker Commission took over from the Barroso Commission. Tertiary regulation seems more robust to changes in political leadership.
A special challenge: Decisions with addressees
All statistics about EU legislation based on the EUR-Lex database come with one important caveat: EUR-Lex is incomplete as regards decisions with addressees. Decisions constitute one of the EU’s three binding legal instruments (directives, regulations and decisions). A decision may be of general application, or it may have a specific addressee. In the latter case, it is binding only on the addressee (TFEU, Article 288). An addressee can be a member state, a company, an individual or a plurality of these actors. Decisions of general application must be published in the EU’s Official Journal, whereas decisions specifying an addressee must be notified to the party concerned. It may also, but need not, be published in the EU’s Official Journal.
Since the Official Journal is the main source of EUR-Lex content, not all decisions with addressees are included in EUR-Lex. Therefore, all statistics on EU legislation that are based on EUR-Lex and that include decisions should come with this proviso. This also applies to the tables presented in the preceding section.
It is difficult to gauge the extent of this problem. Some might argue that decisions with addressees are largely irrelevant for political scientists because they deal with mundane details. However, is this true? There are at least three ways to address this question. The first is to look at substance. Do these decisions deal with important issues? Sometimes, the answer is clearly yes. One example is the Commission’s decision 2017/1283 addressed to Ireland. Here, the Commission declared Ireland’s tax breaks to the multinational company Apple illegal under EU state aid rules and ordered Ireland to recover the lost tax revenue; a record sum of €13 billion. This was an important decision not only because of the sums involved but also because it may pave the way for tax evasion to be addressed by the EU’s competition rules.
The second way is to look at the addressees. On average, decisions to a single addressee are probably not so important for political scientists. However, sometimes, decisions are addressed to several parties. For example, a decision may be addressed ‘to the Member States’, in which case the difference to a decision with general application is blurred.
The third way is to ask how numerous decisions with addressees are. If there were only few such decisions, the problem would be manageable. Based on data made available by the Commission, this question can be evaluated for the years 2015–2017 as far as Commission decisions are concerned. 7 Table 7 shows that the nature of the problem varies across types of Commission decisions. The most dramatic discrepancy between published and not-published decisions concerns Commission implementing decisions. Every year, about 200–300 of these decisions are published in the EU’s Official Journal (and then included in EUR-Lex). However, as Table 7 shows, there are almost 10 times as many Commission implementing decisions with addressees. For the other types of Commission decisions in Table 7, the problem is also evident, but it is not as serious.
Number of Commission decisions, 2015–2017.
aCommission decisions with a legal basis in treaties or Commission decisions taken under the regulatory procedure with scrutiny or Commission decisions adopted without comitology control and based on pre-Lisbon secondary acts.
Source: Lines 1, 3 and 5: EUR-Lex (see the Online appendix for the exact EUR-Lex search commands used to identify the acts). Lines 2, 4 and 6: Numbers kindly made available by the Commission upon request by the author.
In sum, decisions with addressees may be substantively important, may apply to a wide range of actors and are numerous. The fact that EUR-Lex only includes a small minority of them is a problem that political scientists should take seriously and handle carefully in their research.
Conclusion: A fertile, but complex, empirical testing ground
The Lisbon Treaty did not make a clean break with the past. This means that it added to, rather than replaced, existing decision procedures and types of acts. This may appear bewildering. However, for political scientists, the post-Lisbon legislative landscape represents a fertile empirical field. The various decision procedures and types of acts combined with the free access to EUR-Lex constitute a unique data source.
It is not difficult to think of questions that may be addressed more systematically with these data. One set of questions deal with the legislative arena. As a baseline, the EUR-Lex database can be used to paint a full-scale descriptive picture of all types of EU legislation and their distribution across, e.g. policy areas. But more analytical questions can also be addressed. What is the relative power position of the European Parliament and Council? The two institutions are commonly thought to be equals because of the expanded use of co-decision. However, as this article has made clear, much legislation is still done according to other procedures. How does the EU’s legislative activity develop over time? Given the multitude of types of acts, this question is more complex to answer than it appears at first sight. How much does decision-making efficiency depend on decision procedure? Speed versus inclusion and deliberation is a classic dilemma in this context. To what extent are the Commission’s proposals changed during the course of the legislative process? The full scale of EU legislation in the EUR-Lex database makes it possible to address all these questions in a large-N set-up.
Another set of questions deal with delegation. When do the member states and the European Parliament delegate power to the Commission? The EUR-Lex database makes it possible not only to paint a full-scale descriptive picture of, say, the use of delegated and implementing acts across policy areas and time, but also to investigate theories of delegation that posit that delegation is caused by factors such as complexity, conflict and decision rules (Franchino, 2007). Other more specific delegation puzzles can also be analysed. Can and does the European Parliament use its hard-won control rights over delegated legislation in practice? The Parliament fought hard for the delegated acts regime, but controlling technically complex delegated legislation is difficult, time-consuming and not always electorally rewarding. Are the Commission’s delegated powers best controlled by ex-ante or ex-post mechanisms? The implementing and delegated acts systems represent two very different approaches to control of executive power. Their relative effectiveness can be evaluated by the EU post-Lisbon legislative experience.
Political scientists already address these questions in skillful and creative ways. However, the post-Lisbon legislative landscape – the full range of legislative procedures and the full range of types of acts – combined with the EUR-Lex database represents an underexploited resource for addressing them even better.
Supplemental Material
Supplemental material for Studying power and influence in the European Union: Exploiting the complexity of post-Lisbon legislation with EUR-Lex
Supplemental Material for Studying power and influence in the European Union: Exploiting the complexity of post-Lisbon legislation with EUR-Lex by Jens Blom-Hansen in European Union Politics
Footnotes
Acknowledgements
The author owes many thanks to Irina Tanasescu, Elizabeth Golberg, Harald Stieber and Gijs Jan Brandsma for helpful comments and discussions. Any remaining flaws are the author’s fault.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research in this article is supported by the Danish Council for Independent Research (grant # 6109-00018B).
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
