European integration assessed in the light of the ‘rules vs. standards debate’

Abstract

The interplay of various legal systems in the European Union (EU) has long triggered a debate on the tension between uniformity and diversity of Member States’ (MS) laws. This debate takes place among European legal scholars and is also paralleled by economic scholars, e.g. in the ambit of the ‘theory of federalism’. This paper takes an innovative perspective on the discrepancy between ‘centralized’ and ‘decentralized’ law-making in the EU by assessing it with the help of the rules versus standards debate. When should the EU legislator grant the national legislator leeway in the formulation of new laws and when should all be fixed ex ante at European level? The literature on the ‘optimal shape of legal norms’ shall be revisited in the light of law-making in the EU, centrally dealing with the question how much discretion shall be given to the national legislator; and under which circumstances. This paper enhances the established decisive factors for the choice of a rule or a standard in a national setting (complexity, volatility, judges’ specialization and frequency of application) by two new crucial factors (switching costs and the benefit of uniformity in terms of information costs) in order to assess law-making policies at EU level.

1 Introduction

A central debate in European Law when it comes to law-making for the whole Union deals with the tension between uniformity and diversity (De Witte et al. 2001; De Búrca and Scott 2000; Dougan 2004; Tuytschaever 1999; Hunt 2010). One policy is the imposition of European-wide equal norms. In contrast to that stands the desire to promote diversity and flexibility throughout the Member States (MS). The central question becomes how far the European Union (EU) legislator is to have the final say on the laws to be adopted in the MS or how much discretion should be given to national legislators to consider national particularities? How can this form of federalism be organized? Depending on the choice of the legal instrument effective implementation of European law can depend to a large extent on the cooperation of the MS (Rehbinder and Stewart 1985). It is for them to decide on the ‘necessary organisational, legal and institutional arrangements’. They furthermore highly differ as to their administrative tradition which leads to different necessities when integrating EU law (Falkner et al. 2005, p. 319; Siedentopf and Hauschild 1990, p. 451; Knill 1998, p. 7).¹

Many economic scholars have dealt with harmonization, centralization and decentralization labeled as the theory of federalism and the theory of regulatory competition (Tiebout 1956; Stigler 1957; Arcuri and Dari-Mattiacci 2010),² and also particularly as to its shape in the EU (Van den Bergh 2002a; b; Weatherill 2004; Ott and Schäfer 2002; Kieninger 2002). This paper will approach the question on the degree of uniformity in particular settings from an innovative new point of view: the rules versus standards debate.

The rules versus standards debate in Law and Economics and legal literature dates back to a seminal paper by Louis Kaplow from 1992 (also Kaplow 1995, 2000).³ Since then this line of research has seen various applications, e.g. to the literature on effective laws in developing countries (Cooter and Schäfer 2008; Faure 2008, 2010; Ogus 2008; Schäfer 2006) or the extensions to behavioural effects (e.g. Korobkin 2000) or e.g. applications in the field of anti-trust (Crane 2007).

While the debate is generally used in a national context as to the relationship between legislature and judiciary, in the context of the EU the debate shall be lifted up to the relationship between two legislators: the European legislator and the national legislator: The literature on ‘rules versus standards’ shall be revisited in the light of law-making in the EU, centrally dealing with the question how much discretion shall be given to the national legislator; and under which circumstances; and further by which type of legal norm this can be reflected? When it comes to legal norms the discussion, as will be motivated, shall focus on the differentiation between the European legal instruments regulation and directive. Notwithstanding this application to the EU context, a more general usage of the theory is possible in any setting where differing legal traditions are organized in a somewhat federal way and some discretion as to law-making can be left to a lower level.

The established decisive factors for the choice of a rule or a standard in a national setting (complexity, volatility, judges’ specialization and frequency of application) will be enhanced by two new crucial factors (switching costs and the benefit of uniformity in terms of information costs) in order to assess law-making policies on European level. After a summary of the ‘rules versus standards’ debate (1), the analogy to EU law making will introduced (2), illustrated and the two new factors applied (3) followed by conclusions (4).

2 The ‘Rules vs. Standards’ debate in the national context

In this section first a characterization of rules and standards is carried out, followed by a more detailed assessment of how to adapt shapes of norms to legal environments.

2.1 General characteristics

One aspect of the debate is a straightforward point: Norms differ as to their degree of precision, detail or complexity. The spectrum ranges from ‘relative simplicity to complexity’.⁴ In balancing the degree of complexity the following holds true: ‘the lower the information costs (for the individual ex ante and the enforcement authority ex post) and the greater the variation in harm caused by different individuals’ actions, the larger will be the net benefit of [high] precision in legal commands’.⁵ This characteristic is discussed as to norms generally. In order to distinguish rules from standards, the notions of clarity and flexibility are crucial though. The basic distinction between rules and standards is the time of promulgation of the respective degree of detail (Kaplow 1992, p. 586).⁶ It concerns whether the law is given content, specification, ex ante or ex post (Kaplow 1995, p. 508; Kaplow 2000), i.e. done before individuals act (rules) or afterwards (standards) (Kaplow 1992, p. 621). Rules are those legal commands which lead to a clear-cut distinction between lawful and unlawful, while standards are general legal criteria which require a more elaborate judicial decision making (Ehrlich and Posner 1974; Diver 1983; Kaplow 1992; Schäfer 2002). The latter requires an application of ‘a background principle or a set of principles to a particularized set of facts in order to reach a legal conclusion’ (Korobkin 2000, p. 23).

As an example with a low degree of detail one can look at a provision setting out a speed limit. A provision stipulated in form of a rule would make the issue clear cut e.g. the speed limit shall be at 60 km/h. In form of a standard, on the other hand, it would simply read that ‘reasonable speed’ should be used. The determination of ‘reasonable’ would take place ex post, while ‘60’ was established as the speed limit ex ante in the first example.

The degree of the precision of laws is at two extremes with how the two law-making techniques are formulated (Parisi 2007, p. 9). This does, however, not exclude that a general standard is applied with a lot of precision by the judiciary. The end result can be the same and in specific situations this would mean different costs have to be incurred.

Both approaches involve costs at two stages: at the law-making and at the adjudication stage.⁷ The size of cost that they lead to, however, varies depending on the moment of measurement. For the decision on the use of rules and standards it has to be considered that in the case of rules the content of the laws has to be determined ex ante. Lawmakers have to carry out studies in advance in order to determine the appropriate rule (Kaplow 1992, p. 569). Therefore, rules are more costly for legislators to promulgate than general standards, which require less specificity beforehand.

‘Standards’, however, while leading to lower costs at the law-making stage, impose higher costs at the implementation and (born by judicial bodies) decision-making stage (Kaplow 1992, p. 570). The determination of each law’s content is carried out by the adjudicators ex post. Standards are formulated in an open way and allow for judges to make a fact-specific determination. They leave them furthermore with discretion. The adaptation to local, varying circumstances is better possible by open standards (Fon and Parisi 2007, p. 4.). Learning from a rule is considered to be cheaper because the content of the laws is extensively set out in it beforehand (Kaplow 1992, pp. 563).⁸ This suggests that rules will tend to produce behaviour that is more in conformity with legal norms than would equivalent standards (Kaplow 2000, p. 510). However, too much detail in setting out the rules will reduce the effect considerably. The language used is furthermore important. Standards are more robust to attempts by subjects of the law to avoid legal consequences (Wegner 1997; Kaplow 2000, p. 514; Fon and Parisi 2007, p. 4).⁹

To sum up standards generally have a broader scope of application than rules (Fon and Parisi 2007, pp. 16). Situations do not have to be thought of beforehand. To adapt a rule is more difficult and might require legal amendments. Whenever the human capital at the ex post stage is better qualified to carry out the detailed interpretation of the law, less complexity beforehand is warranted.

As already mentioned both ways can lead to the same result: Option one are ‘rules’ as laws that are specified up front with a greater level of detail and that thus leave a lesser margin of discretion in the implementation of such norms and option two, a standard, can be applied with a high degree of detail by the judiciary.

There is thus an area of conflict between initial specification costs and enforcement and compliance costs (including specification at this stage) (Posner 1998a, b).¹⁰ In the choice between specificity and generality at the moment of promulgation, the overall goal should be the minimization of total costs. The suitability of a standard or a rule depends on some main factors of a legal environment that are set out under the next aspect. Depending on these factors the optimal moment of specifying the provisions (right at the moment of enactment in case of rules or later on in the adjudication and implementation process in the case of standards) can be identified.¹¹

2.2 Adaptation to specific circumstances

Next, factors facilitating the decision-making process on a law’s shape in a given area are discussed. In theory, provided that lawmakers and the judiciary are benevolent, the same content and thus result for the penalty that someone has to face and low or high degree of detail can be reached independently of the choice of standards (ex post) or rules (ex ante). The optimum position is dependant on the circumstances since the costs and benefits of using the various shapes of legal instruments vary. For the choice between rules and standards the following main criteria are decisive for a reduction of total costs (promulgation, adjudication and compliance costs) (Fon and Parisi 2007, pp. 6).¹² More plainly, the allocation of tasks for the law-maker or the judiciary has to be organized efficiently depending on the field of law at hand.

• (1) volatility

Volatility, i.e. changes over time in the regulated environment, lead to legal obsolescence. In an area of law that is subject to a frequent change of economic and social conditions leading to an ever new assessment of the optimal set of legal decisions standards are more efficient than rules (Schäfer 2002, p. 2; Kaplow, 1992). Details and specifications that are set out ex ante become obsolete at a faster rate. Therefore the use of more specific rules might be warranted when there is a stable environment and general standards when there is a fast rate of change (Kaplow 1992, p. 621). Standards are less affected by changes as they indicate only the type of circumstance that is relevant and not the specific circumstance. Detailed rules are more sensitive to exogenous, unforeseen changes. With standards the possibility of a detailed analysis by the judiciary ex post is left open.

The range of potential situations to be covered should be formulated in a very open way to possibly cover future situations, too. This can work best if ex post the norm is given content. In cases of a discrepancy between the ex ante legal norm and the particular circumstances of the case there is a potential for societal losses.

• (2) judges’ specialization

Judges’ specialization affects the level of guidance that they need.¹³ Specialized judges are better in applying complex laws and interpreting them because they are very familiar with the laws in their specialized field. In order to decrease adjudication costs it therefore seems to hold that the optimal level of specificity increases the more specialized the courts are that are to apply and interpret them (Fon and Parisi 2007, p. 8). Therefore a greater use of detailed legal provisions would be expected that cover a wide range of situations and variations (Fon and Parisi 2007, p. 14). Due to their specialization they can, however, also deal with vague norms and do not need clear-cut advice on how to decide a case. A lot of interpretation of the norms could thus take place ex post. Judges can create and use case-law.

Fon and Parisi generalize it to ‘greater use of rules will be warranted when legislative costs are lower relative to judicial costs. An increase in judicial human capital, on the other hand, would lower judicial costs and thus justify the use of less specific laws in response to an increase in complexity of the regulated environment (2007, p. 15)’.

• (3) frequency of application

A common point is the finding that the more often a norm is applied, the more a rule with a higher degree of specificity (beforehand) is desirable as these costs then only have to be born once (Kaplow 1992, p. 563).¹⁴ Frequency speaks thus in favour of rules, as adjudication costs would otherwise most likely be higher than promulgation costs. Standards work best when behaviour varies a lot and there is little repetition in case scenarios (Kaplow 2000, p. 510). To set out the legal consequences of a frequent behaviour in detail beforehand seems warranted. The case for standards would be clearly that of situations which arise rarely, or varying circumstances, or a low frequency.

The relationship between volatility and frequency can be analysed in the following way: If volatility is a given in the environment and consequently obsolescence is a big issue in the legal order, lower levels of specificity should be chosen, basically opposite to frequency.

• (4) the complexity of reality

A last main issue is the complexity of reality in the field in which the norm is to be passed. Generally the more complex an environment is, the more costly it is to develop norms that cover every possible variation. Particularly, promulgation costs increase with the complexity of the environment that is to be regulated as it is difficult to specify these contingencies (Fon and Parisi 2007, p. 7). This situation calls for a general standard that can be filled with the details for many possible contingencies ex post. When deciding on the costs of adjudication there is an interrelation with the frequency of application.

3 An analogy: legal norms enacted by the European Union

Next, the forms of legal instruments that are available in the EU and the law-making process are illustrated and the potential of the analogy to directives and regulations is explored.

The essence of the supranational nature of European Law is the fact that MS have transferred some legislative competences to the EU and by this limited their own sovereign rights to adopt legislation in the areas in question.¹⁵ A general point to make is that the scope of competences of the EU differs as to various legal fields. While exclusive competences have been granted to the EU according to art. 3 TFEU in some areas, only so-called ‘shared’ or ‘supporting’ competences (arts. 4 and 5 TFEU) are available in others.¹⁶ This is apparently a source of uniformity and diversity among legal orders in the EU.

From an economic perspective this current state of the law can be subject to change—as also various Treaty amendments in the past have shown. From an economic efficiency perspective law has not necessarily struck the right balance yet. This paper shall abstract from the de facto allocation of competences and assess along the lines of various criteria when there are reasons to favour uniformity or diversity.¹⁷

3.1 Law-making in the EU

In the EU there is an inherent tension between uniformity and diversity as to the legal orders of the 27 MS currently and potentially after implementing new European laws (De Witte et al. 2001; Dougan 2004; Tuytschaever 1999; Hunt 2010). The central question is how law-making powers are to be shared. When weighing the balance between the European legislator and the national legislator, a closer look at legal instruments and law-making policies within the EU is essential.

With regard to available legal instruments at European level in 2003 it was noted that the ‘theory of instruments has yet to establish its place within Union constitutional law’ (Bast 2003, p. 13).¹⁸ Now it is generally specified in the Treaties which instrument shall be used for which area of law. The legal base as set out in TEU and TFEU is to establish the scope for EU legislation, the procedure and type of law, but regularly the choice of legal base will not be straightforward (Chalmers et al. 2010, p. 95). Furthermore in cases where it is not specified art. 296 TFEU is applicable that reads:

“Where the Treaties do not specify the type of act to be adopted, the institutions shall select it on a case-by-case basis, in compliance with the applicable procedures and with the principle of proportionality…”¹⁹ There is thus still discretion left to the Union institutions (Bast 2003, p. 40)²⁰ which is why different institutions at times seek to use the legal basis that provides the procedure most advantageous to them (Chalmers et al. 2010, pp. 95).²¹ This apparently leads to vigorous litigations (Cullen and Charlesworth 1999). Obviously regulations simplify life for the European Commission and there might thus be a preference for this type of instrument (Bast 2003, p. 13). MS, on the other hand, seem to prefer a two-step procedure that a directive provides for in which they can seem active, even if in substance this might not even be the case (Koopmans 1995, pp. 695; Bast 2003, p. 70).

As to the number of different legal acts, a recent change has occurred. Before the Lisbon Treaty entered into force in 2009 European policy was characterized by the ‘famous’ 3 pillar structure, introduced with the Treaty of Maastricht in 1993.²² This meant that in some areas the entity was of supranational nature, MS referred parts of their competences to the EU and limited their own sovereignty, and in others, where the MS were not ready to grant the EU wide law-making and scrutinizing powers, namely the second and the third pillar, the nature of EU law was intergovernmental. These two different natures of EU law led to a wide variety of legal instruments, binding to different degrees, differing as to addressees etc. Thus with some a high degree of uniformity could be achieved, while by others this was impeded and diversity among the MS was the (desired) result.

With the entering into force of the Lisbon Treaty on 1st December 2009 and the abolishment of the pillar structure the number of legislative instruments has been reduced to five (Piris 2010, pp. 92).²³ The main legislative instruments as to law-making in the EU are directives and regulations (Craig and de Búrca 2008, pp. 82).²⁴ Besides regulations and directives, there are decisions, recommendations and opinions (the latter two are not legally binding) (Chalmers et al. 2010, p. 101).²⁵ The legal basis for their enactment is Article 288 of the Treaty on the Functioning of the European Union (TFEU) which reads:

To exercise the Union’s competences, the institutions shall adopt regulations, directives, decisions, recommendations and opinions.

A regulation shall have general application. It shall be binding in its entirety and directly applicable in all MS.

A directive shall be binding, as to the result to be achieved, upon each MS to which it is addressed, but shall leave to the national authorities the choice of form and methods.

A decision shall be binding in its entirety. A decision which specifies those to whom it is addressed shall be binding only on them.²⁶

Recommendations and opinions shall have no binding force.

The new Treaties furthermore introduce a hierarchy of legislative and non-legislative acts (Chalmers et al. 2010, p. 92; Piris 2010, p. 97).²⁷ While the former are those mentioned above that are adopted by an official law-making procedure, the power to issue non-legislative instruments can be granted to the Commission by Council (according to arts. 290 and 291 TFEU).²⁸

The legal instruments lead to different levels of uniformity and diversity of laws among the MS. Another trigger of diversity in the EU is the principle of ‘enhanced co-operation’ that has existed on paper since 1997 and was modified with the Lisbon Treaty (Piris 2010, pp. 89).²⁹ Where the EU does not have exclusive competences those MS, not all, that wish to ‘enhance cooperation’ between themselves can do this under certain conditions. As a measure of last resort, this will inevitably lead to varying laws among the MS.

With varying legal instruments the outcome of the design of the national law can be pre-defined and controlled to varying degrees. It is set out below how the rules versus standards debate can to a large extent be compared to the choice between the two main legislative instruments, directives and regulations, by applying the legal definitions as set out in the second and third sentence of Article 288.

Lastly in this introductory part some main features of the European TFEU law-making process shall be illustrated that is independent of the choice of the legal instrument.³⁰ At European level the law-making process can take various forms (Loos 2006, p. 4). In general, the European Commission has the right of initiative and the Council and the European Parliament take the final decisions following an ordinary or a special legislative procedure (Chalmers et al. 2010, p. 61; Piris 2010, pp. 95).³¹ The law-making process can be complicated because at several stages a political compromise between the various MS and EU institutions has to be found. Also the number of different legislative procedures was reduced with the Treaty of Lisbon (Chalmers et al. 2010, p. 94). The most common form is the ordinary legislative procedure as set out in arts. 289, 294 TFEU which was previously called co-decision procedure according to which Parliament can veto any proposal as it sets out a joint adoption by European Parliament and Council. As ‘special legislative procedures’, mentioned in art. 289.2 TFEU, there is the consultation procedure and the assent procedure. Legal acts are either adopted by the Council with the participation of the Parliament (through consultation or consent) or by the European Parliament with the consent of the Council.

3.2 A promising application: directives versus regulations

In the traditional rules versus standards debate the comparison on costs and benefits takes into consideration two different stages—that of law-making and that of adjudication. When assessing the effectiveness of European law scholars follow the implementation theory, dividing the political process up in policy formulation, implementation and impact (Snyder 1993, p. 26). When making the analogy to European law-making it is striking that the chain of law-making is longer, basically complemented by the law-maker at European level. The upcoming analysis primarily looks at the European decision-maker and the national law-maker as the decisive two interacting phases. Costs (whose minimization is desirable) are incurred at both stages, by the European institutions and by the national legislator. As a further step the interaction with the executive or the judiciary in the MS and the Court of Justice of the European Union (ECJ) shall be commented on.

A Directive lays down certain end results that must be achieved in the MS addressed by it (Craig 2009, p. 6).³² National authorities have to adapt their laws to meet these goals, but are free to decide how to do so. Each directive generally specifies a deadline by which the national laws must be adapted—giving national authorities the room for manoeuvre within the deadlines necessary to take account of differing national situations. As they are objectives they have to be given content at a lower level of governance (Härtl 2006, pp. 174).

They are a typical example of an instrument that leaves flexibility in the way that final laws are implemented in the country and they leave it open as to how exactly to adapt, e.g. the manner in which way enforcement is delivered. The exact content of the new law that is added to the national legal system is given ex post. They are designed to harmonize to some degree rather than to make the laws uniform. Like a standard that gives a framework to the later adjudication, a directive sets out general policies and leaves discretion, however, within the parameters of the directive itself (Hinton 1999, p. 311). In some instances the freedom of the national legislator is very wide: they can go beyond the requirements of the directive. Likewise at times there is room for them to take advantage of the discretionary provisions to minimize a directive’s impact (De Smets 2009).³³ It can be called a ‘compromise’ which unifies law to some extent but also takes account of national particularities (Härtl 2006, p. 173).

From a perspective of legal certainty for the citizens the necessity of the implementation phase has led to some inconsistencies that have consequently been solved by the ECJ. While generally as soon as the implementation phase ends citizens rely on the national legislation (Snyder 1993, p. 24),³⁴ there are some exceptions to this rule (e.g. Hinton 1999, pp. 311):

The court developed the doctrine of ‘direct effect’ to stipulate that unimplemented or badly implemented directives can actually have direct legal force depending on certain conditions.³⁵ On top of that in Francovich v. Italy, ³⁶ the court found that MS may have to compensate individuals and companies that have been adversely affected by the non-implementation of a directive. The rationale for attributing direct effect to directives was to secure the ‘useful effect’ (effet utile) of EU legislation. While e.g. primary EU law sources have direct effect,³⁷ horizontal direct effect of directives is excluded, and individuals can only rely on directives against the state or an emanation of the state and not against other individuals.³⁸

Furthermore following the doctrine of ‘indirect effect’, national provisions always have to be interpreted in conformity with European Law, arguably even before the implementation period ends. For directives this can be a fallback option in cases of the missing horizontal direct effect.

Besides if a MS fails to pass the required national legislation on time, if the national legislation does not adequately comply with the requirements of the directive or if the MS fails to systematically implement a directive it can be sued by the European Commission before the ECJ.

While the passing of a directive is basically a two-step process that leads to the refinement of national laws similar to a standard given content by a judge, a typical one-step process is illustrated next³⁹: A regulation is the most direct form of EU law. As soon as it is passed, is has binding legal force throughout every MS, on a par with national laws. The exact formulations that will later on be found in every MS are fixed ex ante at European level. National governments do in theory not actually have to take action themselves to transpose and implement these regulations. They are part of the national legal systems without the need for transformation or adoption by separate national legal measures. Regulations are analogous to domestic legislation. This is why they are a good illustration of a legal instrument that can lead to uniformity of laws throughout the EU, like ‘rules’ that set out the precise content. In the adoption there is no flexibility left for the MS. Variations in MS are by definition unwanted. Other than directives, there is thus no two-phase legislative procedure consisting of enactment and implementation. Regulations simplify matters in that they enable immediate application and can be directly invoked before courts by interested parties (Twigg-Flesner and Metcalfe 2009). They confer rights on individuals that the national courts are bound to protect.⁴⁰ Unlike a directive it is vertically and horizontally directly effective (Craig 2009, p. 9).⁴¹ For those who are bound by or benefit from a regulation, they guarantee union-wide equal laws.

In this originally intended form regulations and directives satisfy the characteristics for an analogy to the rules versus standards debate rather well. However, the ‘division of labour’ between the two in practice over recent years has become less clear-cut (Král 2008, pp. 252). This development was triggered with the introduction of the notion of ‘limited direct effect of directives’. In their theoretical original usage directives are an instrument for the harmonization of laws and regulation an instrument to create unitary laws (Härtl 2006, pp. 174). However, regulations can now equally be an instrument for harmonization.

It applies to both forms of instrument that the degree of freedom left to the MS as to EU legislation differs clearly when it comes to the issues of minimum or maximum harmonization (Thomson 2009, p. 5).⁴² Minimum harmonization sets out a threshold that MS must meet with their national laws. National laws may, however, exceed the terms of the legislation if desired. In contrast to this maximum harmonization contained in a European norm establishes both a ‘floor and ceiling for domestic regulations’. Here, national laws may not exceed the terms of the EU legislation. A norm can be a mix of maximum and minimum harmonization clauses. Likewise legislation in one legal area can be mixed as to the approaches. Which degree of harmonization is desired is typically expressed within an act.

While it is typical for directives to offer MS a number of options that they can choose when transposing, there are also those that offer few alternatives (Thomson 2009, p. 8). The ECJ has furthermore restricted the discretion with which MS can implement directives by ruling that ‘it is essential that each MS should implement directives in a way which fully meets the requirement of clarity and certainty.’⁴³ There are likewise regulations that leave alternative choices to the MS and so also account for diversity in legal systems. Regulations have moved away from being 100% clear and thus immediately applicable (Bast 2003, p. 38).⁴⁴ They often need implementing legislation and can turn out to lead to harmonization rather than unification (Král 2008, pp. 244).⁴⁵ The instruments can also be interrelated (Craig 2009, pp. 23). The ‘parent’ provision can be either a directive or a regulation, followed by further legislation as either regulations or directives. If the directive is the ‘parent’, regulations can e.g. serve to specify in more detail aspects of the regulatory scheme. Instances can be found where regulations make reference to directives and vice versa.

As discussed in the first section the distinction between rules and standards is one of degree, although it often is convenient to discuss it as an all-or-nothing choice (Kaplow 2000, p. 508). The same applies to the legal reality in the EU.⁴⁶ The rationale of the different legal bite in different policy fields that the two legal instruments were to have was, indeed, undermined by legislative instruments having become substitutable for one another (Chalmers et al. 2010, p. 99). The discussion in this paper will take the original intention as a starting point. The two instruments shall furthermore only be selected as indicators or proxies for the broader picture: the optimal tension between uniformity and diversity.

4 Assessment

Once it was established that the tension between diversity and uniformity can be compared to the choice between a rule and a standard and that using the examples of directives and regulations can be illustrative, it is now assessed in how far the criteria that were developed in the literature are applicable and helpful as to the decision on when to formulate laws in which shape.

4.1 Applying the existing factors

As was set out, the discussion in the context of the EU is lifted to the relationship between the EU and the national legislator. In a given area one can ask if it is better for the European legislator to decide on the exact laws or for the national legislator in the implementation phase. Based on this, when deciding between regulations (rules) and directives (standards), flexibility for the national legislator can be extensively created or completely excluded. It will become clear rather quickly that the established factors of the rules versus standards debate do not suffice to comprehensively assess EU law-making and that additional factors are needed to capture the particularities of the interaction between European Institutions and MS.⁴⁷ Still, first a look shall be taken at the established factors and how these can be applied to this particular scenario:

When it comes to volatility it has to be kept in mind that there is a cost to excluding flexibility by formulating rigid legislation that then has to be amended in a long legislative process. The passing of EU legislation is an even more time consuming process than that on national laws. As there is a necessity to implement directives it will take even more time until the (new) provisions will be set in place compared to legislation that is directly applicable.⁴⁸ Abolishing outdated legislation creates extra costs, too. However, the decision on how rigidly a norm will be formulated will be taken at different levels according to the different legal instruments. When it comes to regulations the legislator at European level will decide on the final formulation—he can opt for shaping it more like a ‘rule’ or a ‘standard’.⁴⁹ With directives it is the national legislator that still has a certain amount of freedom in how to ultimately formulate and implement the directive (Crettez et al. 2009, p. 141).⁵⁰ It can possibly still be formulated in terms of rules or standards by the national legislator. In one scenario (regulations) it is thus the European legislator deciding on the formulations and in the other scenario (directives) it is the national legislator that has some leeway—it follows that the European legislator cannot fully control what national legislators in the various MS will do (Chalmers et al. 2010, p. 64).⁵¹

To some extent the implementation period for directives means that decisions on the ultimate formulations are decided on later than in the case of regulations that take immediate effect and thus an adaptation to changes that have happened between the beginning of the implementation phase and the passing of the law is to some extent possible.

Volatility could also be a reason to reduce harmonizing efforts organized at EU level to a minimum and leave the competence completely to the MS. Obviously this will most likely retain their diversity but it would guarantee a faster legislative process and leave the decision on how to shape the national laws (according to these criteria preferably as a standard) firmly in their hands. MS sovereignty is basically retained, thus diversity favoured over uniformity.

Whilst the relationship between the EU and the national legislator is primary, later in the chain judges also come into the picture as they shape the impact of legislation. Judges, as stated before, do need lesser guidance the more they are specialized; likewise they can cope with far more complicated details. Having to deal with EU-legislation on top of national legislation is a challenge for domestic judges, particularly with regulations that are formulated in ‘Brussels wording’ or when they are to apply the concept of ‘indirect effect’. With European Law in general, though, it is an advantage that domestic courts can ask the ECJ whose interpretation is then binding for the national court for a preliminary ruling via Art. 267 TFEU when doubts exist concerning the interpretation or validity of EU legislation (Chalmers et al. 2010, pp. 149). Once a directive is implemented, the wording as it is then formulated by the national legislator should be less of an obstacle to the application.

The more national judges are to apply laws of other MS according to private international law in various scenarios, the more a uniform approach as to union-wide laws can be preferred. Having said that, except for cases where there was no previous law in the country as to the particular area in which a European norm is passed, a judge confronted with a European norm that would adapt his legal system’s laws would first have to learn how to apply new concepts. This could be made dependant on sectors that regularly involve cross-border disputes and others that do not.

As to the frequency of application there might be a benefit of having a wording that is familiar to the national judge (i.e. an implemented directive) in case of frequent use of a provision. The decision on how complex these national laws would be rests to some degree, actually to the extent set out within the directive, with the national legislator. In contrast where there is a high degree of cross-boarder interaction a benefit can be found in a norm that unifies the corps of law (this can easily be achieved via a regulation).

A high frequency could, in contrast to the volatility scenario, speak in favour of increasing the European influence, as here learning (by citizens, judges, and executive) of new laws will be fast due to a regular application.

The complexity of reality: If many provisions in a directive, (i.e. numerous and possibly complex) have to be considered by the national legislator in the implementation process these are likely to cause delay or even mistakes in the implementation phase (Thomson 2009, p. 9; Loos 2006, p. 3).⁵² Also the transposition checks that are carried out by the European Commission are cumbersome. It can be interpreted as an advantage for regulations since these can be implemented faster when there are many details. Also as to complexity of reality one can again wonder if this speaks in favour of increasing or decreasing EU law-making efforts in general and there might be arguments to favour national approaches, giving less or no competences to the EU in the Treaties in the first place.

This is the last of the established criteria for the national context. As the institutional setting, when it comes to coordinating law-making between the EU and the national legislator, is obviously different, it has become clear that the comparison is currently too weak and the existing criteria have to be refined. In the EU law context it seems initially more appropriate to refer to reality differently, i.e. to basically refer to the complexity of the national legal order that the EU legislation is to affect.⁵³ This leads to the first of the new factors.

4.2 Introducing two new factors

• (1) Switching costs

The starting point is the degree of interference between the new EU-made legislation and the previous national legislations. It is typical for European law that generally, in an area where a European norm is passed, law is already in existence in the MS. The European law basically comes on top, which is different from the case in a national setting where the rules versus standards issue is usually discussed. Alternatively the EU norm will enhance or amend the national law. In cases where there has not previously been any national law in a field, it will fill this gap. All MS in these situations are required to change the laws (Legrand 1997; Legrand and Munday 2003; Ogus 2002). The law that has to be integrated is ‘made in Brussels’. Even though MS will put forward their interests e.g. in Council meetings the result will render amendments of the national laws necessary that vary throughout the MS.⁵⁴ When laws have to be coordinated with the pre-existing legal norms ‘switching costs’ have to be born by the national legislator for adapting the national law to the new EU law. These will vary throughout the MS. Switching costs are the higher the more the shape of the national law differs from that of the new EU norm and the more discrepancies there are between what is set out in the law and what already exists in the national legal system. The diversity of national laws in the area at stake (e.g. huge discrepancies in the common law and the civil law approach) and how conscious the EU act is of this determine overall switching costs in the EU. Single MS might benefit from similarities that their laws already have to the European norm,⁵⁵ likewise they might benefit where they had no previous laws in a particular area. Where there is no previously existing law, at first glance there are no switching costs to be incurred. However, as soon as one realizes how different fields of law are connected to each other and that it is difficult to change traditions, one has to acknowledge that even in this situation adaptation is required and thus switching costs will have to be born.⁵⁶

In this instance freedom in how exactly to make these changes can thus reduce switching costs. If the national legal orders are heterogeneous there will always be MS incurring high and low switching costs and the overall costs thus have to be assessed. This will result in the optimal flexibility that has to be granted to national legislators. It determines in which instances uniformity can be achieved and in which instances it would be very costly.

Expressing the choices of the legislator in terms of the European main legal instruments, the following might hold: Rules, i.e. regulations, add the same law to all MS. In areas of law where there already is a lot of harmony among the MS or no pre-existing laws, regulations can work well as switching costs are minor for all. However, the more national legal systems’ styles diverge from that of the new law, the more will switching costs increase due to the directly effective regulation. The use of ‘standards’ is thus advisable when switching costs are relevant. Impeding switching costs calls for diversity among MS which can be achieved with directives leaving the MS flexibility as to how to achieve certain end results. Where there is a high degree of harmony already among the MS regulations can work well, unless they go against what is considered to be the established harmony. The term ‘gradual’ can be used as to directives since national legislators can introduce a new concept to a legal system in such a way that it is successfully integrated, e.g. make necessary changes to various previous acts instead of just add. Likewise freedom might be given to the MS partly to introduce a new concept at first and at a later stage move to completion. A directive seems to be appropriate to set out such a stepwise integration.

One form that the costs of switching can take,—integrating a law that stands in contrast to the general national law,—is that of ‘fragmentation costs’ (e.g. Cuijpers and Koops 2008). This is a potential consequence of implementing regulations which are rigidly the same in all the MS but might lead to a high degree of fragmentation of national laws. The original harmonious legal order becomes fragmented by the European ‘intruder’. Strict objectives within a directive can also lead to a similar problem, though.

As set out the distinction between directives and regulations is to some extent blurred but it shall be pointed out that some difference will always remain: With a directive a transposition of its content into national law is required (Král 2008, pp. 252). With a regulation by its nature the content is pre-defined, except for cases where national implementing measures are necessary. The task here is rather to concretize etc. Therefore with directives one gets a more national character in the implementation phase while with regulations the community character is retained. On top of this regulations may or may not require transposition measures, whereas a directive, by its nature, must require such measures.⁵⁷

With very complex norms there will always be higher costs involved with both instruments, directives or regulations. The question is how high the costs of legislative coordination are and how they can be reduced. The degree of harmonization set out in a norm is another crucial factor to establish how much has to be added to pre-existing laws and how drastically this has to be carried out.

The next point that particularly concerns laws spread across different, but interacting, legal systems is the question of the effects of differing degrees of uniformity on information costs.

• (2) Benefit of uniformity in terms of information costs⁵⁸

When contrasting uniformity and diversity, there is a clear advantage to uniformity: aligning laws in various MS reduces information costs. It reduces the costs of finding out how laws are implemented and applied in the other MS, and this can be very beneficial for citizens, companies or judges. Again the distinction between regulations and directives shall be used as an indicator of this divide. Generally the two instruments lead to varying degrees of uniformity. A regulation in general automatically leads to the exact same law in all the MS (the aim: unification). There is thus no additional information requirement to find out how a norm is implemented at domestic level. This is true for the citizens or companies of a MS of the EU. When such a regulation is interpreted in national courts the judgments should furthermore not differ.⁵⁹ In order for this to work the regulation should be a rather pure type and leave no or few necessities for coordination with national legal provisions as otherwise the law in the area to be regulated would not be unified.

Directives give MS discretion to achieve the aims set out in it. The uniformity depends on how strictly these objectives are formulated and also varies with the degree of harmonization (e.g. mininum or maximum harmonization). In some cases flexibility is given to implement provisions that deviate from the original text, in others deviations are forbidden. In extreme cases a norm leaves no opportunity for MS to consider cultural, historical, commercial and legal differences (Tesauro and Russo 2008, p. 198).⁶⁰ MS, however, still have some freedom to decide on the form and methods to achieve the directive’s objectives. The uniformity of national laws that implement directives is lower than that resulting from regulations. How much lower depends on how strict the objectives to be achieved are set out in the directive.

Thus a desired high level of uniformity requires the use of regulations if this is beneficial from the point of view of information costs. With directives the effect is less and often less clear. The stricter a directive is about the aims to be achieved the closer it would get to being likewise effective. One might for example wonder if it can be beneficial for a consumer to know what the substantive law (cooling off or prescription periods) and the procedural law (who can address which court?) for buying a certain product are like in a MS. Information costs could be reduced if this was aligned.⁶¹ The case of the entrepreneur might be even more obvious: in order to establish a business in various MS unified law would simplify his life considerably. In other instances reducing information costs might be much less crucial. The freedom that is left to the MS increases the number of mistakes in the implementation phase. Variations, wanted and unwanted, increase information costs. In extreme cases only clarifications in ECJ case law, e.g. when the European Commission sues a MS for failure to fulfill an obligation, could lead to a high degree of uniformity. Due to the limited direct effect of directives, full reliance on the provisions is possible only at a far later stage than with regulations.

When giving the national law-maker freedom in how to implement a directive, there can at times be only little prior assessment of how many extra details, variations etc., a national legislator might add or which pre-existing legal documents the new law will connect to in order to integrate it.

Besides the issues mentioned in the first part (like volatility, complexity of the environment etc.) these two aspects are the new and most decisive factors for the decision as to which instrument to use. They set out the main criteria as to when uniformity or diversity can be appropriate in the EU as regards the costs incurred with both strategies. The factors established in the previous literature leave crucial distinctions unexplained. To some considerable degree the examples of directives and regulations could be used as a convincing illustrative proxy for the policy decision on uniformity versus diversity among European legal orders.

The analysis was carried out as if the decision of which legal instrument to use was purely dependent on the assessment of these costs and abstracted from the reality according to which the EU has differently wide competences in various legal fields. The established factors give guidance to the legislator as to which instrument to choose looking at EU-wide welfare considerations. The results of this paper do, even though some discretion is left to the European law-makers as to the legal instrument to choose, actually concern the moment of formulating the Treaties and granting different instruments in the ambit of different concepts of competence (exclusive, shared, supporting, none).

The assessment comes down to the trade-off between the legal reality in the EU—and in how far which type of instrument would be an intruder to a legal system—balanced with the benefits, possible long-term benefits, of unification in terms of information costs. The evaluation depends on the power that one awards to the concept of ‘path dependence’.

While this is one reasoned set of advice that is given to the legislator, in reality different motivations might rest with law-makers.

5 Conclusion

The central question for this research was how to coordinate law-making between the EU and the national legislator—federalism—in the ambit of uniformity and diversity of legal orders. How much leeway shall be given to the national legislator when it comes to law-making in the EU? This paper approached this topic differently to previous literature from the point of view of the rules versus standards debate. It was set out that in order to assess the procedures in the EU one has to go beyond the established criteria in the rules versus standards debate, particularly because it deals with the interrelation of the European and the national legislator and not just a pure national context. The current set of criteria was refined and two new factors added: switching costs and the benefit of uniformity in terms of information costs. These are the decisive cost indicators for advocating uniformity or diversity.

While they are not a complete match, the instruments regulation and directive were chosen as convincing proxies to illustrate this decision on the law-making level. From the completed set of factors of the rules versus standards debate one can deduce in general in which context a legal instrument should aim at uniformity and in which a flexible approach can be warranted. This can be applied in any context where differing legal traditions are organized in a somewhat federal way and some discretion as to law-making can be left to a lower level.

The indicators that were set out can inspire considerable future research: Firstly the theory can be positively tested with a real example drawn from existing European law. There are reasons why the field enforcement of consumer laws can be a fruitful starting point. Various directives in the field of consumer law in the past were very conscious of the different enforcement traditions in the MS, that can be dominated by a strong focus on public or private law enforcement, by leaving the choice of the enforcement mechanisms to the MS.⁶² More recently the regulation on consumer protection cooperation created a EU-wide network of national enforcement authorities with similar investigative and enforcement powers.⁶³ This meant that MS where such an authority did not exist were induced to establish one (Stuyck 2009, p. 79). This e.g. led to the Netherlands that were originally purely relying on private law enforcement setting up a Consumer Authority. There has been both, a change in the legal instrument and the approach as regards favouring uniformity and diversity in terms of enforcement traditions. By thoroughly assessing this field and others it could be illustrated in how far the reality complies with the theory and in which aspects it does not. From this one could try to come to normative conclusions and set out more clearly when which instrument should be used. A third paper could integrate this approach with the established literature on the economics of federalism and regulatory competition.