Summary

This study concerns the principle of legality (nullum crimen sine lege) in the field of European Union criminal law. It analyses the way in which the principle of legality is interpreted by European and Dutch legislatures and judges, and how this interpretation affects legal certainty and the distribution of powers.

The legality principle expresses conditions for the exercise of the state’s right to punish (jus puniendi), one of the most intrusive powers of the state. The legality principle as enshrined in article 1 of the Dutch Penal Code (dpc) entails that (i) only written laws can establish criminal liability (the lex scripta principle), that (ii) these laws must be clearly drafted (the lex certa principle), that (iii) courts may not extend the scope of the criminal law (the limits to judicial interpretation, often referred to as the prohibition of analogy), and that (iv) laws may not have retroactive effect, unless (v) this is favorable to the accused (the lex mitior principle). The legality principle governs legislation and adjudication, and demands that these activities are carried out with utmost regard for legal certainty and distribution of powers, and ultimately for individual liberty and autonomy.

The legality principle was first formulated during the Enlightenment in the context of the sovereign nation state. During the second half of the twentieth century, however, the formation of the European Community – later the European Union – profoundly changed the European landscape. This new situation has had consequences for the functioning of the principle of legality in the member states. Whereas enactment of criminal legislation and adjudication of criminal cases were once exclusive prerogatives of the sovereign state, these legal actions are increasingly being governed by European Union law. Indeed, European Union law exercises influence on the scope of criminal liability in the national legal orders, both directly (through the mechanism of direct effect) and indirectly (through obligations to criminalize certain conduct and through the duty of consistent interpretation). At the same time, the legality principle constrains the ways in which European law can affect national substantive criminal law.

This study analyzes the role of the legality principle within the context of ‘European criminal law’, a core notion which for the purpose of this book refers to the multilevel legal field (composed of several autonomous legal orders, not a legal order in itself), in which European Union law has influence on national criminal law. The analysis covers five elements or components which the legality principle entails: (i) the demarcation of sources of criminal liability; (ii) the lex certa principle; (iii) the limits to judicial interpretation; (iv) the prohibition of retroactivity; and (v) the lex mitior principle. Two research questions guide the analysis, the first of which is broken down into three sub-questions. The research questions read:

The overall aim of the book is to better understand how the aims of the principle of legality can be realized within the pluralistic context of European criminal law. Also, the book aims to promote understanding of the interaction between European law and national criminal law, and between legal systems more generally.

Chapters two and three introduce the principle of legality and the effects of European law on national substantive criminal law. Chapter two, ‘The Principle of Legality’, presents a view on the principle of legality and its foundations. The chapter outlines the history of the principle within the history of ideas, with a focus on the Enlightenment conception of legality. The Enlightenment conception of legality took the theory of legal positivism as a point of departure and maintained that all laws are man-made (‘posited’). Related to that point of departure, this conception of legality required that all laws establishing criminal liability must be written. Historically, this view has rather been the exception than the rule, however. Both before and after the Enlightenment, it has been more common to assert that the legality principle applies only to offences that are ‘posited’ (such as tax fraud), and not to offences that are part of natural law (such as murder). Even though the Enlightenment conception was a historical exception, modern thought on the principle of legality has been heavily influenced by it. Chapter two illustrates this by showing how all five components of the modern legality principle which are included in this research can be deduced from the writings of Beccaria and Montesquieu.

In chapter two I also develop an analytical and normative framework for the subsequent chapters. This is done by describing and defining the concepts of legal certainty and distribution of powers. Legal certainty is seen as the prime foundation of the legality principle, and the analysis of its conceptualization in European criminal law focuses on three issues. The first concerns the question whether legal certainty is mainly understood as subjective legal certainty (i.e. the certainty of the individual about his or her legal position, which is a trait of the legal subject), or as objective legal certainty (i.e. the conditions within the law itself that enable subjective legal certainty, which is a trait of the law itself). Second, I ask whether the demands posed by the principle of legal certainty are defined by context or not: is the need for legal certainty dependent upon the nature of the legal relationship, the nature of the rule, and/or the addressee? The third issue is whether legal certainty applies to the individual as an accused person, or whether other persons, such as (potential) victims are also entitled to legal certainty, i.e. the certainty that the law protects them against other individuals. In addition to legal certainty, the distribution of powers is studied in this book as a foundation of the principle of legality. The reason behind this choice is that the distribution of powers is of particular relevance to European criminal law, since powers are not only distributed horizontally between legislatures and courts, but also vertically between the European Union and its member states. This situation necessitates an analysis of how competences and responsibilities are distributed among the different actors, and between the European and national level.

In order to complete the analytical framework, chapter two furthermore introduces the distinction between the legal positivist and the rights conception of legality and develops this distinction into a tool for describing and analyzing the European criminal law conception of legality. The legal positivist and rights conception of legality were originally described as the rule book and the rights conception by Dworkin, and have subsequently been applied to criminal law by – amongst others – Rozemond, Claes, and Brouwer. Following Brouwer, I refer to the two conceptions as the legal positivist conception (rather than the rule book conception) and the rights conception of legality, because in my view, the core of the distinction lies in different theories of law underpinning the two conceptions, legal positivism and natural law. This difference between the legal positivist and rights conception has practical consequences. The rights conception requires of the legislature and the courts that they balance the interests of the suspect with other interests, mainly those of the (potential) victims – and this balancing act takes place within the principle of legality. In a legal positivist understanding, by contrast, the principle of legality always pulls decisions in the direction of more legal certainty for the accused. The conceptions of legality also lead to different understandings of the concept of legal certainty. The distinction between the two conceptions of legality is used mainly in chapter five on lex certa and in chapter six on the limits to judicial interpretation, and it is explained in more depth in those chapters.

Whenever normative views on legal certainty are defended throughout the book, these are based on the assumption that legal certainty is best understood as objective legal certainty. Besides, the demand for legal certainty should not be contextually defined, and should only refer to accused persons only and not to victims. This view on legal certainty is underpinned by the legal positivist theory of law. The normative analysis of the vertical distribution of powers is based on the viewpoint that vertical distribution of powers must be transparent and rational. These normative considerations align with the legal positivist conception of the principle of legality.

Chapter three, ‘The Effects of European Law on Substantive Criminal Law’, describes how European law influences substantive criminal law in the member states. The choices made in this chapter define the scope of the analysis in subsequent chapters. The relation between the European and national legal order is discussed from the perspective of constitutional pluralism. This theoretical framework is chosen because it provides the most accurate description of the current relationship between the different legal orders in the European Union. This is not to say, however, that the theory of constitutional pluralism serves as a justification for the current state of affairs. In the context of this book, the use of constitutional pluralism simply implies that not only the European perspective on the interaction between European and national law is worth studying, but also the Dutch perspective.

One factor that determines the possible effects of European law on national criminal law is the scope of legislative competences of the European Union regarding substantive criminal law. Since the entry into force of the Treaty of Lisbon in 2009, an important question in this regard has been whether the legislative competence of the European Union concerning substantive criminal law is limited to the legal basis provided by article 83 TFEU. I argue that while not yet the case, harmonization of substantive criminal law on other legal bases is likely to occur in the future, and that even unification of substantive criminal law using regulations cannot be ruled out as a future possibility. A second question is whether directives can prescribe full harmonization of criminal offences. It is argued in chapter three that it must be ascertained for each directive separately whether and how much discretionary leeway it leaves to the member states. Total harmonization cannot not be ruled out as a possibility.

European law affects national substantive criminal law through primacy leading to the disapplication of national law, direct effect, and consistent interpretation. These concepts and mechanisms were developed by the Court of Justice in the context of the European Communities. How these must be applied to criminal law is not self-evident: the Area of Freedom, Security and Justice is marked by an inherent tension between effectiveness and the protection of fundamental rights. In the context of the Area of Freedom, Security and Justice, European law often requires member states to take repressive measures (e.g. the criminalization of certain types of conduct). Direct and indirect effects of these measures may thus come at the expense of individual rights, most prominently the right to legal certainty. The tension between effectiveness and the protection of fundamental rights manifests itself in the case law of the European Court of Justice on the applicability and boundaries of primacy, direct effect, and consistent interpretation in the field of substantive criminal law.

In chapters four through eight, the concrete norms of the legality principle are analyzed in turn.

Chapter four, ‘Sources of Criminal Liability in European Criminal Law’, aims at clarifying the type and origin of norms that can establish criminal liability. The type of sources refers to the question whether only written laws may establish criminal liability (as in systems recognizing the lex scripta principle), or whether also other sources, for example case law, may do so (as in systems using a substantive definition of ‘law’). The origin of sources refers to the question whether criminal liability must have a basis in the national law, or whether European norms may be used directly as base for establishing criminal liability – that is, without implementation in the national legal order. Against the backdrop of constitutional pluralism, these questions are discussed and answered both from a European and a Dutch perspective.

From the European perspective, European law sets no formal limit on the types of legal sources that can establish criminal liability. The concept of ‘law’ in article 49 of the Charter can refer to both written and unwritten sources of criminal law, as long as these meet the criteria of accessibility and foreseeability. Hence, the Court of Justice uses substantive criteria for the demarcation of ‘law’, thereby following the case law of the European Court of Human Rights (ecthr). This is particularly relevant for the field of competition law, as competition rules are interpreted and applied at the European level. For the criminal law systems of the member states, however, the concept of law used in the Charter is of rather limited importance, because member states are obliged to transpose directives leading to the criminalization of certain conduct into written (national) criminal provisions. They may not use policy guidelines or case law to widen the scope of criminal liability in accordance with the directive, even if their national system allows the development of criminal law through case law.

Limits to the origin of sources emanating from European law were not found either, even though at present there are no directly applicable criminal provisions in place at the European level. Directives are unable to serve as such because of their nature – they need to be transposed and can only produce direct effect under limited circumstances. Regulations, on the contrary, are in principle able to produce the effect of establishing criminal liability. However, thus far, establishing criminal liability has been an exclusive power of the member states. The regulations currently in place only contain the constituent elements of the offence, but no criminalization, nor a sanction. The criminalization of the offence, as well as the sanction, is always part of national law. Neither the case law of the European Court of Justice, nor the Treaties, contain formal barriers to establishing criminal liability in directly applicable regulations. If a regulation were to provide not only the constituent elements of the offence, but also a criminalization and a sanction, these would be directly applicable in national criminal courts. However, the need for establishing criminal liability in directly applicable regulations does not seem to be widely felt.

After discussing the European point of view, the chapter shifts focus to the Dutch perspective. Regarding the type of sources, article 1 of the Dutch Penal Code enshrines the lex scripta principle. Hence, only written laws are accepted as a source of criminal liability. The origin of laws that establish criminal liability are more diffuse. The Dutch Supreme Court (Hoge Raad) has ruled that article 1 dpc requires a written provision in Dutch criminal law. This rules out the use of European regulations as a source of criminal liability. However, the Supreme Court has also accepted that European law itself defines the ways in which it can affect national law. In my opinion, the major shifts that have taken place both in the development of the international and European legal order, as well as in the Dutch understanding of those legal orders and their relationship with national law, warrant a renewed debate on the origin of sources of criminal liability, primarily in the light of democratic legitimation of criminal provisions.

To sum up: at this moment, legislation has primacy over other sources of criminal law from a European perspective, which means that the Dutch lex scripta principle is not currently under pressure. Other types of sources are not formally ruled out by European law, but their practical significance is limited, and mainly confined to competition law. National legislation is pivotal to European criminal law, as thus far a national criminal provision is always needed in order to establish criminal liability. There is however nothing that necessitates this from a European perspective.

Chapter five, ‘The Lex Certa Principle in European Criminal Law’, discusses whether vague legislation at the eu level leads to vague legislation at the qnational level, and, related to this question, whether eu legislation is and should be bound by the lex certa principle. In order to answer these questions, two types of uncertainty are distinguished: semantic uncertainty, which occurs when the text of the law itself is unclear; and systematic uncertainty, which occurs when a law is unclear as a result of its relation to other laws. The most important form of semantic uncertainty is vagueness, which can be further dissected into criterial vagueness and gradual vagueness. Criterial vagueness exists when it is unknown which necessary and sufficient conditions must be met in order for an phenomenon to fall under the scope of a predicate, e.g. ‘an item of property’; and gradual vagueness exists when it is unknown which part of a scale is covered by a predicate, e.g. ‘large quantities’. The category of systematic uncertainty is broken down into the types of inconsistency and complexity.

Both semantic and systematic uncertainty are widespread in European law. This gives rise to the following questions: may the validity or applicability of directives and regulations be reviewed in light of the lex certa principle? If so, by whom? Whenever regulations are used in a national criminal case as the description of the constituent elements of the offence, the national court must determine whether its provisions are clear enough to be used as such. The Court of Justice has formulated a minimum standard for this review, which is that individuals must be able to foresee, if necessary with the help of relevant case law, which acts will lead to criminal liability. The Court of Justice has ruled that directives are, in principle, subject to review by the Court of Justice on grounds of lex certa. However, in all cases so far, the Court found that the directives in question do not violate the lex certa principle, because they do not establish criminal liability. This means that it is the duty of the national legislatures to make sure that the national criminal provisions transposing the directive are sufficiently clear. Usually, member states can choose between transposing a directive verbatim, or adapting the formulations of the directive to their national legal system, but if the directive is too vague, European law requires member states to make their national law more concrete. The responsibility for safeguarding legal certainty is thus imposed on the national legislature and national courts. Here, a comparison with competition law is illustrative, as the demands imposed on member states in criminal law are formulated more strictly than those imposed on European actors in the context of competition law. In competition law, the requirements of clarity and precision are interpreted differently according to the contextually defined need for legal certainty, and the Court of Justice engages in balancing several interests involved in the case – not only the need for legal certainty of the suspect. Hence, in core criminal cases, the lex certa principle is constructed in a manner consistent with the legal positivist conception of legal certainty, whereas in competition law cases, the approach is inspired by the rights conception of the principle.

The previous paragraph concluded that the lex certa principle is currently not applied directly to European law – at least not at the European level, as European law contains no directly applicable criminal provisions. But how is this situation to be appreciated? On the one hand, many of the directives and framework decisions aiming at the harmonization of substantive criminal law have been heavily criticized for being vague. The vague formulations in directives and framework decisions, it is assumed, will translate into vague criminal provisions at the national level. On the other hand, vagueness on the European level has been defended with reference to the desirable margin of discretion such vagueness creates for member states when transposing these measures. The fifth chapter maintains that such a defense mistakes criterial vagueness for discretion. Indeed, the effect of criterial vagueness can be the exact opposite of what proponents have argued, as criterial vagueness often makes it impossible for member states to discern whether and how much discretion they have. This statement is based on the observation that several of the heavily criticized, criterially vague formulations have been transposed word for word into national law, including Dutch law, rather than being made more concrete. Whereas criterial vagueness clearly poses problems, the examined instances of gradual vagueness in directive and framework decisions have been clarified in Dutch law. This implies that criterial vagueness is the more problematic of the two types, whereas gradual vagueness can more easily be resolved. Based on this relation between criterial vagueness of European law and criterial vagueness of Dutch law, this chapter argues that the lex certa principle is not only relevant on the national level, but can also be seen as relevant to the European legislature, even though it cannot function in such a clear-cut manner as it does for the national legislature. The lex certa principle compels the European legislature to create the circumstances that enable the national legislature to enact clear and precise criminal provisions. One of these circumstances is that member states can know their obligations and are not confronted with uncertainty regarding the existence and scope of their margin of discretion.

The lack of certainty in European law, in the form of both semantic and systematic uncertainty, not only poses challenges to national legislatures, but also to criminal courts. Even so, all Dutch case law revolves around semantic uncertainty. Much of this case law concerns the regulation of professional activities using criminal law. As in the European Convention on Human Rights (echr) and in eu law, a contextually defined understanding of legal certainty leads to lower standards of clarity. Complaints about vagueness only very rarely lead to the disapplication of a national provision. The chapter concludes with an examination of whether other mechanisms, such as the mistake of law exception or prospective (over)ruling, are used to compensate for this rather lenient review, but no such effect could be found. One last observation from the analysis of Dutch case law is that when Dutch criminal courts refer to the lex certa principle, they refer to its codification in article 1 dpc and article 7 ECRH, but not to article 49 of the Charter.

Chapter six, ‘The Limits to Judicial Interpretation in European Criminal Law’, explores the limits to judicial interpretation of both the European Court of Justice and the national criminal courts. A core question in this chapter is: how does the Court of Justice delineate the duty of consistent interpretation imposed on national criminal courts in light of the legality principle? To help answer this question, I develop a typology of three manners in which judicial discretion can be ‘limited’ (the word ‘limiting’ should not be taken too literally, as the way in which judges reach a decision cannot really be limited or governed by norms; rather it is the justification of choices as laid down in the judgment which is subject to norms and which can hence be limited). This typology is in part based on the well-known observation that the concept ‘interpretation’ suffers from process-product ambiguity: interpretation can both refer to an activity (the process) and to the result (the product) of that activity. In a similar fashion, limits to the judicial function can refer both to the process and to the product of interpretation, as well as to the application of a norm in a concrete case. The process of interpretation (again: in reality the justification of the outcome) can first be limited by stating that courts are bound to use (or may not use) certain methods of interpretation (for example: ruling out analogical interpretation as a method), or by establishing a formal hierarchy of methods (for example: the outcome may never be in contradiction with the text of the law). Second, limits can relate to the product of interpretation (for example: an interpretation may not violate legal certainty). Third, the application of an interpretation in an individual case may be subject to limits (for example: an extensive interpretation may not be applied retroactively).

Similar to the lex certa principle, the Court of Justice has thus far not applied the limits to judicial interpretation to adjudication at the European level, as no criminal liability is established at this level. It is up to the national criminal court to decide whether it was foreseeable for the accused that he would be held liable for his actions. The interpretation of a provision by the Court of Justice does not depend on the way in which it is enforced in the member states. The Court does, however, respect the limits set by the text of secondary legislation and its ‘natural and usual meaning’. This is especially the case when European legislation is enforced directly, using punitive sanctions. However, restrictive interpretations are not explicitly motivated by the principle of legality. This may change in the future with the growing significance of European law for substantive criminal law.

As stated above, an important issue in chapter six is the duty imposed upon national courts to interpret national law that falls under the scope of European law in light of the wording and the purpose of the relevant European instrument. This obligation may lead to tension with the legality principle whenever inconsistencies exist between a directive requiring criminalization of certain acts or omissions, and a national provision criminalizing some, but not all of those acts (or omissions). According to the Court of Justice, the duty of consistent interpretation is limited in a strict manner by the wording of the national provision, especially in the field of substantive criminal law: a consistent interpretation may not lead to determining or aggravating criminal liability without a proper basis in national criminal law. Using the typology of ‘limits to interpretation’ introduced above, chapter six shows that the limits to judicial interpretation posed by European law relate to both the process of interpretation (grammatical interpretation takes primacy over other interpretive methods) and to the product of interpretation (aggravating criminal liability is prohibited), but not to the application of interpretations. When compared to the limits flowing from article 7 echr, it appears that the Court of Justice is stricter than the European Court of Human Rights, as in the case law of the latter only the application of a certain interpretation is seen to violate the principle of legality. The ECtHR does not oppose a gradual extension through case law as such, but only states that an extensive interpretation may not always be applied in the concrete case. The Court of Justice opposes reaching an extensive result at all, which implies that whenever a consistent interpretation cannot be reached in an individual case, the national legislature will have to amend its legislation.

Because of the strict limitations, the duty of consistent interpretation does not force national courts to transgress the interpretive leeway they usually enjoy in the Dutch legal culture. This is all the more so because the Court of Justice explicitly states that courts must exercise their duty of consistent interpretation in accordance with their discretion under national law. However, the horizontal distribution of powers is not untouched by the duty of consistent interpretation. Dutch criminal courts can be forced to overrule contrary interpretations given by the legislature in preparatory works, and hence must take a critical stance towards their national legislature and national law. This does not, however, make interpretation by national courts more autonomous, as they must use their discretion completely at the service of European law. The relationship between the national court and the national legislature is loosened; but the relationship with the European legislature that comes in its place is a strict hierarchical one.

Chapter six furthermore contains a study of the way in which Dutch courts deal with their obligation of consistent interpretation – and their duty to safeguard the principle of legality while doing so. The Dutch Supreme Court has given some rather doctrinal judgments in private law and tax law cases, in which it interprets and explains the European case law regarding conforming interpretation. In these judgments, a product-oriented approach is prescribed: courts must determine whether national law precludes reaching a consistent result. This is the case only when the text of national law contravenes European law, but not when an historical interpretation points in another direction than a consistent interpretation. No such standard ruling has been given in a criminal case. In criminal law, the role which European law plays for the interpretation of national law often remains doubtful, and therefore no hard conclusions can be drawn yet about whether the duty of interpretation has different implications for criminal law than for private law. Lower courts sometimes seem compelled to steer close to the limits to their judicial discretion and opt for a consistent interpretation in order to resolve discrepancies between national and European law. In some cases, they use the mechanism of consist ent interpretation whereas the disapplication of national law would actually seem more appropriate; however, in these cases only the method, but not the outcome is problematic. In other cases, a consistent interpretation arguably equals an extensive interpretation. This could lead to a violation of European law, as safeguarding the principle of legality is one of national courts’ duties under European law. A final conclusion that could be drawn from the case law is that criminal courts are very reluctant to interpret national law in light of European law when a directive leaves discretion to the member states, irrespective of whether this discretion has actually been used by the legislature. It is contended in chapter six that it is only the legislature who is granted discretion when transposing a directive. When this discretion is not explicitly used to pose stricter norms, the presumption of the legislature wanting to transpose correctly leads to the presumption that the Dutch offence has the same scope as the offence as prescribed in the directive. In other words: discretion stands in the way of a consistent interpretation only when used explicitly by the legislature.

Chapter seven, ‘The Prohibition of Retroactivity in European Criminal Law’, is a concise one. This is because European norms do not establish criminal liability, and, as a rule, only apply to future situations. There are exceptions to this rule, but these do not apply to instruments that may form the basis for a conviction in a criminal case – such as the inclusion of an organization or a group on a list of terrorist groups. Also, because of the prohibition of retroactivity, European law never requires member states to criminalize conduct retroactively. All legislation transposing and implementing European law (both directives and regulations) must respect the prohibition of retroactivity, which thus applies to all norms possibly influencing the scope of criminal liability. This is in accordance with the Dutch understanding of the legality principle.

Chapter eight, ‘The Lex Mitior Principle in European Criminal Law’, deals with the possibility of granting retroactive effects to changes in legislation that benefit the defendant. In my view, this principle is, strictly speaking, not a part of the principle of legality, as it has a different rationale. However, it is often associated with the principle of legality as it forms the main principle governing changes of the criminal law and as such forms a tandem with the prohibition of retroactivity. Furthermore, the application of the lex mitior principle in European law is fascinating from the perspective of interaction between legal systems: it may lead to direct retroactive effect being given to directives, and in some cases it can possibly even require retroactive effect. In other words, the lex mitior principle is not a constraint to the impact of European law, but a legal basis for it. Little is known about when and under what conditions the principle of lex mitior applies and what it entails. For all these reasons, it is included in the scope of this book.

Because the lex mitior principle is different from the other principles studied, chapter eight contains a tailor-made analytical framework for studying different conceptions of the lex mitior principle. Lex mitior principles in different legal orders differ on roughly three points: their scope, their rationale, and their status. The scope of the principle refers to the changes in legislation that may be given retroactive effect. The scope can include all legislative changes that directly or indirectly change the scope of criminal liability or the sanction (the content doctrine, or materiële leer); or it can include only changes in the criminal law (the formal doctrine); or be restricted to only changes regarding sanctions (the sanctions doctrine). Next to these formal criteria, substantive criteria can be used to restrict the scope of the lex mitior principle; for example, the principle may allow retroactive effect of changes in legislation only if they were brought about because of a revised opinion on the part of the legislature, but not if they were caused by a change in circumstances. Second, the rationale of the principle can be understood as being based in (i) clemency; (ii) a revised assessment regarding culpability of the conduct (i.e. a change in opinion regarding whether certain conduct is deserving of punishment); or (iii) the principle of proportionality of sanctions. Third, the status of the principle varies greatly between different legal orders: it can be a fundamental right that can be invoked against the will of the legislature, or it can be a policy principle that can be used at will by the legislature.

The analysis of the European lex mitior principle is hampered by a lack of sound data. The principle has been recognized in case law of the European Court of Justice, and was later codified in article 49(1) of the Charter of Fundamental Rights. The scope of this principle is highly uncertain, but there are good reasons to believe that it applies at least to changes in sanctions and changes in the constituent elements of the offence (the formal doctrine) – and possibly even to all legislative changes (the content doctrine). The case law contains some indications that substantive restrictions apply as well, but so far these restrictions are highly unclear. As to the rationale of the principle, the Court of Justice has not explicitly justified why the lex mitior principle is a fundamental principle of European law, but it has implicitly renounced the suggestion that the assessment of culpability should be leading. Considering the supranational character of the principle, it seems most consistent to see the proportionality of sanctions as the foundation for granting legislative changes retroactive effect in the context of European criminal law. What is clear, is its status: the principle is recognized as a fundamental principle of European law which must be observed by national authorities. Exceptions are only possible if they meet the criteria set in article 52(1) of the Charter.

The Dutch lex mitior principle has traditionally known no formal restrictions of its scope, but it has been limited in a substantive manner, as it is applied only to legislative changes stemming from a revised opinion regarding culpability (gewijzigd inzicht in de strafwaardigheid van de gedraging). Knigge has drawn attention to the fact that, when understood like this, the Dutch lex mitior principle may be in conflict with supranational obligations. This is so because while the Dutch principle is a policy principle to be used at the discretion of the legislature, supranational rights may be invoked against the national legislature. Recognition of the lex mitior principle in supranational contexts, mainly article 7 echr, has led the Dutch Supreme Court (instigated by advocate general Knigge) to modify its settled case law. Changes regarding sanctions that are beneficial to the defendant need no longer stem from a revised opinion regarding culpability, but are applied indiscriminately to all pending cases. This change in the Supreme Court’s case law is based upon the – questionable – assumption that supranational renderings of the lex mitior principle are all limited in scope to changes regarding sanctions (the sanctions doctrine). For all other changes in the criminal code or otherwise, the Supreme Court has maintained its criterion of a revised legislative opinion regarding culpability.

Case law of Dutch courts shows differences amongst courts in their approach to the lex mitior principle in the context of European law. This can be explained by the fact that article 1(2) dpc provides no clear-cut instructions as to the application of the principle in a multilayered context. In the current situation, courts must take a stance on two different issues in every case. First, courts must decide whether European directives may or must lead to the disapplication of national law in retrospect, before there was direct effect or before they even entered into force. Second, courts must decide whose opinion regarding the culpability of the conduct they accept as decisive: that of the European or the national legislature? An examination of case law shows that criminal courts are highly divided. Retroactive effects of European law have, albeit rather implicitly, been accepted by the Supreme Court as a possibility. It is nevertheless likely that changes in provisions forming part of European criminal law are more seldom applied retroactively compared to changes in purely national provisions. This is so because the answer of the Supreme Court to the second question listed above. In an important case, the Supreme Court did not consider the later (more lenient) provision to be applicable, as the provision was amended with the purpose of harmonizing legislation upon its entry into force. That reasoning has the potential of being applied to many other cases of transposing legislation. Lower courts have several times complemented this argument with an examination of the reasons behind the legislative change at the European level as well. The purposes of article 1(2) dpc call for an inquiry into the motives behind legislative changes at both the European and Dutch level. This approach, I argue, is more in line with the fact that where the European Union has legislative competence, it is the assessment of the European legislature that is decisive for the scope of criminal liability at the national level. The chapter includes a set of recommendations for Dutch criminal courts that should result in a European law-oriented, realistic approach to these matters.

Chapter nine, the conclusion, recapitulates the findings of chapters four through eight, in order to answer the research questions at an aggregated level that goes beyond the level of the five elements of the principle of legality discussed throughout the book.

The first research question was: how is the legality principle applied by European and Dutch legal actors (legislatures and courts)? As no criminal offences are enacted or applied at the European level, European actors currently do not deem their actions to be governed by the legality principle. It is at the national level that this principle is safeguarded. The Court of Justice has articulated strict limits to the effect European law can have in national substantive criminal law because of the legality principle: member states must create an anterior, clear, and precise legal basis for criminal liability in national written law, and courts are bound by these written offences and may not extend criminal liability beyond the written law. Under some circumstances, national authorities can be obliged to apply changes in legislation retroactively. Even though the tension between effectiveness and the protection of the legality principle is resolved by granting primacy to the legality principle, national actors are still confronted with many uncertainties relating to their duties under European law. The Dutch legislature struggles with uncertainty vis-à-vis the latitude of their discretionary powers when transposing directives. Dutch criminal courts, in turn, seem increasingly aware of their mandate to give full effect to European law, but are less aware of their mandate to protect the legality principle – which is just as much their duty under European law. When referring to the legality principle, they refer to article 1 dpc and article 7 echr, but not to article 49 of the Charter.

Regarding the application of the legality principle in European criminal law, the general conclusion can be drawn that it is mainly aimed at protecting the legal certainty of accused persons. This means that individuals may not be held liable if liability was not foreseeable at the time they committed the act. Whether criminal liability was indeed foreseeable, depends heavily on the context: persons engaging in professional activities which are usually strictly regulated, are required and deemed able to assess which rules apply to their activities. Similar to the echr context, the concept of legal certainty which is used in European Union criminal law focuses on subjective legal certainty. Preventing arbitrary convictions plays no significant role in the case law of the Court of Justice. The same holds true for the distribution of powers.

Even though this application of legal certainty is rather narrow, protecting legal certainty is a powerful obligation for member states. Whenever tension exists between the effectiveness of European law and legal certainty of suspected persons, legal certainty must prevail. The Court of Justice formulates these obligations without legal certainty being weighed against effectiveness; there is no balancing act. Obligations deriving from the legality principle for member states, which they must observe when they implement European law – hence, whenever the multilayeredness of European criminal law manifests itself – thus display characteristics of the legal positivist conception of legality. This is remarkable, as one would expect the application of the legality principle to be in line with that of the corresponding article 7 ECHR, which is characterized by the rights conception of legality. When looking at the case law of the European Court of Justice on the legality principle in competition law, the Court seems to be guided much more by the ecthr case law. Apparently, it chose a different approach for its case law on the legality principle whenever the legality principle functions as a limit to the effects of European law on national substantive criminal law. This difference is speculatively explained in chapter nine by the wish of the Court of Justice to promote the legitimacy of European criminal law and by the need for transparency in the field, so that effective supervision of member states’ compliance can be carried out by European institutions.

The conclusion also provides an answer to the second research question: how does the way in which the legality principle is applied affect legal certainty and the distribution of powers in the field of European criminal law? It would seem that the strict conception of the legality principle used by the Court of Justice ensures that no problems arise regarding legal certainty and the distribution of powers, but this is not the case. In fact, legal uncertainty is a widespread and pervasive problem for European criminal law that manifests itself in three different manners. First, it occurs often in the form of vagueness, a type of semantic uncertainty. Criterial vagueness of European law becomes problematic when it leads to vague national criminal provisions, and this indeed appears to happen in practice. Second, the scope of criminal liability can become obscured by inconsistencies between European and national law; we are then dealing with systematic uncertainty. This too occurs frequently. Third, European criminal law is complex, which forms the second manifestation of systematic uncertainty. Hence, legal uncertainty does not only appear in the form of vague European legislation, but also, or perhaps even primarily, as a result of the interplay between the European and national legal orders. To phrase it differently: systematic uncertainty may pose a bigger challenge to legal certainty in European criminal law than semantic uncertainty, even though the latter is the more easily recognized and better known form of legal uncertainty. Looking at the two forms of systematic uncertainty, it appears that in particular inconsistencies are problematic in the field of European criminal law as the rules of conflict governing inconsistencies in European law do not produce foreseeable outcomes. This is mainly true of consistent interpretation, the main mechanism of decentralized enforcement when direct effect is ruled out – which it often is in the field of substantive criminal law. The outcome of a case in which courts are called upon to interpret national law in conformity with European law in order to overcome minor discrepancies between the two is not easily predicted. This uncertainty adds to the lack of a clear conception of the lex mitior principle, which leads to uncertainty about when changes in legislation favorable to the accused may or must be applied retroactively. As a result of all of this, the applicability and scope of national criminal provisions is often subject to doubt, a fact leading to an uncertain scope of criminal liability.

It was emphasized above that the responsibility for safeguarding the legality principle is put almost exclusively in the hands of national courts. However, invoking the principle of legality before a Dutch criminal court is hardly ever successful from the perspective of the suspect. Similar to strictly national cases, the court will usually be of the opinion that criminal liability was indeed foreseeable for the defendant. Hence, the provisions were not too vague to be applied and there was no excusable mistake of law on the side of the defendant. This finding is influenced by the fact that an important share of European criminal law lies not at the core of criminal law, but concerns criminal regulation of economic affairs. In that context, persons carrying out professional activities are expected to carefully assess the rules regulating their activities.

Legal uncertainty is caused in part by an opaque and at times not fully rational distribution of powers between European and national actors. I argue that the responsibility for the quality of criminal legislation should not lie exclusively at the national level, as well-drafted European legislation is of paramount importance to the drafting of clear criminal provisions at the national level. The European legislature should at least strive to fully demarcate the scope of obligations vis-à-vis the member states. Member states must be able to discern which acts must be or may not be criminalized, and if and to what extent they are allowed to make autonomous policy choices when implementing European legislation into the national legal order. National legislatures, in turn, must not pass uncertainty on to their citizens and their criminal courts; they must take full responsibility for the quality of national legislation. They are, however, at times hampered in this effort by their limited discretion. Hence, ensuring that criminal provisions live up to the standards of the legality principle requires extra efforts on the parts of all actors involved.

The conclusion proceeds to offer recommendations for the different actors involved in European criminal law. Most of these recommendations aim primarily at creating awareness of the fact that in order to ensure that criminal provisions live up to the standards of the legality principle, extra efforts of all actors involved are required. Related to this awareness is a plea for viewing legal certainty not only as a trait of legal subjects, but as an objective trait of the law in place – hence, a plea for a greater emphasis on objective legal certainty. Respecting citizens as autonomous individuals necessitates foreseeability of the possible attribution of criminal liability, also for those who are not subjected to criminal law enforcement. This means that drafting clear criminal provisions remains a high priority, and that the responsibility for safeguarding legal certainty may only to a limited extent be passed on to national criminal courts.