I.  The Idea behind the New “Category Model”

Criminal sanctions are the harshest weapons in the arsenal of a state when reacting to misconduct. The existing national regimes of sanctions and the means of their imposition and execution still differ considerably from country to country as they are closely tied to national cultural, social and historical roots. Indeed, the coherence of the national sanctioning systems and the values and ideas on which they are based should not be destroyed by EU harmonisation directives. The fundamental principles of the criminal law sanctioning systems form part of the “national identities” of the Member States, which shall – according to Art. 4 (2) of the Treaty on European Union (TEU) – be respected by the Union. Nevertheless, the Treaty on the Functioning of the European Union (TFEU) confers on the EU a competence to harmonise the national sanctioning systems. Art. 83 (1) and (2) TFEU, in particular, empower the EU to establish “minimum rules concerning the definition of criminal offences and sanctions”.

A thorough study conducted by the members of the European Criminal Policy Initiative (ECPI) – a research group of 20 academics¹ from 12 Member States of the European Union – leads to the finding that previous attempts to harmonise criminal sanctions on the basis of Art. 83 TFEU, especially by using “minimum maximum penalties”, could not achieve their aim of an efficient approximation of the law on sanctions. They even resulted in implementational difficulties in many of the national legal orders.² The study furthermore revealed that, as minimum maximum penalties rely on numeric values which are identical for all states (e.g. a minimum maximum of two years of imprisonment), their use generally leads to an increase in punitivity, tends to destroy the coherence of national systems and affects the proportionality of national penal concepts.

The newly developed model should be thought of as an alternative. On the one hand, it guarantees the protection of proportionality and coherence of national sanctioning systems. On the other hand, the interest of the EU in harmonising national criminal sanctions can be satisfied to the utmost in order to enable an efficient judicial cooperation within the EU in accordance with the principle of mutual recognition. The point of departure is the gradation of criminal sanctions pursuant to their severity, as common in all EU Member States. The new model uses national grades of severity as a “bridge” between the interests of the EU and the interests of the Member States: The EU legislator shall be put in a position to develop a coherent sanctioning system on EU level, tiered only by the degree of severity – the category – with regard to the offences being harmonised, whereas the Member States specify the concrete penalties by assigning sanctions from their own national systems to the categories used by the EU. In this way, a genuine balance of interests can be achieved: The EU can determine the seriousness of harmonised offences by classifying them into categories, while the Member States only “fill” these categories with sanctions familiar to their national systems, thus ensuring the coherence of their own legal order.

II.  The Background: Unsatisfactory Status Quo

1.  Minimum triad

The legal acts in force so far aiming at the harmonisation of criminal sanctions apply different methods: The traditional formula, introduced by the European Court of Justice (ECJ) itself in its “Greek Maize Scandal” jurisprudence,³ requires the Member States to take “effective, proportionate and deterrent sanctions” and is often referred to as the “minimum triad”. Although these specifications are rather vague, in respect to some fields of crime it may indeed be sufficient that the EU restricts itself to require the Member States to provide for these only imprecisely outlined sanctions by their own.

In accordance with such understanding, the new model seeks to retain the “traditional” and minimally invasive approach. But as far as most offences, and above all the most severe of them, are concerned, the EU will understandably und rightly want to go further and deploy additional requirements, a tendency also observable in previous attempts at harmonisation of national laws as to sanctions.⁴

2.  Minimum maximum penalties and minimum maximum penalty ranges

a) The most detailed provisions of the EU so far, which concern consequences of criminal misconduct, stipulate so-called “minimum maximum penalties” which obligate the Member States not to fall below a certain maximum penalty (e.g. “Member States shall take the necessary measures to ensure that an offence referred to in […] is punishable by a maximum penalty of at least five years of imprisonment.”). A less precise stipulation is the so-called “minimum maximum penalty range” which does not prescribe a specific value in relation to the lowest maximum penalty allowed, but instead grants Member States a scope within which the maximum penalty to be imposed may range (e.g. “[…] is punishable by a maximum penalty of at least one to three years of imprisonment.”). As both instruments only provide minimum rules, they do not hinder Member States from choosing sanctions, which exceed the (minimum) maximum penalty, hence going further than the (minimum) maximum penalty range contained in the harmonisation directive. Thus, in the end, a “minimum maximum penalty range” essentially is nothing but a minimum maximum penalty, as it only obligates the Member States to impose not less than the bottom threshold of the given range as a maximum penalty.

b) According to the thorough comparative study of the ECPI, which provides the basis for the hereinafter-presented new model, the use of minimum maximum penalties as well as minimum maximum penalty ranges is far from being efficient in terms of bringing about effective harmonisation. Minimum harmonisation concerning only the maximum penalty has proven to be disappointing. One reason is that in most countries, the maximum penalty is much less relevant for the individual punishment than the minimum threshold. Furthermore, the legal comparison revealed a large discrepancy between Member States as far as the dimension of statutory maximum penalties is concerned. They are, for example, limited to twelve years by the Finnish constitution, whereas in France and Italy, maximum penalties of up to thirty years are possible.

c) By stipulating minimum maximum penalties, the EU intends to reflect the different degrees of severity inherent to the criminal behaviours the Member States wish to penalise. These efforts for a greater systemisation of the sanctioning requirements peaked in the Council document from 27^th May 2002,⁵ which stated that one of four minimum maximum penalty ranges needed to be chosen in every act of harmonisation to come, depending on the severity of the crime. Although this approach clearly demonstrates the EU legislator’s large interest in classifying the severity of the offences both at EU level and in a coherent manner, a convincing systematic classification could not be achieved.

d) The analysis of relevant legislation enacted so far (agreements, framework decisions and directives) concerning requirements of criminal sanctions revealed eight different minimum maximum penalties – 1, 2, 3, 4, 5, 8, 10 and 15 years – and three minimum maximum penalty ranges – 1–3, 2–5 and 5–10 years.⁶ But the use of these penalties and penalty ranges varies greatly. Whereas the highest and lowest minimum maximum penalties were only used once each (1 year stipulated in the directive on “child abuse”⁷ and 15 years in the framework decision and directive “on combating terrorism“⁸), the minimum maximum penalty of 8 years⁹ and the range of 1–3 years have already been used five times, and the minimum maximum penalty of 5 years was used four times in total. Notably, the threshold ranges were not used in the last years;¹⁰ the most recent acts dominantly apply the minimum maximum penalties of 2, 3, 5 and 8 years. In case of legal acts being substituted after the entry into force of the Lisbon Treaty, a double tendency can clearly be observed: The requirements in the replacing acts become more specific and the minimum maximum penalties tend to be more severe.

e) The comparative study also showed that not only do the maximum sanctions differ greatly among the Member States (cf. above, b)),¹¹ but also the minimum maximum penalties stipulated by the EU do not find a numerical equivalent in the national legal systems. Most Member States basically apply a system of gradation of maximum punishments, but the grades used do differ considerably. As a result, not all of the “maximum penalty” values referred to by European minimum maximum penalties are familiar to all European Member States, thus resulting in implementational problems.

If Member States want to stick to “their system”, they select levels (values) that are already known to their legal system. As harmonisation de lege lata always is a minimum harmonisation, Member States are forced to choose the next higher level; they cannot, however, go back to a lower level. This effectively means that states that – for legitimate reasons – do not want to give up the coherence of their sanctioning system have to follow the deplorable trend of increasing punitivity!¹²

3.  Specific EU penalty ranges

But, if minimum maximum penalties in harmonisation directives are inefficient, what could be done? The stipulation of specific penalty ranges by the EU or at least minimum penalties can, of course, be discussed:

• Specific EU penalty ranges which consist in prescribing a combination of a minimum and a maximum penalty (e.g. “between 3 to 5 years imprisonment”), are politically unenforceable at this point in time and – de lege lata – not covered by Art. 83 TFEU. The competence in the treaties only allows minimum harmonisation and therefore does not empower the EU to introduce binding penalty ranges, which necessarily imply an upper limit (in the example: 5 years imprisonment as the maximum penalty allowed).

• Although the introduction of “minimum penalties” is not excluded by the wording of Art. 83 TFEU, the accompanying effect of limitation of judicial discretion is either completely unknown or at least totally contrary to the legal system of many countries (especially Denmark and France). In order to avoid grave conflicts with those legal systems, their peculiarities, which can even be considered to be part of the identity of the state (“national identities”) in the sense of Art. 4 (2) TEU, must be taken seriously.

III.  The New Approach: Relative Comparability

The principal innovation of the new model is to depart from concrete figures and values and render superfluous the use of minimum maximum penalties, which have proven to be unsatisfactory. It instead relies on the idea of “relative comparability”. The aim is to achieve an adequate balance between the interests of the EU legislator as to effective harmonisation on the one hand and the interests of the Member States with regard to maintaining the internal coherence of their law on sanctions on the other hand, hereby respecting the fundamental principles of subsidiarity and proportionality.

1.  Imposition of categories of severity by the EU

When describing the legal consequences of a harmonised offence, the EU legislator has a legitimate interest in expressing the gravity of the offence also in relation to other offences, which have already been harmonised. The result is a hierarchy of offences, expressed by a classification of sanctions in a number of categories (hence: “category model”).

Thus, the EU legislator is in the position to create a “nucleus of an EU sanctioning system”. The stipulation of further, more detailed specifications in relation to the sanctions in each category is – for that purpose and at least for the time being – not necessary. This step is left to the Member States, which have a legitimate interest in not losing their basically conclusive and coherent law on sanctions and their sanctioning traditions. Moreover, it is only by this means that all further national provisions on specification, adaptation and enforcement of sanctions, which differ considerably from state to state and which cannot realistically be approximated, are able to effectively survive and apply coherently. As all legal systems of the Member States are familiar with a hierarchy, or graduation of criminal sanctions due to their severity, it should in principle be possible for them to classify the sanctions which already exist in their national systems in such a manner that they can be assigned to a specific, EU-stipulated number of severity categories.

2.  Main features of the category model

Broken roughly down to the core, it can be said that the model consists in a two-step process. In a first step, the EU legislator stipulates categories that – according to the EU – express the (also in relation to other harmonised offences) relative severity of the criminal misconduct. In accordance with the underlying division of responsibilities between the EU and the Member States, it is the EU legislator who finally decides on the number of categories to be used. According to the ECPI’s study, a number of 5 categories will be necessary (to differentiate sufficiently between the gravity of offences actually or potentially subject to harmonisation under the TFEU), but also sufficient (to reflect adequately the gravity steps as they can be ascertained in all Member States according to their character and severity – using a number of criteria). Choosing Roman numerals for the individual categories, “category I” shall designate the relatively mildest and “category V” the relatively most severe sanctions.

Example: In the harmonising directive the EU could provide that the basic act of trafficking human beings (as defined in Art. 2 of the Directive on Human Trafficking¹³) is a category II offence. If the perpetrator deliberately or by gross negligence (cf. Art. 4 [2] [c] of the Directive on Human Trafficking) endangered the life of the victim, the EU could consider the act a category IV offence.

In a second step, it is the task of the Member States to provide for a sanction specific to their own national legal order, which – in relation to severity – corresponds to the chosen EU category. It could prove useful if the EU provided some guidance as to the criteria to apply when categorising the national sanctions, e.g. in the form of setting up a “manual”. In its study, the ECPI exemplifies some of the useful criteria, which could appear in such a manual – they are the result of the underlying comparative study. However, apart from this “soft guidance“, the EU must not interfere in the process of categorisation itself.

Example: Each Member State classifies its own sanctions into five categories and chooses from its toolbox of sanctions a penalty which, in its view, corresponds to category II (for the basic offence of trafficking in human beings) or category IV (for qualified trafficking in human beings). A state such as Germany could - on the basis of a number of factors (e.g. character of the offence as a crime or misdemeanour, limitation period, possibility of conditional sentence) – decide from the 18 penalty ranges known in German law to assign e.g. the penalty range of imprisonment for up to 3 years to the European category II and to provide for this penalty range when implementing the Directive on the harmonisation of sanctions with regard to the basic offence of trafficking in human beings. All internal legal consequences linked to this type and amount of penalty remain applicable. For example, the penalty framework chosen here in our example constitutes a misdemeanour (and not a felony) under the German Criminal Code, which in turn has substantive and procedural consequences: The attempt is basically non-punishable, as is the attempted instigation. A discontinuation of the criminal proceedings is possible on the basis of opportunity grounds according to Sec. 153, 153a German Code of Criminal Procedure.

Although the Member States are granted a wide “filling discretion”, their leeway is not without limits. Whenever a specific legal act aiming at the harmonisation of criminal sanctions is implemented, the Court of Justice of the EU retains a certain degree of control power if the European Commission triggers infringement proceedings according to Art. 258 TFEU. If this happens, the Member State in question has to demonstrate transparently the reasonableness and coherence of its categorisation. Because of this control mechanism, it makes sense from the point of view of the Member States to follow a consistent strategy of categorisation right from the start which – and this is obviously another advantage of the model – requires critical self-reflection on their own law on sanctions.

In a nutshell: With regard to harmonisation purposes, it is not the denomination and quantification of the criminal sanction that is important, but its effect and consequences for the person affected, which can only be inferred from the overall context with the national law (on sanctions etc.) as a whole. It is therefore essential to maintain the coherence of the Member States’ systems; the formal comparability of penalties, by contrast, is not conclusive from the outset.

IV.  An Outlook into the Future

Apart from reconciling national and EU interests in an approach that attaches great importance to national identities and to the principle of subsidiarity, the category model is obviously “fit” for further advancements – dependent on the political climate between the Member States, of course.

It seems conceivable that – beyond the basic model presented here – the Member States’ filling discretion may be reduced by increasing the category requirements set out by the EU. In particular, it seems feasible that the EU may introduce certain characteristics the sanctions of a specific category must fulfil (e.g. “Category IV sanctions […] must be sanctions which are regularly no longer subject to suspension.”).

Should the EU in the long run be granted wider competences not only to harmonise national criminal sanctions but also to create supranational penalties for supranational offences,¹⁴ the use of the category model may be considered as a constructive interim model as it may be an interesting and promising point of departure for the development of a coherent supranational sanctioning system. Figuratively speaking, it is the EU that then fills its own (future) arsenal of supranational sanctions into the abstract categories that are otherwise filled by the states with their own national sanctions. The experience gained from applying the category model will make it easier to develop a basic structure for coherent supranational criminal law.