JOACHIM VOGEL

HOW TO DETERMINE INDIVIDUAL CRIMINAL RESPONSIBILITY IN SYSTEMIC CONTEXTS: TWELVE MODELS 

JOACHIM VOGEL
Professor and Holder of the Chair of Criminal Law and Criminal Procedure
University of Tuebingen 
Justice at the High Court of Appeal, Criminal Division,
Stuttgart, Germany

I. Introduction

It is agreed that criminal responsibility is individual responsibility. But who is an individual? And who exactly is responsible? The traditional answer to the first question is: a natural person; the traditional answer to the second question: a natural person who voluntarily did an act and knowingly contributed to the criminal result. However, the problem arises whether these traditional answers reflect the systemic structures of modern society. To quote my esteemed colleagues Simone Rozès and Edmondo Bruti-Liberati (Provisional Programme):

Starting from the division of labour …, there is no longer a ‘central’ figure to charge with liability within the company, since the company worker does not meet the legal requirements to qualify as the author of a crime (producer, exporter, etc.), and company directors or executives themselves are not the material authors of the facts contested in law. Hence the idea of not (only) referring natural persons, but (also) corporate entities to a criminal court.

Indeed, there are at least three areas of crime where such problems arise: economic and environmental crime which is typically committed within the framework of companies; organized crime which is committed within the framework of criminal organisations; and last but not least «state crime» which is committed within the framework of governments, armies, police bodies, bureaucracies etc.

In these areas many legal systems all over the world have developed special legal regimes to cope with the problem of how to determine individual criminal responsibility in systemic contexts. These legal regimes are partly based on statute law – e.g. the statutory provisions on criminal responsibility of legal persons in the French Nouveau Code Penal – and partly on case law – e.g. the German figure of «Organisationsherrschaft» (dominion of an organisation), the French «responsabilité pour fait d’autrui» or the common law «strict» or «vicarious liability» and, in military contexts, the «respondeat superior»-principle. It is absolutely impossible to present even a small fraction of these many and different legal regimes during a 30 minutes lecture. Therefore, I would like to go to a more theoretical, more abstract level and present twelve models of how we can possibly determine individual criminal responsibility in systemic contexts. My analysis differs significantly from the traditional analysis of legal regimes on individual criminal responsibility. Therefore, I would like to start with a short criticism of the traditional analysis and only afterwards turn to my own, new analysis.

II. The traditional analysis: «monistic» vs. «dualistic/pluralistic» model

In Germany as well as in other Member States of the European Union, we find a classical analysis of the law of individual responsibility, parties to the offence and participation in the offence. According to that analysis, there are only two possible models of legal rules in this area^([1]):

Either a criminal law system is based on a «monistic model» (in German legal terminology: «Einheitstätersystem»). Its basic idea is that each individual contributing to an offence is liable as perpetrator and responsible as such – regardless of the significance of the contribution. Its significance may of course influence the sentencing process but not the responsibility as such. A monistic system can be found, for instance, in Austrian criminal law and – although contested in the legal discussion – in Italian criminal law (Art. 110 Criminal Code).

Or a legal system makes use of the «dualistic/pluralistic model» (in German legal terminology: «Beteiligungs» or «Teilnahmesystem»). The basic idea is to distinguish at least two ty­pes of parties to the offence: main offenders or perpetrators or authors on the one hand; mere accomplices, instigators, aiders and abettors on the other hand. Such a system can be found, for instance, in German, Spanish or English criminal law. The dualistic/pluralistic model is said to be closer to social reality where primary and secondary responsibility are distinguished. ^([2]) However, the question arises which are the legal criteria to distinguish between main offenders and secondary parties. Traditionally, the distinction is based on subjective elements: the intention of the person contributing to the offence as an author or only as an accomplice («animus auctoris», «animus socii»^([3])). Today, legal scholars try to identify more objective criteria such as the significance of the contribution and the individual position of the contributor. In German doctrine, the figure of «Tatherrschaft» (dominion of the act) has been developed by my famous Munich colleague Claus Roxin and has been, to some extent, accepted by German jurisprudence. Anyway, it is clear that the liability of mere accomplices, instigators, aiders and abettors is derived from the liability of the main perpetrator which is called in German «Akzessorietät», in French «emprunt de la criminalité», in English «derivative responsibility».

Although the traditional analysis is well established, I feel that it is quite insufficient. It focuses on active and voluntary participation in the offence. But participation by omission – breach of a duty to act^([4]) – is not covered; neither are negligent offences^([5]) and offences which can only be committed by qualified persons^([6]) (what is called in German doctrine «Sonderdelikte»). The analysis gives rise to pointless questions such as the Austrian discussion on whether Austrian criminal law really belongs to the monistic model or to a functionally dualistic model^([7]). It does not integrate many new forms of individual responsibility and participation we find in comparative criminal law, for instance vicarious liability, responsabilité pour fait d’autrui ou de chef d’entreprise, conspiracy, complicity after fact and responsibility of legal persons. But above all, the traditional analysis is based on a very limited, very technical concept of participation in the offence: the situation that an individual participates in a well-defined, consummated or at least attempted offence committed by another individual. In social life, we often encounter very different situations: more or less structured organisations, even loose networks, which are fully or partly or only exceptionally involved in criminal activities, and individuals which contribute to these activities in various forms – actively or passively – and on various levels. That is what I mean by «systemic contexts». So I feel that the base of an adequate theory of participation in the offence in systemic contexts must be a non-technical, functional understanding of participation as any contribution to criminal activities. Such a comprehensive understanding enables us, for instance, to realize that various new offences such as membership in a criminal organisation, money laundering, financing of terrorism or incitement to genocide are – of course – forms of participation. And it is not surprising that we find a lot more models of legal rules on participation than the «monistic» and «dualistic/pluralistic» model.

III. A new analysis: «naturalistic» vs. «normative» model; «individual» vs. «systemic» model; «sentencing orientated» vs. «responsibility orientated» model; «special part» vs. «general part» model; «imputation» model; «supervision» model; «conspiracy» model; «collective responsibility»  model

Indeed, we can identify not less than twelve possible models (which do not exclude each other but can be found, in positive criminal law, in various combinations). Although I am not quite sure whether the English terminology is convincing, I would like to call them: «naturalistic» vs. «normative» model; «individual» vs. «systemic» model; «sentencing orientated» vs. «responsibility orientated» model; «special part» vs. «general part» model; «imputation» model; «supervision» model; «conspiracy» model; «collective responsibility» model. Let us have a look at what is behind these terms:

1. and 2. «Naturalistic» vs. «normative» model

Traditionally, participation is a naturalistic concept and means «causal con­tri­bu­tion to the offence» or «an act which results in an offence being committed». The causal chain is essential, but also (setting mens rea aside) sufficient. If you apply such a naturalistic concept to complex systemic structures, for instance a company or an army, you will end up at what has been called in German legal discussion «bottom up approach»: You will start with the immediate actors, e.g. workers assembling a dangerous machine or soldiers shooting civilians; proceed to the middle level, e.g. construction managers or field officers; and finally reach the top level, e.g. the board of directors or the general staff. The longer the causal chain, the more difficult it is to establish individual responsibility, for instance by proving a command chain (a problem very clearly to be seen in The Hague at the moment). On the other hand, the immediate actors often simply do not know enough to be guilty of an offence, or they can plead necessity, order, mistake of law etc.

These problems can be overcome if the naturalistic model is replaced by a normative model. In a normative model, the leading concept is neither «participation» nor «causality» but «responsibility». Responsibility does not necessarily require positive and full causality^([8]). A co-perpetrator or accomplice is responsible even if he had not had any decisive influence on the offence being committed or not. A person with supervisory authority and duty is responsible if he could have influenced the offence. There are inchoate offences – e.g. incitement to commit genocide – where criminal responsibility exists even if the main offence has not (yet) been committed. If you apply a normative model in a systemic environment, you will end up at what has been called in German legal discussion as «top-down-approach»: You will start with those who have the «main responsibility», e.g. directors, generals, political leaders, and work your way down the ranks until you finish with the «small fish»^([9]). You will argue that, if the «small fish» is responsible as perpetrator, the «big fish» must be responsible as perpetrator, as well. In effect, nearly all will be treated as perpetrators (so that a normative model leads to some kind of monistic model in the sense explained before^([10])).

The present tendency is clearly in favour of the normative model. It lies at the very heart of international criminal law and can be studied in the jurisprudence of the I.C.T.Y. and I.C.T.R. and also in the Rome Statute of the I.C.C. But also national legislations and national courts prefer normative approaches. For instance, the German Federal Supreme Court has developed a doctrine of «Organisationsherrschaft» (dominion of an organisation) to hold leaders of an organisation – a state, a company, a criminal organisation – liable as main perpetrators even if they neither planned nor explicitly ordered the offence. ^([11]) In Japanese criminal law, we find special forms of co-perpetration which are tailored to mafia bosses («Komplott-Mittäterschaft»^([12])). In English and American law, we find even a strict or vicarious liability of directors of companies for offences committed by subordinates. «Respondeat superior» is an ancient principle of military criminal law; it has now developed into the well-known command responsibility of international criminal law ^([13]) .

However, a normative model has problems of its own. The term «responsibility» is quite vague, and the term «normative» implies that there are norms or standards or principles which can serve as guidelines to measure responsibility. Such a principle is the principle of personal fault or culpability. Mens rea, knowledge and intention, are still key factors of responsibility. ^([14]) However, knowledge and intention are not sufficient to make a person individually responsible, and the Latin saying «cogitationis nemo poena patitur» is still valid (and still a limit to conspiracy responsibilities). You also need some contribution in form of an act or an omission in breach of a legal duty. Therefore, another key factor of responsibility is the power to act in an organisation or to control it^([15]). Furthermore, the principle of proportionality limits responsibility. In particular, problems of proportionality arise in case adequate behaviour is labelled as criminal behaviour^([16]) – e.g. financial services which may constitute money laundering, financing of terrorism etc. – and if behaviour in the very pre- and post-zone of the offence is criminalized^([17]) – e.g. conspiracy or complicity after fact.

Of course, these are vague principles. However, I do doubt whether we will be able to develop clearer, more distinct principles. For example, the German legal discussion has identified quite a lot of quite diverse principles: «Tatherrschaft» (dominion of the act); «Sonder­pflichtigkeit» (special duty)^([18]); «Risikoerhöhung» (creating or increasing the risk that an offence will be committed)^([19]); «Solidarität» (solidarity with crime)^([20]); «Normerschütterung» (destabilization of norms) ^([21]); «Autonomie» (auto­nomy)^([22]); «Regressverbot» (rupture of imputation). It is unclear how these principles relate to each other, and also whether they are internationally applicable or restricted to German law. A possible general theory might be the so-called «Lehre von der objektiven Zurechnung» (theory of objective imputation) where risk-creation and risk-realization are central concepts which might be very useful for a normative model of participation^([23]).

3. and 4. «Individual» vs. «systemic» model

Traditional criminal law focuses on the individual, not on society or systems. The responsibility of society as a whole (or of parts of society) is not – or only exceptionally – taken into consideration in a criminal process. In this sense, traditional criminal law is based on an «individual» model.

But such a model is certainly inadequate if we look at criminal structures and institutions and at offences committed in an organised and structured manner^([24]), for example organised crime in economy, public administration or politics and the so-called «Makrokriminalität» (macro crime) such as genocide and other crimes against humanity or war crimes. The German criminal law scholar Klaus Marxen^([25]) has shown that we need a new and «systemic» model to cope with the problems of individual responsibility for offences committed a structured and organised environment and manner. Here, individual responsibility is generated by three elements: The first element is a specific behaviour of a person within the system, an act or an omission in breach of a duty to act. The second element is the system itself, the criminal organisation and criminal structure, the criminal group and/or the criminal program. And the third element is an offence committed somewhere in the system. The second element – the systemic context – «links» the first element, the act or omission, to the third element, the offence, regardless whether the act or omission was the proximate cause of the offence. That means: a person who knowingly and willingly contributes to the functioning of the system will be responsible for the offence even if he or she does not immediately participate in the commission of the offence, if the offence can be foreseen in the framework of the system. For example, a person who organizes concentration camps where a genocide is being committed will be individually responsible for each single killing even if he or she does not have specific knowledge about place, time and manner of specific killings.

In present criminal law, we find such a reasoning in the framework of the conspiracy responsibility: a person who engages in a criminal conspiracy will individually be made responsible for offences being committed subsequently, at least if these offences could have been reasonably foreseen. Extensive forms of co-perpetration in systemic patterns are also recognized in international criminal law, e.g. in the well-known Tadic decision of the I.C.T.Y, and in Japanese criminal law («Komplott-Mittäterschaft»)^([26]).

Another important feature of the systemic model is the irrelevance of alternative causalities^([27]). In particular, it is no defence that the offence would have been committed anyway and that the perpetrator would have be probably or even certainly replaced by another person who was willing and capable to commit the offence. Even if such a reasoning is hardly compatible with the causality concept of «condicio sine qua non»^([28]), it reflects a fundamental ethical principle: He who runs with the pack is responsible for what the pack will do.

5. and 6. «Sentencing orientated» vs. «responsibility orientated» model

In the first Tadic decision, the prosecution at the I.C.T.Y. cites «the modern trend to move away from very technical definitions about the degree of responsibility, and instead move us to focus on whether the action in any way incurred criminal liability … The relative degree of responsibility is a matter for sentencing»^([29]). Indeed, it is the very idea of «monistic» models (in German legal terminology: «Einheitstätersysteme»^([30])) that we should not ask whether a person is liable as main perpetrator (author, primary party) or accomplice (instigator, aider and abettor, secondary party) but weigh the degree of participation in the sentencing process. Therefore, «monistic» models may also be called «sentencing orientated».

In contrast, it is still the majority of the criminal law systems in the world which make a legal distinction between primary and secondary liability – which may be called a «responsibility orientated» model. However, a closer look reveals that these systems are indeed moving towards a «sentencing orientated» model^([31]). In many legal systems, there is no difference between the penalties provided for main perpetrators on the one hand and secondary parties on the other hand. At best, legal systems provide for (mandatory or facultative) mitigation of penalties for accomplices (aiders and abettors)^([32]). An organizer who is technically only inciter or instigator will often be punished more severely than the persons who actually commit the offence and is technically main perpetrator.

Indeed, the practical value of the distinction between primary and secondary parties may well be doubted. Often – and setting symbolic reasons aside – nothing will depend on whether the accused is instigator or accomplice by encouraging. On the other hand, a complex system of degrees of responsibility will lock up precious legal resources. In Germany, books and articles have been written about whether heads of organisations responsible for offences are «indirect» perpetrators under Art. 25 par. 1 German Criminal Code, co-perpetrators under Art. 25 par. 2 or instigators under Art. 26 – a question of hardly any practical importance because the same penalties apply. ^([33]) There is a clear and present danger that real and really important questions are neglected. For instance, a real problem is to define the «minimum threshold» of participation and responsibility, in particular in systemic contexts. Is a person who simply passes on an order to commit an offence responsible and participant? What about a person who is engaged in per se legitimate, usual business which, as a matter of fact, furthers criminality, e.g. sells certain chemical substances to a potential terrorist organisation? And which degree of authority is necessary to trigger liability in systemic contexts – is the truck driver who knowingly brings innocent victims to a concentration camp (co-) responsible for the atrocities committed there?

7. and 8. «Special part» vs. «general part» model

The principle of legality («nullum crimen sine lege») requires continental criminal law systems to lay down the rules on responsibility and participation by written law. The very minimum is to declare that any participant is liable (see Art. 110 Italian Criminal Code). Due to the «lex certa»-principle, clear and restrictive definitions of the possible forms of participation are preferable. In modern criminal codes, they are laid down in the general part, they apply – in principle – to any specific offence and are spelt out in an abstract, general and not casuistic way which may be called the «general part» model. In contrast, a «special part» model provides for specific participation rules in specific offences. Historically speaking, the law on participation in the offence started from specific and casuistic rules in the special part – e.g. as annex to the murder prohibition in ancient criminal law – and developed into the abstract, general rules in the general part which are now the international standard.

However, we can identify counter movements. Many «new» offences such as money-laundering, financing of terrorism, membership in a criminal organisation etc. are, as a matter of fact, nothing else than special forms of participation, in the case of money-laundering complicity after fact. ^([34]) Recently, criminal legislators tend to mention certain forms of participation explicitly in the text of the specific offence^([35]), e.g. in the field of drug trafficking offences where certain forms of aiding and abetting are explicitly mentioned so that – technically speaking – a mere aider and abettor will be perpetrator of the offence.

The recent trend gives rise to interesting legal questions which are not sufficiently discussed. For instance, the problem arises whether specific rules on participation in a specific offence exclude the application of general rules of participation^([36]). An example drawn from German law where it is a separate criminal offence to «aid» a receiver of stolen goods: Is it also a criminal offence to aid an aider of a receiver of stolen goods under the general rule on aiding or abetting^([37])? Another question is whether specific rules on participation in specific offences should be interpreted in accordance with the interpretation of the respective general rules.

9. «Imputation» model

If each participant realizes the offence him- or herself – e.g. if a group of soldiers loots a house and each soldier steals valuable objects –, legal problems of participation and responsibility do not arise^([38]). It is only when a participant does not realize all the constituent elements of the offence in person that it becomes necessary to impute to him another person’s act. Holding an officer who orders a civilian to be shot by a subordinate liable for murder, we impute the subordinate’s act to the officer. Therefore, we can say that an «imputation» model lies at the heart of any law of participation in the offence. Depending on the participants’ position we may distinguish between «horizontal» imputation among peers and «vertical» imputation in hierarchies. ^([39]) Paradigm of a «horizontal» imputation is co-perpetration or complicity, paradigm of a «vertical» imputation indirect or mediate perpetration by means of an innocent agent.

As a rule, imputation requires fault (intention, mens rea, at least negligence) although there are exceptions, in particular in systemic contexts where a «strict» or «vicarious» liability of superiors, a «responsabilité pour fait d’autrui» is well known in various criminal law systems. However, fault is only the basis and not the object of imputation. What is imputed are acts – either acts as such or illegal or punishable acts. An imputation which starts from a illegal or punishable act done by another person results in a «derivative responsibility», in French «emprunt de criminalité», in German «Akzessorietät». It is, as a rule, the basis of secondary forms of participation such as counselling or procuring, encouraging, provocation, instruction, instigation, aiding or abetting. However, we can see in many legal systems that the principle of derivative responsibility, emprunt de criminalité or Akzessorietät is increasingly under­mined^([40]). In theory, a derivative responsibility cannot exist if the main offender is not or only partly responsible, e.g. a child, a mentally ill person or a person acting in good faith. But exactly these situations are classical examples of «indirect» or «mediate» perpetration by means of an innocent agent. Once again, we notice that the traditional «dualistic» model moves towards a «monistic» model.

10. «Supervision» model

Imputation of another person’s illegal, and criminal, behaviour can be based on activity, for instance on instigating, aiding or abetting the other person. However, many criminal law systems acknowledge an imputation based on passivity, on the failure to prevent and control offences committed by subordinates – which may be called the «supervision» model of participation. It is one of the most important instruments to determine individual responsibility in systemic contexts.

In international criminal law, the «supervision» model is the basis of the so-called «command» or «superior responsibility». In many national criminal law systems, we find comparable rules in the national military or administrative law, but increasingly also in economic criminal law where directors and managers are bound to prevent offences commited by their subordinates and otherwise are criminally or at least administratively liable if a subordinate commits an offence that would not have been committed if the superior had exercised due care – which is, under German law, an administrative offence of «Aufsichtspflichtverletzung».

In a more general perspective, the «supervision» model belongs to the difficult area of participation by omission. It is an accepted principle of criminal law that omissions can be punished if there was a duty to act. So, the key question of the «supervision» model is therefore to establish a duty to prevent subordinates from committing offences and to control them. German legal doctrine has developed a quite unique theory according to which legal duties are derived from factual or social positions, the so-called «Garantenpositionen»^([41]), and it is highly debated whether superiors are in a position which requires them to prevent and control crime committed by subordinates. However, it is clearly impossible to derive norms from facts. Therefore, we must turn to positive law – military, administrative, economic law – in order to identify obligations to prevent and control crime. As mentioned before, these obligation do indeed exist, also with regard to subordinates who are fully responsible.

The practical problems of a superior responsibility are twofold: On the one hand, prosecution must establish that the superior was in effect in power and could have prevented the offence from being committed. In order to ease the burden of proof, many legislations – among them Germany – require not more than the proof that it would have been more difficult for the subordinate to commit the offence if the superior would have exercised due care. ^([42]) In international criminal law, it is even sufficient that the superior did not prosecute the subordinate after the offence had been committed. Here, we see again that modern criminal law replaces causal by statistical or risk relations. On the other hand, the superior’s mens rea must be proven. However, mens rea does not necessarily mean knowledge or intention. Under many national legislations, negligence may be sufficient, and there are even forms of «strict» or «vicarious» liability where not even negligence is required^([43]).

11. «Conspiracy» model

The responsibility under the «imputation» or «supervision» model implies that an offence has really been committed or at least attempted. However, many criminal law systems recognize a criminal liability for conspiring to commit an offence even before (and regardless of) the commission of the offence – which can be called «con­s­pi­racy» model. It is a traditional model of criminal liability in common law countries but is also advancing in continental countries where risk management has become an idée directrice of modern societies. The traditional criminal law approach – to intervene only after an offence has been committed and after harm has been done – becomes less and less acceptable, also with a view to the extent and the quality of the harm which criminals can cause in a modern technical society.

In detail, we can identify four legal techniques based on a «conspiracy» model: firstly, many legislations incriminate the conspiracy itself, i.e. the planning or plotting of an offence by at least two persons^([44]). Secondly, there are what has been called by German doctrine «climate offences» (Klimadelikte), i.e. inchoate offences like attempted instigation, private oder public incitement^([45]), certain forms of hate crime etc. Thirdly, many legislations make it a separate criminal offence to be member of a criminal or terrorist organisation or to support or aid a criminal or terrorist organisation. The European Union even has made such an criminal offence obligatory for the Member States^([46]). And fourthly, collective violent action may be punished as such regardless of results, e.g rioting in football stadiums or civil disorder ^([47]).

Of course, such incriminations are far-reaching. In the German discussion, several voices – many of them belonging to the so-called «Frankfurt school» around the German legal scholars Winfried Hassemer, Klaus Lüderssen and Wolfgang Naucke – have called for a «core» or «classical criminal law» which does without such incriminations of mere «risks» before actual harm has been done. But which are the alternatives? It seems clear to me that a State has the right, and also the obligation, to intervene if a terrorist organisation plans a terrorist attack. Certainly, it is possible to call such an intervention a «police» or «risk preven­tion act». But such a juggling with words is not really convincing and even dangerous because the legal consequence would be that essential judicial guarantees – e.g. the right to be silent – would not apply. So I hold that the «conspiracy» model is a possible model of criminal law intervention – provided that fundamental rights and guarantees are respected.

12. «Collective responsibility» model

A last attempt to cope with responsibility in complex system contexts is to make the system itself, the collective entity, the organisation responsible – which may be called «collective responsibility» model. However, collective responsibility is not an alternative to individual responsibility but an addition: There is not a single legal system known to me where punishing an individual is excluded if the collective entity is punishable or after the collective entity has been punished. Indeed, «dual» responsibility is the rule – also because most legal systems require an offence committed by an individual member of the collective entity which is imputed to the entity as such. In legal doctrine, it has often been said that collective criminal responsibility serves as substitute where individual criminal responsibility cannot be proven. At least the German experiences do not support such a suspicion.

It is well known that the criminal responsibility of organisations, companies, legal persons or other collective entities is not generally accepted although more and more legislations have introduced or are introducing such a responsibility. On the other hand, practically all legal systems do intervene if members of a collective entity commit offences on its behalf or in its favour and make use of administrative measures, administrative sanctions (in particular fines like the German «Geldbuße») or civil liability for harm resulting from the offence^([48]).

The doctrinal debate started from the fundamentalist argument that «societas non delinquere potest» because a collective responsibility would run against the very idea of cri­minal law, punishment and culpability. Meanwhile, it is well established that collective cri­minal responsibility is possible although its structures differ from the structures of in­dividual criminal responsibility. Important concepts are: derivative or vicarious lia­bility; alter-ego-model; aggregation theory; doctrine of self-identity; corporate attitude / climate. Therefore, the discussion turns to real questions: Which is the «trigger mechanism» for collective responsibility (ratione materiae, ratione personae)? Which are the sanctions, and which the criteria for sanctioning? Which is the standard of proof? Which procedural rules apply? And how do you enforce sanctions against collective entities which may change their legal status, may be bought by another entities and even dissolve themselves any time? A rich reservoir of answers can be found in the E.C. law: Collective responsibility is triggered rationae materiae by acting on behalf or in favour of the collective entity, ratione personae by persons with power to represent or to decide or to control. Sanctions are fines, but also exclusion from public subsidies or contracts, judicial supervision, dissolution. Criteria for sanctioning are gravity of the offence, proceeds from the offence, annual turnover. The requirements of responsibility must be proven beyond reasonable doubt; however, it must not be proven which person exactly committed the offence, and negligence may be proven by comparing the company with a careful company. Certain procedural guarantees tailored to natural persons – e.g. the right to be silent – do not or only partly apply to legal persons. Changes in the legal status as such do not matter.

III. Final Remarks

Simone Rozes and Edmondo Bruti-Liberati point out a «dilution, or broadening of the scope of classic criminal liability (…) in response to the phenomenon, often described by modern sociology, of the collectivisation of social life today».

Basically, I do agree: There is indeed a change in the «classic» law of individual responsibility, of participation in the offence: «Modern» participation law is based on a «normative» rather than a «naturalistic» model, has certain «systemic» features, makes use of the «supervision» and «conspiracy» model and recognizes a collective criminal responsibility. Therefore, the scope of criminal liability is indeed «broadened». But I would not say that criminal responsibility is «diluted». Rather, it is focused in a normative – non-naturalistic – perspective on persons with power to represent, to decide and to control. I fear that those who favour a «classic» and naturalistic criminal law intend, or at least accept, that persons with power to represent, to decide and to control enjoy impunity because they – as Simone Rozès and Edmondo Bruti-Liberati – note        «are not the material authors of the facts contested in law» in a naturalistic sense. But in a normative sense, those who decide and control are material authors and should be made responsible.

Bibliography

Amelung Knut, Die "Neutralisierung" geschäftsmäßiger Beiträge zu fremden Straftaten im Rahmen des Beihilfetatbestands, in Samson Erich (ed.), Festschrift für Gerald Grünwald zum siebzigsten Geburtstag (Baden-Baden 1999), pp. 9 et seq.

Baumann Jürgen, Weber Ulrich, Mitsch Wolfgang, Strafrecht Allgemeiner Teil, 10. edition (Bielefeld 1995).

Bloy René, Die Beteiligungsform als Zurechnungstypus im Strafrecht (Berlin 1985).

de Doelder Hans,Tiedemann Klaus, (eds.), La criminalisation du comportement collectif (The Hague et al. 1996).

Eser Albin, Huber Barbara, Cornils Karin (eds.), Einzelverantwortung und Mitverantwortung im Strafrecht (Freiburg im Breisgau 1998).

Eser Albin, Heine Günter, Huber Barbara (eds.), Criminal Responsibility of Legal and Collective Entities (Freiburg im Breisgau 1999).

Eser Albin, Yamanaka Keeichi, Einflüsse deutschen Strafrechts auf Polen und Japan (Baden-Baden 2001).

Hassemer Winfried, "Professionelle Adäquanz. Bankentypisches Verhalten und Beihilfe zur Steuerhinterziehung", Teil I, wistra, Zeitschrift für Wirtschaft. Steuer. Strafrecht (1995), 41 et seq.

Jähnke Burkhard et al. (eds.), Leipziger Kommentar zum Strafgesetzbuch, 11. edition (Berlin 1992 et seq.).

Jakobs Günther, Strafrecht Allgemeiner Teil, 2. edition (Berlin 1991).

Kienapfel Diethelm, Der Einheitstäter im Strafrecht (Frankfurt a.M. 1972).

Klip André et al. (eds.), Annotated leading cases of international criminal tribunals (Antwerpen 1999).

Kreuzer Arthur, (ed.), Handbuch des Betäubungsmittelstrafrechts (München 1998).

Kühl Kristian, Strafrecht Allgemeiner Teil, 3. edition (München 2000).

Lackner Karl/Kühl Kristian, Strafgesetzbuch, 24. edition (München 2001).

Langer Winrich, Das Sonderverbrechen  (Berlin 1972).

Lesch Heiko, Das Problem der sukzessiven Beihilfe (Frankfurt a.M.1992).

Lesch Heiko, "Die Begründung mittäterschaftlicher Haftung als Moment der objektiven Zurechnung", ZStW, Zeitschrift für die gesamte Strafrechtswissenschaft, 105 (1993), 271 et seq.

Lüderssen Klaus,  Zum Strafgrund der Teilnahme (Baden-Baden 1967).

Lüderssen Klaus (ed.), Aufgeklärte Kriminalpolitik oder Kampf gegen das Böse?, Bd. III (Baden-Baden 1998).

Murmann Uwe, Die Nebentäterschaft im Strafrecht. Ein Beitrag zu einer personalen Tatherrschaftslehre (Berlin 1993).

Neumann Ulfried, Nomos-Kommentar zum Strafgesetzbuch, Baden-Baden, Updated: 11. Lieferung (2002).

Otto Harro, «Vorgeleistete Strafvereitelung» durch berufstypische oder alltägliche Verhaltensweisen als Beihilfe, in Eser Albin et al. (eds.), Festschrift für Theodor Lenckner zum 70. Geburtstag (München 1998), pp.194 et seq.

Paul Carsten, Der Begriff des Handeltreibens nach dem Betäubungsmittelgesetz, StV (Strafverteidiger 1998), 623 et seq.

Ransiek Andreas, Unternehmensstrafrecht (Heidelberg 1996).

Renzikowski Joachim, Restriktiver Täterbegriff und fahrlässige Beteiligung (Tübingen 1997).

Rogall Klaus, Bewältigung von Systemkriminalität, in Widmaier Gunter, Roxin Claus (eds.), 50 Jahre Bundesgerichtshof, Festgabe aus der Wissenschaft, Bd. IV (Heidelberg 2000 ).

Roxin Claus, Strafrecht Allgemeiner Teil, 3. edition (München 1997).

Roxin Claus, Täterschaft und Tatherrschaft, 7. edition (Berlin 2000).

Schmid Niklaus, Strafverfahren und Strafrecht in den Vereinigten Staaten, 2. edition (Heidelberg 1993).

Schönke Adolf/ Schröder Horst, Strafgesetzbuch. Kommentar, 26. edition (München 2001).

Schmidt-Salzer Joachim, Massenproduktion, lean production und Arbeitsteilung - organisationssoziologisch und -rechtlich betrachtet. Haftungs- und Strafrecht als Mittel zur Konkretisierung und Verdeutlichung der Organisation von Unternehmen und Unternehmensgruppen, BB (Betriebsberater 1992), 1866 et seq.

Schumann Heribert, Strafrechtliches Handlungsunrecht und das Prinzip der Selbstverantwortung der Anderen (Tübingen 1986).

Schünemann Bernd, "Strafrechtsdogmatische und kriminalpolitische Grundfragen der Unternehmenskriminalität", wistra, Zeitschrift für Wirtschaft. Steuer. Strafrecht (1982), 41 et seq.

Stein, Ulrich, Die strafrechtliche Beteiligungsformenlehre (Berlin 1988).

Triffterer Otto (ed.), Commentary on the Rome Statute of the International Criminal Court (Baden-Baden 1999).

Triffterer Otto, «Command responsibility« Grundstrukturen und Anwendungsbereiche von Art. 28 des Rom-Statutes; Eignung auch zur Bekämpfung des internationalen Terrorismus?, in Prittwitz, Cornelius et al. (eds.), Festschrift für Klaus Lüderssen (Baden-Baden 2002).

van der Wyngaert Christine (ed.), International Criminal Law: A Collection of International and European Instruments, 2. rev. edition (The Hague 2000).

Vogel Joachim, "Individuelle Verantwortlichkeit im Völkerstrafrecht. Zugleich ein Beitrag zu den Regelungsmodellen der Beteiligung", ZStW, Zeitschrift für die gesamte Strafrechtswissenschaft, 114 (2002), 403 et seq.

Vogel Joachim, Elemente der Straftat, GA (Goltdammer’s Archiv) (1998), 127 et seq.