Since the start of this century, penal law, in Europe and beyond, has become more severe in the name of the increasing security risks for the State and the safety of people. In some approaches, security has become the ‘foremost freedom’ and the precautionary principle has moved from hazardous products to dangerous people, whilst security measures are becoming independent of sentences. This represents a real metamorphosis of criminal justice, and more broadly social control that is taking place around the world.
This metamorphosis, which results in a confusion of legal categories, marks a move from punitive justice to predictive justice, at the risk of seriously contravening the rights guaranteed by the European Convention on Human Rights.
The move from punitive justice to predictive justice
In France, the ‘new Criminal Code’ of 1994 was designed to counter the so-called ‘Security and Liberty’ Law of 1981. It resulted in a reaffirmation of the principle of responsibility, the search for alternative forms of punishment to imprisonment and a refusal to distinguish security measures from sentences. This sometimes misunderstood refusal represented an intention to take into account the fact that, even if they contribute to defending society, so-called security measures are always felt like punishment by the convict, most particularly when the measure is coercive and restrictive, or even depriving them of the freedom to come and go.
The French Law of 2008 on Security Detention (rétention de sûreté) overturned this notion of sentencing. By allowing a convict to be incarcerated by a judge after the end of their sentence, for indefinitely renewable periods of one year if they are considered dangerous, the law renounces the principle of responsibility. This law represents a development that we could qualify as “anthropological” as it disassociates the danger posed by a person from their guilt and the security measure (even depriving someone of their freedom) from punishment, making it possible to imprison somebody not for the crimes that they have committed, but for those that they could commit. By moving from punishment to prediction, this justice, which is still called ‘penal’ could openly challenge the very human notion of responsibility in favour of dangerousness, which would remove the boundaries between humans and non-humans. In other terms, by substituting ‘anticipating’ for ‘punishing’, predictive justice would lead to a process of dehumanisation. All the more so, as, by abandoning the premise of free will, we negate the presumption of innocence, contrary to Article 6 of the European Convention on Human Rights. In essence, such an approach substitutes the proof of guilt that underlies the pairing of ‘responsibility/punishment’ with the assessment of danger and the likelihood of repeat offending that is characteristic of the other pairing: ‘danger/security measure’.
Security detention, therefore, marks a transformation of retributive justice (the pairing of guilt-responsibility) for predictive justice (danger-neutralisation). It leads to imposing an assessment of the danger that an individual represents, which replaces proof of guilt and challenges the presumption of innocence, making it impossible to prove the contrary, as it requires proof of the absence of any possibility of repeat offending. Such an approach ignores the notion of self-determination specific to humankind and forgets that it is the foundation for personal responsibility and free will. How can this measure foster the feeling of responsibility that is essential to reintegration into society?
Confusion between punishment and security measures
If we want to make criminal legislation consistent once again and ensure that it is harmonised at the European level, we need to clarify the notions of punishment and security measures that are not only confused but are also interpreted differently.
French lawmakers make no distinction between punishment and security measures. It is however due to this distinction that the law of 2008 avoided being censured by the French Constitutional Court, despite the imprecision of the notion of dangerousness and the indeterminate nature of the measure, which is clearly contrary to the principle of the lawfulness of offences and punishments. According to the French Constitutional Court, Security Detention is neither a punishment nor a penalty; but it acknowledges at the same time that this logic is difficult to support in view of its nature of deprivation of liberty, the duration of the deprivation, its possible unlimited renewal and the fact that it is declared by a court after a sentence, which led the Constitutional Court to deduce that its retroactive application would be contrary to the Constitution. By doing so, the Constitutional Court implicitly recognises a confusion between crime and deviancy, which the ECtHR has not failed to observe. Even if it claims to separate guilt (crime) from danger (deviancy), and therefore punishment from the security measure, it finally combines them by applying the same principle of non-retroactivity.
The ECtHR clarified this point in 2009 in the case of M. v. Germany. With regard to the German law on security detention, which dates from the Nazi period and inspired the French Law of 2008, the Court recalled that “the Convention does not lend itself to a general policy of prevention directed against a person or a category of people who are considered to be dangerous due to their continuous propensity to delinquency” (§82). It recalled that it is not bound by the classification stipulated by domestic law, that the notion of punishment determined by Article 7 has an “autonomous scope” and that one must look “beyond appearances”. It considered that due to its nature of depriving somebody of their liberty and its unlimited duration, security detention could “absolutely be understood as a punishment” (§130). By re-qualifying what the Germans call “preventive detention” as a punishment, the Court issued a single response assimilating the security measures to punishment and requiring compliance with the fundamental principle of non-retroactivity. It imposes realistic conditions and guarantees: to achieve the objective of preventing criminal behaviour, there must be “a high level of care, a multi-disciplinary team, intensive work of an individual nature with the detainees” with a view of liberation that “must constitute a real possibility”.
The Court’s case law has nevertheless allowed for a certain mixing of the categories and neutralised the safeguard that it had consecrated with this case. Through the rulings on Bergmann v. Germany (Application No. 23279/14) and Ilnseher v. Germany (No. 10211/12 and 27505/14), the Court indeed allowed that security detention could be justified with regard to article 5 § 1 e) for the detention of a person of unsound mind and could be extended after a detainee suffering from mental illness and presenting a threat to the public had finished their sentence, when they benefited from an appropriate therapeutic environment. It added that in cases where security detention was extended due to the necessity of treating mental issues, the nature and the objective of this measure had changed to the point where this could no longer be qualified as ‘punishment’ in the sense of Article 7, which therefore justified the retroactive application of the security measure.
In other terms, there is a change in the treatment: the person of unsound mind having committed a criminal offence for which they were judged to be responsible and that has progressively served their sentence with the guarantees provided for by Article 7, switches to the regime of Article 5 that serves as the basis for potentially unlimited preventive deprivation of liberty. This development results from a form of à la carte psychiatric transformation of the repressive apparatus authorised to detain people of unsound mind to punish them, before detaining them to care for them. The risk with which the Court is faced is opening the possibility for States to subvert the goal of Article 5 of the Convention, which is protecting against the arbitrary, by using purely formal internal classifications, which represents a reversal of the harmonisation of the guarantees that have been patiently built until now with the determination of the Court (“independent notions”, general directive for the interpretation of the examination “beyond appearances”, etc.), including in the most sensitive political contexts.
The dangers of ‘dangerousness’
Dangerousness (dangerosité), as defined by French law as “a very high probability of re-offending”, therefore a probability of which the effects cannot be predicted with certainty, raises the question of how dangerousness can be assessed. Another point of confusion with worrying consequences is a tendency to assimilate psychiatric dangerousness, which results from a diagnosis that can be made by a medical expert, with so-called criminal dangerousness that merges with a prognosis of repetition or re-offending, for which the type of experts who could make the assessment is unknown and that even seems impossible in principle. Despite a highly critical opinion from the National Human Rights Commission and the fierce opposition of a number of professionals, both doctors and lawyers, the French law of 2008 on security detention still created strange institutions known as ‘social-medical-legal security centres’ and the ‘multi-disciplinary commissions to assess dangerousness’. Behind the complexity and opacity of the procedures provided for, it is above all a matter of the denaturing of expertise; a confusion between psychiatric dangerousness related to mental illness feeding into the illusion that there is an expertise that makes it possible to assess the level of criminal dangerousness. The possibility of even having an effective defence could even be challenged in the event that a multi-disciplinary commission, concluding on the dangerousness of a convict, suggested that a person be made subject to security detention.
Indeed, it is hard to see how the sentencing to security detention could be challenged as it would be difficult, if not impossible, to demonstrate a posteriori that the convict was not dangerous when the decision was taken and that the measure was not ‘the only means‘ of prevention. The result is the unheralded situation of a legal decision that excludes any possibility of legal error, therefore any possibility of reparation for a person who is unjustly detained, even though their detention may be perpetual. Granting, in the name of collective security, almost unlimited coercive powers, is exactly what characterises the totalitarian model of criminal policy as opposed to the liberal model.
The issues in terms of criminal policy and safeguarding the rule of law
Predicting dangerousness, which negates the concept of free will, challenges the equal dignity that underlies Article 3 of the European Convention for the Protection of Human Rights and Fundamental Freedoms. When it results in this extreme version of unlimited or even perpetual incarceration, the so-called ‘security’ measure goes completely against the ‘right to security’ that includes protecting individuals with regard to the State and characterises the rule of law. When disassociated from the sense of punishing a wrong, this deprivation of liberty has no educational or therapeutic role that legitimises, according to Article 5 of the European Convention on Human Rights, the incarceration of minors or the mentally ill; but it is all the same not comparable with the short-term detention that is permitted for foreigners subject to expulsion or extradition. It is a measure of neutralisation, of ‘social defence’ in the broadest and the most general sense of the term.
The whole construction of modern penal law in democratic countries is based around a ‘liberal’ ideology, in the sense that, in the name of individual liberties, it limits the State’s right to punish transgressions defined as crimes (discrepancies from standards, from an obligation), reserving deviancies (discrepancies from normality) for civil society that alone may take corrective measures. It is true that the liberal model has never operated in its purest state. Therefore, the European Convention on Human Rights, not only permits security measures consisting of depriving people of certain rights or restricting liberty (social-legal supervisors), but also legitimises the deprivation of liberty without a prior crime, on the condition that it makes it possible to educate minors, or for the treatment of the mentally ill. The idea of dangerousness is underlying, but it is not consecrated as such with such a broad and ill-defined meaning.
If it goes as far as authorising the retention of foreigners in an illegal position, or (even more disputably) that of vagabonds, Article 5 cannot be extended to dangerousness and legitimise neutralisation measures, a fortiori when they are of an undetermined length: even when it includes an educational or therapeutic role of normalising deviancies, the rule of law recognises that it is subject to limits in form and substance.
Deviations for security reasons are incompatible with the rule of law, which implies that the State is subject to the limits of the law. To be effective, without negating fundamental rights, democracy must avoid a certain number of traps, and in particular that of the illusion of ‘zero risk’. The illusion is believing, and making others believe, that it is possible to remove all risks. The European Court of Human Rights identified this trap in the “Klass case” of 1978, a case of security interceptions with regard to terrorism. At the time, the Court also specified that “States may not, in the name of the struggle against espionage and terrorism, adopt whatever measures they deem appropriate”. The danger would be, to repeat the expression used by the European Court, “undermining or even destroying democracy on the ground of defending it”.
To protect itself, democracy must clarify the choices and explain, beyond criteria of seriousness, the conditions for the acceptability of risk; which risks are we prepared to accept to preserve our rights and freedoms? At a time when security detention appears to be the symbol of the security that would become an absolute protection value, relativising all other human rights, the difficult mission for the European Court of Human Rights consists precisely of recalling the principles laid down by the European Convention on Human Rights.
At a time when the world, which is overrun by passions for identity and security, appears to vacillate in every direction, all eyes are on these Human Rights Judges, the last ramparts against the steady erosion of the rule of law that is specific to democratic societies that already appears to be taking shape. Protecting human rights is not a utopian dream. It is the only realistic answer to the dangers that threaten the survival of humanity from all sides.
 Article 706 -53-13 (1) of the French Code of Criminal Procedure.
 CNCDH, Avis sur le projet de loi relatif à la rétention de sûreté et à la déclaration d’irresponsabilité pour cause de trouble mental et réponses du gouvernement, 7 February 2008.
 J. Danet, “Droits de la défense et savoirs sur le crime”, in Défendre, Dalloz, 2004.
 J. Danet and C. Saas, “Le fou et sa dangerosité, un risque spécifique pour la justice pénale”, RSC, 2007, p. 779; ECHR, Rivière v. France, 24 Oct. 2006, RTDH 2007, p. 261-268, note J-P. Céré.
 About a member of the Mafia, see CEDH, Guzzardi v. Italy, 6 Nov. 1980.