Awakening the Leviathan through Human Rights Law – How Human Rights Bodies Trigger the Application of Criminal Law

This article analyses the role that human rights bodies play in triggering the application of criminal law. By examining the jurisprudence of the Inter-American Court of Human Rights, the European Court of Human Rights, the UN Human Rights Committee, as well as other human rights bodies, the article discusses how these institutions have started imposing on states positive obligations to criminalise, prosecute and punish serious human rights violations. While criminal law has traditionally been seen as a threat to fundamental rights, human rights bodies have contributed to presenting criminal law in a positive vein, as an essential instrument of human rights protection. The mainstream of the human rights movement has largely lauded the trend. This article challenges this view, by presenting the pitfalls of using human rights law to extend the reach of criminal justice in order to ensure that perpetrators are held accountable. Not only the imposition of duties to criminalise and punish ends up restricting the accused’s fundamental rights and neglecting the conceptualisation of criminal law as ultima ratio, but the invocation of criminal-law enforcement might also contribute to enhancing the coercive power of the state and, therefore, make state abuses more likely.


I. Introduction
'The objective of international human rights law is not to punish those individuals who are guilty of violations, but rather to protect the victims and to provide for the reparation of damages resulting from the acts of States responsible'. 1 These lines, pronounced by the Inter-American Court of Human Rights (IACtHR) in Velàsquez Rodríguez, distinguish the role of international human rights law (IHRL) from that of criminal law. The former is designed to protect individuals from state abuse, whilst the latter has the antithetical aim of employing state power to punish those who commit a crime. Yet, it is somehow ironic that these words were pronounced in the judgment considered as the precursor to a trend that has made criminal punishment one of the main objectives of IHRL.
Velàsquez Rodríguez is at the crossroads of a paradigmatic change in the relationship between human rights and criminal law. Until the mid-1980s, the large majority of human rights lawyers did not favour using criminal law for fear that employing penal powers might lead to abuses. 2 However, in the years immediately following, the end of the Cold-War inaugurated a period of democratic transitions and attempts to advance accountability for serious human rights violations. 3 In this context, criminal punishment gradually started being considered as the decisive instrument for banning practices of oppression and violence, as well as for promoting justice and peace. 4 Today, criminalisation, prosecution and punishment have assumed a central Pinto 163 push to criminalise and punish coming from the IACtHR is becoming a form of "neo-punitivism" that elevates the penal law to a level of a social painkiller and confuses civil and constitutional protection with criminal law itself. 17

B. Purposes of the Article
The contemporary turn to criminal law within IHRL has arisen with no systematic consideration about the aims and the risks of criminal law. It is, therefore, paramount to scrutinise such embracing of penal measures to advance human rights protection. The role played by human rights bodies in this development is particularly important. Although some research has been carried out into the role of human rights bodies in imposing state obligations on penal matters, it generally focuses on a single institution without a comprehensive review and comparison of jurisprudence with other human rights bodies. 18 Moreover, while attention has been devoted to the immediate effects of the doctrine of positive obligations, 19 only few scholars have analysed the wider implications this trend can have on the development of criminal law itself. 20 Specifically, what remains understudied is whether the shift towards a criminal justice increasingly focused on human rights protection may result in an overturning of the liberal model of criminal law. 21 This article aims to fill both gaps. Firstly, the case-law of human rights bodies is scrutinised and compared to trace the current tendencies and standards in human rights protection through criminal justice mechanisms. Secondly, the implications and the pitfalls of the aforementioned jurisprudence is investigated. The central thesis is that the uncritical imposition of duties to criminalise and punish is likely to produce unwanted consequences.
Firstly, there is a growing leaning to require prosecution and punishment in the interests of the victim. Yet, if criminal law is exercised in the name of the victim and no longer with an exclusive focus on the defendant, the door is open for placing limitations on the rights of the accused in order to better pursue the mandate of giving justice to the victim. 22 Furthermore, in the case of grave abuses of human rights, the tendency is to assume an outright duty to employ criminal measures, without giving much weight to the justification for that call. However, IHRL may eventually require alternative, non-criminal responses for addressing past abuses. Finally, the obligations to use criminal law have the potential to enhance the state's coercive reach. When fundamental rights are no longer used as a constraint for the state's control and dominance, but become the key drivers of such power, there is a very real risk of coercive overreach upon individuals subject to the state's penal power. In this context, the question that arises is whether there is still room for a liberal model of criminal justice within the jurisprudence of human rights bodies.

C. Course of Analysis
An examination of the role of human rights bodies in triggering the application of criminal law presents an important opportunity to analyse the dynamic relationship between human rights and criminal law. Part I seeks to capture the fundamental ambiguity human rights present: both as limiting state power and as requiring it. In the context of criminal justice, while human rights have traditionally been used to limit the excesses of security of the state, they are increasingly presented as positive obligations, contributing to the extension of penal power. Part II analyses the case-law of the IACtHR, the European Court of Human Rights (ECtHR) and the UN Human Rights Committee (HRC), as well as that of other human rights bodies. It outlines current trends in human rights protection through criminal law-mechanisms. The purpose is to take stock of existing standards within the different institutions and to consider to what extent duties to prosecute and punish are seen as means of human rights protection. Part III contends that the failure of human rights bodies to assess the implications of their resort to criminal justice is likely to have negative consequences, both for effective human rights protection and for the appropriate use of penal measures. In this regard, the "liberal theory of criminal law" is applied in reviewing the case-law of human rights institutions. 17 Pastor (n 11). In his piece, Pastor criticises the Argentine Supreme Court, the IACtHR, and the Argentine human rights movement for their "neo-punitivism", arguing that increased prosecutions violate prohibitions against double jeopardy, as well as other due process rights. 18 E.g. Tulkens (n 11); Basch (n 11); Mégret and Calderón (n 11); Pastor (n 11). 19 E.g. Tulkens (n 11); Mavronicola (n 16); Basch (n 11). 20 E.g. Pastor (n 11); Sorochinsky (n 9). 21 The terms "liberal model of criminal law" and "liberal theory of criminal law" are used to refer to a defensive criminal law policy, with a strong defence of the presumption of innocence and due process, as well as a normative preference for the defendant as opposed to other interests in punishment. 22 Sorochinsky (n 9) 207.

II. The Liberal Model of Criminal Law and Its Relationship with Human Rights
Traditionally, criminal law has been conceived as the state's prerogative to preserve public order. 26 In classical liberalism, inspired by the will to limit and separate state power to enhance the individual's freedom, 27 it is conceived as an "odious law" given its interference with the autonomy of the person. 28 Accordingly, the history of the twentieth century is marked by the development of national constitutions and international instruments aimed at limiting the coercive power of the state for the sake of personal liberties and political dissent. 29 In the words of von Liszt, the constitutional aspects of criminal law are 'the criminal's Magna Charta': they ' are the citizen's bulwark against the State's omnipotence; they protect the individual against the ruthless power of the majority, against Leviathan'. 30 The last century also saw the growing recognition, development and enforcement of human rights law through domestic and international instruments. 31 In a context in which human rights have increasingly claimed greater enforcement, many have started seeing criminal law as the decisive instrument for banning practices of oppression and violence. 32 Criminal law has gradually ceased to be seen with disfavour and criminal trials have progressively become the pivotal places in which to protect individuals rights and freedoms. While IHRL has traditionally had a role of limiting the state's coercive power, today it is progressively employed to trigger criminal law in order to promote accountability and uphold the rule of law. 33 In other words, in a 'paradoxical relationship', 34 human rights appear as both the "shield" to and the "sword" of criminal law. 35 On the one hand, by providing due process rights, human rights are a defence against an authoritarian use of criminal law. 36 35 Tulkens ascribes to Christine Van den Wyngaert, judge at the ICC, the first reference to an opposition between the "shield" function and the "sword" function of human rights with regard to criminal law ((n 11) 577). 36 Safferling (n 9) 1472. 37 ibid.

A. The Defensive Function of Human Rights in Criminal Justice
Criminal law is the body of law that interferes the most with the individual's private sphere and liberties. It expresses the highest legal censure of acts within society and it provides for the harshest sanctions, including incarceration and possibly loss of life. 38 Not only may criminal punishment provoke great suffering, but also the entire criminal trial entails subjection to an authority and exposure to public stigmatisation. 39 If employed arbitrarily, criminal law may become the state's main instrument of violence and oppression. 40 On the other hand, IHRL has traditionally been concerned with neutralising the state's power when it may be employed in a manner detrimental to individual rights. 41 Consequently, the main concern of IHRL in criminal justice has long been the protection of the accused's rights and the limitation of the state's coercive power. 42 Hence, since its origin, IHRL has provided procedural and substantive principles for protecting the individual from state's arbitrary use of the criminal mechanisms. 43 Substantive principles include, for instance, the provisions that crimes and punishment must be established by law, 44 the prohibition of inhuman and degrading treatment or punishment, 45 and the prohibition upon unlawful or arbitrary detention. 46 Due process rights oppose state repression in the prosecution and adjudication of crimes, including the right to a fair trial, the presumption of innocence and the need to prove the guilt of the accused beyond reasonable doubt. 47 From the above there emerges a particular conception of criminal law, that can be defined as a liberal model. 48 Firstly, even though it may be necessary to have criminal-law provisions, the fact that they threaten fundamental rights entails that criminalisation should be minimised and used only as the last tool among many offered by social policy. 49 The second element is a higher protection granted to the defendant as opposed to the state's or victim's interests in punishment. In this regard, whenever the defendant's rights conflict with other interests or in the event of ambiguities, the legal conflict should be resolved in favour of the former. 50

B. The Offensive Function of Human Rights in Criminal Justice
In recent years, a new function of human rights in criminal matters has arisen: human rights are invoked to trigger the application of criminal law and criminal measures are increasingly seen as the preferred means of ensuring the protection of IHRL. 51 The basic idea is that perpetrators of human rights abuses deserve criminal punishment, and impunity is a violation of human rights per se. Recourse to criminal solutions by means of IHRL generally has two limbs. 52 Firstly, IHRL requires states to criminalise conduct that endangers an individual's liberties. Secondly, effective criminal law-mechanisms (investigation, prosecution and punishment) should be put in place to enforce human rights provisions.
This new trend in the relationship between human rights and criminal law is the result of three developments. Firstly, since the 1980s, the traditional conception of human rights as a state's negative duty not to interfere with individuals' freedom has shifted to a "positive" approach to IHRL, where the latter requires the public authorities to act positively to remove barriers and ensure the exercise of freedoms. 53 Hence, for instance, the prohibition of torture and the right to life now entail the obligations of the state to investigate breaches of these provisions and to hold perpetrators accountable. 54 A second development is related to the growing relevance of prosecution and punishment in international humanitarian law and transitional justice. 55 The idea of individual accountability for grave violations of international law originated after the Second World Word with the Nuremberg trials. Nuremberg established the principle that serious acts of violence perpetrated against civilians should be prosecuted and punished. 56 Although criminal punishment was incorporated into the four Geneva Conventions of 1949, the legacy of Nuremberg remained a dead letter until the end of the Cold-War, when a new wave of transition raised the opportunity to impose accountability through criminal law for the abuses of past regimes. 57 From Latin-America to Africa, from Asia to Eastern Europe, young democratic governments supplanted long-rooted dictatorships. Debates on how to deal with past abuses gave rise to a new field of scholarship, policy and practice called "transitional justice". 58 From the beginning, the field was marked by the tension between punishment and amnesty. 59 On the one hand, there was a moral obligation to prosecute and punish past human rights abuses. 60 On the other, there emerged the practical and political difficulties in holding perpetrators accountable at the domestic level while promoting peaceful transitions and political stability. 61 Yet, over time, civilian demands for justice, fostered by NGOs and international institutions, have driven the approach towards retributive responses for human rights violations. 62 In the same period, the atrocities that occurred in the former Yugoslavia and in Rwanda prompted the United Nation Security Council (UNSC) to establish ad hoc criminal tribunals for the former Yugoslavia (ICTY) in 1993 and Rwanda (ICTR) in 1994. 63 The latter led the way to the establishment of the International Criminal Court (ICC) in 1998. 64 The convergence of international crimes with serious human rights violations, as well as the fact that increasingly more international offences are committed in peacetime, have eventually contributed to strengthening the offensive function of human rights in criminal justice. 65 In particular, IHRL is increasingly used as a surrogate of international criminal law, when the latter is structurally unable to intervene. 66 Since international criminal law has failed to establish criminal responsibility for all serious human rights violations, 67 it has become more and more natural to resort to IHRL and its offensive function of promoting punishment of grave breaches of international law.
Thirdly, the new role of human rights as a driver of punishment has grown along with a general shift in focus in criminal justice. 68 Since the early 1980s, first in academia and then in practice, there has been a growing attention to the victims of crimes. 69 As a result, under the pressure of the victims' rights movement, several international instruments, as well as domestic reforms dealing with victims' rights, have been adopted. 70 The criminal-law perspective has moved from the exclusive relationship between the state and the accused to a three-party relationship that also includes the victim. 71 Moving the emphasis from the defendant to also include the victim's interests has also resulted in a new attitude toward criminal law, namely no longer being solely a tool to preserve public order, but also an instrument for empowering and protecting the oppressed. 72

III. Criminal Justice as an Instrument of Human Rights Protection
The affirmation of the role of criminal law in promoting accountability for human rights violations and the struggle against impunity emerges in particular from the jurisprudence of human rights bodies, through what has been defined by Huneeus as 'international criminal law by other means'. 73 A growing number of cases before the IACtHR, the ECtHR and the HRC, as well as before other human rights bodies, show how human rights instruments are used for triggering the criminal law. 74 Criminal justice is increasingly seen as a necessary instrument of human rights protection, especially in situations of serious human rights violations (e.g. torture, murder, enforced disappearance, human trafficking). Accordingly, in order to protect victims' rights and freedoms, positive obligations to criminalise, prosecute and punish certain offences or negative obligations not to use "criminal clemency" (decriminalisation and amnesty laws) have been imposed on states. Within the context of the Organisation of American States (OAS), since Velàsquez Rodríguez 75 the IACtHR has upheld that states have a duty to refrain from violations, to prevent and to punish them. 76 Punishment, in particular, is regarded not only as a form of deterrence, but also as a necessary remedy for the victims. 77 A slightly similar approach is visible in the case-law of the ECtHR and the HRC, although neither of these two bodies has come to recognise a victims' right to have offenders punished. 78

A. Inter-American System of Human Rights
The efforts of the Organization of American States (OAS) bodies, the Inter-American Commission on Human Rights (IAComHR) and the IACtHR, in advancing criminal responsibility for human rights violations derive from a context of widespread impunity in Latin-America and massive repressions perpetrated during the authoritarian regimes in the 1970's and 1980's. 79 Pursuant to IACtHR's case-law, each Member State 'has the obligation to use all the legal means at its disposal to combat' situations of impunity. 80 Accordingly, the Inter-American institutions have progressively relied on criminal punishment as the preferred means for ending human rights abuses. In this regard, the OAS bodies have developed a rigorous jurisprudence requiring states to punish almost every human rights violation, including by invalidating or refraining from adopting amnesty laws. 81

Duty to Punish
The IACtHR's doctrine on state obligations in criminal matters has its starting point in Velásquez Rodríguez, the first decision of the Court in a contentious case. 82 Relying on Article 1(1) of the American Convention on Human Rights (ACHR), the Court held that State Parties have two obligations. First, they must respect the 'rights and freedoms recognized by the Convention'. Second, they must '"ensure" the free and full exercise of the rights recognized by the Convention to every person subject to its jurisdiction'. 83 As a consequence of this second obligation, 'the States must prevent, investigate and punish any violation of the rights recog- nized by the Convention', regardless of whether they are involved in the violation at issue. 84 Therefore, in addition to abuses perpetrated by public officials, States have a duty to punish human rights violations that are committed by private individuals and, as such, are not clearly attributed to the State. 85 Having found Honduras in breach of Article 1(1), the Court did not order the state to adopt criminal measures. Rather, it requested the latter to compensate the victim's next-of-kin. 86 However, in following cases from the mid-1990s onwards, the Court began to require the state to effectively punish individual perpetrators. 87 In Velásquez Rodríguez and other cases in the 1990s, the IACtHR sustained, pursuant to Article 1, the call for punishment as a means to protect human rights in general. 88 Moreover, the Court also derived from Article 1(1) both a duty to prosecute and a duty to criminalise serious violations of human rights. 89 On the other hand, since the decision in Panigua-Morales, the IACtHR has started relating the state obligations in criminal matters to the fulfilment of victims' rights. 90 Here, the Court justifies the duty to criminalise, prosecute and punish as a means to guarantee the victims 'free and full exercise' of their human rights. 91 In this way, the prosecution and punishment of the perpetrator become ways to ensure a retrospective restoration of victims' rights. 92 Following decisions of the Inter-American Commission, 93 the IACtHR established the idea that the state's engagement with criminal law is necessary to protect victim's rights to a fair trial (Article 8) and to judicial protection (Article 25). Article 8 was originally provided to protect the accused's right to a due process within domestic criminal systems. 94 Yet, its broad scope, that also encompasses '[every person's] right to a hearing […] for the determination of his rights and obligations of a civil, labor, fiscal, or any other nature', 95 has allowed the Court to derive the victim's right to a fair trial during the prosecution of the offender. 96 According to the IACtHR, this provision covers the victim's right 'to have [the crimes] effectively investigated […], those responsible prosecuted for committing said unlawful acts; to have the relevant punishment, where appropriate, meted out'. 97 Accordingly, the right to a fair trial is read as it provides for a right of victims to criminal justice. This does not include only a right to participation, but also the right to have the offender prosecuted and punished by a court. 98 The second source of victims' rights is Article 25, which guarantees to everyone the right to an effective remedy in the event of violations of his/her rights. In Loayza-Tamayo, the IACtHR held that every state must 'guarantee to every individual access to the administration of justice […] so that, inter alia, those responsible for human rights violations may be prosecuted and reparations obtained for the damages suffered'. 99 Currently, the IACtHR tends to combine Article 8 with Article 25 as providing a victim's right to criminal justice. 100 In Goiburú, the IACtHR held that in cases of violations of particular gravity (including torture and forced disappearance of persons) the state obligation to punish those responsible has attained the status of jus cogens. 101 The offensive function of human rights in criminal justice emerges clearly from the jurisprudence of the IACtHR: the use of criminal courts and procedure appears at times to be no longer motivated by the need to protect the accused right to a fair trial, but by the need to advance victims' rights. 102 In particular, should the interests of the accused's and those of the victims collide, for the Court the latter must prevail. 103 In this regard, when rules in favour of the accused risk preventing the prosecution or the punishment of perpetrators of serious human rights violations, the Court has often asked for the rights of the accused to be restricted or even neutralised. 104 There are two clear examples of this trend. Firstly, Article 8(4) provides for the principle of ne bis in idem for the accused acquitted by final judgment. However, the IACtHR stated that investigation can be re-opened 'even if the case ended in an acquittal with the authority of a final judgment, since the dictates of justice, the rights of the victims, and the spirit and the wording of the ACHR supersedes the protection of the ne bis in idem principle'. 105 Secondly, the Court has maintained that 'the State may not invoke […] the non-retroactivity of criminal law […] to decline its duty to investigate and punish those responsible' of the victim's death, 106 even though Article 9 provides no exception for the prohibition of ex post facto law.
To date, the IACtHR's doctrine on the duty to punish has been maintained in many cases. For the court, this duty applies to situations of serious human rights violations, as well as to other violations of rights enshrined in the Convention. 107 The rationale behind the Court's jurisprudence is that criminal prosecution and punishment are necessary mechanisms to end impunity and ensure human rights protection. 108 Ultimately, the development of this jurisprudence is a prominent example of the new role human rights have in criminal justice, as a driving force of punishment and criminalisation rather than as protection against state coercive power.

Prohibition of Amnesties
The Inter-American institutions have dealt with amnesty laws since the early 1990s. 109 In a series of cases in 1992, the IACommHR stated that amnesties for serious crimes in force in Argentina, El Salvador and Uruguay were incompatible with the ACHR since they had prevented the possibility of a criminal investigation and prosecution. 110 In 1996, the IACommHR stated that the victim's 'right to justice' stemming from Article 8(1) and 25 ruled out amnesties for serious human rights abuses. 111 Furthermore, according to the Commission, the application of amnesties eliminated 'the most effective means for protecting [human] rights, which is to ensure the trial and punishment of the offenders'. 112 Yet, throughout the 1990s the IACommHR did not refer any case concerning amnesties to the Court.
In 2001, in Barrios Altos, the IACtHR pronounced its first decision on amnesty laws. 113 In what is considered the leading case on amnesties, the Court held that 'self-amnesty laws lead to the defenselessness of victims and perpetuate impunity; therefore, they are manifestly incompatible with the aims and spirit of the Convention'. 114 Then, the IACtHR found specific violations of the duty to prosecute and punish derived from Article 1(1) and 2, as well as of victims' rights stemming from Article 8(1) and 25. 115 In 2006, the Court confirmed its jurisprudence in Almonacid. 116 Here, the IACtHR ruled that the Chilean amnesty law constituted a violation of the Convention since it prevented the punishment of crimes against humanity. 117 Although the legislation concerned was not technically a self-amnesty because it applied also to political opponents, for the Court it acted as if it were, 'since it was issued by the military regime to avoid judicial prosecution of its own crimes'. 118 Moreover, the IACtHR excluded the possibility of truth and reconciliation commissions being an admissible 'substitute for the duty of the State to reach the truth through judicial proceedings', since they are unable to satisfy the victims' right to have the offender punished. 119 The third landmark case is Gomes Lund. 120 The IACtHR found Brazil in breach of the Convention, because its amnesty law prevented criminal prosecution and punishment for enforced disappearance. Specifically, the Court established that only criminal prosecution was ' an appropriate instrument to prevent future violations of human rights' of jus cogens nature, including the crimes covered by the Brazilian amnesty law. 121 Most importantly, the IACtHR clarified that 'the non-compatibility with the Convention includes amnesties of serious human rights violations and is not limited to […] "self-amnesties"'. 122 In 2011, in Gelman the Court invalidated Uruguay's amnesty law. 123 This legislation differed from the other amnesties scrutinised by the IACtHR because it had been enacted during the peace negotiations that led to the end of a civil war and it had been upheld twice by popular referenda in 1989 and in 2009. Nevertheless, the Court ruled that all ' amnesty laws are, in cases of serious violations of human rights, expressly incompatible with the letter and spirit of the [Convention]'. 124 The most recent judgment on amnesty laws is El Mozote. 125 In this case, the IACtHR did not limit itself to upholding its doctrine on an absolute ban on amnesties shielding human rights abuses, but it went as far as to interpret Article 6(5) of the 1977 Additional Protocol II to the 1949 Geneva Convention 126 as excluding amnesties that prevent prosecution and punishment of war crimes. 127 To summarise, according to the Inter-American institutions, amnesties for serious human rights violations are always in contrast with the ACHR, regardless of their nature, purposes, and the context in which they have been adopted. 128 They are in breach of the general and absolute duty to ensure human rights through investigation, prosecution and conviction, as well as violating the victim' rights to a fair trial and judicial protection. 129 In this regard, criminal accountability ceases to be an instrument whose justification in terms of necessity is constantly required and becomes the preferred and unquestioned method for promoting justice.

B. European Court of Human Rights
In the context of the European human rights system, there has also been a growing tendency to demand the intervention of criminal law to counter human rights violations, especially in response to the structural deficiency in accountability during and after the Kurdish and Chechnyan conflicts. 130 By imposing positive obligations on states, the ECtHR has increasingly triggered the application of domestic criminal law to protect the rights and freedoms enshrined in the European Convention (ECHR). 131 In particular, the Court tends to support several forms of positive obligations in criminal matters. 132 Firstly, states have the duty, in cases of violations of Convention rights, to ensure effective investigations capable of leading to the identification and punishment of the offender. 133 Secondly, states are asked to criminalise serious offences. 134 Another obligation is the duty to establish an effective criminal-law system to deter the commission of offences and prosecute those responsible. 135 Moreover, the Court requires states to punish serious human rights violations, including torture, domestic and sexual violence as well as systematic killings. 136 Finally, there is the duty to refrain from adopting measures that would prevent criminal justice. 137

Positive Obligations to Criminalise, Prosecute and Punish
The jurisprudence of the ECtHR concerning the use of criminal law to protect human rights results from the theory of "horizontal applicability of human rights" and the doctrine of "positive obligations". 138 Since the decision in Marckx, the Court has established that not only do states have "negative obligations" not to interfere with individuals' rights, but they also have "positive obligations" to ensure effective protection of the provisions of the ECHR. 139 States are not only responsible for violations of Convention rights attributed to their agents, but even abuses committed by private individuals where the state failed to prevent the violations due to negligence or tolerance. 140 Both the horizontal effect and the theory of positive obligations emerge from X and Y, the leading judgment concerning the duty to protect human rights by means of criminal law. 141 The point at issue was the inability of the applicant to open a criminal prosecution against the alleged rapist of his daughter, a minor with learning disabilities. Having criticised the shortcomings in the national legislation, the ECtHR held that in the case of wrongdoing against sexual integrity ' effective deterrence is indispensable' and 'it can be achieved only by criminal-law provisions'. 142 The requirement that states must have in place criminal laws securing compliance with Article 8 in the sphere of sexual life has subsequently been confirmed in subsequent cases. 143 The duty to create particular criminal offences so as to protect fundamental rights has also been announced in regard to other rights. Firstly, in Kiliç the ECtHR stated that states have an obligation to secure the right to life (Article 2) 'by putting in place effective criminal-law provisions'. 144 Yet, in Cavelli and Ciglio the Court stated that 'if the infringement of the right to life […] is not caused intentionally, the positive obligation imposed by Article 2 to set up an effective judicial system does not necessarily require the provision of a criminal-law remedy in every case'. 145 Secondly, in A v UK the Court concluded that the protection of criminal law is required to prevent torture or inhuman or degrading treatment (Article 3). 146 In MC, in particular, the ECtHR pointed out that 'the member States' positive obligations under Articles 3 […] must be seen as requiring the penalisation and effective prosecution of any non-consensual sexual act'. 147 Furthermore, in Cestaro the Court ordered Italy 'to introduce into the Italian legal system legal mechanisms capable of imposing appropriate penalties on those responsible for acts of torture'. 148 Finally, positive obligations to criminalise are provided in relation to the prohibition of slavery (Article 4). 149 Notably, in Siliadin the ECtHR held that 'states have positive obligations […] to adopt criminal-law provisions which penalise the practices referred to in Article 4'. 150 From the case-law it emerges that, unlike the IACtHR, the recourse to criminal law is prescribed by the ECtHR only in the most serious violations (i.e. murder, torture, sexual and domestic violence) and only when the core of a fundamental right is at stake. 151 However, the duty to criminalise certain human rights breaches is not limited to cases where such duty already exists in customary or conventional international law (i.e. Article 4(1) of the Convention Against Torture), but it has progressively been extended to other violations where the opportunity of criminalisation is more questionable. 152 Another obligation imposed by the ECtHR is the duty to enforce criminal law. In Öneryildiz the Court expressed the view that states should establish a criminal-law machinery ' capable of ensuring that criminal penalties are applied' when the right to life is breached, 153 while in Ali and Ayşe Duran the ECtHR stated the same principle in relation to Article 3. 154 Generally, the Court does not provide for an express duty to criminal prosecution, only requiring the state to conduct ' effective investigation capable of leading to the identification and punishment of those responsible'. 155 Yet, in recent cases, the Court has gone as far as to maintain that in cases of serious violence (e.g. assaults for racial reasons or police ill-treatments), measures that could lead only to civil compensation, but not to the prosecution of the offender, do not fulfil the State's procedural obligations under Article 2 or 3. 156 Some authors have discussed whether the ECtHR has ever affirmed a duty to punish serious human rights violations. 157 In most cases, the Court only affirms a general obligation to set up ' a law-enforcement machinery for the prevention, suppression and punishment' of breaches of Convention rights, without providing for an express duty to sanction the individual offender. 158 Nevertheless, there are a few decisions where the Court has been more willing to assert an obligation of punishment. 159 In some cases of torture by state officials the ECtHR has highlighted the necessity of punishment, 160 but in relation to the right to life the Court has recognised that the obligation to punish would apply only 'if appropriate'. 161 From the case-law it emerges that the ECtHR generally imposes state obligations in criminal matters with the aim of ensuring the effective enjoyment of Convention rights. 162 In this regard, criminal law is employed for prevention and deterrence. 163 Moreover, when the Court has to deal with extreme situations of human rights violations, criminal justice is considered as the preferred means to ensure ' adherence to the rule of law', 164 to avoid impunity 165 and to restore 'public confidence in the State's monopoly on the use of force'. 166 Accordingly, the need for criminal-justice responses to human rights abuses is justified by the Court in order to safeguard society as a whole from future human rights violations.
Indeed, unlike the IACtHR, the obligation to employ criminal law does not derive from an individual right of the victim to have the offender prosecuted or punished as a form of restoration. 167 The individual victim is only recognised as having a right to criminal investigation being instituted and, if appropriate, punishment of the perpetrator. 168 Yet, according to the Court, this is an obligation of means, not of result. 169 Nevertheless, in a handful of cases, the ECtHR has shown that future developments in line with the IACtHR cannot be excluded. 170 In Nikolova and Velichkova the Court went so far as to consider criminal proceedings and sanctions as remedial measures for the victim. 171 Similarly, in Al Nashiri v Romania and Abu Zubaydah v Lithuania, the ECtHR held that 'the notion of an "effective remedy" entails, in addition to the payment of compensation where appropriate, a procedure enabling a thorough and effective investigation capable of leading to the identification and punishment of those responsible'. 172 Most importantly, the ECtHR, along the same lines as the IACtHR, has shown no hesitation in considering the criminal-law mechanism as the best means to protect fundamental rights. This tendency has been noticed by Judge Tulkens who, in several concurring or dissenting opinions, felt free to criticise such human rights development. 173 As a former professor of criminal law, she warned her colleagues of the risk from using criminal punishment to protect rights and freedoms, namely that of losing 'sight of the subsidiarity principle, which is a basic axiom of criminal law'. 174 Indeed -she added -'use of the weapon of punishment is acceptable only if there are no other means of protecting the values or interests at stake'. 175

Amnesties
Throughout its jurisprudence, the ECtHR has taken a censorious approach towards states' use of measures that hinder criminal justice, including amnesties, pardons and statutes of limitations. 176 In 1991, the European Commission of Human Rights was called upon to review an amnesty law that had been enacted by France for New Caledonia with the aim of resolving conflicts in the island. 177 The Commission declared the application inadmissible, by stating that 'the crime of murder may be covered by an amnesty […] unless [the amnesty] can be seen to form part of a general practice aimed at the systematic prevention of prosecution of the perpetrators of such crimes'. 178 This reasoning, based on the idea that amnesty laws are permissible provided that they are exceptional in character and necessary for pursuing a legitimate aim, was followed by the Court in Tarbuk. 179 This approach differs from the absolute ban provided by the IACtHR, but is limited to complaints in relation to the right to life. When the ECtHR is called upon to deal with alleged violations of Article 3, it tends to adopt a stricter approach. In Abdülsamet Yaman and Yeşil and Sevim, for instance, the Court declared that statutes of limitations, amnesties and pardons are inadmissible in cases of torture or ill-treatment. 180 This idea was restated in Ould Dah. 181 Notably, the Court held that '[t]he obligation to prosecute criminals should not […] be undermined by granting impunity to the perpetrator in the form of an amnesty law that may be considered contrary to international law'. 182 In 2014 the ECtHR pronounced its most recent and authoritative decision on amnesties in Marguš. 183 The applicant, a Croatian national, complained that his 2007 conviction for war crimes was in violation of the prohibition of double-jeopardy since he had previously been subjected to proceedings for the same charges and released by application of an amnesty law. Having reviewed the state of international jurisprudence on amnesties, the ECtHR stated that a 'growing tendency in international law is to see amnesties [for war crimes] as unacceptable because they are incompatible with the unanimously recognised obligation of States to prosecute and punish grave breaches of fundamental human rights'. 184 For these reasons, according to the Court, granting amnesty to the applicant amounted to a fundamental defect. Yet, the European judges left open the possibility that amnesties for war crimes might be acceptable in 'some particular circumstances, such as a reconciliation process and/or a form of compensation to the victims'. 185 In summary, according to the ECtHR, measures that would prevent criminal justice are generally incompatible with the ECHR since they foster impunity and hinder accountability. Ultimately, from the Court's point of view, it is no longer the employment of a criminal-law mechanism that should always be justified, but rather the decision not to make use of it.

C. Other Human Rights Bodies
In recent years, there has been a general rising tendency to favour criminal accountability for those responsible for serious human rights violations. 186 Beyond the IACtHR and the ECtHR, other international and regional human rights bodies have begun to order criminal prosecutions and punishment when states failed to act. 187

UN Human Rights Committee
Since the beginning of the 1990s, the HRC has developed a significant jurisprudence focused on the duty of states to bring to justice and punish perpetrators of serious human rights abuses, including arbitrary killing, enforced disappearance, torture or ill-treatment, sexual and domestic violence, as well as human trafficking. 188 In early cases the Committee merely used to order states 'to bring to justice any persons found to be responsible' for human rights violations. 189 Conversely, in more recent views, the HRC has explicitly prescribed prosecution and punishment. 190 The underlying rationale is that impunity fosters human rights violations and, accordingly, punishment is necessary to deter future violations. 191 According to the HRC, the criminal sanction is not only an option for the state, but it is mandatory in order 'to respect and to ensure to all individuals' the rights recognised in the International Covenant on Civil and Political Rights (ICCPR). 192 Until 2008, the position of the HRC was that the ICCPR ' does not provide that private individuals have the right to demand that the State criminally prosecute another person'. 193 However, in more recent cases, the Committee has started pointing out that the failure to prosecute and punish violates both the victim's individual rights (especially the right to life and the prohibition of torture and mistreatment) and the right to an effective remedy under Article 2(3) (a). 194 According to the HRC, ' criminal investigation and consequential prosecution are necessary remedies for violations of human rights'. 195 Therefore, 'the State party is under an obligation to provide the [victims] with an effective remedy, including initiation and pursuit of criminal proceedings'. 196 This implies that the duty to prosecute and punish gross human rights abuses may no longer be conceived as a due diligence obligation, but as mandatory in all circumstances. 197

Pinto 175
The Committee has also supported a general prohibition of amnesties for grave human rights violations. Amnesty laws are generally seen as 'incompatible with the duty of States to investigate' serious abuses, 'to guarantee freedom from such acts', as well as 'to ensure that they do not occur in the future '. 198 Consequently, states should ensure that amnesties are 'not applied or utilized for granting impunity to persons accused of serious human rights violations'. 199 Thus, the HRC has repeatedly rejected the claim that amnesties might be necessary for reconciliatory purposes. 200

Other Bodies
There are several human rights conventions that stipulate an obligation to prosecute and, where appropriate, to punish. 201 Notably, Article 4(1) of the UN Convention Against Torture (CAT) requires State parties to ensure that ' all acts of torture are offences under [their] criminal law'. The UN Committee Against Torture, the CAT monitoring body, has repeatedly declared that states must investigate and punish acts of torture and ill-treatment. 202 The same body has also expressed negative opinions towards amnesties. 203 Neither the CAT nor the Committee Against Torture grant an individual right of the victim to have the perpetrator punished. 204 The African Commission on Human and Peoples' Rights has also criticised amnesties granting immunity from prosecution to violators of human rights on the basis that they infringe the victims' right to legal protection enshrined in Article 7(1) of the African Charter on Human and Peoples' Rights. 205 In the Arab world and in Asia, mechanisms such as the Arab Committee of Human Rights and the ASEAN Intergovernmental Commission on Human Rights have not yet produced relevant case-law in relation to positive obligations in criminal matters.

D. Conclusion
The overview of the practice of human rights bodies shows an ever-rising tendency to order states to employ their domestic criminal law in the context of human rights violations. It should be noted that this practice has largely been confined to serious human rights violations that involve an absolute right, 206 but it has started creeping into somewhat lesser human rights breaches. Criminal prosecution and punishment are increasingly seen as necessary tools to advance the promoting and safeguarding of human rights, as well as to deter future abuses. Although the duty to employ criminal law is often interpreted as a due diligence obligation, there is growing case-law requiring states to prosecute perpetrators of gross human rights abuses in all circumstances.
Yet, there are still important differences within the jurisprudence of different bodies. The OAS bodies have the most radical approach. Not only is criminal law considered as a necessary mechanism to end human rights violations, but it is also seen as indispensable to protect the rights of the victims and give them restoration. In this respect, the IACtHR has long established the victim's right to have the offender tried and punished. Given the conception of criminal law as the undisputed and essential tool in safeguarding human rights, every measure that hinders criminal justice is declared by the Court of no legal effect, including amnesty laws, statutes of limitations, the principle of non-retroactivity and the prohibition against double-jeopardy. 207 From the ECtHR practice it also emerges a growing tendency to require criminal-law enforcement, as part of states' duties to investigate, prosecute and punish those responsible for human rights violations, as well as part of the obligation to criminalise serious offences. However, this has been affirmed only with regard to serious violations, namely the right to life, acts of torture and ill-treatment, domestic and sexual violence, and systematic breaches of Convention rights. For now the ECtHR has yet to go as far as the IACtHR to require punishment as a measure of individual redress and satisfaction. Nevertheless, there are signs of a potential development in this direction. 208 Regarding the HRC, prosecution and punishment are required in case of gross human rights abuses. In addition, the HRC has started considering the criminal sanction as a remedy but it has remained more cautious on conceptualising a victim's right to criminal justice.
Finally, there are also common traits among different human rights bodies. One is the idea that a greater use of criminal law will benefit society as a whole and provide general human rights protection. Another shared feature is the aversion towards amnesty laws, especially for international crimes, because they prevent criminal prosecutions, trials and punishments from taking place. Yet, while the ECtHR has to date accepted that amnesties for serious crimes might be acceptable in exceptional circumstances, the IACtHR and the HRC appear to have imposed an absolute ban on amnesties shielding gross human rights abuses.

IV. The Impact of the Human Rights Treaty Bodies' Jurisprudence
In the development of positive obligations in criminal matters, human rights bodies appear to have often failed to assess the potential implications of their doctrine. While it is always important to consider the impact of a decision, this is especially true in the context of criminal law as a threat to an individual's freedoms. 209 Indeed, human rights bodies are mostly concerned with individual accountability and protection of victims in the particular case rather than with the impact of their outcomes on domestic criminal justice systems, as well as on the conception of criminal law. There is no serious assessment concerning the necessity of criminal law, nor on what purposes criminal accountability may serve or whether such may be achieved. 210 Even in case of violations of absolute human rights (arbitrary killings, enforced disappearances, torture), uncritical reliance on criminal-law mechanisms may become problematic if it comes to disregard relevant considerations on the opportunity of criminalisation. 211

A. Allegedly Positive Effects
Human rights bodies generally share a basic assumption: prosecuting and punishing perpetrators of human rights violation is the optimal method for combating impunity. 212 Consequently, if prosecution and punishment for those who committed atrocities is "intrinsically good", little or no justification is needed for imposing state obligations in criminal matters. 213 However, although there is no doubt that human rights violations should be condemned and prevented, the idea that criminal law best serves the purpose of addressing these abuses is more controversial.
The aims pursued by human rights bodies when they turn to criminal law are mostly deterrence, restoration of the rule of law and, for the IACtHR, redress for victims. 214 Yet, whether these purposes are achieved by criminal prosecutions and punishment is a matter of great debate. According to Orentlicher ' criminal punishment is based, above all, on a deterrence rationale'. 215 In this respect, -she continues -'states' duty Pinto 177 to punish human rights crimes is owed as much to society as to individual victims'. 216 Kim and Sikkink have supported the idea that human rights prosecutions have, at least, ' a strong and statistically significant impact' on decreasing the level of abuses. 217 This view is shared by Safferling, who maintains that criminal prosecutions may effectively contribute to reducing human rights violations, strengthening the legal system, and helping individual victims to bear their suffering. 218 Yet, these claims are rejected by other scholars. In his response to Orentlicher, Nino contends that an indiscriminate duty to punish human rights abuses 'would probably provoke, by a causal chain, similar or even worse abuses'. 219 Koskenniemi has highlighted the inability of criminal law to deal with events of "extraordinary evil", given their ' enormous moral, historical, or political significance'. 220 Similarly, Drumbl argues that '[c]riminal law […] is inherently limited as a response to mass atrocity and as a device to promote justice in its aftermath'. 221 According to Engle, criminal law's focus on individual accountability risks overlooking the root causes of human rights violations. 222 Finally, Aukerman has shown that, in cases of serious human rights abuses, not only are prosecutions 'ineffective deterrents compared to other political and diplomatic tools'. 223 but also the merits of prosecution in establishing the rule of law and providing restorative justice are unclear. 224 Because it is controversial that prosecution and punishment are the most effective means to protect human rights, human rights bodies should not blindly turn to criminal law nor should they rule out noncriminal-law alternatives as less effective.

B. Weakening the Rights of the Accused
If not adequately justified and pondered case-by-case, the human rights bodies' mandate requiring states to use criminal law to protect human rights has the potential to lead to undesirable outcomes. Although this practice concerns mainly serious abuses, the human rights bodies have laid the groundwork to extend it to other violations. Should this approach become more pervasive, there may be troubles to come. Concerns arise especially regarding the consequences that the doctrine of positive obligations in criminal matters may produce for the liberal model of criminal law, which has to date characterised human rights bodies' approach to crime and punishment. 225 The first cause for concern is that decisions invoking indiscriminately the duty to prosecute and punish might negatively affect defendants' rights in concrete criminal cases.
Firstly, criminal punishment is increasingly presented as the only effective method to secure accountability for grave human rights breaches. As illustrated by Mégret and Calderón,226 this approach may encourage a "culture of conviction" where punishment is seen as the end to pursue whatever the cost. While this risk is present but yet limited within the context of the ECHR by a jurisprudence that refers to the duty to punish as an obligation of means, 227 the same cannot be said with regard to the IACtHR, where failure to punish is increasingly seen as a violation of human rights per se. 228 Secondly, while the focal point for human rights courts in relation to criminal law has long been the rights of the accused to due process, the greatest attention is now devoted to the need to satisfy the rights of the victim. 229 This is not only the case of the IACtHR (where the Court has recognised the victim's right to have perpetrators of human rights violations properly tried and punished), but also of the ECtHR, where criminal proceedings are at times seen as remedial measures. 230 The structural framework of IHRL contributes to placing the victim rather than the defendant in the spotlight. In a system where cases are brought to court by victims or their next-of-kin and where the respondent is the state and not the accused, it seems inevitable that the focus is on the victim rather than on the criminal defendant, to whom human rights bodies do not even give standing. 231 This situation is not problematic per se. The victim's right to justice and the defendant's due process guarantees are not mutually exclusive and ordered prosecution may of course be undertaken in compliance with the latter rights. Nonetheless, the accused's and the victim's interests may collide.
In a situation of conflict between the call for criminal punishment and the protection of due process for the accused, the temptation may be to give precedence to the former, whilst restricting the fundamental rights of the latter. In La Canuta, for instance, the IACtHR stated that the victim's right to an expeditious trial should trump the defendant's right to have adequate time for his defence. 232 This normative preference for the victim is particularly acute where rules in favour of the accused might culminate in shielding perpetrators of human rights abuses from criminal accountability. In this regard, human rights bodies appear to be willing to construct various limitations to the principle of fair trial, the prohibition against double-jeopardy and retrospective law when they are perceived to hinder criminal responsibility. 233 Ultimately, since in several legal systems such principles are not yet settled enough, should this trend become predominant, domestic authorities may interpret the decisions of human rights bodies as enabling them to disregard the fundamental rights of defendants. 234 As shown by Tittemore, this may already be the case in South America, where national courts in Argentina, Chile, Peru and Colombia have declared the prohibition against applying provisions designed to shelter criminal defendants in cases involving serious human rights abuses. 235 In the Simón Case, for instance, the Argentinian Supreme Court relied on Barrios Altos to prevent national legal provisions such as amnesty, statutory limitations and the legality principle from hindering criminal accountability for gross human rights violations. 236 In addition, there is the risk that in the long-run the courts may pass the idea that due process rights are applicable only to defendants charged with crimes that do not constitute a human rights violations, while others deserve a less robust form of protection from punishment. As observed by Basch, 237 such an approach is not dissimilar to the "criminal law of the enemy" of Jakobs. 238 Yet, this model, based on the idea that not all human beings possess the dignity of humanity, is very far from a liberal model of criminal law, as well as from the concept of human rights as rights of every man and woman.
At the same time, the gravity of harm caused and the seriousness of the abuses committed in the context of grave human rights violations is not helping to maintain the primacy of the accused's rights. The temptation is to apply the medieval principle "in atrocissimis licet iudici iura transgredi". 239 The statement in Opuz, where the ECtHR underlined that 'in domestic violence cases perpetrators' rights cannot supersede victims' human rights to life and to physical and mental integrity', 240 appears to go in this direction. 241 However, the rights of the defendant should be protected regardless of the gravity of the harm in question, and, when not absolute, balanced with competing rights and interests, but they must never be suspended. Every individual, whatever the allegation of crime, is entitled to equal rights when facing criminal trial. Ultimately, the risk is not just of weakening the position of the accused, but of jeopardising the core function of human rights as shield and shelter against the arbitrariness of power.

C. Disregarding the Principle of Subsidiarity
The second reason of concern is that, as long as prosecutions and punishments are presented as unquestionably positive methods of human rights protection, the conceptualisation of criminal law as ultima ratio may cease to have relevance. 242 The indisputable ideal, that human rights violations do not go unanswered, may become questionable if accountability and justice are limited to criminal punishment. 243 Throughout their case-law, both the ECtHR and the IACtHR have maintained that criminalisation of certain abuses is required to effectively secure certain human rights. 244 In particular, from this jurisprudence there emerges an endorsement of criminal-law mechanisms -as opposed to other forms of legal regulation -as a means of deterring future crimes, promoting the rule of law and restoring public confidence in government, as well as providing justice to the victims. As shown by Engle, 'the correspondence between criminal prosecution and human rights has become so ingrained that expressing opposition to any particular international prosecution is sometimes seen as anti-human rights'. 245 In this context, the principle of subsidiarity of the criminal law finds it difficult to resist such impulses of criminalisation that arise from the human rights framework. 246 The principle that criminal law must be regarded as a last resort is one of the fundamental pillars of a liberal model of criminal law. 247 It postulates that criminal law should only be applied if less intrusive measures are inadequate to achieve the aims of security. 248 This principle is based on the recognition that criminal law entails the harshest restrictions on individual liberties and, therefore, it should be employed with restraint. 249 Its practical application presumes that, before resorting to criminal law, an evaluation should be carried out in order to ascertain whether criminal law should be applied or if less intrusive measures (including civil reparations, loss of rank and monetary fines) would be sufficient. 250 However, should human rights bodies consider criminal law as the only adequate method for addressing serious human rights violations, any different form of accountability is excluded. 251 Yet, if we accept that finding those responsible for abuses, preventing future crimes, restoring the rule of law and giving redress are the goals to be pursued with a view to human rights protection, it appears that, beside criminal prosecution, there are other possible means, including civil trials, 252 truth and reconciliation commissions (TRCs), and other non-legal measures (such as legislative reparations, public inquiries, transparency, and the politics of commemoration). 253 Actually, as argued by Aukerman, the superiority of criminal law in pursuing those aims is controversial and, rather, in most cases prosecution and punishment 'have no clear advantages over other mechanisms'. 254 Criminal trials are focused on individual cases and inevitably fail to investigate the roots of systematic abuses as a bigger social phenomenon. In this regard, TRCs are more suitable institutions, being a forum for discussion and confrontation between victims and offenders. Moreover, economic, political, social, and cultural factors are to be taken into account whilst assessing the efficiency of criminal law in comparison to other measures. Clearly, what works well in one country might not work as well in others. It is for this reason that the resorting to criminal law at the international level might be especially problematic without a careful examination of its local suitability. 255 Ultimately, instead of accepting uncritically that serious human rights abuses are to be addressed primarily through criminal law, human rights bodies should first look at the specifics of the case to determine which measures are adequate and necessary to secure effectively the rights at issue. Penal instruments might indeed prove to be necessary, but a compelling justification is constantly needed for their employment.
Most importantly, such justification cannot only be built upon the gravity of the offence or the importance of the rights violated, but also upon the actual need for criminal law in concrete cases. 256

D. Enhancing the State's Coercive Power
Another cause for concern arises regarding the future application of the human rights courts' doctrine on penal obligations within domestic systems. In particular, when human rights bodies impose on states duties to resort to criminal law, human rights protection is advanced to demand more criminalisation and, ultimately, more coercive power on the part of the state. 257 Yet, criminal law is the most disruptive instrument a state can apply against individuals. 258 Therefore, should the call for penal measures not properly balance the need to fulfil the rights of individuals at harm whilst safeguarding the perpetrators' rights of defence, there is a serious risk of justifying the state's arbitrariness through the very same IHRL. In other words, the potential for the state to exploit the doctrine on positive obligations in order to enhance its power to the detriment of individuals' autonomy becomes high. 259 Security legislation that has been adopted all around the world in the aftermath of the 9/11 attacks clearly illustrates the potential for negative spill-overs in requiring states to adopt coercive protective measures. 260 In particular, the right to security and the correlated duty to protect people's lives have been employed to justify increasing police power, to allow complete surveillance over citizens, and to authorise indefinite detention without trial. 261 Similarly, the idea of human rights security through criminal law might offer domestic authorities the opportunity to invoke human rights in defence of various penal measures that restrict freedom. 262 Encouraging states to make more extensive use of penal measures might have troublesome consequences especially (but not only) in young and fragile democracies where there is a history of abuses committed by public authorities in their alleged fight against crime and terrorism. As noted by some authors, the 'risk of coercive overreach' is already high both in Latin-American and European countries. 263 Engle reports, for instance, that, in the wake of the international pressure on the Mexican government to prosecute those responsible for the murders of some women in Ciudad Juárez, the state authorities arrested and detained a number of innocent people, and also tortured them to coerce confessions. 264 As a matter of fact, an uncritical invocation of criminal justice could mean in practice large-scale incarceration, harsher punishments and wider powers to arrest and detain suspects. Conversely, criminal justice reform initiatives, directed at implementing alternatives to prison and humanising imprisonment, risk being completely neglected.
Human rights have long served to protect the individual from governmental interference. Ironically, here the situation is reversed: IHRL end up legitimising and even requiring the use of the most intrusive forms of governmental interference. 265 That is not to say that the intervention of the state may not be advisable and even necessary to secure fundamental rights. 266 Yet, such intervention should not violate, or appear to violate, the very essence of human rights. 267 Therefore, any extension of the state's power needs to be adequately justified and carefully pondered, and not simply triggered by the finding of a positive obligation.

E. The End of a Liberal Model of Criminal Law?
The resort to criminal law in securing human rights protection that emerges throughout the jurisprudence of human rights bodies is a prime example of the new, offensive role of human rights in criminal justice. However, the emergence of the latter function appears to be accompanied by a general weakening of the traditional, defensive role of human rights in relation to criminal punishment. While the IACtHR is clearly moving in this direction, the ECtHR's approach, though not yet reaching the OAS bodies' levels, is taking the European Court along a similarly problematic path. 268 Alongside this, one wonders whether human rights bodies are abandoning a liberal model of criminal justice to embrace a more authoritarian approach.
The very concept of human rights, permeated by the idea of limiting the state's power, has always prompted liberal models of criminal justice, and human rights bodies had long been the leading supporter of defensive criminal-law policies. 269 A liberal system of criminal justice (also known as defensive 270 or due process model) 271 is based on the idea that criminal repression should be constrained in respect of the autonomy of the individuals who are subject to the system. 272 Therefore, principled restraints and compelling justifications on criminalisation and on the infliction of punishment are to be provided. 273 Many scholars have elaborated theories on limiting the reach of criminal law. Feinberg's work on the harm principle as a limit to criminalisation is one of the most familiar examples. 274 Inspired by JS Mill, 275 Feinberg established that liberal democracies ought to criminalise behaviours only if they are harmful to others. Yet, harm is not sufficient: in order to resort to criminal law, there should also not be ' other means that [are] equally effective at no greater cost to other values'. 276 Another influential doctrine is the German Rechtsgut ("legally protected good") theory. 277 On the base of this theory, a state can properly apply criminal law against behaviours that violate a determined Rechtsgut. According to Roxin, the only Rechtsgüter worthy of criminal-law protection are the personal freedoms mentioned in the Constitution. 278 The harm principle, but also the Rechtsgut, acts as justifications for the state's ability to criminalise and punish, as well as marking the scope for the greatest possible expansion of criminal law within liberal states. 279 It should be stressed that such principles serve as a limitation and not as a driver of criminal law.
Another important constraint is that human rights must be respected. 280 Three fundamental principles are put to work: necessity (restrictions to human rights are to be necessary to prevent harm), subsidiarity (less intrusive measures should be assessed as inadequate), and proportionality (criminal measures must be equally and proportionately applied to the actor's wrongdoing). 281 These principles, along with due process rights, are also what give legitimacy to criminal trials and convictions. Indeed, the failure to adhere to such safeguards may contribute to the perception that criminal justice is unfair and arbitrarily exercised. 282 Finally, in a penal system embracing liberal principles, the punishment of an individual can never be a means to pursuing other goals, including the well-being of other persons or the solving of emerging social problems. 283 While a liberal system conceives criminal law as a 'necessary evil' 284 that can be employed only as ultima ratio, the jurisprudence of all human rights bodies increasingly shows an opposite trend towards a conceptualisation of criminal justice as the pivotal remedy for gross human rights violations. In this regard, the liberal model of a criminal law justified only when directed to prevent "harm to others" is moving to the idea that criminal law must always intervene in protection of fundamental rights. Paradoxically, the "harm principle" ends up promoting rather that restraining the application of criminal law. 285 To a limited degree, authoritarian traits are visible throughout the jurisprudence of human rights bodies. Firstly, human rights bodies' prime concern in relation to criminal justice is the efficiency of the penal system authorities to enhance their coercive power, thus making state abuses more likely. Ultimately, the shift toward an offensive function of human rights in criminal justice may be resulting in an overturning within the liberal model of criminal law. While the "harm principle" begins promoting -instead of restraining -the application of criminal law, there is the risk that emerging authoritarian tendencies will be covered by the human rights veil.
What remains to be asked is whether the expansive dynamics of positive obligations in criminal matters is leading to an irreversible overturning of the relationship between human rights and criminal law, or whether the punitive wishes stemming from the jurisprudence of human rights bodies are nothing but further evidence of the fundamental ambiguity that has always characterised criminal law: aimed at protecting the right to security, it culminates in limiting other fundamental rights such as privacy and personal liberty. Several elements support the second scenario. Punitive wishes have always been present in liberal democracies, where the fundamental principles of criminal law have constantly been in tension with other societal aims, including the reduction of crime. A real liberal system of criminal justice has always been nothing more than an "ideal-type" in a Weberian sense. 296 Yet, the situation analysed in this article is new because the threats to the liberal model of criminal law do not come from the usual quarters. They come from human rights bodies that traditionally have a moderating influence on states' punitive policies. Also, they come hidden behind the morality of human rights language. This makes them less visible and more devious than a threat from "law-and-order" politicians or crime control advocates.