Coercive Human Rights: Positive Duties to Mobilise the Criminal Law under the ECHR 9781509937875, 9781509937905, 9781509937899

Citation preview

COERCIVE HUMAN RIGHTS Traditionally, human rights have protected those facing the sharp edge of the criminal justice system. But over time human rights law has become increasingly infused with duties to mobilise criminal law towards protection and redress for the violation of rights. These developments give rise to a whole host of questions concerning the precise parameters of coercive human rights, the rationale(s) that underpin them, and their effects and implications for victims, ­perpetrators, domestic legal systems, and for the theory and practice of human rights and criminal justice. This collection addresses these questions with a focus on the rich jurisprudence of the European Court of Human Rights (ECtHR). It explores four interlocking themes surrounding the issue of coercive human rights: First, the key threads in the doctrine of the ECtHR on duties to mobilise the criminal law as a means of delivering human rights protection. Secondly, the factors that contribute to a readiness to demand coercive ­measures, including discrimination and vulnerability, and other key justificatory reasoning shaping the development of coercive human rights. Thirdly, the most pressing challenges for the ECtHR’s coercive duties doctrine, including: –– how it relates to theories and rationales of criminalisation and criminal punishment; –– its implications for the fundamental tenets of human rights law itself; –– its relationship to transitional justice objectives; and –– how (far) it coheres with the imperative of effective protection for persons in precarious or vulnerable situations. Fourthly, the (prospective) evolution of the coercive human rights doctrine and its application within national jurisdictions. Volume 7 in Hart Studies in Security and Justice

Hart Studies in Security and Justice Series editor: Liora Lazarus The interplay between security and justice has always featured prominently in legal scholarship, but it has taken on a particular urgency since the new Millennium. The new scholarly questions that arise are theoretical, doctrinal and empirical, cutting across a range of traditional sub-disciplines within the legal academy. They address some of the most pressing legal issues of our time, such as the legal status of the ‘the war on terror’, the nature of states of exception, targeted killing, preventive pre-trial detention, mass surveillance and the numerous other threats that security poses to human rights, the rule of law and liberal democracy. The purpose of this series is to engage with security and justice scholarship broadly conceived, and to promote a sophisticated and complex understanding of the important challenges it faces. The series is inclusive, promoting new and established scholars from a range of disciplines. It covers doctrinal, empirical, historical and theoretical work, as well as studies which focus on domestic, comparative and international dimensions of emerging security and justice fields. The series also strives to promote the most inclusive range of politics and methodologies, scrutinizing received wisdom and established paradigmatic approaches, and promoting an intellectual dialogue between its authors and the wider field of law as a whole. Recent titles in this series: Surveillance, Privacy and Trans-Atlantic Relations Edited by David Cole, Federico Fabbrini and Stephen Schulhofer Parliament’s Secret War Veronika Fikfak and Hayley J Hooper Permanent States of Emergency and the Rule of Law Alan Greene The National Security Constitution Paul F Scott The Constitutional Structure of Europe’s Area of ‘Freedom, Security and Justice’ and the Right to Justification Ester Herlin-Karnell Security and Human Rights Benjamin Goold and Liora Lazarus

Coercive Human Rights Positive Duties to Mobilise the Criminal Law under the ECHR

Edited by

Laurens Lavrysen and

Natasa Mavronicola

HART PUBLISHING Bloomsbury Publishing Plc Kemp House, Chawley Park, Cumnor Hill, Oxford, OX2 9PH, UK 1385 Broadway, New York, NY 10018, USA HART PUBLISHING, the Hart/Stag logo, BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc First published in Great Britain 2020 Copyright © The editors and contributors severally 2020 The editors and contributors have asserted their right under the Copyright, Designs and Patents Act 1988 to be identified as Authors of this work. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. While every care has been taken to ensure the accuracy of this work, no responsibility for loss or damage occasioned to any person acting or refraining from action as a result of any statement in it can be accepted by the authors, editors or publishers. All UK Government legislation and other public sector information used in the work is Crown Copyright ©. All House of Lords and House of Commons information used in the work is Parliamentary Copyright ©. This information is reused under the terms of the Open Government Licence v3.0 (http://www.nationalarchives.gov.uk/doc/ open-government-licence/version/3) except where otherwise stated. All Eur-lex material used in the work is © European Union, http://eur-lex.europa.eu/, 1998–2020. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication data Names: Lavrysen, Laurens, editor.  |  Mavronicola, Natasa, editor. Title: Coercive human rights : positive duties to mobilise the criminal law under the ECHR / edited by Laurens Lavrysen and Natasa Mavronicola. Description: Oxford, UK ; New York, NY : Hart Publishing, an imprint of Bloomsbury Publishing, 2020.  |  Series: Hart studies in security and justice; volume 7  |  Includes bibliographical references and index. Identifiers: LCCN 2020027671 (print)  |  LCCN 2020027672 (ebook)  |  ISBN 9781509937875 (hardback)  |  ISBN 9781509937899 (ePDF)  |  ISBN 9781509937882 (Epub) Subjects: LCSH: Human rights.  |  International law and human rights.  |  Convention for the Protection of Human Rights and Fundamental Freedoms (1950 November 5)  |  International criminal law. Classification: LCC K3240 .C5933 2020 (print)  |  LCC K3240 (ebook)  |  DDC 345.24/05—dc23 LC record available at https://lccn.loc.gov/2020027671 LC ebook record available at https://lccn.loc.gov/2020027672 ISBN: HB: 978-1-50993-787-5 ePDF: 978-1-50993-789-9 ePub: 978-1-50993-788-2 Typeset by Compuscript Ltd, Shannon To find out more about our authors and books visit www.hartpublishing.co.uk. Here you will find extracts, author information, details of forthcoming events and the option to sign up for our newsletters.

Foreword

T

his excellent book is the perfect example of the indispensable and fruitful relationships between the case law of the European Court of Human Rights and legal doctrine, and between the judges and the scholars. From both sides, we are fulfilling our social role. Applying the European Convention on the basis of individual applications, the judges have to decide on a case-by-case basis, whereas the scholars have and should have a general overview of the judgments of the Court in order to put them into context and to scrutinise their coherence and potential as well as their limits and shortcomings. Reliable legal commentary draws our attention to evolutions or discrepancies that are not always noticeable with the same acuity when we are at the core of a case. That is the reason why scholars and judges are indispensable allies for ‘the maintenance and further realization of human rights and fundamental freedoms’ (Preamble of the Convention). In this volume, which tackles a very difficult and complex topic, the scientific quality of the various reports is outstanding, even remarkable. They offer a substantial and unique contribution towards clarifying what is at stake in the growing development on what is now called ‘coercive human rights doctrine’. The Court’s judgment MC v Bulgaria1 of 4 December 2003 is highly significant since it did not give weight – or at least was not sensitive – to the idea that recourse to criminal law is not necessarily the only answer, that criminal proceedings should remain a last resort or subsidiary remedy and that their use calls for a certain degree of restraint. This concern about a sort of evident need for criminal law triggered the need for an analysis.2 Over time, the ‘discomfort’ affecting the relationship between criminal law and human rights, under both its offensive and defensive face, which is very well documented by the authors of the different contributions, became obvious. In this respect, the book is very well built, as the sociologists like to say. Indeed, it is really illuminating and also useful to have a general panorama, a complete view from a far distance covering a wide area of issues. Across the 13 chapters, we are provided with serious tools for a clear understanding of the criminalisation of human rights: a full picture of the history of this trend which finds its main source in the positive obligations and horizontal applications of the Convention, alongside the international

1 MC v Bulgaria ECHR 2003-XII. 2 F Tulkens and M van de Kerchove, ‘Les droits de l’homme: bonne ou mauvaise conscience du droit pénal ?’ in F Verbruggen, R Verstraeten, D van Daele and B Spriet (eds), Strafrecht als roeping. Liber amicorum Lieven Dupont (Leuven, Universitaire Pers Leuven, 2005) 949 ff.

vi  Foreword criminal law context; a comprehensive presentation of the case law of the Court (including migrant labour case law), with a discussion of the elements adduced to explain/justify this development or to run counter to it; and a concrete analysis of the implications and repercussions of this evolution. As a matter of fact, it transpires from many contributions in this book that all the main core elements of both criminal law and human rights are at stake in ‘coercive human rights’ and give rise to fundamental questions which are analysed in depth. From the criminal law side, the deterrence argument seems to be among the most important ones, in the pure tradition of the utilitarian penal t­heories of Beccaria and Bentham. The KU v Finland judgment of 2 December 2008 is a perfect illustration of this.3 As to the assumption by the Court that c­ riminal remedies are, in any event, the most effective in terms of deterrence, the Report on Decriminalisation of the Council of Europe clearly shows that the ­effectiveness of general deterrence based on criminal law depends on various factors and that such an approach ‘is not the only way of preventing undesirable behaviour’.4 If deterrence could be seen as a common-sense basis for criminalisation, it should also be recalled that worldwide substantial empirical research has shown its weak and dubious effect, except for people who do not need it.5 Under the umbrella of Durkheim’s work, the need to reinforce the symbolic function of criminal law is also invoked by some in support of the use of it to enforce human rights, especially when it comes to Articles 2 and 3 of the Convention. For others, this approach disregards the wrongdoings of the enforcement of criminal law. Symbolism versus instrumentalism? In the same vein, faced with gross human rights violations, some consider retributive responses as reparations. From the human rights side, the protection of the victims and the effectiveness of human rights are supposed to be the main reasons to choose the recourse to criminal law. As observed by many authors, this emphasis put on the offensive role of human rights through the intervention of criminal law, whether we want it or not, is accompanied by a weakening of its defensive role at the national and European levels in terms, for instance, of legality, proportionality and subsidiarity, which represent real threats to rights and liberties. The Grand Chamber judgment of Austin and Others v UK6 of 15 March 2012 is an example of this. So what does effectiveness mean? Moreover, pursuing along this avenue, the state

3 KU v Finland ECHR 2008, para 46: ‘States have a positive obligation inherent in Article 8 of the Convention to criminalise offences against the person including attempts and to reinforce the deterrent effect of criminalisation by applying criminal-law provisions in practice through effective investigation and prosecution.’ 4 European Committee on Crime Problems, Report on Decriminalisation (Strasbourg, Council of Europe, 1980) 75–78. 5 F Tulkens and M van de Kerchove, Introduction au droit pénal. Aspects juridiques et ­criminologiques, 10th edn (Waterloo, Kluwer, 2014) 609 ff. 6 Austin and Others v UK ECHR 2012.

Foreword  vii is now accountable not for having used criminal law, but for not having used it or having used it with too much leniency,7 which is a reversal of p ­ erspective. In this respect, among others, the concept of impunity should be revisited, especially in the context of transitional justice, as is well explained in the very interesting contribution devoted to this topic. Finally, it is greatly welcome that the book has devoted analysis to the cascade implications of coercive human rights not only within the Court’s case law itself, but also in the outside world, ie, in the national criminal law systems, for instance, on a legislative level (the Modern Slavery Act 2015 in the UK) or on a procedural level (the dubious treatment of unlawfully obtained evidence). In particular, I fully agree with the observation that human rights institutions and judges are sometimes acting with insufficient attention to the potential ‘pitfalls’ of their decisions on the life of people all over Europe. To conclude, let me quote H Packer in his influential book of 1968, The Limits of the Criminal Sanction: The criminal sanction is at once prime guarantor and prime threatener of human freedom. Used providently and humanely it is guarantor; used indiscriminately and coercively, it is threatening. The tensions inherent in the criminal sanction can never be wholly resolved in favour of guaranty and against threat.8

Personally, at the risk of going against the mainstream, I am reluctant to accept a ‘culture of conviction’ in the name of human rights. I continue to be convinced that human rights are the last rampart against any abuse of power and that this unique role must be preserved at any cost in a democratic society. In any case, this seminal book of great intellectual honesty is certainly an invitation, as proposed by its co-editors, ‘to treat coercive human rights as an object of continued critical scrutiny’.9 Françoise Tulkens Former Judge at the European Court of Human Rights and Emeritus Professor at UCLouvain 1 May 2020

7 F Tulkens and S van Drooghenbroek, ‘La clémence pénale et les droits de l’homme. Réflexions en marge de la jurisprudence récente de la Cour européenne des droits de l’homme’ in F Tulkens, Y Cartuyvels and C Guillain (eds), La peine dans tous ses états. Hommage à Michel van de Kerchove (Brussels, Larcier, 2011) 125 ff. 8 HL Packer, The Limits of the Criminal Sanction (Stanford, Stanford University Press 1968) 366. 9 See Natasa Mavronicola and Laurens Lavrysen, ch 1 in this volume.

viii

Contents Foreword�������������������������������������������������������������������������������������������������������v Françoise Tulkens Notes on Contributors��������������������������������������������������������������������������������� xi 1. Coercive Human Rights: Introducing the Sharp Edge of the European Convention on Human Rights�������������������������������1 Natasa Mavronicola and Laurens Lavrysen PART I KEY THREADS IN ECtHR DOCTRINE 2. Positive Obligations and the Criminal Law: A Bird’s-Eye View on the Case Law of the European Court of Human Rights���������������������29 Laurens Lavrysen 3. Positive Obligations and Coercion: Deterrence as a Key Factor in the European Court of Human Rights’ Case Law�������������������������������55 Paul Lemmens and Marie Courtoy PART II PERSPECTIVES ON VICTIMS’ PROTECTION AND REDRESS 4. Retribution through Reparations? Evaluating the European Court of Human Rights’ Jurisprudence on Gross Human Rights Violations from a Victim’s Perspective�����������������������������������������������������71 Alina Balta 5. Shaping Coercive Obligations through Vulnerability: The Example of the ECtHR�������������������������������������������������������������������93 Corina Heri 6. Criminal Law Responses to Hate Speech: Towards a Systematic Approach in Strasbourg?���������������������������������������������������������������������� 117 Stephanos Stavros

x  Contents PART III CRITICAL REFLECTIONS: THEORY, IMPACT, LIMITATIONS 7. Positive Obligations in View of the Principle of Criminal Law as a Last Resort����������������������������������������������������������������������������������� 141 Nina Peršak 8. Sowing a ‘Culture of Conviction’: What Shall Domestic Criminal Justice Systems Reap from Coercive Human Rights?���������������������������� 161 Mattia Pinto 9. Coercive Overreach, Dilution and Diversion: Potential Dangers of Aligning Human Rights Protection with Criminal Law (Enforcement)�������������������������������������������������������������������������������������� 183 Natasa Mavronicola 10. Separating Protection from the Exigencies of the Criminal Law: Achievements and Challenges under Article 4 ECHR���������������������������� 203 Vladislava Stoyanova 11. The Limitations of a Criminal Law Approach in a Transitional Justice Context������������������������������������������������������������������������������������ 223 Brice Dickson PART IV UNCHARTED WATERS FOR THE ECtHR’S COERCIVE DUTIES DOCTRINE 12. Preventive Obligations, Risk and Coercive Overreach��������������������������� 249 Liora Lazarus 13. Coercive Human Rights and Unlawfully Obtained Evidence in Domestic Criminal Proceedings�������������������������������������������������������� 267 Kelly M Pitcher Postscript: Coercive Human Rights in Times of Coronavirus���������������������� 293 Natasa Mavronicola and Laurens Lavrysen Index��������������������������������������������������������������������������������������������������������� 297

Notes on Contributors Alina Balta is a PhD researcher with Tilburg Law School, Tilburg University. She has a background in law and victimology, having obtained an LLM in International and Human Rights Law and an MSc in Victimology and Criminal Justice. She currently specialises in reparations and reparative justice for victims of international crimes and gross human rights violations. To this end, her PhD project is concerned with reparations in the context of the European Court of Human Rights, the Inter-American Court of Human Rights, the International Criminal Court and the Extraordinary Chambers in the Courts of Cambodia, evaluating their contribution to reparative justice for victims. Marie Courtoy completed a bachelor’s degree in sociology and anthropology (Université Saint-Louis – Bruxelles) before embarking on a second bachelor’s degree (Université Saint-Louis – Bruxelles) and a master’s degree (Université catholique de Louvain) in law. Since graduating, she has acquired practical experience in the field of human rights and migration. She completed internships at the European Court of Human Rights and at the United Nations High Commissioner for Refugees in Brussels. However, she was already active in the voluntary sector within various associations and non-governmental organisations (NGOs). She is a member of the European Rights and Migration Team, a research team that belongs to the Charles de Visscher Center for International and European Law at UCLouvain. She is completing a PhD under a FRESH grant there, which focuses on environmental migration in the broader context of environmental justice. She combines law and anthropology, and her research takes place in collaboration with the Max Planck Institute for Social Anthropology. Brice Dickson is Emeritus Professor at Queen’s University Belfast. He has written extensively on international human rights law, comparative law and the development of human rights standards by national supreme courts, as well as on the interplay between human rights and conflicts and transitions. He is a specialist in the European Convention on Human Rights, on which he edited a book in 1997 (Human Rights and the European Convention, Sweet & Maxwell). His monograph entitled The European Convention on Human Rights and the Conflict in Northern Ireland (Oxford University Press, 2010) is the leading text on that topic. He has written several other leading texts in human rights and public law, including Human Rights and the United Kingdom Supreme Court (Oxford University Press, 2013), and has published seminal ­articles in major peer-reviewed journals. He served as the first Chief Commissioner of the Northern Ireland Human Rights Commission (1999–2005).

xii  Notes on Contributors Corina Heri is a postdoctoral researcher at the Amsterdam Center for International Law at the University of Amsterdam. Her postdoctoral research, which is supported by the Jaarsma-Adolfs Fund, focuses on the proposal of a human right to land and the emergence of new human rights claims more ­generally. She completed her PhD research at the University of Zurich in 2017. Her PhD research, which concerned the concept of vulnerability in the Article 3 case law of the European Court of Human Rights, was funded by the Swiss National Science Foundation and the Janggen-Pöhn Foundation. Laurens Lavrysen is a postdoctoral researcher (funded by the FWO – Research Foundation Flanders) connected to the Human Rights Centre of Ghent University. He holds a PhD in law from Ghent University (2016). He is the author of Human Rights in a Positive State: Rethinking the Relationship between Positive and Negative Obligations under the European Convention on Human Rights (Intersentia, 2016). He has written on various topics in the area of ECHR law, including on the positive obligations doctrine, the principle of proportionality, the notion of procedural justice, and poverty as a human rights issue. He also has a research interest in the area of the history of the ECHR, in particular in Belgium. He has contributed to various third-party interventions submitted to the European Court of Human Rights by the Ghent University Human Rights Centre. He is also the co-editor of the Strasbourg Observers blog and editor-in-chief of the Flemish Journal for Human Rights (Tijdschrift voor Mensenrechten). Liora Lazarus, BA (UCT), LLB (LSE), DPhil (Oxon), is an Associate Professor in Law at the University of Oxford and a Fellow of St Anne’s College, Oxford. Her primary research interests are in comparative human rights, security and human rights, comparative theory and comparative criminal justice. After publishing Contrasting Prisoners’ Rights (Oxford University Press, 2004), she widened her research field and co-edited Security and Human Rights (1st edn, Hart Publishing, 2007; 2nd edn, Hart Publishing, 2019) and Reasoning Rights (Hart Publishing, 2014) alongside a range of articles and chapters. She has a particular interest in the right to security and coercive duties, and is also now exploring the relationship between security and the rule of law. Throughout her career, she has sought to balance academic research with public interest work. She has completed two public reports for the UK Ministry of Justice on balancing public protection with human rights, and on the relationship between rights and responsibilities. She also produced a major report, with her Oxford colleagues, on the evolution of fundamental rights for the European Parliament. She has acted as an advisor to the UK Stern Review on Rape Complaints (2013), was asked to give evidence to the Parliamentary Modern Slavery Bill Committee (2014) and was part of an International Committee of Experts to the Basque Working Group on Treatment of Political Detainees (2014). Paul Lemmens has been the Belgian Judge at the European Court of Human Rights in Strasbourg since 2012. He also is a professor at KU Leuven. He studied

Notes on Contributors  xiii law at the University of Antwerp and at KU Leuven, and obtained his law degree in 1976 (magna cum laude). He subsequently obtained an additional master’s degree in law at Northwestern University, Chicago (1978). In 1987, he obtained his PhD at KU Leuven on Article 6(1) of the European Convention. He began his academic career as a research assistant in constitutional law (1976–77) and civil procedure (1979–86). He was subsequently appointed professor (1986) and in this capacity has taught international human rights law (since 1986) and also civil procedure (1986–95), administrative procedure (1995–97) and constitutional law (1997–2008). He was also a judge at the Belgian Council of State (Supreme Administrative Court) (1994–2012) and was a member of the UNMIK Human Rights Advisory Panel in Kosovo (2007–12). Natasa Mavronicola is Reader in Law at Birmingham Law School, University of Birmingham. She holds a PhD in law from the University of Cambridge (entitled ‘Delimiting the Absolute: The Nature and Scope of Article 3 of the European Convention on Human Rights’), a BCL from the University of Oxford and an LLB (Hons) from University College London. Her research interests are in human rights law, with a particular focus on the right not to be subjected to torture or related ill-treatment and the right to life, intersections between human rights and criminal justice, and particular features of supranational human rights interpretation. She has published work on human rights law in journals such as the Human Rights Law Review and the Modern Law Review. In 2017–19, she served as Thematic Advisor to Professor Nils Melzer, the United Nations Special Rapporteur on Torture, and she has previously conducted research for the Council of Europe. Nina Peršak holds a doctorate in law from University of Ljubljana and an LLM and MPhil in social and developmental psychology, both from the University of Cambridge. She is a Visiting Full Professor at the Faculty of Law, University of Ljubljana, teaching comparative human rights law, and is a Scientific Director of the Institute of Criminal Law Ethics and Criminology in Ljubljana. She also holds a law professorship habilitation from the University of Maribor. Previously, she has worked as a research professor at the Faculty of Law of Ghent University in Belgium, as a senior fellow of the Institute for Advanced Study at CEU in Budapest, and held several research and professorial posts in Ljubljana in Slovenia and visiting scholarships at Cambridge, Helsinki, Sofia and UC Berkeley. Her main research interests lie in the areas of criminal law and criminalisation, human rights, the rule of law, victimology, social control, sociolegal studies and social psychology. Her publications include Criminalising Harmful Conduct (Springer-Verlag, 2007), Legitimacy and Trust in Criminal Law, Policy and Justice (Routledge, 2014) and Regulation and Social Control of Incivilities (Routledge, 2016). She serves on editorial boards of several international journals and is a member of the European Commission’s Expert Group on EU Criminal Policy.

xiv  Notes on Contributors Kelly M Pitcher is Assistant Professor of Criminal Law and Criminal Procedure at Leiden University. She holds an LLB in English law from the University of Kent and an LLM in criminal justice and human rights from the University of Aberdeen. Following her studies, she worked with counsel representing accused at the United Nations International Criminal Tribunal for the former Yugoslavia, thereafter studying Dutch law at the University of Amsterdam. From 2009 to 2015, she worked at the University of Amsterdam, where, in addition to undertaking doctoral research, she taught subjects in Dutch criminal law and criminal procedure. In September 2016, she defended her doctoral thesis entitled ‘Judicial Responses to Pre-trial Procedural Violations in International Criminal Proceedings’, which has since been published (TMC Asser Press/Springer, 2018). Her research interests lie in the fields of criminal procedure and criminal evidence, with an emphasis on comparative law, human rights and theoretical perspectives. Mattia Pinto is a PhD candidate in the Law Department at the London School of Economics and Political Science (LSE), where he also teaches criminal law and legal research and writing skills. His PhD research analyses the role that human rights play in fostering and legitimising penal expansion. He holds degrees from the University of Bologna (Five-Year Master of Law) and King’s College London (LLM). Prior to joining LSE, he worked as a legal intern at the European Court of Human Rights (Registry) and the International Criminal Court (Office of Public Counsel for the Defence), and as a research assistant in international criminal law at King’s College London. Stephanos Stavros is a human rights lawyer with extensive experience in international courts and independent monitoring bodies. His academic interests range from criminal law to freedom of religion and expression, and minority and migrants’ rights. He has worked for the Council of Europe (the European Commission and Court of Human Rights, the legal department and central administration of the Council of Europe and the European Commission against Racism and Intolerance (ECRI)), the Greek Ministry of Foreign Affairs (dealing with EU legal issues) and the Greek Council for Refugees. He holds a PhD in law from the University of London and an MBA from Henley Management College. Vladislava Stoyanova is Associate Professor at the Faculty of Law, Lund University, and a Wallenberg Academy Fellow (2019–24). Her research interests are in the area of human rights, migration and refugee law. She is the author of Human Trafficking and Slavery Considered: Conceptual Limits and States’ Positive Obligations in European Law (Cambridge University Press, 2017). She is also the co-editor of Seeking Asylum in the EU: Selected Protection Issues Raised by the Second Phase of the Common European Asylum System (Martinus Nijhoff, 2015), The New Asylum Countries in South-East Europe: Realities and Legal Challenges (Martinus Nijhoff, 2018) and International Law and Violence against Women: Europe and the Istanbul Convention (Routledge, 2020).

1 Coercive Human Rights Introducing the Sharp Edge of the European Convention on Human Rights NATASA MAVRONICOLA AND LAURENS LAVRYSEN

I. INTRODUCTION

I

n the context of the criminal justice system, human rights law has ­traditionally been associated with the protection of the rights of suspects and defendants against the exercise of coercive and carceral state power.1 Under the European Convention on Human Rights (ECHR), criminal investigations must comply with human rights, including the right not to be subjected to torture or ill-treatment (Article  3) and the right to respect for private life (Article  8), safeguards must be in place against arbitrary deprivations of liberty (Article  5) and criminal trials must comply with the right to a fair trial (Article  6), encompassing, inter alia, the presumption of innocence and the right to legal assistance. In addition to this function, which can be referred to as the ‘shield’ function of human rights in the application of criminal law, human rights law has increasingly also acquired a ‘sword’ function.2 The European Court of Human Rights (ECtHR) has developed coercive duties,3 requiring 1 See, for example, L Doswald-Beck, ‘Fair Trial, Right to, International Protection’ in Max Planck Encyclopedia of Public International Law, 2nd edn (Oxford, Oxford University Press, 2012); N Rodley and M Pollard, The Treatment of Prisoners under International Law, 3rd edn (Oxford, Oxford University Press, 2009). On this dimension of the ECHR, see R Goss, Criminal Fair Trial Rights: Article 6 of the European Convention on Human Rights (Oxford, Hart Publishing, 2014); S Snacken, ‘Resisting Punitiveness in Europe?’ (2010) 14 Theoretical Criminology 273. 2 F Tulkens, ‘The Paradoxical Relationship between Criminal Law and Human Rights’ (2011) 9 Journal of International Criminal Justice 577. 3 L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce?’ in L Zedner and J Roberts (eds), Principles and Values in Criminal Law and Criminal Justice (Oxford, Oxford

2  Natasa Mavronicola and Laurens Lavrysen states to mobilise the criminal law in order to protect against or provide redress for human rights violations. In particular, the ECtHR has established that states are under obligations to criminalise certain human rights violations, such as human trafficking, torture and rape, and may be under an obligation to criminally investigate human rights violations, to prosecute offenders and to impose criminal sanctions.4 These strands of case law have now crystallised into what could properly be described as the ECtHR’s coercive human rights doctrine, which is the focus of the present collection. This evolution raises a whole range of challenges that go to the heart of the purpose and function of human rights law. In a landmark article on the subject, former ECtHR Judge Françoise Tulkens has described the relationship between the ‘shield’ and ‘sword’ functions of human rights law in the criminal justice system as a paradox that one must come to terms with.5 The ECtHR’s emerging coercive human rights doctrine gives rise to a range of questions, such as: how to properly balance the ‘shield’ and ‘sword’ functions, and how to address potential tensions between the human rights of (potential) victims and those of suspects and defendants; what may be the dangers (if any) of mobilising the criminal law through coercive human rights doctrine; and what are the limitations (if any) of a criminal law approach from the viewpoint of (prospective) victims’ human rights and broader interests. It also raises the question as to whether it is feasible, desirable and institutionally appropriate for the ECtHR in particular to make such demands – sometimes referred to as the exercise of ‘quasi-criminal jurisdiction’6 – and, if so, to what extent. While similar questions have already been the subject of extensive debate in the context of other jurisdictions,7 the ECtHR’s coercive human rights doctrine has only relatively recently attracted growing academic attention.8 This can be explained by the fact that the development of the doctrine is itself a relatively recent phenomenon.9 Two monographs, both adopting a doctrinal approach, have been written on the subject by Anja Seibert-Fohr (2009) and Krešimir

University Press, 2012) 136. Lazarus uses the notion of ‘coercive duties’ broadly, as she also refers to duties to take preventive operational measures to protect against human rights violations; see, eg, Osman v UK ECHR 1998-VIII. 4 For an overview, see Laurens Lavrysen, ch 2 in this volume. 5 Tulkens (n 2). 6 A Huneeus, ‘International Criminal Law by Other Means: The Quasi-criminal Jurisdiction of the Human Rights Courts’ (2013) 107 American Journal of International Law 1. 7 On the international context, see section II.B below. 8 In addition to the scholarship in English, see also, eg, in French, Y Cartuyvels, ‘Les droits de l’homme, frein ou amplificateur de criminalisation?’ in H Dumont, F Ost and S van Drooghenbroeck (eds), La responsabilité, face cachée des droits de l’homme (Brussels, Bruylant, 2005); Y Cartuyvels, H Dumont, F Ost, M van de Kerchove and S van Drooghenbroeck (eds), Les Droits de l’Homme, Bouclier ou Épée du Droit Pénal? (Brussels, Presses de l’Université Saint-Louis, 2007); and M Delmas-Marty, ‘Le paradoxe pénal’ in M Delmas-Marty and C Lucas de Leyssac (eds), L ­ ibertés et droits fondamentaux (Paris, Seuil, 1996). See also, in Dutch, P van Kempen, Repressie door ­Mensenrechten (Nijmegen, Wolf Legal Publishers, 2008). 9 On the historical context, see section II.A below.

Coercive Human Rights  3 Kamber (2017). In her study, Seibert-Fohr explores the most prominent dimensions of the obligation to prosecute serious human rights violations under international law, focusing on different branches of international law, including the International Covenant on Civil and Political Rights (ICCPR), the American Convention on Human Rights (ACHR), customary international law, as well as the ECHR.10 Being primarily concerned with examining the extent to which international human rights law in its coercive dimension is able to complement international criminal law, her doctrinal analysis takes stock of the state of international law in the area. Kamber, in turn, provides an overview of the ECtHR’s coercive human rights doctrine and focuses on assessing how the ‘sword’ and ‘shield’ functions of human rights law ought to be reconciled in the ECtHR’s legal reasoning.11 In addition to these monographs, in recent years, the ECtHR’s coercive human rights doctrine has also been the subject of a number of articles that have begun to address some of the more critical questions set out above.12 There has been a particular concern regarding the risk that the coercive human rights doctrine could lead to what Liora Lazarus has labelled ‘coercive overreach’.13 Such risk would materialise if coercive duties were developed ‘in a way which demands the penalisation of acts or omissions which might, as a matter of principle or policy, not necessarily warrant penal sanction’.14 Authors have warned that coercive human rights may lead courts to downplay their traditional commitment to due process rights of defendants15 or that they may even foster a ‘culture of conviction’.16 In addition, there has been a preoccupation regarding the extra-legal implications of coercive human rights doctrine, in particular the risk that it may permeate political rhetoric orientated at legitimising the expansion of coercive measures,17 or that it may 10 A Seibert-Fohr, Prosecuting Serious Human Rights Violations (Oxford, Oxford University Press, 2009). 11 K Kamber, Prosecuting Human Rights Offences (Leiden, Brill, 2017). 12 In addition to doctrinal scholarship that has either a general scope or that focuses on coercive human rights under a specific Convention provision. For an example of the former, see Huneeus (n 6), adding a comparative dimension; and A Ashworth, Positive Obligations in Criminal Law (Oxford, Hart Publishing, 2013), focusing on the ECHR exclusively in ch 8. For an example of the latter, see V Stoyanova, ‘Article  4 of the ECHR and the Obligation of Criminalising Slavery, Servitude, Forced Labour and Human Trafficking’ (2014) 3 Cambridge Journal of International and Comparative Law 407. For a recent monograph (which emerged as the book went to print) that attempts to bridge international human rights law and criminal theory, see S Malby, Criminal Theory and International Human Rights Law (Abingdon, Routledge, 2019). 13 Lazarus (n 3). 14 N Mavronicola, ‘Taking Life and Liberty Seriously: Reconsidering Criminal Liability under Article 2 of the ECHR’ (2017) 80 Modern Law Review 1026, 1037. 15 M Sorochinsky, ‘Prosecuting Torturers, Protecting “Child Molesters”: Towards a Power Balance Model of Criminal Process for International Human Rights Law’ (2009) 31 Michigan Journal of International Law 157, 228. 16 M Pinto, ‘Awakening the Leviathan through Human Rights Law: How Human Rights Bodies Trigger the Application of Criminal Law’ (2018) 34 Utrecht Journal of International and European Law 161, 182. 17 Lazarus (n 3) 149 and 151.

4  Natasa Mavronicola and Laurens Lavrysen serve to cover authoritarian tendencies with a human rights veil.18 On the other hand, there has also been a concern that the ECtHR’s coercive human rights doctrine may result in the conflation of standards to determine state liability for human rights violations and the narrower parameters of wrongs attracting individual criminal liability, which may ultimately lead to the dilution of the substance of ECHR rights.19 The present volume tackles the ECtHR’s coercive human rights doctrine in a concerted manner through a selection of expert analyses of the content, promises and pitfalls of the doctrine. It combines theoretically informed reflections, from a variety of perspectives, with rigorous analysis and critical assessment of the doctrine and its operation in national jurisdictions. The ultimate aim of the collection is to allow for a meaningful re-assessment of the doctrine and of the supportive and critical discourses surrounding it. This volume has been developed out of a seminar on ‘Positive Obligations under the ECHR and the Criminal Law: Towards a Coercive Human Rights Law?’, which took place on 25 May 2018 at Ghent University, Belgium. Afterwards, further papers were sought to complement those presented at the seminar, with a view to securing the coherent exploration of key issues arising out of the ECtHR’s coercive human rights doctrine. This chapter first discusses the development of the ECtHR’s coercive human rights doctrine in context, adopting a historical and international perspective (section II). Next, it introduces the different chapters (section III) and identifies the central themes that emerge from this collection (section IV), both in terms of the critical approaches adopted and the suggestions made to re-orient coercive human rights. Finally, it proposes directions for future research in the area (section V). II.  COERCIVE HUMAN RIGHTS IN CONTEXT

The debate surrounding coercive human rights under the ECHR did not arise overnight, nor does it take place in isolation from similar discussions regarding coercive human rights in other jurisdictions. To better grasp the significance of the debate(s) taking place within the present volume, it is therefore necessary to place the ECtHR’s coercive human rights doctrine in its (a) historical and (b) international context. A.  Historical Context: The Legacy of X and Y v The Netherlands The text of the ECHR, which was adopted in 1950, makes no mention of obligations to mobilise the criminal law with a view to protecting against or

18 Pinto

(n 16) 183. (n 14).

19 Mavronicola

Coercive Human Rights  5 providing redress for human rights violations. The development of coercive human rights has thus exclusively taken place via the ECtHR’s jurisprudence, starting from the X and Y v The Netherlands case of 1985. This case concerned the failure to hold the perpetrator of the rape of a girl with a mental disability criminally liable. Her father had lodged a criminal complaint, which had been dismissed because Dutch law required such a complaint to be lodged by the victim herself, which had been impossible since the law at the same time rendered her legally incapable of doing so based on her disability. According to the Court, this amounted to a violation of Article 8 (the right to respect for private life).20 The significance of the case must be considered against the background of the academic debate that had raged since the early days of the Convention system on the relevance of the ECHR in the context of horizontal relations between individuals. While only states are formally party to the Convention and applications before the Convention system can only formally be directed against states,21 a number of proposals were made in the literature to defend the extension of the scope of the ECHR to horizontal relations. A first proposal to purposively interpret the Convention to also impose obligations on individuals was notably defended by Marc-André Eissen, Deputy Registrar and later Registrar of the European Commission on Human Rights, from the early 1960s onwards. Eissen argued that it does not necessarily follow from the aforementioned formal limitations that individuals do not have obligations under the Convention, which may be binding at the domestic level.22 A second proposal, discussed by Andrew Drzemczewski in an influential article from 1979, was to draw inspiration from the doctrine of Drittwirkung or third-party effect, which had been developed within the context of German constitutional law theory and which requires courts to apply fundamental rights to horizontal relations.23 In the context of the ECHR, Drzemczewski, unlike Eissen, considered that the Convention can only create ‘indirect obligations for individuals in that it may oblige the legislature or the courts to protect individuals from one another’.24 Notably, he already contemplated that legislative measures could be required to safeguard ECHR rights in horizontal relations, referring explicitly to criminal law, alongside civil and administrative law.25 The route taken by the Court in X and Y in substance boils down to Drzemczewski’s narrow version of the Drittwirkung doctrine. Instead of using 20 X and Y v The Netherlands Series A No 91 (1985). 21 Article 25 of the original text of the ECHR, now art 34 ECHR. 22 eg, M-A Eissen, ‘The European Convention on Human Rights and the Duties of the Individual’ (1962) 32 Nordisk Tidsskrift for International Ret 230. 23 A Drzemczewski, ‘The European Human Rights Convention and Relations between Private Parties’, (1979) 26 Netherlands International Law Review 163. See also, eg, E Alkema, ‘The Third-Party Applicability or “Drittwirkung” of the European Convention on Human Rights’ in F Matscher and H Petzold (eds), Protecting Human Rights: The European Dimension (Cologne, Carl Heymanns Verlag KG, 1988). 24 Drzemczewski (n 23) 176. 25 ibid 179.

6  Natasa Mavronicola and Laurens Lavrysen the notion of Drittwirkung, the Court chose to build further on the concept of positive obligations, which requires states to take action to ensure human rights rather than merely abstaining from interfering with them.26 This concept achieved a breakthrough in the 1979 Marckx case,27 in which the Court found, among other things, that Belgium had violated its positive obligation under the right to respect for family life (Article 8 ECHR) because its family law did not ensure that the legal bond between a mother and a child born out of wedlock was established from the mere fact of birth.28 Building upon the Marckx case, in X and Y, the Court held that positive obligations ‘may involve the adoption of measures designed to secure respect for private life even in the sphere of the relations of individuals between themselves’ (emphasis added).29 The Court then found that the case concerned ‘fundamental values and essential aspects of private life’, requiring ‘effective deterrence’, which ‘can be achieved only by criminal-law provisions’.30 Ever since X and Y, the Court has continued to develop coercive human rights obligations within the framework of the positive obligations doctrine.31 At the time, the Court’s development of positive obligations was criticised from an originalist perspective on the basis that it could be assumed that the drafters had not contemplated the possibility that the Convention might imply such obligations.32 While, at present, ‘the idea that human rights can be adequately protected if states content themselves with standing by and doing nothing has become patently absurd’,33 discussion continues on the exact scope of positive obligations.34 As the present volume highlights, coercive human rights is one of the areas of contention in this regard, which exemplifies the tension between human rights law on the one hand constraining state power through negative obligations, and simultaneously mandating and legitimizing the exercise of state power through positive obligations on the other hand – a tension that is, for

26 D Harris, M O’Boyle, E Bates and C Buckley, Law of the European Convention on Human Rights, 3rd edn (Oxford, Oxford University Press, 2014) 23. 27 The Court had already mentioned this concept in passing in the Belgian Linguistics Case Series A No 5 (1968), para 4. 28 Marckx v Belgium Series A No 31 (1979), paras 36–37. In addition, the Court also found a number of violations of art 14 (the prohibition of discrimination) in conjunction with art 8. 29 X and Y v The Netherlands (n 20) para 23. 30 ibid para 27. 31 See Lavrysen, ch 2 in this volume. 32 Most famously voiced by Judge Fitzmaurice in his dissenting opinion to the Marckx judgment. See also P van Dijk, ‘“Positive Obligations” Implied in the European Convention on Human Rights: Are the States Still the “Masters” of the Convention?’ in M Castermans, F van Hoof and J Smith (eds), The Role of the Nation-State in the 21st Century (The Hague, Kluwer Law International, 1998) 18. 33 B Dickson, ‘Positive Obligations and the European Convention on Human Rights’ (2010) 61 Northern Ireland Legal Quarterly 203, 204. 34 eg, M Pitkänen, ‘Fair and Balanced Positive Obligations: Do They Exist?’ (2012) 5 European Human Rights Law Review 538.

Coercive Human Rights  7 instance, equally present in the context of policing35 or the taking of children into care.36 It took some time before the seeds of the coercive human rights doctrine that were planted in X and Y began to grow, with no further case law developments until the late 1990s.37 During this period, most commentators merely took note of the X and Y case, without further probing into the potential farreaching implications thereof.38 In an important monograph on human rights in the private sphere, Andrew Clapham did highlight that, by broadening the scope of positive obligations to horizontal relations, the judgment could have implications for the protection of marginalised and stigmatised groups more generally, but without paying specific attention to the instrumentalisation of the criminal law for these purposes.39 An exception in being more critical are Harris, O’Boyle and Warbrick, who in the first edition of their standard work on the ECHR, discussing the X and Y case, noted that it was ‘unusual, of course, for the Court to be so specific about the content of a state’s obligation’, but considered that the procedural rather than the substantive nature of the gap in the law ‘might militate against too wide an obligation being imposed upon states to criminalise private activities’.40 While the potential of expanding coercive human rights at the level of the ECtHR remained untapped until the late 1990s, by then the context had significantly changed, becoming more conducive towards subsequent case law developments, which in themselves will not be dealt with extensively here, as they are discussed throughout this volume. First, international human rights law as a whole had become more receptive to coercive human rights – the case law of the Inter-American Court of Human Rights, which from the late 1980s onwards was confronted with the aftermath of gross human rights violations that had taken place in the context of Latin American dictatorships, being an important catalyst in this regard.41 Second, from the late 1990s onwards, the ECtHR itself had to deal with increasingly serious cases of human rights violations, in particular

35 eg, Osman v UK (n 3) para 116, in which the Court itself acknowledges the need to discharge the positive obligation to take operational measures to protect the right to life under art 2 in a manner that is compatible with the requirements of art 5 (the right to liberty) and art 8 (the right to respect for private life). 36 While such measures need to be justified in the light of the right to respect for family life under art  8 (eg, in K and T v Finland ECHR 2001-VII), in certain circumstances they may be required under the state’s positive obligations under art 3 to protect children from serious neglect and abuse (eg, in E and Others v UK App No 33218/96 (ECtHR, 26 November 2002)). 37 See Lavrysen, ch 2 in this volume. 38 eg, JG Merrils, The Development of International Law by the European Court of Human Rights, 2nd edn (Manchester, Manchester University Press, 1993) 104–05; F Jacobs and R White, The European Convention on Human Rights, 2nd edn (Oxford, Clarendon Press, 1996) 174. 39 A Clapham, Human Rights in the Private Sphere (Oxford, Clarendon Press, 1993) 213–14. 40 D Harris, M O’Boyle and C Warbrick, Law of the European Convention on Human Rights (London, Butterworths, 1995) 323. 41 See section II.B below.

8  Natasa Mavronicola and Laurens Lavrysen those resulting from the Kurdish and Chechen conflicts.42 These cases formed a fertile ground for the Court to further develop its positive obligations doctrine, in particular by holding states to be under an obligation to investigate serious human rights violations, whether they are committed by state agents or by private individuals.43 It was within this framework of investigative obligations that the Court later recognised that the state is under an obligation to criminally investigate, identify and, if appropriate, prosecute and punish perpetrators of human rights violations.44 Third, the feminist critique of the public/private divide, criticising human rights law for being insufficiently responsive to human rights violations in the private sphere,45 had begun to gain traction. Between the late 1990s and the early 2010s, the ECtHR’s increasing attention to human rights violations taking place in the ‘private’ sphere resulted in the expansion of its coercive human rights doctrine to new domains, such as child abuse,46 rape,47 domestic violence,48 domestic slavery49 and human trafficking.50 These significant case law developments that took place in the area of coercive human rights in the ‘long 2000s’, triggered the emergence, from around 2010 onwards, of the more critical scholarship mentioned above,51 to which this volume aims to contribute. B.  The International Context i.  Coercive Human Rights within a Broader Anti-impunity Agenda The ECtHR’s coercive human rights doctrine is part of a broader international trend. International human rights norms have been increasingly infused with duties to mobilise criminal law (enforcement) towards the protection, and redress for violation, of rights. This evolution in human rights law has taken place against the backdrop of a broader ‘anti-impunity’ agenda in international law. ‘Anti-impunity’ is characterised by an emphasis on holding perpetrators

42 Seibert-Fohr (n 10) 7 and 111. 43 A Mowbray, The Development of Positive Obligations under the European Convention on Human Rights by the European Court of Human Rights (Oxford, Hart Publishing, 2004) 227. Mowbray’s monograph provides an excellent overview of the Court’s early case law regarding investigative obligations under arts 2, 3 and 5. 44 See Lavrysen, ch 2 in this volume. 45 See, eg, C Bunch, ‘Women’s Rights as Human Rights: Towards a Re-Vision of Human Rights’ (1990) 12 Human Rights Quarterly 486; H Charlesworth, C Chinkin and S Wright, ‘Feminist Approaches to International Law’ (1991) 85 American Journal of International Law 613. See also section II.B.ii for a discussion of the relationship of feminism with coercive human rights. 46 A v United Kingdom ECHR 1998-VI. 47 MC v Bulgaria ECHR 2003-XII. 48 Opuz v Turkey ECHR 2009. 49 Siliadin v France ECHR 2005-VII. 50 Rantsev v Cyprus and Russia ECHR 2010. 51 See section I above.

Coercive Human Rights  9 of serious violations of international law criminally accountable. From the Nuremberg trials and the Genocide Convention52 to the development of the ad hoc international criminal tribunals and the International Criminal Court,53 a legal and institutional framework took shape over the course of the twentieth century around the premise that grave wrongdoing should not be left unpunished.54 Although the international prosecution of individual persons may go as far back as the trial of Peter von Hagenbach in 1474,55 the Nuremberg and Tokyo International Military Tribunals marked an important turning point in the gradual solidification within the twentieth century of the imperative of holding individuals criminally accountable for atrocities amounting to serious violations of international law.56 Already, international humanitarian law was demanding the punishment of individuals for specific breaches of jus in bello norms.57 The Genocide Convention, which was drafted shortly after the Second World War, explicitly required States Parties to criminalise and punish genocide, in times of war and peace, and provided for both international and domestic criminal prosecution.58 Individual criminal punishment also prominently appeared in the four Geneva Conventions of 1949, which impose the duty to criminalise and punish grave breaches of the Conventions.59 Over time, express duties on states to prosecute and punish certain acts proliferated across a number of multilateral treaties aimed at matters such as offences in the context of aviation, terrorist bombings and hostage situations.60 Ultimately, the late twentieth and 52 UN General Assembly, Convention on the Prevention and Punishment of the Crime of G ­ enocide, 9 December 1948, 78 UNTS 277. 53 See W Schabas, ‘International Justice for International Crimes: An Idea Whose Time Has Come’ (2006) 14 European Review 421. 54 Seibert-Fohr (n 10) 1. For a critical discussion of this idea, see S Moyn, ‘Anti-impunity as Deflection of Argument’ in K Engle, Z Miller and DM Davis (eds), Anti-impunity and the Human Rights Agenda (Cambridge, Cambridge University Press, 2016). 55 E Greppi, ‘The Evolution of Individual Criminal Responsibility under International Law’ (1999) 81 International Review of the Red Cross 531; see also E van Sliedregt, Individual Criminal Responsibility in International Law (Oxford, Oxford University Press, 2012) ch 1. 56 Greppi (n 55) 536 ff. 57 See, eg, art 28 of the Convention for the Amelioration of the Condition of the Wounded and Sick in Armies in the Field, 6 July 1906, 11 LNTS 440; art 21 of Hague Convention X – Adaptation to Maritime War of the Principles of the Geneva Convention, 18 October 1907, 15 LNTS 340: art 29 of the Convention for the Amelioration of the Condition of the Wounded and Sick in Armies in the Field, 27 July 1929, 118 LNTS 303. 58 UN General Assembly (n 52) art VI. 59 Article 50 of the Geneva Convention for the Amelioration of the Condition of the Wounded and Sick in Armed Forces in the Field, 12 August 1949, 75 UNTS 31; art 51 of the Geneva Convention for the Amelioration of the Condition of Wounded, Sick and Shipwrecked Members of Armed Forces at Sea, 12 August 1949, 75 UNTS 85; art 130 of the Geneva Convention Relative to the Treatment of Prisoners of War, 12 August 1949, 75 UNTS 135; art 147 of the Geneva Convention Relative to the Protection of Civilian Persons in Time of War, 12 August 1949, 75 UNTS 287. See R O’Keefe, ‘The Grave Breaches Regime and Universal Jurisdiction’ (2009) 7 Journal of International Criminal Justice 811. 60 See, among many others, the Convention for the Suppression of Unlawful Acts against the Safety of Civil Aviation, 23 September 1971, 974 UNTS 177; the International Convention for the

10  Natasa Mavronicola and Laurens Lavrysen early twenty-first centuries saw the re-invigoration of the pursuit of individual criminal accountability by means of international prosecutions, with the operation of the ad hoc criminal tribunals for the Former Yugoslavia and Rwanda,61 and the coming into effect of the Rome Statute of the International Criminal Court.62 The anti-impunity orientation in international law came to play a key part in human rights law both through legislative instruments and through the pronouncements of key judicial and oversight bodies. States have explicitly undertaken duties to criminalise, prosecute and punish various human rights violations through widely ratified legal instruments such as the United Nations Convention against Torture,63 the Convention on the Elimination of All Forms of Racial Discrimination64 and the Convention for the Protection of All Persons from Enforced Disappearances.65 But anti-impunity also gradually permeated the doctrine of various human rights courts and oversight bodies, which, in their interpretation of human rights provisions that do not explicitly contain duties to mobilise the criminal law, have established a body of positive obligations to criminalise, prosecute and punish human rights violations. Prominent among these has been the Inter-American Court of Human Rights.66 In its first judgment in a contentious case, Velásquez-Rodriguez v Honduras, the Court underlined that states would be in breach of their duty to ensure human rights if the ‘State apparatus acts in such a way that the violation goes unpunished’.67 Since then, as Alexandra Huneeus has observed, the Court ‘made national prosecution of gross, state-sponsored crimes a centerpiece of its regional agenda’,68 influencing other bodies – notably the African Suppression of Terrorist Bombings, 15 December 1997, 2149 UNTS 256; and the International Convention against the Taking of Hostages, 17 December 1979, 1316 UNTS 205. 61 Both tribunals were established by resolutions of the UN Security Council. The International Criminal Tribunal for the former Yugoslavia was established by SC Res 827, UN Doc S/RES/827 (25 May 1993); the International Criminal Tribunal for Rwanda was established by SC Res 955, UN Doc S/RES/955 (8 November 1994). 62 Rome Statute of the International Criminal Court, 17 July 1998, 2187 UNTS 90. 63 Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, 10 December 1984, 1465 UNTS 85. 64 International Convention on the Elimination of All Forms of Racial Discrimination, 21 December 1965, 660 UNTS 195, Article 4. 65 International Convention for the Protection of All Persons from Enforced Disappearances, 20 December 2006, 2716 UNTS 3. For an account of human rights treaties incorporating duties to mobilise the criminal law in respect of human rights violations, see Seibert-Fohr (n 10) ch 5. 66 FF Basch, ‘The Doctrine of the Inter-American Court of Human Rights Regarding States’ Duty to Punish Human Rights Violations and its Dangers’ (2007) 23 American University International Law Review 195; Seibert-Fohr (n 10) ch 3; F Mégret and J-P Calderón, ‘The Move towards a Victim-Centred Concept of Criminal Law and the “Criminalization” of Inter-American Human Rights Law: A Case of Human Rights Law Devouring itself?’ in Y Haeck, O Ruiz-Chiriboga and C Burbano Herrera (eds), The Inter-American Court of Human Rights: Theory and Practice, Present and Future (Cambridge, Intersentia, 2015). 67 Velásquez Rodríguez v Honduras, Merits, Inter-Am Ct HR (Ser C) No 4 (1988), 29 July 1988, para 176. 68 A Huneeus, ‘International Criminal Law by Other Means: The Quasi-criminal Jurisdiction of the Human Rights Courts’ (2013) 107 American Journal of International Law 1, 1.

Coercive Human Rights  11 Commission on Human and Peoples’ Rights and the European Court of Human Rights – along the way.69 The Human Rights Committee, too, ‘has repeatedly called upon States parties to hold accountable perpetrators of serious human rights violations’.70 From its early jurisprudence, it urged states to bring perpetrators of certain human rights violations ‘to justice’.71 To take a more recent example, in its General Comment No 36 on the right to life issued in 2019, the Committee indicated that states are required to ‘enact a protective legal framework that includes effective criminal prohibitions on all manifestations of violence or incitement to violence that are likely to result in deprivation of life’ and that ‘criminal sanctions attached to these crimes must be commensurate with their gravity’.72 It also stressed that: ‘States parties must generally refrain from addressing violations of article 6 merely through administrative or disciplinary measures, and a criminal investigation is normally required, which should lead, if enough incriminating evidence is gathered, to a criminal prosecution.’73 Examples of human rights doctrine requiring the deployment of the criminal law as a means of protection from, and redress for, human rights violations abound.74 Accordingly, a formidable corpus of jurisprudence and other normative outputs has developed on the international and regional plane delineating duties to criminalise, prosecute and punish.75 The coercive duties doctrine of the ECtHR, which goes as far back as the X and Y v The Netherlands judgment of 1985,76 is an important part of this corpus and has contributed to advancing the anti-impunity norm(s) on the international legal arena.77 For Karen Engle, who has sought to chronicle the ‘turn’ to criminal law in human rights, the culmination of anti-impunity’s hold on human rights is that ‘[t]oday, to support human rights means to favor criminal accountability for those individuals who have violated international human rights or humanitarian law’ and ‘to be against amnesty laws that might preclude such accountability’.78 Increasingly, as Seibert-Fohr has observed: ‘Efforts have been made to refer to [human rights law] in order not only to interpret existing crimes but also to extend the catalogue of international crimes.’79 Coercive human 69 See K Engle, ‘Anti-impunity and the Turn to Criminal Law in Human Rights’ (2015) 100 Cornell Law Review 1069, 1103–06 (see especially fn 139). 70 Seibert-Fohr (n 10) 12. 71 See, eg, Guillermo Ignacio Dermit Barbato et al v Uruguay, Communication No 84/1981, UN Doc CCPR/C/OP/2, para 11. 72 Human Rights Committee, General Comment No 36, UN Doc CCPR/C/GC/36, 3 September 2019, para 20 (citations omitted). 73 ibid para 27. 74 See the thorough analysis (as at 2009) in Seibert-Fohr (n 10). 75 See Malby (n 12); Seibert-Fohr (n 10); Huneeus (n 68); Basch (n 66); Engle (n 69); Engle et al (n 54) ch 1. 76 X and Y v The Netherlands (n 20). See also section II.A above. 77 Seibert-Fohr (n 10) ch 4. 78 Engle (n 69) 1070. 79 Seibert-Fohr (n 10) 3.

12  Natasa Mavronicola and Laurens Lavrysen rights doctrine is therefore part of a multi-layered development of duties to mobilise the criminal law with a view to vindicating and securing compliance with international legal norms. ii.  Critical Perspectives on the Anti-impunity Agenda There has been growing interest in the promise and pitfalls of the antiimpunity agenda in general and within human rights in particular. There are a number of perspectives and prominent strands of support for, and critique of, anti-impunity that are of relevance to the discussions found in this volume. Proponents of anti-impunity view the pursuit of individual criminal accountability – as distinct from state accountability80 – for serious wrongdoing as a crucial mechanism for advancing justice, both by way of redressing past atrocity and as a means of averting further wrongs through deterrence and the entrenchment of the rule of law.81 Kathryn Sikkink, for example, has traced and lauded what she terms the ‘justice cascade’: a ‘shift in the legitimacy of the norms of individual criminal accountability for human rights violations and an increase in actions (such as trials) on behalf of those norms’.82 Kim and Sikkink have indicated that ‘prosecutions are associated with improvements in human rights conditions’,83 particularly in respect of the crime of torture. Without necessarily abandoning a feminist ambivalence about allying with (international/criminal) law,84 a number of feminist activists and scholars have seen promise in the anti-impunity agenda and have sought to harness it in the fight against gender-based violence and other gendered wrongs,85 including by means of a feminist reconfiguration of traditionally patriarchal international legal norms and through the strategy of ‘norm transfer’ from the international to the domestic plane.86

80 K Sikkink and HJ Kim, ‘The Justice Cascade: The Origins and Effectiveness of Prosecutions of Human Rights Violations’ (2013) 9 Annual Review of Law and Social Science 269, 270. 81 Seibert-Fohr (n 10) 283. 82 K Sikkink, The Justice Cascade: How Human Rights Prosecutions are Changing World Politics (New York, WW Norton & Company, 2011) 8. 83 Sikkink and Kim (n 80) 281, citing HJ Kim and K Sikkink, ‘How Do Human Rights Prosecutions Improve Human Rights after Transition?’ (2012) 7 Interdisciplinary Journal of Human Rights Law 69. 84 C O’Rourke, ‘Feminist Strategy in International Law: Understanding its Legal, Normative and Political Dimensions’ (2017) 28 European Journal of International Law 1019, 1045. 85 For a nuanced account of this, predominantly focusing on international criminal law, see E Dowds, Feminist Engagement with International Criminal Law: Norm Transfer, Complementarity, Rape and Consent (Oxford, Hart Publishing, 2020) chs 1–3; see further, eg, R Copelon, ‘Gender Crimes as War Crimes: Integrating Crimes against Women into International Criminal Law’ (2000) 46 McGill Law Journal 217. 86 See, eg, C O’Rourke, ‘International Law and Domestic Gender Justice: Why Case Studies Matter’ in MA Fineman and E Zinsstag (eds), Feminist Perspectives on Transitional Justice: From International and Criminal to Alternative Forms of Justice (Cambridge, Intersentia, 2013); E Dowds,

Coercive Human Rights  13 More critical perspectives, on the other hand, have questioned the pursuit of protection or liberation through coercion and punishment. Karen Engle has captured some of the prominent issues arising out of anti-impunity, and the turn to the criminal law in human rights in particular, by identifying four thematically distinct concerns. The first relates to the individualisation and decontextualisation of wrongdoing and injustice that comes about by focusing on holding individual perpetrators to account and thereby omitting the wider collectives and structures that enable or inflict grave harm and injustice.87 The second concern Engle highlights relates to the ways in which anti-impunity moderates conceptions of (economic) harm and consequent remedies through the modest aim of redressing abuses rather than the bolder aim of restructuring the economic system within which they occur.88 The third concern is that anti-impunity demands alignment with the carceral state, which may unleash some of the carceral state’s worst excesses, including torture and miscarriages of justice.89 Finally, Engle is preoccupied with the way in which a criminal law focus can influence the ‘production of history’: ‘much of the story might be lost’,90 she warns, if the archive is determined by the (criminal) legal process and items’ legal admissibility. Moreover, access to archival materials may be denied if criminal proceedings are a live prospect. Not dissimilarly, from a feminist perspective, there has been scepticism about seeking emancipation through the criminal law and the carceral state apparatus. The entanglement of feminist reform agendas with this apparatus has been described as ‘governance feminism’91 or ‘carceral feminism’.92 Particular concerns regarding the feminist anti-impunity agenda relate to anti-impunity’s capacity to embolden a gendered and racialised carceral apparatus, as well as to the ways in which a focus on criminal accountability, by isolating and individualising gendered wrongs, can entail diversion from, and failure to address, the structural conditions in which such wrongs are embedded. For example, writing on the ‘carceral feminism’ surrounding human trafficking and modern slavery, Elizabeth Bernstein has elaborated on how the demand for protection ‘Conceptualising the Role of Consent in the Definition of Rape at the International Criminal Court: A Norm Transfer Perspective’ (2018) 20(4) International Feminist Journal of Politics 624. 87 Engle (n 69) 1120–22. See also V Nesiah, ‘Doing History with Impunity’ in Engle et al (n 54) 75. 88 Engle (n 69) 1122–24. 89 ibid 1124–26. 90 ibid 1126–27. 91 J Halley, P Kotiswaran, H Shamir and C Thomas, ‘From the International to the Local in Feminist Legal Responses to Rape, Prostitution/Sex Work and Sex Trafficking: Four Studies in Contemporary Governance Feminism’ (2006) 29 Harvard Journal of Law and Gender 335. See also J Halley, P Kotiswaran, R Rebouché, and H Shamir, Governance Feminism: An Introduction (Minneapolis, University of Minnesota Press, 2018). 92 E Bernstein, ‘The Sexual Politics of the New Abolitionism’ (2007) 18(3) Differences 128. See also E Bernstein, ‘Carceral Politics as Gender Justice? The “Traffic in Women” and Neoliberal Circuits of Crime, Sex, and Rights’ (2012) 41 Theory and Society 233; A Gruber, ‘The Feminist War on Crime’ (2007) 92 Iowa Law Review 741; A Gruber, ‘A “Neo-feminist” Assessment of Rape and Domestic Violence Law Reform’ (2012) 15 Journal of Gender, Race and Justice 583.

14  Natasa Mavronicola and Laurens Lavrysen through the criminal law has helped bolster the racialised US criminal justice system, recasting ‘the masculinist institutions of big business, the state, and the police’ as allies and protectors of marginalised persons, and effectively absolving neoliberal and state institutions by pinning responsibility for social harm on ‘deviant’ persons.93 Similarly, Kiran Grewal has warned that some international criminal law ‘attempts to respond to sexual and gender-based violence have reinforced rather than challenged patriarchal structures’.94 Other commentators have explored how far mandating the pursuit of criminal accountability in all circumstances of serious wrongdoing may undermine, or be incompatible with, other transitional justice or conflict resolution processes, such as truth and reconciliation commissions, that arguably align with human rights-compatible aims such as peace and truth.95 More broadly, echoing some of Engle’s reflections, scholars have raised the concern that the dominance of criminal punishment in international visions of justice – including ‘transitional justice’ or ‘global justice’ – for grave wrongdoing might eclipse or obscure alternative conceptions of justice, such as the pursuit of equality or redistribution.96 Such debates are distinct, but not wholly dissimilar, from enduring debates within criminology, penology and criminal theory concerning the tensions between criminal, individualised approaches to administering justice, and structural, redistributive models of justice favouring collective protection, equality and liberation.97 These debates form an important backdrop to the analysis and discussion of the ECtHR’s coercive human rights doctrine that is the focus of this collection. III.  STRUCTURE OF THE BOOK

This volume examines the ECtHR’s coercive human rights doctrine through a focus on: (a) the contours of the Court’s reasoning; (b) the interests and needs of 93 Bernstein, ‘The Sexual Politics of the New Abolitionism’ (n 92) 144. 94 KK Grewal, ‘International Criminal Law as a Site for Enhancing Women’s Rights? Challenges, Possibilities, Strategies (2015) 23 Feminist Legal Studies 149, 150. 95 See, eg, two recent contributions on political settlements, amnesties, and the demands of international law: S MH Nouwen, ‘Is There Something Missing in the Proposed Convention on Crimes against Humanity? A Political Question for States and a Doctrinal One for the International Law Commission’ (2018) 16 Journal of International Criminal Justice 877; M Jackson, ‘Amnesties in Strasbourg’ (2018) 38 Oxford Journal of Legal Studies 451. But see also the nuanced discussion in C Hillebrecht, A Huneeus and S Borda, ‘The Judicialization of Peace’ (2018) 59 Harvard International Law Journal 279. 96 SMH Nouwen and WG Werner, ‘Monopolizing Global Justice: International Criminal Law as Challenge to Human Diversity’ (2015) 13 Journal of International Criminal Justice 157. For a nuanced take on international criminal law and transformative – including structural – reparations, see A Durbach and L Chappell, ‘Leaving Behind the Age of Impunity: Victims of Gender Violence and the Promise of Reparations’ (2014) 16 International Feminist Journal of Politics 543. 97 See, eg, Loïc Wacquant, Punishing the Poor: The Neoliberal Government of Social Insecurity (Durham, NC, Duke University Press, 2009). The latter debates are briefly discussed by Mattia Pinto in ch 8 in this volume.

Coercive Human Rights  15 (potential) victims of human rights violations; (c) the most pressing challenges raised by, and for, the ECtHR’s coercive duties doctrine; and (d) under-explored aspects of the doctrine and the doctrine’s implications. The present chapter, offered by way of introduction and overview, synthesises and contextualises key elements of the layered and intricate input provided by the authors, and contemplates the future of (research on) the doctrine. Part I of the book gives an overview of key threads in the ECtHR’s coercive duties doctrine and the justification(s) put forward in the ECtHR’s elaboration of these duties. In Chapter 2, Laurens Lavrysen provides a bird’s-eye view of positive obligations to mobilise the criminal law, and probes key angles and grounding principles in the doctrine’s development. In Chapter 3, Paul Lemmens and Marie Courtoy re-examine the ECtHR’s coercive duties doctrine through the lens of ‘effective deterrence’. In Part II, Alina Balta in Chapter 4 locates strands of reparative justice within the ECtHR’s coercive duties doctrine and assesses them in light of victimological perspectives. In Chapter 5, Corina Heri distils the ways in which vulnerability reasoning contributes to, and shapes, the ECtHR’s coercive duties doctrine, focusing on the protection of children, victims of sexual and domestic violence, and members of the LGBTIQ+ community. In Chapter 6, Stephanos Stavros also considers vulnerability and the ways in which phenomena such as discrimination and marginalisation link to the development of coercive duties in relation to hate speech, and contemplates the future of duties to criminalise and punish in this context. Fundamental challenges for coercive human rights doctrine are considered throughout this collection, but are particularly probed in Part III. The chapters in this part provide theoretically informed as well as doctrinally grounded analysis and critical reflection on the challenges, dangers and limitations inhering in the doctrine. In Chapter 7, Nina Peršak scrutinises the development of duties to mobilise the criminal law in light of criminalisation theory, focusing on the principle of ultima ratio. In Chapter 8, Mattia Pinto examines the capacity of the Court’s coercive human rights doctrine to bolster domestic cultures of penality and sow a ‘culture of conviction’. With a focus on Articles  2 and 3 ECHR, in Chapter 9, Natasa Mavronicola contemplates potential dangers inhering in coercive human rights doctrine, focusing on three: coercive overreach, the weakening of human rights standards and diversion from alternative means of protection. In Chapter 10, Vladislava Stoyanova assesses coercive human rights doctrine under Article 4 ECHR in light of the imperative of effective protection for persons in precarious situations (notably in circumstances of irregular migration) in the context of slavery, forced labour and human trafficking. Finally, in Chapter 11, Brice Dickson explores the extent to which the Court’s doctrine challenges or accommodates (potential) transitional justice objectives. In Part IV, the chapters focus on hitherto under-explored areas in which the ECtHR’s coercive human rights doctrine is likely to affect the activities

16  Natasa Mavronicola and Laurens Lavrysen of domestic authorities. In Chapter 12, Liora Lazarus focuses on how liability attached to the discharge of operational duties by criminal justice agencies may raise a prospect of coercive overreach in shaping policing operations. In Chapter 13, Kelly Pitcher examines how coercive human rights may influence the conduct of the criminal trial, with a particular focus on the doctrine’s implications in respect of the admission of unlawfully obtained evidence in criminal proceedings. IV.  CENTRAL THEMES

Together, the chapters that have been introduced above engage with the ECtHR’s coercive human rights doctrine in a concerted and comprehensive manner. As explained in section I above, the objective of the present volume is to provide the basis for a meaningful re-assessment of the doctrine and of the supportive and critical discourses surrounding it. With this objective in mind, the present section identifies the three central themes of the book, weaving together common threads, where possible, and juxtaposing differences in perspective, in respect of (a) justifications for, (b) implications, and (c) limitations of the doctrine. A. Justifications Most authors in this volume engage with the justification of the ECtHR’s coercive human rights doctrine. On the one hand, a number of authors have addressed the question of the adequacy of the justifications offered by the Court to support the need for criminal law interventions. In the context of criminalisation, Nina Peršak, for instance, argues that in the light of the principle of ultima ratio from criminal law theory, a special justification is required for any form of coercion. She argues that the burden of persuasion rests on those who wish to criminalise, in this case on the ECtHR, which should therefore provide an extensive justification in its judgments when it imposes obligations to criminalise. In this regard, Laurens Lavrysen criticises the Court for often putting more effort into articulating a justification for exceptions to the need for criminal law interventions than in justifying the initial obligation to criminalise. He is also critical or the Court’s failure to substantiate the key assumption underlying its coercive human rights doctrine: that the criminal law actually provides effective deterrence against human rights violations. Mattia Pinto, in turn, argues that the Court demonstrates excessive confidence in the deployment of penal powers, unduly treating the criminal law as some kind of cure for social and moral harms. According to Peršak, there is little evidence to support such a view – she contemplates that the reference to effective deterrence might even be

Coercive Human Rights  17 a rhetorical device for the Court to allow it explicitly to adhere to a preventive instead of a retributivist theory of criminal punishment. A different position is taken by Paul Lemmens, acting Judge at the Court, and Marie Courtoy, who argue that it is not the role of the Court to undertake an empirical study as to whether criminal law can provide effective deterrence against certain acts. In other words, Lemmens and Courtoy accept that the Court can proceed on the assumption that it does. On the other hand, a number of authors have themselves proposed justifications that may serve either to support the Court’s current practice or to provide guidance for future case law developments. While some authors are sceptical on the degree to which criminal punishment is what victims of human rights violations primarily want or need (see section IV.C below), other authors have more positive views on the added value of the Court’s coercive human rights doctrine from what they consider as a victim-oriented perspective. Alina Balta, for instance, scrutinises the reparations ordered by the Court in cases involving gross violations of human rights, in particular those involving Chechnya and the Turkish south-east. In a number of cases, the Court has required retributive responses as part of these reparations, in particular calling for the truth to be established and for the identification and punishment of perpetrators. Emphasising the symbolic function of the criminal law, Balta argues that such retributive responses may be a necessary response to the experience of victimisation, which has placed the victim outside the sphere of the perpetrator’s moral obligations. For this reason, retributive responses may, on her account, re-emphasise victims’ membership of the moral sphere and thereby contribute to the restoration of their social standing and worth. Corina Heri, in turn, considers that the use of vulnerability reasoning, which is often invoked to justify coercive human rights duties in Article 3 cases, can be conducive to the adoption of a victim-oriented perspective. According to her, such perspective implies that the lived reality of a victim is taken seriously in the context of the law’s formulation and implementation, and public authorities’ interaction with victims of human rights violations. She argues that the concept of vulnerability could thereby serve as a lens to focus attention on those victims who are least likely to be treated respectfully, believed, heard or granted justice. Finally, Stephanos Stavros endorses the added value of coercive human rights from the viewpoint of victims’ rights, arguing in favour of extending positive obligations under the ECHR to require the criminalisation of hate speech. In order to do so, he stresses the need to ensure the coherence of the international legal framework, arguing that the development of coercive human rights duties in this area would be in line with the work of the European Commission against Racism and Intolerance (ECRI), the Council of Europe’s specialised mechanism for the fight against racism and intolerance, and with the International Convention on the Elimination of Racial Discrimination as well as the International Covenant on Civil and Political Rights.

18  Natasa Mavronicola and Laurens Lavrysen B. Implications One of the central themes throughout this book concerns the question of the potential coercive overreach98 stemming from the Court’s coercive human rights doctrine. Most critical of this is Mattia Pinto, who considers that the ECtHR has shown a sort of ‘criminal fetishism’, overlooking the risks associated with penal expansion and under-estimating the fact that the criminal law always poses a threat to the individual’s freedoms. Nina Peršak warns of the risk that the Court may open the door for obligations to criminalise more widely, under pressure from non-liberal dispositions or even merely as compassionate responses to applicants’ suffering. Natasa Mavronicola is particularly concerned about the danger that lies in equating circumstances amounting to a breach of the negative obligation under Articles 2 and 3 with circumstances demanding the imposition of criminal liability on individual persons, and disregarding factors that might be reasonably considered as a basis for exculpatory defences under domestic criminal law or that might diffuse the culpability of individuals in the particular context. She also raises the inverse concern that conflation of criminal law and human rights law may result in the dilution of human rights standards through the application of a criminal lens in determining what amounts to a human rights violation. For instance, she criticises the ‘honest belief’ test used in cases of death resulting from excessive use of force. She argues that this test, which echoes a criminal liability standard and which has emerged due to the tendency to align violations of the right to life with criminal liability, is insufficiently strict in holding states responsible for human rights violations, and advocates for an objective standard of absolute necessity where state agents resort to lethal violence. In contrast with these critical voices, other authors consider that the risk of coercive overreach in the ECtHR’s case law has so far not materialised. Corina Heri, for instance, admits that the case law does have ‘quite extensive coercive reach’, but emphasises that coercive human rights obligations are generally only ‘deployed in cases concerning heinous human rights violations, supported by national and international law’. According to Paul Lemmens and Marie Courtoy, the Court applies implicit consensus-like reasoning in order to limit its coercive duties to what is supported by comparative or international law, restricting the need to criminalise to human rights violations that are considered to be ‘“normally” regulated by criminal law’. Lemmens and Courtoy conclude that: ‘The Convention sets certain limits, but generally leaves a wide margin of appreciation to the competent authorities.’ This aligns with the findings emerging from Laurens Lavrysen’s chapter, who similarly argues that the Court is generally quite respectful of the state’s margin of appreciation. Lavrysen considers in particular that, where criminalisation is concerned, the Court has



98 See

section I above.

Coercive Human Rights  19 mainly stepped in to tackle the under-inclusiveness of existing criminal law protection. For instance, in MC v Bulgaria,99 the Court emphasised the need for states to define the crime of rape in a way that focuses on the lack of consent, including in the absence of physical resistance by the victim. Where the obligation to prosecute and criminally punish is concerned, both Kelly Pitcher and Laurens Lavrysen find that the Court typically only intervenes where the failure to convict a person or the application of particular defences has not taken place in good faith, which may foster a climate of impunity where exemption from punishment might be considered as leading to further human rights violations. According to Pitcher, this is consistent with the primary rationale of the procedural coercive human rights obligations being that of contributing to general human rights protection. However, in order to assess the implications of the ECtHR’s coercive human rights doctrine, one cannot stop at the level of what is strictly required by the case law. In her influential contribution on the topic in 2012, Liora Lazarus already drew attention to the fact that coercive overreach may also result from the ‘rhetorical assertion of coercive duties’, which needs to ‘be read against the background of an increasing rhetorical use of rights to justify coercive measures by politicians and state actors internationally’.100 In their chapters, Kelly Pitcher, Mattia Pinto and Liora Lazarus focus on under-explored areas in which the ECtHR’s coercive human rights doctrine may have an unintended (or insufficiently evaluated) impact on the activities of domestic authorities. Pitcher discusses the extent to which coercive human rights may influence the question of how to address pre-trial procedural violations in criminal proceedings. She critically discusses the Dutch system under which evidence obtained unlawfully can and in some conditions must be admitted with a view to protecting the rights of the victims of a crime or their next of kin. Pitcher considers that this system seems to go further than what the ECtHR strictly requires, as decisions on the exclusion of evidence taken in good faith and aimed at advancing legitimate goals of criminal procedure or justice seem to be allowed (and sometimes required) under the ECHR, even if this may result in the acquittal of the suspect. However, she suggests that the Dutch system may represent the undue coercive implications of the ECtHR’s coercive human rights doctrine, as the doctrine may incentivise states to use ‘end justifies the means’ reasoning in order to secure criminal punishment. Discussing the UK’s Modern Slavery Act, Mattia Pinto similarly argues that the ECtHR’s case law in this area101 might be read as supporting a ‘culture of conviction’, in which harsh prison sentences, greater powers to prosecute and convict, and extensive preventive measures have come to be seen as desirable and benign when advanced in the name of human rights.



99 MC

v Bulgaria (n 47). (n 3) 149. 101 Which is also discussed at length by Vladislava Stoyanova in ch 10 in this volume. 100 Lazarus

20  Natasa Mavronicola and Laurens Lavrysen Finally, while most of the present volume focuses on criminalisation, prosecution and criminal punishment, Liora Lazarus extends the debate on coercive overreach by exploring the coercive implications that may flow from the ECtHR’s preventive obligations case law. In cases like Osman v UK,102 the Court has held the state liable for failing to take operational measures to prevent human rights violations from taking place. According to Lazarus, the example of so-called Osman warnings and preventive risk orders under the Modern Slavery Act in the UK illustrate that this case law puts considerable pressure on law-enforcement agencies to act pre-emptively in order to avoid being held liable, which feeds into a risk-averse operational culture within policing. Most authors offer proposals with a view to mitigating the risk of coercive overreach or dealing with tensions between the interests of victims and those of defendants. Some authors advocate for more caution and clarity in the Court’s case law. Kelly Pitcher, for instance, underlines that there is a need to carefully frame the scope of coercive human rights in order to avoid coercive overreach, while Natasa Mavronicola cautions against statements appearing to conflate state responsibility for human rights violations with individual criminal liability. Other authors, such as Stephanos Stavros, and Paul Lemmens and Marie Courtoy, consider that a balanced approach coupled with the margin of appreciation may be an appropriate method by which to defuse tensions between victims’ and defendants’ rights. Corina Heri, in turn, suggests that the concept of vulnerability can be used as a tool to steer further case law developments by requiring criminalisation and criminal law responses in a more targeted manner, prioritising the need for effective deterrence through the criminal law to protect those considered vulnerable. In his chapter, Stavros also invokes vulnerability reasoning in order to justify particular protection of groups that are typically targeted or particularly affected by hate speech. Lemmens and Courtoy, in turn, emphasise the purpose of effective deterrence. In their view, where lack of criminal punishment does not undermine the deterrent effect of the law, criminal punishment should not be required under the Convention. Finally, in the conclusion of her chapter, Kelly Pitcher calls for discussion on what is lost and gained when human rights are invoked to justify a coercive approach that is already based on instrumentalist objectives which aim to ensure that those who have committed a criminal offence are convicted. This gives rise to the broader question of what is lost as a result of the ECtHR’s coercive human rights doctrine. Some authors in the present volume have begun to address this question. Liora Lazarus, for instance, is particularly concerned about how coercive human rights may contribute to the prioritisation of a culture of risk over a rights culture. Mattia Pinto, in turn, considers that by providing a human rights rationale for crime-control measures, coercive human rights delegitimise calls



102 Osman

v UK (n 3).

Coercive Human Rights  21 for criminal justice reforms aimed at reducing custodial sentences or decriminalising less serious infractions. He also considers that the doctrine could lead to neutralising critique about conditions of imprisonment or – in the context of human trafficking – the current policing of borders. C. Limitations Some authors have voiced concerns regarding the limitations of a criminal law approach to addressing human rights violations. The criminal law may, for instance, be limited in the extent to which it can provide what (potential) victims of human rights violations want or need. In this regard, with a particular focus on human trafficking, Vladislava Stoyanova criticises the Court’s case law as being too narrowly focused on the scope and application of the criminal law. Such an approach may be detrimental for trafficking victims, as there is a risk that they only receive assistance when they are useful for criminal proceedings against traffickers. She argues that the case law needs to be re-oriented in order to explicitly impose obligations on states to identify and provide assistance to trafficking victims independently of any criminal proceedings. More broadly, viewing criminal law as insufficient to protect victims, Natasa Mavronicola has called for a protective and preventive re-orientation within positive obligations doctrine, which may warrant the prioritisation of noncriminal tools of protection, such as effective access to justice and refuges for victims of domestic violence, and methods of investigation and redress that are capable of focusing on ‘rotten orchards’ and not merely on ‘bad apples’. In this context, it is worth noting Mattia Pinto’s warning that the insistence on coercive human rights tends to pull ‘non-punitive remedies into the criminal justice orbit’, which suggests that non-criminal protective measures should be more purposefully conceived and operationalised outside a criminal law paradigm. Finally, in the specific context of transitional justice, Natasa Mavronicola also particularly warns that the coercive human rights doctrine may privilege a thin notion of justice, in which a focus on criminal investigations and prosecutions risks crowding out other processes, such as those oriented at truth and reconciliation. Brice Dickson goes deeper into the limitations of coercive human rights in the context of transitional justice. Focusing in particular on the conflict in Northern Ireland, Dickson argues that tampering with the criminal law may be essential if conflict resolution processes are to succeed. Reflecting on the extent to which such tampering may be accommodated by the ECtHR’s coercive human rights doctrine, with particular focus on amnesties, he considers that amnesties may be acceptable from the viewpoint of the ECtHR provided that they do not concern ‘grave breaches of fundamental rights’ and that the state does not in some way collude in covering up what happened.

22  Natasa Mavronicola and Laurens Lavrysen V.  FUTURE DIRECTIONS

It goes without saying that all those interested in the intersections between human rights and criminal law or criminal justice should continue to keep a close eye on the development of coercive human rights doctrine. This collection offers a strong basis on which to contemplate concrete incremental developments as well as pressure points that may produce more dramatic developments in the legal parameters of duties to mobilise criminal law (enforcement) to protect against and/or provide redress for human rights violations. Beyond this, there are many directions that scholarship can take in future engagements with coercive human rights doctrine. These directions encompass empirical, comparative and conceptual investigations of the doctrine’s (actual or potential) implications and/ or opportunity costs. Suggested areas for future research concern in particular (a) the reception, perception and effects, as well as (b) the systemic and institutional implications of the ECtHR’s coercive human rights doctrine, and (c) the roads not taken by the doctrine, and how its contours might be re-imagined. A.  The Reception, Perception and Effects of the ECtHR’s Coercive Human Rights Doctrine Systematic empirical examination of the way in which the ECtHR’s coercive human rights doctrine is understood, pursued and received would offer fruitful avenues for further evaluating its justifications, effects, limitations and dangers. This could include empirical research identifying how key actors and stakeholders – including judges, victims of human rights violations, civil society and legal practitioners – view the functioning and telos of the ECtHR’s coercive human rights doctrine, and attempting to measure the actual functioning and effects of the doctrine in practice. This might encompass an empirical investigation of how (far) coercive human rights doctrine maps onto the wishes or needs of (potential) victims of human rights violations, or onto ‘what works’ in terms of deterrence and effective prevention. Furthermore, while the capacity of the doctrine to embolden and enlarge the carceral state has been widely contemplated, not least in this collection,103 largescale empirical research could establish how far it has actually done so. There are difficult questions to be addressed in this context, such as: when coercive human rights doctrine is invoked, does it in itself serve to enlarge penality or is it justifying coercion that would, more likely than not, have still occurred, but been justified with reference to a more overtly coercive agenda? How far are appeals to human rights (perceived to be) more (or less) potent tools than more overtly

103 See section IV on the central themes in this collection. See especially Mattia Pinto and Liora Lazarus, chs 8 and 12 in this volume respectively.

Coercive Human Rights  23 coercive discourses in widening the net of coercion and control? Alternatively, what legitimacy or rhetorical power might appeals to coercive human rights add in practice to more overtly coercive and carceral discourses in terms of expanding the carceral state? Conversely, a number of authors in this volume have seen criminal law (enforcement) tools not only as capable of affording effective redress for, and protection from, human rights violations, but also as capable of becoming more attuned to particular needs and vulnerabilities of (potential) victims of human rights violations and accordingly delivering more effective and comprehensive protection.104 In this context too, there is scope for empirical investigation, to explore how far coercive human rights doctrine is having the effect of reshaping the criminal law and its operationalisation in the criminal justice and law enforcement system by tying it more closely to the protection of human rights.105 Could coercive human rights doctrine, by treating protection as central to criminal law and its enforcement, operate to enhance the protective orientation and effectiveness of the criminal law and criminal law enforcement system? And might a human rights orientation in the sword dimension of the criminal law apparatus – that is, in the suppression of crime – also imbue a human rights orientation within this apparatus more generally, thereby also enhancing the shield function of human rights? In other words, might a criminal justice apparatus – including police and prosecutorial authorities – that is increasingly more attuned to positive duties to mobilise criminal justice processes under human rights law also come to be more holistically respectful of human rights (such as the right to a fair trial and the prohibition of torture) in the discharge of criminal justice? These speculative questions warrant systematic empirical examination. As Anja Seibert-Fohr highlighted in 2009: ‘While initially the human rights institutions were concerned with large-scale impunity there is a growing body of cases, especially under the European Human Rights system, dealing with inadequacies in criminal legislation and in the conduct of criminal proceedings in individual cases’ (emphasis added).106 The fact that the ECtHR’s coercive human rights doctrine is permeating ‘everyday’ criminal law (enforcement) entails that there is fruitful scope for systematically tracing this doctrine’s ‘cascade’107 to domestic jurisdictions and the precise contours of its immediate, but also medium-term and longer-term impact on domestic criminal law and criminal justice processes. 104 See section IV on the central themes in this collection. See especially Alina Balta and Corina Heri, chs 4 and 5 in this volume respectively. 105 The related idea that aligning the substance of the criminal law with the protection of human rights can form the basis of a liberal criminal theory is explored in T Hörnle, ‘“Rights of Others” in Criminalisation Theory’ in A P Simester, A du Bois-Pedain and U Neumann (eds), Liberal Criminal Theory: Essays for Andreas von Hirsch (Oxford, Hart Publishing, 2014). 106 Seibert-Fohr (n 10) 8. 107 The term is borrowed from Sikkink (n 82) above.

24  Natasa Mavronicola and Laurens Lavrysen B.  Systemic and Institutional Implications The doctrinal developments critically examined in this collection carry enormous significance for the international system of human rights protection and the international legal system more broadly. The ECtHR’s coercive human rights doctrine has developed against the wider backdrop of an anti-impunity agenda that has taken hold within international human rights law and global justice more broadly. The implications of the ECtHR’s doctrine for this wider global context should continue to be reviewed,108 and it is hoped that this volume will be valuable in this regard. Particularly in view of the doctrine’s continuing development in areas beyond the contexts of conflict, transition and mass atrocity, scholars should continue to probe the ways in which the doctrine is migrating, cross-fertilising and evolving across regional, international, and domestic jurisdictions, and across regimes of international (human rights) law, the precise parameters of this cross-fertilisation and the key actors involved. The examination of the doctrine’s evolution and impact in systemic terms should encompass an inquiry into what coercive human rights doctrine entails for the integrity of human rights law itself. The tensions between human rights’ ‘sword’ and ‘shield’ functions and their implications have been well-rehearsed prior to, and within, this collection, and should continue to be explored. Yet, beyond this, there are questions about how an alignment with criminal law might reshape the way in which human rights standards delineate state responsibility for human rights violations. In her chapter in this collection, Natasa Mavronicola highlights the danger that applying a criminal lens to human rights norms may serve to weaken or narrow the demands they make of the state apparatus. The extent to which the coercive shift may take human rights doctrine in more minimalist directions warrants further investigation. More generally, there is considerable scope for contemplating how the substance of coercive human rights doctrine, but also the method and reasoning employed in its elaboration and application, may come to reconfigure the international (human rights) legal system. Moreover, in light of the counter-carceral and ‘pro’-carceral facets of its jurisprudence, scholarship may more earnestly come to grapple with the conceptual parameters of the ECtHR’s penology, understood as its philosophical stance on criminal punishment. A key question in this regard is whether there can be a unifying narrative from the rehabilitative approach of Vinter v UK,109

108 For a recent example of the examination of the ECtHR’s coercive human rights doctrine’s significance in relation to international criminal law, see JD Ohlin, ‘The Right to Punishment for International Crimes’ in F Jeßberger and J Geneuss (eds), Why Punish Perpetrators of Mass Atrocities? Purposes of Punishment in International Criminal Law (Cambridge, Cambridge University Press, 2020). 109 Vinter v UK (2016) 63 EHRR 1.

Coercive Human Rights  25 which rejected irreducible whole life terms of imprisonment and put forward rehabilitation as the prime rationale of carceral punishment, to the retributivist and deterrent stance characteristic of case law that effectively demands the implementation of carceral sanctions for certain human rights violations, such as Derman v Turkey.110 Even if a unifying narrative is not forthcoming, the fragmentation itself offers a site for fruitful reflection on what the ECtHR’s plurality of approach to penality signifies and how it interacts with the penal structures of Contracting States. Yet another set of questions arises in respect of the institutional implications of the ECtHR’s coercive human rights doctrine. It is generally acknowledged, not least by the Court itself, that the ECtHR is not institutionally tasked with, or equipped for, determining individual criminal liability.111 Yet the increasing demands it makes in respect of criminalisation and commensurate punishment raise questions about whether an institutional transformation may be required to accommodate the increasingly coercive and punitive role it has sought to play. If the ECtHR were more explicitly to assume criminal jurisdiction, the practical challenges and implications, as well as wider promises and pitfalls, of such a prospect warrant assessment. In this context, it is worth noting the ongoing efforts to imbue the African Court on Human and Peoples’ Rights with criminal jurisdiction through the Malabo Protocol.112 Relatedly, more scrutiny is warranted of the coercive function of the Committee of Ministers and the Department for the Execution of Judgments of the European Court of Human Rights, which make concerted efforts to secure the implementation of duties to criminalise and punish by Contracting States,113 against the backdrop of what is often portrayed as a ‘crisis’ of compliance or implementation for the ECHR system.114 It would be interesting to explore the implications of these bodies’ criminal justice role in terms of how they are – or ought to be – institutionally equipped, supported and, perhaps, monitored.

110 Derman v Turkey (2015) 61 EHRR 27, paras 28–29; see also, eg, Okkalı v Turkey (2010) 50 EHRR 43, paras 63–65; Nikolova and Velichkova v Bulgaria (2009) 48 EHRR 40, para 63. 111 Tanlı v Turkey App No 26129/95 (ECtHR, 10 April 2001), para 111; see also Tekin and Arslan App No 37795/13 (ECtHR, 5 September 2017), paras 81 and 109. 112 See S Nimigan, ‘The Malabo Protocol, the ICC, and the Idea of “Regional Complementarity”’ (2019) 17 Journal of International Criminal Justice 1005; F Aja Agwu, ‘The African Court of Justice and Human Rights: The Future of International Criminal Justice in Africa’ (2014) 6 Africa Review 30. 113 See Huneeus (n 6) 25. See, for example, on the ‘Makaratzis’ group of cases in relation to Greece: Decision of Ministers’ Deputies, Document no CM/Del/Dec(2018)1331/H46-13 (Committee of Ministers, 6 December 2018), available at: https://search.coe.int/cm/Pages/result_details.aspx?Obje ctID=09000016808fddda#globalcontainer. See also Guidelines and Reference Texts on Eradicating Impunity for Serious Human Rights Violations, adopted by the Committee of Ministers on 30 March 2011 (Council of Europe 2011), available at: https://rm.coe.int/1680695d6e. 114 G Stafford, ‘The Implementation of Judgments of the European Court of Human Rights: Worse Than You Think – Part 2: The Hole in the Roof’, EJIL: Talk!, 8 October 2019, https://www. ejiltalk.org/the-implementation-of-judgments-of-the-european-court-of-human-rights-worse-thanyou-think-part-2-the-hole-in-the-roof.

26  Natasa Mavronicola and Laurens Lavrysen C.  Roads not Taken Besides the concrete or potential implications of the ECtHR’s coercive human rights doctrine, there is room for contemplating the roads not taken by the doctrine and the arguments that have shaped it. Samuel Moyn has argued that ‘anti-impunity should invite not mindless allegiance, but exploration of what it is trying to do and actually achieving, compared with hypothetical alternatives with which it competes or which it even rules out’.115 There is therefore scope for contemplating what is excluded or missed when individual criminal accountability is seen – by the Court or by litigants – as the means of protection against, and redress for, human rights violations. Reflection on roads not taken can amount not merely to a study in counterfactuals, but also to an exercise in re-imagining. Might the ECtHR’s robust positive obligations doctrine be (re-) orientated towards addressing the structures and systems – beyond the penal frame – that are facilitating or inflicting harm, or that are failing to empower and protect (potential) victims of human rights violations?116 This search for opportunity costs and meaningful alternatives may be tied more broadly to a quest for the absent and unspoken in dominant human rights discourse. Accordingly, an amplification of non-hegemonic or counter-hegemonic discourses117 and marginalised voices could help identify what might be missing and what might be re-imagined. What might happen if thinking in this area foregrounded the views and interests of persons of colour, women, children, migrants, asylum-seekers, sex workers, persons from poorer backgrounds, queer and trans persons? Questions surrounding what is generated and what is foreclosed by the ECtHR’s coercive human rights doctrine will undoubtedly endure and proliferate. While opinion can and, most likely, will continue to reasonably differ on the desirability of the ECtHR’s coercive interventions, we hope that this collection serves as an invitation to treat coercive human rights as an object of continued critical scrutiny.

115 S Moyn, ‘Anti-impunity as Deflection of Argument’ in Engle et al (n 54) 87–88. 116 See, in this respect, Stoyanova, ch 10 in this volume, who distinguishes the protection of trafficking victims in irregular migration circumstances from the prosecution of their traffickers. 117 On this, see Richard Falk, ‘The Power of Rights and the Rights of Power: What Future for Human Rights?’ (2008) 1 Ethics & Global Politics 81; see also T Basok, ‘Counter-hegemonic Human Rights Discourses and Migrant Rights Activism in the US and Canada’ (2009) 50 International Journal of Comparative Sociology 183.

Part I

Key Threads in ECtHR Doctrine

28

2 Positive Obligations and the Criminal Law A Bird’s-Eye View on the Case Law of the European Court of Human Rights LAURENS LAVRYSEN

I. INTRODUCTION

T

he European Court of Human Rights (ECtHR), using its positive obligations doctrine, increasingly requires states to mobilise the criminal law in order to prevent or redress violations of the European Convention on Human Rights (ECHR). On the one hand, the substantive aspect of certain Convention rights may oblige the state to criminalise particular human rights violations, including those committed by private actors (‘obligation to criminalise’). In this context, I will use the term ‘human rights offences’, which are defined by Kamber as ‘all criminal breaches of human rights, irrespective of whether they are committed by the state or by a private party, which attain the minimum level of severity necessary to attract the specific heightened protection under international human rights law and which should therefore constitute a criminal offence under the relevant domestic criminal law’.1 On the other hand, increasingly the Court demands under the procedural aspect of certain Convention rights that states criminally prosecute and punish perpetrators of human rights offences (‘obligation to punish’). In this chapter, I will map the Court’s case law on positive obligations and the criminal law: with respect to what kind of human rights violations does the Court require the mobilisation



1 K

Kamber, Prosecuting Human Rights Offences (Leiden, Brill, 2017) 19.

30  Laurens Lavrysen of the criminal law? In doing so, I will also scrutinise the legal reasoning used by the Court to justify the development of these kinds of substantive and procedural obligations. While not necessarily claiming to be exhaustive, this chapter is based on the in-depth study of a large corpus of 126 cases2 in which the Court has discussed questions of criminalisation or criminal punishment as an element of the state’s positive obligations.3 This corpus has been collected on the basis of the snowball method, conducting searches in the Court’s HUDOC database to identify cases using relevant keywords distilled from other cases (such as ‘manifest disproportion’), complemented by an additional check of the literature on the topic4 to avoid overlooking relevant cases. Before turning to a separate discussion of the Court’s case law on the obligations to criminalise and to punish, I will first briefly discuss how these are embedded in the respective substantive and procedural aspects of the state’s positive obligations under the ECHR, as well as how these are interrelated. II.  SUBSTANTIVE VERSUS PROCEDURAL POSITIVE OBLIGATIONS

In law in general, the distinction between substance and procedure is not always an easy one to draw. Cook has, for example, observed that ‘there is no “line” already in “existence” which can be “discovered” by analysis alone … but rather a “no-man’s land”’.5 Instead of considering the distinction between substance and procedure as a bright line, he considers it more appropriate to recognise that: [T]he ‘substantive’ shades off by imperceptible degrees into the ‘procedural’ and that the ‘line’ between them does not ‘exist’, to be discovered merely by logic and analysis, but is rather to be drawn so as best to carry out our purpose.6

With respect to the Court’s positive obligations case law, I have argued elsewhere that it makes most sense to explain the distinction between substance and procedure by reference to the aim of the respective obligations. Whereas substantive positive obligations aim to promote particular states of affairs, procedural positive obligations are primarily concerned with considerations of fairness, even though they may indirectly contribute to a state of affairs of better substantive human rights protection.7 The latter is particularly evident as far as the

2 Up to 31 October 2019. 3 Excluding cases that merely touch upon the compatibility of the investigation with the procedural positive obligations. 4 See especially Kamber (n 1). 5 W Cook, ‘“Substance” and “Procedure” in the Conflict of Laws’ (1932–33) 42 Yale Law Journal 333, 335. 6 ibid 343. 7 L Lavrysen, Human Rights in a Positive State (Cambridge, Intersentia, 2016) 52–53.

Positive Obligations and the Criminal Law  31 obligation to punish is concerned, which, as we will see, is on the one hand developed within the framework of the investigative obligations that primarily aim at the fair treatment of victims, but on the other hand also aims for a deterrent effect that may contribute to a state of better protection against similar human rights violations. In the context of human rights offences, substantive positive obligations (including the obligation to criminalise) primarily serve a preventive function, whereas procedural positive obligations (including the investigative obligations, which encompass the obligation to punish) primarily serve a remedial function.8 However, as with law in general, there is no-clear cut distinction between substance and procedure in the Court’s positive obligations case law. The study of the cases from the corpus has in particular shown that it may sometimes be difficult to disentangle these obligations given their interrelated character. Take the example of domestic violence cases. Most of these cases have been examined by the Court from a substantive perspective, resulting in the finding of a violation of the substantive positive obligations to protect an individual at risk of being subjected to domestic violence as a result of a failure of the state authorities to take adequate operational measures in response to allegations thereof.9 However, in Irene Wilson v UK, the Court examined the case exclusively through a procedural lens, focusing on whether the state had complied with its obligation to investigate the incident complained of and, in particular, to punish the perpetrator thereof. The Court seemingly justified this approach by holding that, in contrast to earlier domestic violence cases, in this case ‘there was no continuing situation of violence against the applicant’ as she had only brought one incident to the attention of the authorities.10 In other words, where a ‘continuing situation of violence’ does exist, the procedural positive obligations are to some extent subsumed by the substantive positive obligations to protect the victim against further violence. For the purposes of this chapter, it is also particularly relevant to highlight the interrelationship between the substantive positive obligations to criminalise certain human rights offences and the procedural obligation to investigate and if appropriate prosecute and punish perpetrators thereof. Take the rape case of MC v Bulgaria, in which the Court found a violation of Articles 3 and 8 ECHR because no criminal proceedings had been brought against the perpetrators due to a substantive problem relating to how the offence of rape was defined under domestic criminal law.11 In its judgment, the Court did not distinguish between procedural and substantive positive obligations, but rather merged both issues

8 However, substantive positive obligations and procedural positive obligations do not necessarily serve, respectively, a preventive and a remedial function in every context; see ibid 50–51. 9 See, eg, Opuz v Turkey ECHR 2009; Kontrová v Slovakia App No 7510/04 (ECtHR, 31 May 2007). 10 Irene Wilson v UK App No 10601/09 (ECtHR, 23 October 2012), para 48. 11 MC v Bulgaria ECHR 2003-XII, para 166.

32  Laurens Lavrysen given their interrelated character in the case at hand. In the Article  3 case of Cestaro v Italy, the Court similarly emphasised this interrelationship: The absence of criminal legislation capable of preventing and effectively punishing the perpetrators of acts contrary to Article  3 can prevent the authorities from prosecuting violations of that fundamental value of democratic societies, assessing their gravity, imposing adequate penalties and precluding the implementation of any measure likely to weaken the penalty excessively, undermining its preventive and dissuasive effect.12

As a logical implication, wherever the Court has imposed an obligation to punish, the state must implicitly be considered to be under a prior obligation to criminalise.13 For the purposes of the following section, cases in which the obligation to criminalise can implicitly be derived from the obligation to criminally prosecute and, if appropriate, punish such offences will be discussed along with cases in which the Court explicitly required such criminalisation under the substantive positive obligations. However, not all findings of a violation on account of the inadequacy of a criminal law response by the state necessarily indicate that such a response is mandated by the Convention. For instance, judgments from which it is evident that non-criminal remedies could in principle be acceptable under the Convention, but are simply not provided for or have not been pursued in the domestic legal system – as a result of which the Court can only examine the adequacy of the measures that have been taken within the framework of criminal proceedings – will not be included.14 In short, the obligation to criminalise and the obligation to punish thus interrelate in the same way that the overarching categories of substantive and procedural positive obligations, in which they are embedded, do. Taking into account this caveat, both obligations will be further discussed separately for analytical purposes. III.  THE OBLIGATION TO CRIMINALISE

In this section, I will map the infringements of human rights, whether committed by state agents or by private actors, that states are explicitly or implicitly required to criminalise according to the Court’s case law, as well as the reasoning provided by the Court to justify this. I will do so on an article-by-article basis, focusing on those articles under which the Court has required states to criminalise particular human rights offences, with a particular focus on Articles 2, 3, 4 and 8 ECHR. 12 Cestaro v Italy App No 6884/11 (ECtHR, 7 April 2015), para 209. See also, under art  4, CN v UK App No 4239/08 (ECtHR, 13 November 2012). 13 See, eg, Öneryildiz v Turkey ECHR 2004-XII; see section III.A below. 14 See, eg, Mučibabić v Serbia App No 34661/07 (ECtHR, 12 July 2016), para 129.

Positive Obligations and the Criminal Law  33 A. Article 2 While the Court had already interpreted Article  2 ECHR (the right to life) as imposing requirements on the state’s legal framework in the McCann case in 1995 – in particular requiring that ‘national law must strictly control and limit the circumstances in which a person may be deprived of his life by agents of the State’15 – it was in the Osman case in 1998 that the Court actually required states to criminalise infringements of the right to life, regardless of whether they are committed by state agents or private actors. The Grand Chamber held that Article  2 requires states ‘not only to refrain from the intentional and unlawful taking of life, but also to take appropriate steps to safeguard the lives of those within its jurisdiction’.16 The Court considered that ‘the State’s obligation in this respect extends beyond its primary duty to secure the right to life by putting in place effective criminal-law provisions to deter the commission of offences against the person’, but ‘may also imply in certain well-defined circumstances a positive obligation on the authorities to take preventive operational measures to protect an individual whose life is at risk from the criminal acts of another individual’.17 What is remarkable is the laconic manner in which the Court, by way of obiter dictum as the case in fact related to an alleged lack of preventive operational measures, indicated – without providing any reasons to explain why – that states should criminalise ‘offences against the person’, seemingly considering this a self-evident ‘primary duty’ derived from Article 2. The Court did link this ‘primary duty’ to the more general obligation ‘to safeguard the lives of those within its jurisdiction’. However, this sentence, derived from the then recently issued LCB judgment,18 is just another way of stating that Article 2 gives rise to positive obligations. However, it fails to explain why such positive obligations should encompass a specific obligation to criminalise. It must also be noted that the Court holds the state to be under an obligation to criminalise ‘offences against the person’, which seems broader than the unlawful taking of a person’s life and arguably encompasses at least life-endangering offences against the person as well.19 In later cases, the Court fine-tuned the circumstances in which protection via the criminal law is required under Article 2. In 2002, in the medical negligence case of Calvelli and Ciglio, the Grand Chamber, for example, clarified that [I]f the infringement of the right to life or to personal integrity 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 15 McCann and Others v UK Series A No 324 (1995), para 151. 16 Osman v UK ECHR 1998-VIII, para 115. 17 ibid. 18 LCB v UK ECHR 1998-III, para 36. 19 Mutatis mutandis, Makaratzis v Greece ECHR 2004-XI, paras 49–55, in which the Court held that art 2 is applicable in cases of life-endangering use of force by state agents, even when the victim survives.

34  Laurens Lavrysen case. In the specific sphere of medical negligence the obligation may for instance also be satisfied if the legal system affords victims a remedy in the civil courts, either alone or in conjunction with a remedy in the criminal courts, enabling any liability of the doctors concerned to be established and any appropriate civil redress, such as an order for damages and for the publication of the decision, to be obtained. Disciplinary measures may also be envisaged.20

Again, the Court failed to provide reasons to justify its position. A likely explanation for this distinction may be that the Court considers non-intentional deaths to typically involve a lesser degree of culpability on the part of the person responsible than intentional deaths. This seems to be confirmed by a line of reasoning from the Öneryildiz case, decided in 2004, in which the Grand Chamber qualified Calvelli and Ciglio by ruling that particular cases of death resulting from gross negligence – arguably involving a higher degree of culpability – could nonetheless still require the application of criminal-law mechanisms: Where it is established that the negligence attributable to State officials or bodies on that account goes beyond an error of judgment or carelessness, in that the authorities in question, fully realising the likely consequences and disregarding the powers vested in them, failed to take measures that were necessary and sufficient to avert the risks inherent in a dangerous activity … the fact that those responsible for endangering life have not been charged with a criminal offence or prosecuted may amount to a violation of Article 2, irrespective of any other types of remedy which individuals may exercise on their own initiative.21

According to the Court, these considerations clearly applied to the case at hand, which concerned a methane explosion on a state-run landfill that resulted in the flooding by waste of the slum swellings situated below the dump, causing 39 fatalities. Here the Court did put some effort into justifying the distinction. By way of analogy, it referred to ‘cases of homicide’, holding that in such cases: [T]he interpretation of Article  2 as entailing an obligation to conduct an official investigation is justified not only because any allegations of such an offence normally give rise to criminal liability … but also because often, in practice, the true circumstances of the death are, or may be, largely confined within the knowledge of State officials or authorities.22

According to the Court: [S]uch considerations are indisputably valid in the context of dangerous activities, when lives have been lost as a result of events occurring under the responsibility of the public authorities, which are often the only entities to have sufficient relevant

20 Calvelli and Ciglio v Italy ECHR 2002-I, para 51. In exceptional cases, ‘where the fault attributable to the health-care providers went beyond a mere error or medical negligence’, a criminal law response may nonetheless still be required in medical negligence cases; see Lopes de Sousa Fernandes v Portugal ECHR 2017, para 215. 21 Öneryildiz v Turkey (n 13) para 93. 22 ibid.

Positive Obligations and the Criminal Law  35 knowledge to identify and establish the complex phenomena that might have caused such incidents.23

The considerations of the Court shed more light on the reasons for more generally distinguishing cases in which a criminal law response to a death is required under Article 2 from cases where this is not the case. As the ‘knowledge of State officials or authorities’ fails to explain the need to criminalise intentional deaths caused by private actors, the second reason seems to be more decisive: states ought to criminalise ‘offences against the person’ that, according to the Court, ‘normally give rise to criminal liability’. The Court has fine-tuned the contours of the Öneryildiz exception in further case law. In Jasinskis in 2010, for example, it has held that criminal law protection is necessary when a person in police custody has died as a result of a negligent failure to seek adequate medical treatment.24 While in principle the Öneryildiz exception only appears to apply to negligence by state agents,25 occasionally the Court has also applied it to negligence by private actors.26 For example, in Sinim in 2017, it held that a criminal law response was necessary, despite the fact that the fatality resulted from negligence by a private actor – the applicant’s husband had died after the truck he had rented had a collision with another vehicle and caught fire, while he had not been informed by the truck owner of the fact that the truck was carrying inflammables. The Court justified the need for a criminal law response by holding that: [T]he death in the instant case resulted from the responsible parties’ voluntary and reckless disregard of their legal duties under the relevant legislation, as opposed to a simple omission or human error, which in the Court’s opinion sets this case apart from other cases of non-intentional deaths where it has found civil remedies to be sufficient.27

B. Article 3 The first case under Article  3 (the prohibition of torture and of inhuman and degrading treatment or punishment) in which the Court explicitly mentioned the obligation to criminalise particular forms of ill-treatment is the MC case, decided in 2003.28 Under Articles 3 and 8 combined, the Court held that: ‘States have a 23 ibid. 24 Jasinskis v Latvia App No 45744/08 (ECtHR, 21 December 2010), para 73. 25 Mučibabić v Serbia (n 14) para 129. 26 Without, however, considering that there is a general obligation to criminalise intentional ­ missions; see Makarová v UK App No 67149/17 (ECtHR, 16 June 2018), para 33. o 27 Sinim v Turkey App No 9441/10 (ECtHR, 6 June 2017), para 63. The Court emphasised that ‘the transportation of certain categories of dangerous goods without the permission of the competent authorities is an offence punishable by imprisonment’ (para 64). 28 Implicitly, this was already apparent from A v UK ECHR 1998-VI; see A Ashworth, Positive Obligations in Criminal Law (Oxford, Hart Publishing, 2013) 199–200.

36  Laurens Lavrysen positive obligation inherent in Articles 3 and 8 of the Convention to enact criminallaw provisions effectively punishing rape and to apply them in practice through effective investigation and prosecution.’29 In doing so, it mainly referred to the principles developed under Article 8 in the earlier case of X and Y v The Netherlands from 1985, in which the Court recognised such an obligation in the context of the protection against sexual violence.30 Both cases will be discussed more extensively in the context of Article 8. The Grand Chamber in the Gäfgen case in 2010, referring back to MC, recognised in more general terms that: ‘For an investigation to be effective in practice it is a prerequisite that the State has enacted criminal-law provisions penalising practices that are contrary to Article 3.’31 Here, the Court fails to explain why the considerations justifying criminalisation in the specific context of protection against rape mutatis mutandis apply to all ‘practices that are contrary to Article 3’. In the above-mentioned case of Cestaro in 2015, the Court found a procedural violation of Article 3 mainly because the domestic courts’ failure to hold police officers accountable for the acts of torture they committed resulted from the absence of a specific criminal offence of torture in Italian legislation. According to the Court: [T]he Italian criminal legislation applied in the instant case … proved both inadequate in terms of the requirement to punish the acts of torture in issue and devoid of any deterrent effect capable of preventing similar future violations of Article 3.32

Here the Court emphasised that the obligation to criminalise is instrumental to both retributive and deterrent purposes. The Cestaro judgment is further relevant because, in the Court’s ruling on general measures under Article 41, the obligation to criminalise obtains a more autonomous character, transcending the framework of procedural obligations. Under Article 41, the Court ‘considers it necessary to introduce into the Italian legal system legal mechanisms capable of imposing appropriate penalties on those responsible for acts of torture and other types of ill-treatment under Article 3’.33 Finally, the Myumyun case decided in 2015 also merits some discussion, as here the Court was most explicit as to the required scope of the criminal provisions protecting against torture. While in police custody, the applicant was subjected to torture for several hours, consisting of repeated blows, numerous kicks and electric shocks. As the treatment did not lead to long-term damage to the applicant’s health, the police officers were only convicted of causing light bodily harm. According to the Court, the problem lay with the fact that in the



29 MC

v Bulgaria (n 11) para 153. and Y v The Netherlands Series A No 91 (1985). 31 Gäfgen v Germany ECHR 2010, para 117. 32 Cestaro v Italy (n 12) para 225. 33 ibid para 246. 30 X

Positive Obligations and the Criminal Law  37 absence of a specific criminal offence of torture, the offence relied on by the domestic authorities was not ‘capable of squarely addressing the full range of issues thrown up by the act of torture to which the applicant fell victim’.34 Criminal law protection against torture must also sufficiently take into account the psychological suffering involved, as: [O]ne of the distinguishing characteristics of torture is that it not only – and not always – seriously damages the physical health of the person subjected to it but also affects in a very serious way that person’s dignity and psychological well-being.35

In other words, the criminal offence should adequately reflect the dynamics involved in the human rights violation of torture. Similarly, in the context of domestic violence, the Court requires that criminal law protection is in place regardless of whether injuries of a certain degree of severity have been inflicted, ‘because domestic violence may take many forms, some of which do not result in physical injury – such as psychological or economic abuse or controlling or coercive behaviour’.36 As is the case under Article 2, where infringements of an individual’s personal integrity that attain the minimum level of severity required to trigger the applicability of Article 3 are caused unintentionally, criminal law protection is not necessarily required.37 Arguably, as under Article 2, this might not apply to cases of ‘gross negligence’.38 C. Article 439 The first case in which the Court required criminalisation under Article  4 (the prohibition of slavery, servitude and forced labour) was the case of Siliadin in 2005, concerning a Togolese minor without residence papers who had been held in a state of servitude by the couple she resided with. The Court found the state to be under a positive obligation to criminalise situations that come within the scope of Article 4: The Court reiterates that Article 4 enshrines one of the fundamental values of democratic societies. Unlike most of the substantive clauses of the Convention and of 34 Myumyun v Bulgaria App No 67258/13 (ECtHR, 3 November 2015), para 77. 35 ibid para 74. 36 Volodina v Russia App No 41261/17 (ECtHR, 9 July 2019), para 81. 37 See, eg, Mitkus v Latvia App No 7259/03 (ECtHR, 2 October 2010), para 76 (medical negligence). In Nicolae Virgiliu Tănase v Romania App No 41720/13 (ECtHR, 25 June 2019), para 123, the Court held that ‘bodily injuries and physical and mental suffering experienced by an individual following an accident which is merely the result of chance or negligent conduct cannot be considered as the consequence of “treatment” to which that individual has been “subjected” within the meaning of Article 3’. 38 See also Kamber (n 1) 242, referring to Denis Vasilyev v Russia App No 32704/04 (ECtHR, 17 December 2009). 39 See also V Stoyanova, ‘Article  4 of the ECHR and the Obligation of Criminalising Slavery, Servitude, Forced Labour and Human Trafficking’ (2014) 3 Cambridge Journal of International and Comparative Law 407.

38  Laurens Lavrysen Protocols Nos. 1 and 4, Article 4 makes no provision for exceptions and no derogation from it is permissible under Article 15 § 2 even in the event of a public emergency threatening the life of the nation … In those circumstances, the Court considers that, in accordance with contemporary norms and trends in this field, the member States’ positive obligations under Article 4 of the Convention must be seen as requiring the penalisation and effective prosecution of any act aimed at maintaining a person in such a situation.40

In other words, given the fact that both Articles  3 and 4 are absolute, nonderogable rights that constitute ‘fundamental values of democratic societies’, the Court has mutatis mutandis applied the principles developed under Article 3 to the context of Article 4. In the absence of a specific offence of servitude or another offence under French criminal law under which the conduct of the couple could be brought, the Court considered that ‘the criminal-law legislation in force at the material time did not afford the applicant, a minor, practical and effective protection against the actions of which she was a victim’.41 Invoking the principle of effectiveness, the Court here seems to consider criminalisation to be instrumental to and essential for the effective protection against situations contrary to Article 4. In its subsequent case law, the Court further specified the kind of criminal law protection required under Article 4. First, in the Rantsev judgment in 2010, in which the Court recognised that human trafficking in itself falls within the scope of Article  4, it recognised ‘the duty to penalise and prosecute trafficking’ as one of the aspects ‘of member States’ general undertaking to combat trafficking’.42 While the Court in Siliadin merged questions of substance and procedure, in Rantsev it explicitly distinguished between the obligation to criminalise, as part of a broader framework of substantive obligations under Article 4 ‘to put in place an appropriate legislative and administrative framework’, and the ‘procedural obligation to investigate trafficking’.43 Second, in the case of CN in 2012, it specified what was required in terms of criminal law protection against servitude. An investigation into the applicant’s complaint that she was held in conditions of servitude was discontinued by the authorities because the circumstances of her case did not appear to constitute the offence of trafficking people for the purposes of exploitation. The Court found a procedural violation of Article 4, resulting from a ‘lacuna in domestic law’.44 According to the Court: [D]omestic servitude is a specific offence, distinct from trafficking and exploitation, which involves a complex set of dynamics, involving both overt and more subtle forms of coercion, to force compliance. A thorough investigation into complaints of such

40 Siliadin v France ECHR 2005-VII, para 112. 41 ibid para 148. See also CN and V v France App No 67724/09 (ECtHR, 11 October 2012). 42 Rantsev v Cyprus and Russia ECHR 2010 (extracts), para 285. 43 ibid paras 290–93 (Cyprus) and 301–03 (Russia), and 299–300 (Cyprus) and 307–09 (Russia), respectively. 44 CN v UK (n 12) para 80.

Positive Obligations and the Criminal Law  39 conduct therefore requires an understanding of the many subtle ways an individual can fall under the control of another. In the present case, the Court considers that due to the absence of a specific offence of domestic servitude, the domestic authorities were unable to give due weight to these factors.45

In other words, as with torture under Article 3, domestic law should provide for a criminal offence that adequately reflects the dynamics involved in the human rights violation of servitude. D. Article 8 The seminal case related to criminal law protection under Article 8 is the case of X and Y, decided in 1985. The case concerned the rape of a mentally disabled girl. The criminal complaint by the victim’s father was dismissed because Dutch law required such a complaint to be lodged by the victim herself, which had been impossible since she was legally incapable of lodging such a complaint as a result of her disability. Countering the government’s argument that civil remedies were open to the applicant, the Court ruled that: [T]he protection afforded by the civil law in the case of wrongdoing of the kind inflicted on Miss Y is insufficient. This is a case where fundamental values and essential aspects of private life are at stake. Effective deterrence is indispensable in this area and it can be achieved only by criminal-law provisions; indeed, it is by such provisions that the matter is normally regulated.46

In other words, the need for criminalisation is justified by reference to three reasons: the gravity of the infringement (‘fundamental values and essential aspects of private life’), the assumption that effective deterrence requires criminal law provisions – embedding the obligation to criminalise in a preventive framework – and the fact that these kinds of infringements ‘normally’ constitute criminal offences. In its subsequent case law, the Court has somewhat clarified the boundaries of what it considers to fall under the notion of ‘fundamental values and essential aspects of private life’ and consequently ought to be protected via the criminal law. Besides rape, the Court has indicated that the sexual abuse of minors clearly falls under this notion,47 as well as ‘attacks on the physical integrity’ more generally48 – where Article 8 arguably applies even when such an attack does not attain the minimum level of severity to trigger the applicability of Article 3.49 45 ibid. 46 X and Y v The Netherlands (n 30) para 27. 47 See, eg, M and C v Romania App No 29032/04 (ECtHR, 27 September 2011). 48 Beganović v Croatia App No 46423/06 (ECtHR, 25 June 2009), para 56. 49 However, ‘not every act or measure of a private individual which adversely affects the physical and psychological integrity of another will interfere with the right to respect for private life guaranteed by Article 8’, for instance, where a person was seriously injured as a result of a non-intentional traffic accident; see Nicolae Virgiliu Tănase v Romania (n 37) paras 128–31.

40  Laurens Lavrysen The Söderman case decided in 2013 was particularly important in determining where to draw this line in the context of offences of a sexual nature. The case concerned the failure to prosecute the applicant’s stepfather for covertly filming her naked in the absence of a statutory provision criminalising such conduct. Discussing the general principles concerning criminalisation under Article 8, the Court held that: [I]n respect of less serious acts between individuals, which may violate psychological integrity, the obligation of the State under Article  8 to maintain and apply in practice an adequate legal framework affording protection does not always require that an efficient criminal-law provision covering the specific act be in place. The legal framework could also consist of civil-law remedies capable of affording sufficient protection.50

The Court illustrated this with the example of the protection of a person’s picture against abuse by others, in which it accepted that civil law remedies in principle provide sufficient protection.51 Applying the general principles to the facts of the case, the Court did not require criminal law provisions to be in place because ‘the offence in question did not involve any physical violence, abuse or contact’ and ‘did not attain the seriousness of the grave acts in the case-law … which concerned rape and sexual abuse of children’.52 In other words, in order to attain the gravity threshold in cases concerning sexual offences under Article 8, acts of physical violence, abuse or contact are required.53 Outside the context of infringements of one’s physical integrity, the Court has also required criminal law protection in the case of Ageyevy, decided in 2013. This case concerned the disclosure of the adopted status of the applicants’ son to the media in the context of a mediatised trial regarding alleged child abuse in which they were eventually acquitted. According to the Court, ‘unauthorised communication of confidential information’ of this nature falls under the notion of ‘fundamental values and essential aspects of private life’.54 However, it is not clear to what extent other infringements of Article 8, outside the context of the protection of one’s physical integrity, may similarly require a criminal law response. As the notion of ‘fundamental values and essential aspects of private life’ is quite vague, since it merely indicates the existence of a threshold of gravity of the infringement without specifying how to determine whether such threshold is attained, this gives rise to some degree of legal uncertainty as to the scope of the obligation to criminalise under Article 8. However, it can be assumed that what holds true for Articles 2 and 3 – that infringements

50 Söderman v Sweden ECHR 2013, para 85. 51 ibid, referring to, inter alia, Von Hannover v Germany ECHR 2004-VI. 52 ibid para 86. 53 This was confirmed by the later case of A, B and C v Latvia App No 30808/11 (ECtHR, 31 March 2016). 54 Ageyevy v Russia App No 7075/10 (ECtHR, 18 April 2013), para 196.

Positive Obligations and the Criminal Law  41 do not necessarily require criminal law protection when they are caused by (mere) negligence – a fortiori applies to Article 8.55 Finally, as with Articles 3 and 4, it is necessary that the criminal law protection is tailored to the specificities of the human rights offence, as becomes evident from the rape case MC. At the domestic level, no criminal proceedings were brought against the perpetrators because the prosecutor considered the use of force or threats not to have been established beyond reasonable doubt. Taking into account ‘a clear and steady trend in Europe and some other parts of the world towards abandoning formalistic definitions and narrow interpretations of the law in this area’, the Court found that the state had failed to comply with its positive obligations under Article 8 combined with Article 3, which ‘must be seen as requiring the penalisation and effective prosecution of any non-consensual sexual act, including in the absence of physical resistance by the victim’.56 E.  Other Convention Provisions Under Articles  9 (freedom of religion), 10 (freedom of expression) and 11 ­(freedom of association and assembly), the Court has occasionally considered that a failure to apply criminal law mechanisms in the event of third-party infringements of these rights may violate the state’s procedural positive obligations under these provisions.57 However, it does not seem that separate criminal offences are required to protect against these kinds of infringements. As the relevant cases have all involved acts of violence targeting the substance of these respective freedoms, it can be assumed that the question of criminal law protection against violent third-party infringements of these freedoms is subsumed under the broader obligation to criminalise violence against the person under Articles 2 and 3.58 The same does not seem to hold true for Article 14 ECHR (the prohibition of discrimination), which the Court, in combination with Articles 2 and/or 3, has interpreted as encompassing a procedural obligation to investigate whether a discriminatory motive has played a role in acts of violence against the person. In the specific context of racist violence, the Grand Chamber in Nachova, decided 55 Similarly, see Kamber (n 1) 336, referring to, inter alia, Georgel and Georgeta Stoicescu v ­Romania App No 9718/03 (ECtHR, 26 July 2011). 56 MC v Bulgaria (n 11) paras 156 and 166. 57 See, eg, Begheluri and Others v Georgia App No 28490/02 (ECtHR, 7 October 2014), paras 163–65 (art 9); Özgür Gündem v Turkey ECHR 2000-III, para 45 (art 10); and Ouranio Toxo and Others v Greece ECHR 2005-X (extracts), para 43 (art 11). 58 See also, with respect to art 10, Kamber (n 1) 346. In Karaahmed v Bulgaria App No 30587/13 (ECtHR, 24 February 2015), para 110, the Court did consider that the investigation should not have been directed only at the physical acts of violence involved in a demonstration at a mosque, although without unambiguously requiring the need for criminal law protection against non-violent acts of disruption. See also Stephanos Stravos, ch 6 in this volume.

42  Laurens Lavrysen in 2005, required from states that ‘a distinction is made both in their legal systems and in practice between cases of excessive use of force and of racist killing’.59 Later, in the case of Abdu in 2014, when assessing Bulgaria’s legal framework in a favourable manner, by way of obiter dictum the Court seemingly suggested that – in line with their obligations flowing from the International Convention on the Elimination of All Forms of Racial Discrimination – states must make violence based on racial grounds a separate criminal offence.60 Again, criminal law protection must be sufficiently tailored to the specificities of the human rights offence of racist violence, as the result of treating racist violence and nonracist violence alike ‘would be to turn a blind eye to the specific nature of acts that are particularly destructive of fundamental rights’.61 Finally, it is also noteworthy that in the somewhat ambiguous judgment in the burglary case of Blumberga in 2008, the Court explicitly considered but ultimately rejected the need for criminal law protection under Article 1 Protocol 1 (the right to property). On the one hand, the Court considered that ‘a positive obligation arises for the State to ensure in its domestic legal system that property rights are sufficiently protected by law’.62 It referred to the MC case in order to mutatis mutandis hold under Article 1 Protocol 1 that ‘where the interference is of a criminal nature, this obligation [to ensure that property rights are sufficiently protected by law] will in addition require that the authorities conduct an effective criminal investigation and, if appropriate, prosecution’.63 Here the reasoning of the Court is somewhat circular, as in the absence of an objective standard as to what amounts to an ‘interference of criminal nature’, this boils down to the Court saying that what is protected by the criminal law ought to be protected by it. On the other hand, in the following paragraph, the Court considers that: [T]he possibility of bringing civil proceedings against the alleged perpetrators of a crime against property may provide the victim with a viable alternative means of securing the protection of his rights, even if criminal proceedings have not been brought to a successful conclusion, provided that a civil action has reasonable prospects of success.64

In other words, criminal law protection is not required under Article 1, Protocol 1 as long as there are effective civil remedies available. The reason for the absence of mandatory criminal law protection seems to lie in the fact that the Court altogether considers that: [T]he obligation to investigate is less exacting with regard to less serious crimes, such as those involving property, than with regard to more serious ones, such as violent

59 Nachova

and Others v Bulgaria ECHR 2005-VII, para 160. v Bulgaria App No 26827/08 (ECtHR, 11 March 2014), para 47. 61 Nachova and Others v Bulgaria (n 59) para 160. 62 Blumberga v Latvia App No 70930/01 (ECtHR, 14 October 2008), para 67. 63 ibid. 64 ibid para 68. 60 Abdu

Positive Obligations and the Criminal Law  43 crimes, and in particular those which would fall within the scope of Articles 2 and 3 of the Convention.65

F.  Conclusion on Criminalisation In sum, with some exceptions, the obligation to criminalise is confined to the context of violence against the person and other intentional infringements of the physical integrity of the person. The reason for this seems to be that the Court’s case law arguably only requires the criminalisation of those human rights violations that are so serious that they are ‘normally’ regulated by the criminal law. One can of course criticise the Court on the ground that there is no adequate vantage point to determine whether the latter is in fact the case. In this regard, it seems as if the Court is applying some kind of European consensus-like reasoning, drawing inspiration from the practices in the majority of Member States, yet without explicitly employing this label or the associated legal methodology.66 A survey of the case law reveals that the Court has mainly stepped in when confronted with cases of under-inclusiveness of the existing criminal law protection. This is often the case where the criminal law provisions in place are not sufficiently tailored to the specificities of the human rights offence concerned, as is evidenced by the case law concerning torture, servitude, rape and racist violence. As emphasised by the Court in the Myumyun case, in the context of Article 3, sufficiently tailored criminal law provisions are particularly important as: [T]he national authorities cannot be expected to discharge their positive obligations under Article  3 of the Convention by acting in breach of the requirements of its Article 7, one of which is that the criminal law must not be construed extensively to an accused’s detriment.67

Besides referring to the fundamental character of the right or the seriousness of the infringement concerned and to the ‘normal’ nature of criminal law protection, the Court has also occasionally invoked the principle of effectiveness as a rationale for criminalisation. Here, the Court proceeds on the assumption that the criminal law is an effective tool of deterrence and thereby can be considered as instrumental to a better state of protection against human rights violations, although without making any effort to support this assumption.68 All in all, the Court could clearly do a better job in providing reasons to justify imposing

65 ibid para 67. 66 On a European consensus, see, eg, K Dzehtsiarou, European Consensus and the Legitimacy of the European Court of Human Rights (Cambridge, Cambridge University Press, 2015). 67 Myumyun v Bulgaria (n 34) para 76. 68 See also Nina Peršak, ch 7 in this volume, questioning this assumption.

44  Laurens Lavrysen an obligation to criminalise on the state. Paradoxically, in its case law, the Court has sometimes put more effort into articulating a justification for exceptions to the need for criminal law protection – as is evidenced by the Söderman case – than in justifying the initial obligation to criminalise in the first place. IV.  THE OBLIGATION TO PUNISH

As we will see, in recent years the Court has been increasingly willing to hold that the state is under an obligation to criminally prosecute and punish the perpetrator of a human rights offence. However, at the outset, it must be emphasised that this evolution is at odds with some of the older general principles concerning how the Court assesses such cases. The following quote from the Tanlı case, decided in 2001, in which the applicant complained under Articles 2 and 3 about the death and alleged ill-treatment of his son by Turkish police officers, is particularly illustrative in this respect: When there have been criminal proceedings in the domestic courts concerning those same allegations, it must be borne in mind that criminal law liability is distinct from international law responsibility under the Convention. The Court’s competence is confined to the latter … The responsibility of a State under the Convention, arising for the acts of its organs, agents and servants, is not to be confused with the domestic legal issues of individual criminal responsibility under examination in the national criminal courts. The Court is not concerned with reaching any findings as to guilt or innocence in that sense.69

In later cases, the Court further clarified that ‘its task is not to deliver guilty or not-guilty verdicts on the individual in question … or to determine the penalty to be imposed, which matters come under the exclusive jurisdiction of the domestic criminal courts’.70 While a first limitation thus stems from the Court’s position as an ­international court (the institutional limitation), a second limitation relates to the scope of the procedural positive obligations themselves (the scope limitation), which are constructed by the Court as obligations of means rather than of result.71 In Öneryildiz, the Court explained that: It should in no way be inferred … that Article 2 may entail the right for an applicant to have third parties prosecuted or sentenced for a criminal offence (see, mutatis mutandis, Perez v France …) or an absolute obligation for all prosecutions to result in conviction, or indeed in a particular sentence (see, mutatis mutandis, Tanlı v Turkey).72



69 Tanlı

v Turkey ECHR 2001-III, para 111. v Italy (n 12) para 207. 71 See, eg, Armani Da Silva v UK ECHR 2016, para 233. 72 Öneryildiz v Turkey (n 13) para 96. 70 Cestaro

Positive Obligations and the Criminal Law  45 With respect to the reference to Tanlı, it is remarkable that the Court here transforms the institutional limitation into a scope limitation. However, the reference to Perez is more striking as in the relevant paragraph from that judgment, the Court merely ruled that ‘the right to have third parties prosecuted or sentenced for a criminal offence’ cannot be considered as falling under the concept of ‘civil rights and obligations’, triggering the application of Article 6(1), unless it is ‘indissociable from the victim’s exercise of a right to bring civil proceedings in domestic law’.73 It is hard to see how this can be meaningfully transposed without further explanation to the context of procedural obligations under substantive human rights. In any event, the turn towards obligations to punish seems at odds with both the institutional and scope limitations. From an institutional perspective, while the Court is of course not competent to formally declare an individual criminally liable or to impose a particular criminal sentence, in substance it boils down to the same thing when the Court finds that a failure by the state to do so may violate the Convention.74 Even more remarkable is the fact that the Court regularly pays lip service to the scope limitation while at the same time nonetheless finding a procedural violation on account of the failure to adequately punish the perpetrator, as it did, for example, in the Öneryildiz case – in this connection, Kamber talks of a ‘right-claim’ of the victim to the effective application of criminal law mechanisms.75 It goes without saying that this kind of internal incoherence within the case law and even within individual judgments is regrettable. While the Court has also occasionally required criminal punishment under Articles 476 and 8,77 it is under Articles 2 and 3 that the obligation to punish has fully matured. It does not seem that the Court makes a clear distinction in terms of the general principles governing the obligation to punish based on the article concerned, which is attested by the fact that these principles are easily transposed from the one provision to the other. Therefore, in what follows, the case law on the obligation to punish will be discussed as a whole rather than separately per article. First, I will explain how the obligation to punish gradually developed within the framework of procedural positive obligations, highlighting the Court’s reasoning to justify these developments. Second, I will give a brief overview of how the Court determines whether or not there has been a violation of the Convention on account of the inadequate criminal punishment of the perpetrator of a human rights offence.

73 Perez v France ECHR 2004-I, para 70. 74 This is sometimes referred to as the exercise of quasi-criminal jurisdiction; see A Huneeus, ‘International Criminal Law by Other Means: The Quasi-criminal Jurisdiction of the Human Rights Courts’ (2013) 107 American Journal of International Law 1. 75 Kamber (n 1) 507. 76 Siliadin v France (n 40) para 145. 77 See, eg, Bevacqua and S v Bulgaria App No 71127/01 (ECtHR, 12 June 2008), para 83.

46  Laurens Lavrysen A.  The Development of the Obligation to Punish The main developments relating to the procedural positive obligations to investigate have taken place under Articles 2 and 3. As a rationale for this development, the Court held in the McCann (1995) and Assenov (1998) cases that the general legal prohibitions enshrined in Articles 2 and 3, respectively, ‘would be ineffective in practice’ without the requirement that ‘some form of effective official investigation’ takes place into arguable claims that these provisions have been violated.78 In the specific context of Article 2, the Court clarified that: The essential purpose of such investigation is to secure the effective implementation of the domestic laws which protect the right to life and, in those cases involving State agents or bodies, to ensure their accountability for deaths occurring under their responsibility.79

While not explicitly requiring the punishment of the perpetrators of human rights offences in its early case law, the Court nonetheless did emphasise that such investigation, in order to be effective, must be ‘capable of leading to the identification and punishment of those responsible’.80 Of course, the possibility of resulting in the imposition of criminal punishment is not necessarily the only reason why a recourse to criminal law mechanisms may be required – often only a criminal investigation will provide the necessary means to establish what has happened and who was responsible.81 However, it does seem that, quite early on, criminal punishment of perpetrators was already at the very least considered as the ‘natural outcome’82 of an effective investigation. Falling short of recognising an autonomous obligation to punish, in the case of Nachova in 2005, the Court held that: Compliance with the State’s positive obligations under Article 2 of the Convention requires that the domestic legal system must demonstrate its capacity to enforce criminal law against those who unlawfully took the life of another.83

The case of Öneryildiz, decided in 2004, constitutes an important development in the case law, as the Grand Chamber here explicitly broadened the scope of the procedural positive obligation to investigate to also encompass the conduct

78 McCann and Others v UK (n 15) para 161; and Assenov and Others v Bulgaria ECHR 1998-VIII, para 102. 79 Paul and Audrey Edwards v UK ECHR 2002-II, para 69. 80 The Court first held so in the context of art 13 in Aksoy v Turkey ECHR 1996-VI, para 98. In Assenov and Others v Bulgaria (n 78) para 102 and Oğur v Turkey ECHR 1999-III, para 88, the Court transposed this line of reasoning to the context of arts 3 and 2, respectively. 81 F Tulkens, ‘The Paradoxical Relationship between Criminal Law and Human Rights’ (2011) 9 Journal of International Criminal Justice 577, 591, with reference to Khashiyev and Akayeva v Russia App Nos 57942/00 and 57945/00 (ECtHR, 24 February 2005), para 121. 82 The Court used this term (‘issue naturelle’) in the case of Nasr and Ghali v Italy App No 44883/09 (ECtHR, 23 February 2016), para 272. 83 Nachova and Others v Bulgaria (n 59) para 160.

Positive Obligations and the Criminal Law  47 of the domestic criminal courts, thereby opening up the possibility to also ­scrutinise their sentencing practices: [T]he requirements of Article 2 go beyond the stage of the official investigation, where this has led to the institution of proceedings in the national courts: the proceedings as a whole, including the trial stage, must satisfy the requirements of the positive ­obligation to protect lives through the law.84

According to the Grand Chamber, ‘the national courts should not under any circumstances be prepared to allow life-endangering offences to go unpunished’.85 The Grand Chamber considers that its task: [C]onsists in reviewing whether and to what extent the courts, in reaching their conclusion, may be deemed to have submitted the case to the careful scrutiny required by Article 2 of the Convention, so that the deterrent effect of the judicial system in place and the significance of the role it is required to play in preventing violations of the right to life are not undermined.86

Around the same time, in the context of Article 3, the Court started to develop a line of case law concerning so-called ‘“measures” which are unacceptable’87 under the Convention. In the context of the right to an effective remedy under Article 13, the Court held in Abdulsamet Yaman in 2004 that: [W]here a State agent has been charged with crimes involving torture or ill-treatment, it is of the utmost importance for the purposes of an ‘effective remedy’ that criminal proceedings and sentencing are not time-barred and that the granting of an amnesty or pardon should not be permissible.88

In the Okkali case in 2006, the Court not only transposed the concept of ‘unacceptable measures’ to the context of the procedural positive obligations under Article  3, but also extended it to cover the unjustified use of mitigating circumstances to reduce and suspend a prison sentence as well, as thereby the domestic courts had ‘exercised their discretion more in order to minimise the consequences of an extremely serious unlawful act than to show that such acts could in no way be tolerated’.89 For some time, this ‘minimise the consequences’ line of reasoning operated as an autonomous ground to find a violation of the procedural positive obligations on account of the conduct of domestic criminal courts – sometimes in connection with the notion of ‘unacceptable measures’90 and sometimes separately.91 Later, the ‘minimise the

84 Öneryildiz v Turkey (n 13) para 95. 85 ibid para 96. 86 ibid. 87 Okkali v Turkey ECHR 2006-XII (extracts), para 76. 88 Abdulsamet Yaman v Turkey App No 32446/96 (ECtHR, 2 November 2004), para 55. On the acceptability of amnesties, see Brice Dickson, ch 11 in this volume. 89 Okkali v Turkey (n 87) paras 75–76. 90 See, eg, Derman v Turkey App No 21789/02 (ECtHR, 31 May 2011), para 28. 91 See, eg, Böber v Turkey App No 62590/09 (ECtHR, 9 April 2013), para 35.

48  Laurens Lavrysen consequences’ line of reasoning was subsumed92 under the now-predominant ‘manifest disproportion’ line of reasoning to eventually wholly disappear in favour of the latter from 2015 onwards.93 While initially the Court thus seemed to struggle to find the ­appropriate language to address questions of inadequate sentencing for human rights offences, this search came to an end with the development of the ‘manifest disproportion’ line of reasoning. The first case in which the Court resorted to this line of reasoning was the Article 2 case of Nikolova and Velichkova in 2007. Taking into account the fact that ‘the Convention is intended to guarantee not theoretical or illusory, but practical and effective rights’, it held that: In cases of deaths occurring as a result of the use of excessive force, it must in particular verify whether the State has complied with its duty under Article 2 to secure the right to life by putting in place effective criminal-law provisions to deter the commission of offences against the person, backed up by law-enforcement machinery for the prevention, suppression and sanctioning of breaches of such provisions, and by not allowing life-endangering offences to go unpunished.94

For this reason, the Court considered that: [W]hile [it] should grant substantial deference to the national courts in the choice of appropriate sanctions for ill-treatment and homicide by State agents, it must exercise a certain power of review and intervene in cases of manifest disproportion between the gravity of the act and the punishment imposed. Were it to be otherwise, the States’ duty to carry out an effective investigation would lose much of its meaning, and the right enshrined by Article 2, despite its fundamental importance, would be ineffective in practice.95

In other words, as in Öneryildiz, the deterrence rationale is an important ­justification for the Court to impose an obligation to punish on the state. In addition, the Court invoked the principle of effectiveness as well as a purposive reading of the obligation to investigate in order to justify the existence of an obligation to punish. Later, this line of reasoning was transposed from Article 2 to Articles 396 and 897 and was also applied in horizontal situations,98 where the perpetrator is a private actor. The ‘manifest disproportion’ line of reasoning became fully established when it was endorsed by the Grand Chamber in the Article 3 case of Gäfgen in 2010.99

92 Kasap and Others v Turkey App No 8656/10 (ECtHR, 14 January 2014), paras 59–60. 93 The last time it was mentioned was in Ateşoğlu v Turkey App No 53645/10 (ECtHR, 20 January 2015), para 28. 94 Nikolova and Velichkova v Bulgaria App No 7888/03 (ECtHR, 20 December 2007), para 61. 95 ibid para 62. 96 See, eg, Atalay v Turkey App No 1249/03 (ECtHR, 18 December 2008), para 40. 97 See, eg, A v Croatia App No 55164/08 (ECtHR, 14 October 2010), para 67. 98 See, eg, Beganović v Croatia (n 48) para 78. 99 Gäfgen v Germany (n 31) para 123.

Positive Obligations and the Criminal Law  49 It is somewhat paradoxical that the obligation to punish has been clearly embedded within a remedial, victim-centred framework of procedural positive obligations – conferring on the victim a corresponding right-claim that the perpetrator of a human rights offence is adequately criminally punished100 – despite the fact that, as we have seen, by and large, general deterrence is the most important rationale adduced by the Court to justify the existence of the obligation to punish. The Öneryildiz case is particularly illustrative in this respect, as the Grand Chamber here, as mentioned above, emphasised the significance of the deterrent effect of the judicial system in preventing human rights offences. Besides being instrumental to the prevention of future human rights offences, the Grand Chamber simultaneously stressed other general social goals of punishment, in particular that it is ‘essential for maintaining public confidence and ensuring adherence to the rule of law and for preventing any appearance of tolerance of or collusion in unlawful acts’.101 For instance, the connection between the obligation to punish and the objective of general human rights protection, via the deterrence rationale, may arguably be stronger in cases involving a culture of impunity than in cases concerning what Seibert-Fohr labels as ‘singular incidents’.102 It must be admitted that, occasionally, the Court has also invoked specific deterrence and retribution as rationales for the obligation to punish. The former is particularly evident in the Court’s case law concerning domestic violence, where, as held above, in cases of a continuing situation of violence, the boundaries between procedural and substantive protection tend to blur. In Beganović, for example, the Court brought the obligation to punish within the framework of substantive protection against acts of domestic violence, holding ‘that the main purpose of imposing criminal sanctions is to restrain and deter the offender from causing further harm’.103 Interestingly, it also held that: [T]he obligation on the State to bring to justice perpetrators of acts contrary to Article 3 of the Convention serves mainly to ensure that acts of ill-treatment do not remain ignored by the relevant authorities and to provide effective protection against acts of ill-treatment.104

Thereby, the Court suggests that a criminal law response may be required in order for the state to adequately recognise the suffering of victims of human rights offences – a more victim-centred rationale than the general deterrence rationale. The retribution rationale, in turn, has been put forward most explicitly by the Court in Jelić, in which it noted ‘that among the main purposes of

100 Kamber (n 1) 5. 101 Öneryildiz v Turkey (n 13) para 96. 102 A Seibert-Fohr, Prosecuting Serious Human Rights Violations (Oxford, Oxford University Press, 2009) 140 and 202. See also Kelly M Pitcher, ch 13 in this volume. 103 Beganović v Croatia (n 48) para 85. 104 ibid para 79.

50  Laurens Lavrysen imposing criminal sanctions are retribution as a form of justice for victims and general deterrence aimed at prevention of new violations and upholding the rule of law’.105 The Court here thus complements general deterrence with a victim-centred conception of retribution – considering criminal punishment as instrumental to the vindication of the victim’s value and dignity, which have been damaged by the human rights offence106 – as a rationale for imposing an obligation to punish on the state. B.  The Scope of the Obligation to Punish In this section, I will provide a brief overview of how the Court has applied the obligation to punish in practice. At the outset, it must be noted that the Court has always emphasised that the determination of whether there was a ‘manifest disproportion between the gravity of the act and the punishment imposed’ requires granting ‘substantial deference to the national courts in the choice of appropriate sanctions’.107 A survey of the Court’s case law allows us to draw a number of conclusions on how it determines whether the state, within its very wide margin of appreciation, has complied with the obligation to punish. The most commonly applied approach by the Court to determine whether a perpetrator was not adequately sentenced is to exercise a procedural review of the reasoning provided by the domestic authorities, as opposed to a more substantive type of review in which the Court steps in to determine for itself what constitutes an appropriate sentence for a particular human rights offence. This is in line with a broader move towards subsidiarity in the case law, in which the Court shows a ‘willingness to defer to the reasoned and thoughtful assessment by national authorities of their Convention obligations’.108 In a number of cases, the Court, for example, found a violation on account of the use of sham reasons by the domestic courts, in particular when applying mitigating circumstances in order to reduce or suspend the sentence or to suspend the pronouncement of the judgment altogether109 – which arguably could be considered as fostering a culture of impunity. This was, for instance, the case when the domestic court held that the defendants had cooperated in the investigation while they had not made any statements other than persistently denying the

105 Jelić v Croatia App No 57856/11 (ECtHR, 12 June 2014), para 90. 106 Kamber (n 1) 57. See also Alina Balta, ch 4 in this volume. 107 See, eg, Nikolova and Velichkova v Bulgaria (n 94) para 62. 108 R Spano, ‘Universality or Diversity of Human Rights? Strasbourg in the Age of Subsidiarity’ (2014) 14 Human Rights Law Review 487, 491. See also L Huijbers, ‘The European Court of Human Rights’ Procedural Approach in the Age of Subsidiarity’ (2017) 6 Cambridge International Law Journal 177. 109 The Court has explicitly acknowledged this in Cestaro v Italy (n 12) para 224.

Positive Obligations and the Criminal Law  51 allegations against them,110 that they had behaved well during the trial while they had not attended a single hearing,111 that they had been awarded medals for excellent police service and had positive references from their superiors while they had committed prolonged acts of torture,112 that the victim’s conduct had provoked the ill-treatment while the abuse continued at the police station,113 and that the acts complained of were ‘manifestly unimportant’ and did not show the ‘requisite social danger’ while they had caused severe injuries to the victim.114 The same applies the other way around, as the Court will not find a violation of the obligation to punish if it considers the reasons provided for a particular sentence to be convincing. For example, in Leparskiene, the Court found no violation of Article 2 chiefly because the domestic courts had given ‘substantial reasoning’115 for only imposing a suspended prison sentence for manslaughter on a police officer who had shot and killed the applicant’s son – in particular, the fact that he had aimed at the tyres of the car in which the son was seated but accidentally hit the son instead because the car was moving fast. Another example is the case of Irene Wilson, in which the Court did not find a violation of Article 8 on account of the fact that only a suspended prison sentence was imposed for an isolated event of domestic violence. According to the Court, this was based on a careful assessment of all the evidence and ‘there was merit in the sentencing judge’s approach’, as suspending the sentence for three years ‘deterred the applicant’s husband from any further violent behaviour towards her’ during this period, which according to the Court ‘arguably gave the applicant longer and better protection from her husband than imprisoning him immediately would have done’.116 The same holds true for the Court’s assessment of prosecutorial decisions, where it has accepted decisions not to prosecute when they were based on convincing reasons, such as those relating to the insufficiency of evidence,117 to the application of the self-defence test under domestic law118 or the fact that a death during a police operation had been caused by institutional failings rather than by conduct for which individual state agents ought to have been held criminally liable.119 It must be stressed that the procedural review exercised by the Court is often infused with certain substantive considerations. In particular, the reasoning of the domestic courts must show that due weight has been given to the gravity 110 Ali and Ayşe Duran v Turkey App No 42942/02 (ECtHR, 8 April 2008), para 68. 111 Zeynep Özcan v Turkey App No 45906/99 (ECtHR, 20 February 2007), para 43. 112 Kopylov v Russia App No 3933/04 (ECtHR, 29 July 2010), para 141. 113 Atalay v Turkey (n 96) para 43. 114 Doiciu v Romania App No 1454/09 (ECtHR, 5 May 2015), para 68. 115 Leparskiene v Lithuania App No 4860/02 (ECtHR, 7 July 2009), para 53 116 Irene Wilson v UK (n 10) para 50. 117 See also Dickson, ch 11 in this volume, with reference to Gürtekin and Others v Cyprus App No 60441/13 etc (ECtHR, 11 March 2014,), paras 27–28. 118 On condition that this test is compatible with ECHR standards; see Armani Da Silva v UK (n 71) paras 244–56. On this case, see also Natasa Mavronicola, ch 9 in this volume. 119 Armani Da Silva v UK (n 71) paras 283–86.

52  Laurens Lavrysen of the human rights offence. This can a contrario be inferred from the case of Cestaro, in which the Court held that it could not ‘criticise the domestic courts for having wrongly assessed the gravity of the charges against the accused’.120 In the torture case of Sidiropoulos and Papakostas, the Court also criticised the domestic courts for having failed to consider whether the imposed sentence was capable of deterring the perpetrator or other state agents from committing similar offences or whether it could be perceived as fair by the victims121 – the latter consideration being a rare example of the Court invoking genuine retributive considerations when assessing the adequacy of criminal sentences. Only rarely has the Court provided clear substantive guidelines concerning the requisite severity of a criminal sentence – it has generally refrained from making abstract statements on these matters, as it considers that the ‘[a]ssessment of the adequacy of the punishment will … depend on the circumstances of the particular case’.122 However, under Article  3, the Court has ruled that the imposition of ‘mere fines’ must be considered as inadequate with respect to acts of torture and inhuman or degrading treatment.123 In addition, it ruled that ‘the imposition of enforceable prison sentences’ is the appropriate response in ‘cases concerning arbitrary and serious acts of brutality by State agents which the latter then attempted to conceal’.124 More generally, the Court also seems to require that the domestic courts are ‘all the more stringent’125 when punishing state agents in comparison to private actors – arguably requiring a sentence that, all things being equal, is more severe in the case of the former in order to discharge ‘the State’s duty to combat the sense of impunity the offenders may consider they enjoy by virtue of their very office and to maintain public confidence in and respect for the law-enforcement system’.126 On an overall assessment of the doctrine, it seems fair to conclude that in general, the Court’s predominantly procedural, yet substance-infused approach to assessing the adequacy of criminal sentences for human rights offences does indeed leave ‘substantial deference’ to the domestic courts. It must be noted that the Court could sometimes do a better job in explaining how it conducts its review, as in a number of cases it simply jumps to conclusions that the assessment of the domestic courts was inadequate in relation to the severity of the criminal sanction without explaining why.127 Preferably, the Court should always sufficiently engage with the reasoning of the domestic courts in order to avoid being criticised for only paying lip service to the state’s wide margin of 120 Cestaro v Italy (n 12) para 224. Similarly, see Myumyun v Bulgaria (n 34) para 75. 121 Sidiropoulos and Papakostas v Greece App No 33349/10 (ECtHR, 24 January 2018), paras 94–95. 122 Cestaro v Italy App (n 12) para 208. 123 Myumyun v Bulgaria (n 34) para 75. 124 Gäfgen v Germany (n 31) para 124. 125 Enukidze and Girgvliani v Georgia App No 25091/07 (ECtHR, 26 April 2011), para 274. 126 ibid. 127 See, eg, Yeter v Turkey App No 33750/03 (ECtHR, 13 January 2009), paras 67–68; Külah and Koyuncu v Turkey App No 24827/05 (ECtHR, 23 April 2013), para 42.

Positive Obligations and the Criminal Law  53 appreciation in this area. Finally, it must be emphasised that a downside of the Court’s deferential approach is that it risks providing little guidance to domestic authorities. The case law shows that the Court more easily holds that a sentence falls below the minimum required under the Convention than that it indicates what that minimum would have been in the circumstances of the case. V. CONCLUSION

One of the main questions addressed in this volume is whether the European Court of Human Rights goes too far in its case law on positive obligations and the criminal law – in particular, whether it commits the sin of ‘coercive overreach’.128 In the absence of an objective standard to determine the appropriate role of the Court in this area, reasonable disagreement on this question may be inevitable. In any event, in this chapter I have attempted to map the Court’s case law in this area in order to enable an empirically informed discussion on these matters. While the Court should be careful not to heedlessly go down a slippery slope in this area, hitherto the coercive obligations derived from its case law are still fairly well contained. On the one hand, as far as the obligation to criminalise is concerned, the Court has by and large confined this to situations that are ‘normally’ regulated by the criminal law. On the other hand, it does seem to grant ‘substantial deference’ to domestic courts in sentencing matters. However, coercive human rights inevitably generate tensions with the orthodox concern of human rights law about the protection of defendants against abuses in the criminal justice system – Tulkens talks about a ‘paradoxical relationship’ between the ‘shield’ and ‘sword’ functions of human rights in the application of criminal law.129 Moreover, given the Court’s tendency to incrementally develop its jurisprudence,130 the risk that standards that have originally been developed in cases involving systemic failings in the criminal justice system or a culture of impunity are ultimately applied to micro-manage criminal processes regarding ‘singular incidents’ is real.131 For these reasons, any development in the case law in this area requires stronger and more principled justifications than those provided so far by the Court132 in order to contain the risk of unwarrantedly widening the web of social control in the name of human rights. 128 L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce?’ in L Zedner and J Roberts (eds), Principles and Values in Criminal Law and Criminal Justice (Oxford, Oxford University Press, 2012). See also Mavronicola, ch 9 in this volume. 129 Tulkens (n 81). 130 J Gerards, ‘Margin of Appreciation and Incrementalism in the Case Law of the European Court of Human Rights’ (2018) 18 Human Rights Law Review 495. 131 See also Seibert-Fohr (n 102) 140. 132 See also Peršak, ch 7 in this volume.

54

3 Positive Obligations and Coercion Deterrence as a Key Factor in the European Court of Human Rights’ Case Law PAUL LEMMENS AND MARIE COURTOY

I. INTRODUCTION

T

hrough its case law, the European Court of Human Rights (hereinafter ‘the Court’) stresses the importance of an adequate response by the competent authorities to serious violations of human rights. Under certain conditions, the required response may imply the criminal prosecution and punishment of the perpetrator.1 For the Court, the question is often whether a reaction of this nature is indeed required, from the point of view of the rights of the victim, or whether a non-criminal response may suffice. This book deals with the scope of this requirement.2 It has been argued that the assertion of ‘coercive duties’ carries the risk of ‘coercive overreach’.3 The Court is sometimes

1 For an excellent analysis of the relevant obligations under the European Convention on Human Rights (hereinafter ‘the Convention’), see K Kamber, Prosecuting Human Rights Offences: Rethinking the Sword Function of Human Rights Law (Leiden, Brill, 2017). 2 A mirror question is to what extent an individual (a suspect, accused or convicted person) is protected against the use of the criminal system. See P Pinto de Albuquerque, ‘The Overuse of Criminal Justice in the Case Law of the European Court of Human Rights’ in C van Kempen and M Jendly (eds), Overuse in the Criminal Justice System. Le recours excessif au système de justice pénale (Cambridge, Intersentia, 2019). 3 See L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce?’ in L Zedner and J Roberts (eds), Principles and Values in Criminal Law and Criminal Justice: Essays in Honour of Andrew Ashworth (Oxford, Oxford University Press, 2012) 135, 149, quoted by Natasa Mavronicola in ch 9 in this volume.

56  Paul Lemmens and Marie Courtoy criticised for not setting proper limits to the duty to prosecute and punish, thus making it possible to ‘abuse’ the Court’s case law for the advancement of a political agenda based on security.4 The criticism addressed to the Court starts with the perceived lack of justification for the requirement of criminalisation. This is not an unusual accusation to be levelled against the Court, which is confronted with similar complaints in other areas as well. Without underestimating these complaints, it might be useful to recall that the Court is a judicial body which reaches its decisions – or at least its published decisions and judgments – in collegiate formations. Its decisionmaking is different from that of an academic who writes an article in which he or she can freely explore various issues and points of view before coming up with a conclusion. A court must usually be more limited in its approach. Furthermore, it might sometimes be preferable not to say too much in a judgment when every argument may open up new discussions. In this chapter, we will try to explain the position of the Court in order to clarify its perspective on positive obligations and coercion. While our focus is on the Court’s case law, our opinions may nonetheless be somewhat subjective, as we are obliged to partly base our findings on an interpretation of the case law. The opinions expressed in this chapter therefore cannot be understood as necessarily representing those of the Court. II.  GENERAL CONTEXT: THE POSITIVE OBLIGATION TO PROTECT

While the object of most articles of the European Convention on Human Rights is essentially that of protecting the individual against arbitrary interference by the public authorities, they do not merely compel the state to abstain from such interferences; in addition to this primarily negative undertaking, there may be positive obligations inherent in an effective respect for the rights that are guaranteed by the Convention.5 As a result, states can be held directly responsible for acts committed by their agents and indirectly responsible for acts committed by private individuals or entities, namely if the competent state organs did not take reasonable measures to prevent the commission of those acts. The obligation to protect against acts by private persons thus implies a general obligation of prevention. This obligation to prevent obviously entails in the first place an obligation of a substantive nature, namely the obligation to set up ‘a legislative framework taking into consideration the various interests to be protected in a particular context’.6 As will be explained below, in some cases the framework must be



4 Lazarus

(n 3) 149–51. Marckx v Belgium Series A No 31 (1979), para 31. 6 Bărbulescu v Romania [GC] App No 61496/08 (ECtHR, 5 September 2017), para 115. 5 See

Positive Obligations and Coercion  57 designed in such a way as ‘to provide effective deterrence against threats’ to fundamental rights,7 whereby deterrence is understood as ‘the action of discouraging an action or event through instilling doubt or fear of the consequences’.8 This obligation may go as far as obliging the legislature to provide for criminal law provisions. In cases where the nature of the threat to fundamental rights is not such that it warrants criminal law provisions, other measures may be more relevant. Whatever the nature of the legislative framework, legislation is not enough. The positive obligation to secure for citizens the right to effective respect for their human rights includes the provision of enforcement machinery protecting their rights.9 In other words, the legal framework must be ‘backed up’ by lawenforcement machinery.10 The ensuing procedural obligation for the competent state authorities to apply this machinery is intended to ensure the effectiveness of the substantive legal rules in practice.11 Deterrent legislation is indeed aimed at generally preventing violations of human rights. However, the deterrent effect of a law will be without substance if the law is not consistently applied whenever there is a serious breach of an individual’s human rights. To the extent that the Convention thus implies an obligation to criminalise certain acts, one should be careful not to misunderstand its rationale. Criminalisation is linked to the prevention of certain serious violations of human rights. It is not to be seen as a form of implementation of the so-called ‘right to security’, based on a sort of presumption that this right must be realised in order to enjoy all other rights, which can therefore be invoked to justify any coercive measure.12 Nor does the Court’s case law call for strict law enforcement in all circumstances.13 The objective is only to have in place legislation with a deterrent effect, followed by concrete measures showing that the authorities are not prepared to allow the offences to go unpunished,14 in order to prevent the commission of particularly wrongful acts. 7 See, with respect to threats to the right to life, Öneryildiz v Turkey [GC] ECHR 2004-XII, para 89; Nicolae Virgiliu Tănase v Romania [GC] App No 41720/13 (ECtHR, 25 June 2019), para 135. 8 See the definition of ‘deterrence’ in the Lexico online dictionary: https://www.lexico.com/ definition/deterrence. 9 A, B and C v Ireland [GC] ECHR 2010, para 245. 10 See, with respect to enforcement machinery of a criminal law nature, among others, Osman v UK Reports 1998-VIII, para 115; Mustafa Tunç and Fecire Tunç v Turkey [GC] App No 24014/05 (ECtHR, 14 April 2015), para 171. 11 Compare Giuliani and Gaggio v Italy [GC] ECHR 2011 (extracts), para 298. 12 See the criticism put forward by Liora Lazarus, who denounces a twofold pernicious movement around the promotion of a ‘right to security’: on the one hand ‘righting security’, where security is used as a means of legitimation for strong coercive measures, and, on the other hand, ‘securitising rights’, where there is a shift in the hierarchy of human rights in favour of the right to security as a basic right conditioning the enjoyment of other rights: L Lazarus, ‘The Right to Security’ in R Cruft, M Liao and M Renzo (eds), The Philosophical Foundations of Human Rights (Oxford, Oxford University Press, 2015). 13 See, with respect to the discretion to prosecute or not to prosecute a given individual, Armani Da Silva v UK [GC] App No 5878/08 (ECtHR, 30 March 2016), paras 258–61. 14 See Öneryildiz (n 7) para 96; Giuliani and Gaggio (n 11) para 306; Armani Da Silva (n 13) para 239.

58  Paul Lemmens and Marie Courtoy Finally, it should be noted that there is an ‘all the more stringent’ obligation to react against crimes committed by state agents.15 Indeed, ‘what is at stake is not only the issue of the individual criminal-law liability of the perpetrators but also the state’s duty to combat the sense of impunity the offenders may consider they enjoy by virtue of their very office and to maintain public confidence in and respect for the law-enforcement system’.16 States must avoid any appearance of tolerance or collusion between security forces and the prosecution, particularly where the investigation is in the hands of the police.17 While there is therefore an additional rationale for the criminalisation of certain acts committed by state agents, the obligation to criminalise remains primarily related to the effectiveness of the legal framework. III.  THE SUBSTANTIVE OBLIGATION TO CRIMINALISE CERTAIN ACTS

Let us return to the substantive obligation to provide for legislation that criminalises certain acts. When does such an obligation follow from the Convention? When the Court describes the circumstances in which such an obligation may exist, it uses restrictive language. This is illustrated by its judgment in X and Y v The Netherlands: This is a case where fundamental values and essential aspects of private life are at stake. Effective deterrence is indispensable in this area and it can be achieved only by criminal-law provisions; indeed, it is by such provisions that the matter is normally regulated. (Emphasis added)18

There are three elements in this statement which deserve a closer look. First of all, in X and Y v The Netherlands, ‘fundamental values’ or ‘essential aspects of private life’ were at stake. If we slightly generalise this aspect, we could say that criminal law provisions are required (only) when fundamental values or essential aspects of human dignity are at stake. On other occasions, the Court has referred to ‘fundamental values of democratic societies’.19 The right to life and the prohibition of torture and inhuman or degrading treatment (Articles 2 and 3 of the Convention) come immediately to mind.

15 Enukidze and Girgvliani v Georgia App No 25091/07 (ECtHR, 26 April 2011), para 274; ­Vazagashvili and Shanava v Georgia App No 50375/07 (ECtHR, 18 July 2019), para 92. See also Armani Da Silva (n 13) para 234. 16 Enukidze and Girgvliani (n 15) para 274; Vazagashvili and Shanava (n 15) para 92. 17 Tahsin Acar v Turkey [GC] ECHR 2004-III, para 224; Öneryıldız (n 7) para 96. 18 X and Y v The Netherlands Series A No 91 (1985), para 27. 19 Siliadin v France ECHR 2005-VII, para 148; Sandra Janković v Croatia App No 38478/05 (ECtHR, 5 March 2009), para 47.

Positive Obligations and Coercion  59 These rights are indeed so fundamental that part of the legislative framework protecting them must consist of criminal law provisions.20 The obligation to criminalise was later also recognised with respect to slavery, servitude and forced labour, which are prohibited by Article  4 of the Convention.21 Likewise, the protection of the physical and moral integrity of a person, covered by the notion of ‘private life’ in Article 8, against grave acts, such as rape, requires the adoption of criminal law provisions.22 Second, criminal law provisions are required in order to provide ‘effective deterrence’. When is a measure a deterrent? The idea is that the measure should be such that potential perpetrators will fear the consequences of their acts.23 Of course, to assess (normally retrospectively) the deterrent effect of a given provision will involve some degree of speculation. The Court is not in a position to undertake any empirical study in this respect.24 Nor can it argue on the basis of a given criminological theory.25 What the Court does in its assessment of whether the state provided for deterrent legislation is a sort of retrospective analysis, on the basis of what actually happened after an incident constituting an alleged violation of a substantive right: how effective was the deterrent effect of protective legislation, knowing that, notwithstanding the existence of that legislation, a violation of fundamental rights occurred? X and Y v The Netherlands was about a young girl with a mental disability, who was sexually abused in the privately run institution where she lived. At that time in the Netherlands, there was legislation criminalising violations of the sexual integrity of a mentally disabled person; the authorities had also carried out a proper investigation of the girl’s case. However, it appeared that there was a gap in the law: the court dealing with

20 See, among many others, with respect to the right to life, Osman (n 10) para 115; Kılıç v Turkey ECHR 2000-III, para 62; Mahmut Kaya v Turkey ECHR 2000-III, para 85; and with respect to protection against grave forms of ill-treatment, Beganović v Croatia App No 46423/06 (ECtHR, 25 June 2009), para 56; Gäfgen v Germany [GC] ECHR 2010, para 117; Đurđević v Croatia ECHR 2011 (extracts), para 51; O’Keeffe v Ireland [GC] ECHR 2014 (extracts), para 148. 21 Siliadin (n 19) para 89. 22 See, among many others, X and Y v The Netherlands (n 18) paras 22 and 27; Söderman v Sweden [GC] ECHR 2013, para 82; López Ribalda and Others v Spain [GC] App Nos 1874/13 and 8567/13 (ECtHR, 17 October 2019), para 113. By contrast, ‘in respect of less serious acts between individuals, which may violate psychological integrity, the obligation of the State under Article 8 to maintain and apply in practice an adequate legal framework affording protection does not always require that an efficient criminal-law provision covering the specific act be in place. The legal framework could also consist of civil-law remedies capable of affording sufficient protection’ (Söderman (n 22) para 85; see also López Ribalda and Others (n 22) para 113). 23 See section II above. 24 It should be noted that in ch 7 in this volume, Nina Peršak invites the Court to be more demanding and to require persuasive evidence of the deterrent effect of criminal law measures. 25 There may be different views concerning criminology and it is not the role of the Court to privilege one of them over the others. As we will see later, penal policy is in principle a matter for the national authorities (see section V below).

60  Paul Lemmens and Marie Courtoy the case held that a victim who was legally incapacitated was unable to file a criminal complaint herself (which would trigger criminal proceedings against the alleged perpetrator) and that there was no possibility for a third party to act on behalf of the victim. It might have been possible for the government to argue that, notwithstanding this gap, the law was sufficiently deterrent in this respect since the gap was unknown before the incident occurred and before the girl’s case was examined by the court; potential perpetrators could therefore be considered to generally believe that if they violated the law, they would be prosecuted. However, the Court held that since no criminal prosecution was possible, the law did not provide the victim with ‘practical and effective protection’.26 In other words, because of the gap with respect to legally incapacitated victims, the law was not sufficiently deterrent. The outcome in this case seems to suggest that the question whether or not a provision is capable of deterring potential perpetrators must be answered by relying not only on that provision’s wording in abstracto, but that account should also be taken of what happened in concreto when the application of that provision has been sought in proceedings against the actual perpetrator. However, since it is deterrence that matters, the fact that the alleged perpetrator has in the end not been convicted does not necessarily mean that the legal framework as such is insufficient. Nevertheless, in such a situation, it remains to be seen whether the response in the particular case was sufficient to maintain the deterrent effect of the law. This issue will be examined below.27 Furthermore, it is important to emphasise that deterrence can, and sometimes should, be achieved through non-criminal law measures, either alone or in combination with criminal law measures. For instance, legislation that provides for the administrative inspection of dangerous activities and for administrative sanctions (such as the withdrawal of a licence) for those who do not comply with the applicable safety rules can be deterrent to such an extent that there appears to be no need for the application of criminal law measures.28 Moreover, a criminal law framework may not always be sufficient. In Rantsev v Cyprus and Russia, the issue was whether Cyprus had an appropriate legislative and administrative framework to prevent human trafficking and sexual exploitation, activities falling within the scope of Article  4 of the Convention. The Court found that there was legislation prohibiting these activities and that it provided for severe penalties.29 One could say that the relevant provisions satisfied the deterrence requirement. The Court nevertheless found that this framework was not sufficient to afford practical and effective protection to victims of trafficking and exploitation. In particular, it pointed to weaknesses in the Cypriot immigration policy with respect to cabaret artistes, following the Council of Europe 26 X and Y v The Netherlands (n 18) paras 29–30. 27 See the analysis of the obligation to investigate in section IV below. 28 This does not mean that criminal law provisions are without importance; the application of these provisions can be required as a measure of last resort. 29 Rantsev v Cyprus and Russia ECHR 2010 (extracts), para 290.

Positive Obligations and Coercion  61 Commissioner for Human Rights in considering that ‘the system whereby ­cabaret managers were required to make the application for an entry permit for the artiste [was] rendering the artiste dependent on her employer or agent and increasing her risk of falling into the hands of traffickers’.30 This undermined the effectiveness in an inter-state context of the measures against trafficking and exploitation.31 One could say that the very purpose of deterrence could not be attained. Third and finally, it should be noted that in X and Y v The Netherlands, the Court indicated that the case concerned a situation which is ‘normally’ regulated by criminal law. It may be possible to read this as an implicit reference to consensus. If there is a consensus among the Member States of the Council of Europe to ­criminalise certain acts that infringe a person’s human rights, this may be considered a strong indication of the existence of an obligation for all states to do so. Conversely, a consensus not to criminalise certain acts is a strong indication of the absence of such an obligation. The existence of a consensus, in one direction or another, can be established by the comparison of domestic laws, as well as by reference to international instruments which are widely adhered to. IV.  THE PROCEDURAL OBLIGATION TO INVESTIGATE, TO PROSECUTE AND TO PUNISH

As indicated above, prevention requires that there is not only an adequate legal framework, but also an effective law-enforcement machinery.32 The procedural obligation to react to violations of human rights – that is, to put the law-enforcement machinery in motion – can thus be seen as an element of the obligation to protect, more specifically as an element of the obligation to prevent the commission of acts infringing the fundamental rights of individuals. In situations where the legislation was apparently unable to prevent the commission of such acts, an ensuing investigation into these acts and any judicial proceedings, including the sanctions imposed, must ensure the deterrent effect of the judicial system in place so that the significance of the role it is required to play in preventing violations of human rights is not undermined.33 The judicial response thus forms part of the so-called ‘procedural limb’ of Convention articles (Articles 2, 3, 4 and 8) which are aimed in the first place at guaranteeing a substantive protection.34 30 ibid para 291. 31 ibid paras 291–93. 32 See the discussion in section II above. 33 Öneryildiz (n 7) para 96; Gäfgen (n 20) para 121; Giuliani and Gaggio (n 11) para 306; Armani Da Silva (n 13) para 239; Jeronovičs v Latvia [GC] ECHR 2016, para 106. 34 A procedural obligation also follows from those articles of the Convention which provide for a right of access to a court in the event of an alleged interference with civil rights (art 6(1)) or for

62  Paul Lemmens and Marie Courtoy The judicial response does not always have to be of a criminal law nature. This issue is obviously linked to the question of whether a legal framework including criminal law provisions is required in order to prevent the act constituting the human rights violation at hand.35 A criminal investigation is generally necessary where death has been caused intentionally (Article 2)36 or where treatment amounting to inhuman or degrading treatment or torture37 has been inflicted (Article 3). In such cases, the authorities must act on their own motion once the matter has come to their attention.38 By contrast, in cases concerning unintentional infliction of death or lives being put at risk unintentionally, the procedural obligation to set up an effective judicial system does not necessarily require criminal proceedings to be brought. In such cases, a remedy for the victim in the civil courts, either alone or in conjunction with a remedy in the criminal courts, may be sufficient. In ­appropriate cases, disciplinary measures may also be envisaged.39 Where a criminal investigation is required, there is in the first place an obligation for the competent authorities to establish the facts. The authorities need to investigate what has actually happened and who was involved in the crime. In this way, the investigation is an implementation of ‘the right to the truth regarding the relevant circumstances of the case’.40 This right does not belong solely to the direct victim and/or his or her family; when the case has attracted a certain level of attention, it also belongs to other victims of similar crimes and the general public, who have the right to know what has happened. If the investigation yields indications that one or more individuals are criminally liable, a second obligation arises: the obligation to prosecute and, where guilt has been established, to punish. In the case law of the Court relating to the obligation to provide for a criminal law response to violations of human rights, most cases deal with the preliminary investigation stage of criminal proceedings. This is so since most problems a right to an effective domestic remedy in case of an alleged violation of human rights (art  13). The procedural obligation under these articles does not have the aim of preventing the commission of any acts; rather, it is about reparation of the consequences thereof (see, with respect to art 6(1), Nicolae Virgiliu Tănase (n 7) para 193). 35 See the discussion in section III above. 36 Mustafa Tunç and Fecire Tunç (n 10) para 170; Nicolae Virgiliu Tănase (n 7) para 158. 37 Only treatment that is the consequence of an ‘intentional act’ can be considered ill-treatment falling under art 3 (Nicolae Virgiliu Tănase (n 7) para 121). 38 ibid paras 164 (art 2) and 115 (art 3). 39 See, with respect to medical negligence, Calvelli and Ciglio v Italy [GC] ECHR 2002-I, para 51; Vo v France [GC] ECHR 2004-VIII, para 90; Šilih v Slovenia [GC] App No 71463/01 (ECtHR, 9 April 2009), para 194; Lopes de Sousa Fernandes v Portugal [GC] App No 56080/13 (ECtHR, 19 December 2017), para 137. See more generally Nicolae Virgiliu Tănase (n 7) para 159. However, it should be noted that where the alleged fault went beyond a mere error of judgement or carelessness, compliance with the procedural obligation must include recourse to criminal law (see, eg, Öneryildiz (n 7) para 93; Lopes de Sousa Fernandes, para 215; Nicolae Virgiliu Tănase (n 7) para 160). 40 El-Masri v Former Yugoslav Republic of Macedonia [GC] ECHR 2012, para 191; Al Nashiri v Poland App No 28761/11 (ECtHR, 24 July 2014), para 495; Husayn (Abu Zubaydah) v Poland App No 7511/13 (ECtHR, 24 July 2014), para 489.

Positive Obligations and Coercion  63 arise at this stage. In particular, the Court often has to examine whether the ­investigation was ‘effective’. The criteria for the effectiveness of an investigation were stated in Mustafa Tunç and Fecire Tunç v Turkey.41 They can be summarised as follows: • the investigation must first be adequate, which means that it must be c­ apable of leading to the establishment of the facts and, where appropriate, the ­identification and punishment of those responsible;42 • it must also be thorough, which means that the authorities must take all reasonable steps available to them to secure the evidence concerning the incident and that the investigation’s conclusions must be based on thorough, objective and impartial analysis of all relevant elements;43 • the persons responsible for the investigation should be independent of anyone implicated or likely to be implicated in the events – this means not only a lack of hierarchical or institutional connection, but also a practical independence;44 • the investigation must be started promptly and conducted with reasonable expedition;45 • finally, the investigation must be accessible to the victim or his or her family to the extent necessary to safeguard their legitimate interests, and there must also be a sufficient element of public scrutiny of the investigation.46 Generally speaking, the investigation must be submitted to ‘the careful s­ crutiny required’.47 This is all the more so when the infringement comes from state agents, as appropriate procedures for ensuring their accountability ‘are indispensable in maintaining public confidence and meeting the legitimate concerns that might arise from the use of … force’.48 41 See also the quotation of these principles in Güzelyurtlu and Others v Cyprus and Turkey [GC] App No 36925/07 (ECtHR, 29 January 2019), para 219. 42 Mustafa Tunç and Fecire Tunç (n 10) para 172. In the same sense, among others, Jeronovičs (n 33) para 103; Nicolae Virgiliu Tănase (n 7) para 165. 43 Mustafa Tunç and Fecire Tunç (n 10) paras 174–75. In the same sense, among others, Armani Da Silva (n 13) paras 233–34. Another way to describe this requirement is ‘that the authorities must always make a serious attempt to find out what happened and should not rely on hasty or illfounded conclusions to close their investigation or to use as the basis of their decisions’ (El-Masri (n 40) para 183; Bouyid v Belgium [GC] ECHR 2015, para 123; see also Mocanu and Others v Romania [GC] ECHR 2014 (extracts), para 325; Nicolae Virgiliu Tănase (n 7) para 166). 44 Mustafa Tunç and Fecire Tunç (n 10) para 177. In the same sense, among others, Armani Da Silva (n 13) para 232; Nicolae Virgiliu Tănase (n 7) para 168. 45 Mustafa Tunç and Fecire Tunç (n 10) para 178. In the same sense, among others, Bouyid (n 43) para 121; Armani Da Silva (n 13) para 237; Nicolae Virgiliu Tănase (n 7) para 167. 46 Mustafa Tunç and Fecire Tunç (n 10) para 179. In the same sense, among others, Armani Da Silva (n 13) para 235. 47 Öneryildiz (n 7) para 96; Giuliani and Gaggio (n 11) para 306; Centre for Legal Resources on Behalf of Valentin Câmpeanu v Romania [GC] ECHR 2014, para 133; Armani Da Silva (n 13) para 239. 48 See, among others, McKerr v UK ECHR 2001-III, para 160; Dimitrov and Others v Bulgaria App No 77938/11 (ECtHR, 1 July 2014), para 149.

64  Paul Lemmens and Marie Courtoy The obligation to conduct an effective investigation goes beyond the stage of the preliminary investigation. Where that investigation leads to the institution of proceedings before a trial court, ‘the proceedings as a whole, including the trial stage, must satisfy the requirements of the positive obligation to protect [certain fundamental values] through the law’.49 A good example of this is Öneryildiz v Turkey. This case concerned the explosion of a municipal rubbish tip, which caused the loss of many lives. The Court noted that the investigating authorities had generally acted as could be expected. However, when the case was subsequently brought before the criminal courts, the latter examined the responsibility of the defendants, two mayors, only from the point of view of the crime of ‘negligence in the performance of their duties’ and sentenced them to ‘derisory fines’, ‘[leaving] in abeyance any question of the authorities’ possible responsibility for the death of the applicant’s … relatives’.50 The Court concluded that the Turkish criminal justice system had not ‘secured the full accountability of State officials or authorities for their role in [the tragedy] and the effective implementation of provisions of domestic law guaranteeing respect for the right to life, in particular the deterrent function of the criminal law’.51 The Court thus exercises a certain power of review on the sentence imposed and it can intervene where there is a manifest disproportion between the gravity of the act and the punishment imposed.52 A disproportionately low sentence may destroy the deterrent effect of the relevant criminal law provision. While punishment, as the last stage of criminal proceedings, may be relevant for the assessment of the effectiveness of a criminal investigation (in the broad sense of the word), the Court adds that it should in no way be inferred from this ‘that Article  2 may entail the right for an applicant to have third parties prosecuted or sentenced for a criminal offence or an absolute obligation for all prosecutions to result in conviction, or indeed in a particular sentence’.53 In fact, the Convention does not exclude that there are circumstances that can be a legitimate bar to prosecution or sentencing. For instance, the decision of the prosecutor not to prosecute an individual offender does not in itself constitute a breach of the procedural obligation under the Convention, provided that there are no ‘institutional deficiencies’ in the 49 Öneryildiz (n 7) para 95; Armani Da Silva (n 13) para 239. 50 Öneryildiz (n 7) para 116. 51 ibid para 117. 52 Gäfgen (n 20) para 123; Armani Da Silva (n 13) para 238. Indeed, if the sanction actually imposed on the perpetrator is manifestly disproportionate to the violation of the victim’s rights, there is reason to believe that the relevant criminal provision and the law-enforcement machinery did not deter the perpetrator and that they are in any event not capable of deterring others in a similar situation (see, eg, Gäfgen (n 20) para 124; Austrianu v Romania App No 16117/02 (ECtHR, 12 February 2013), para 74; Sidiropoulos and Papakostas v Greece App No 33349/10 (ECtHR, 25 January 2018), para 95). 53 Öneryildiz (n 7) para 96; Giuliani and Gaggio (n 11) para 306; Armani Da Silva (n 13) para 238.

Positive Obligations and Coercion  65 criminal justice or prosecutorial system.54 A prosecutor can thus decide not to prosecute – and to spare the suspect the financial and emotional costs of a trial – where there is not enough evidence to provide a realistic prospect of conviction.55 A sometimes delicate question is to what extent the discontinuation of criminal proceedings because of the entry into force of a statute of limitations affects the effectiveness of the investigation. It is obvious that where the prosecution has become time-barred because of the authorities’ failure to show diligence, the purpose of effective deterrence is not achieved and the procedural obligation is not fulfilled.56 Where the prosecution becomes time-barred without there being any failure on the part of the authorities, this does not necessarily affect the deterrent effect of the applicable criminal law. The answer to the question thus depends on the circumstances. Particular attention must be paid to situations involving state agents. As stated by the Court, ‘where a State agent has been charged with crimes involving torture or ill-treatment, it is of the utmost importance that criminal proceedings and sentencing are not time-barred and that the granting of an amnesty or pardon should not be permissible’.57 With such a statement, the Court stresses that it is important that the competent authorities do not give the impression that they are willing to allow grave violations of human rights to go unpunished.58 Where the offence committed by the state agent has become time-barred before the end of the proceedings, the authorities’ response will therefore be considered unsatisfactory in view of the gravity of the facts.59 An amnesty or a pardon would be even worse: when a state agent is convicted of a crime that violates Article  2 or 3 of the Convention, ‘the granting of an amnesty or pardon can scarcely serve the purpose of an adequate punishment’.60 Indeed: [T[he Court expects States to be all the more stringent when punishing their own law-enforcement officers for the commission of such serious … crimes than they are with ordinary offenders, because what is at stake is not only the issue of the individual criminal-law liability of the perpetrators but also the State’s duty to combat the sense of impunity the offenders may consider they enjoy by virtue of their very office, and to maintain public confidence in and respect for the law-enforcement system.61

54 Armani Da Silva (n 13) paras 259–61. 55 ibid paras 265–76, in particular paras 265–66. 56 See, among many others, Beganović (n 20) paras 85–86; PM v Bulgaria App No 49669/07 (ECtHR, 24 January 2012), paras 66–67; VK v Russia App No 68059/13 (ECtHR, 7 March 2017), para 189. 57 See, eg, Abdülsamet Yaman v Turkey App No 32446/96 (ECtHR, 2 November 2004), para 55; Ali and Ayşe Duran v Turkey App No 42942/02 (ECtHR, 8 April 2008), para 69; Marguš v Croatia [GC] ECHR 2014 (extracts), para 126. 58 Egmez v Cyprus ECHR 2000-XII, para 71; Turan Cakir v Belgium App No 44256/06 (ECtHR, 10 March 2009), para 69; Marguš (n 57) para 126. 59 Cestaro v Italy App No 6884/11 (ECtHR, 7 April 2015), paras 218–26. 60 Enukidze and Girgvliani (n 15) para 274. 61 Enukidze and Girgvliani (n 15); Vazagashvili and Shanava (n 15).

66  Paul Lemmens and Marie Courtoy V.  THE CONVENTION AS A LIMITATION OF NATIONAL CRIMINAL POLICY?

Under the Convention, the obligation to have in place criminal law provisions and to provide for an effective criminal investigation, including punishment of the perpetrator, is a specific aspect of the general obligation to prevent serious human rights violations. Prevention may be achieved through the deterrent effect of criminal law. The Convention may be violated where there are no criminal law provisions prohibiting a given conduct or where the sanctions are insufficient. The Convention may also be violated where there is an ineffective criminal law response to an incident involving the violation of some core fundamental rights. Does the Court’s case law give rise to a fear of ‘coercive overreach’ or ‘expansive criminalisation’? We leave it to others to give an answer to this question. It may be sufficient for us to repeat that the obligation to provide for the criminalisation and punishment of certain acts is linked to the prevention of such acts. It is not a free-standing obligation. Moreover, the Court’s findings very often depend heavily on the particular circumstances of the case. One should therefore be cautious not to draw unduly broad conclusions from the Court’s judgments. There is also another question, namely whether the Convention sets limits to the power of the states to make certain acts illegal and to punish those who have committed such acts. As in other areas covered by the Convention, the domestic authorities enjoy a margin of appreciation in designing their penal policy,62 in particular in deciding whether a given conduct must be criminalised and whether certain acts must be investigated, prosecuted and punished. The Court’s task is not to take their place, but only to review the decisions adopted by them. The national margin of appreciation extends to determining what is the appropriate sentence for a particular crime.63 The legislature can in principle opt for tougher criminal law provisions by increasing the penalties applicable to criminal offences. Likewise, the prosecuting authorities and the courts can opt for stricter law enforcement. A harsh criminal policy does not, as such, raise an issue under the Convention.64 However, the power of the states is not unlimited. The Convention comes into play, in particular, when a sanction is imposed on an individual who has exercised a human right guaranteed by the Convention, such as the right to 62 See, among others, Clift v UK App No 7205/07 (ECtHR, 13 July 2010), para 73; Khoroshenko v Russia [GC] ECHR 2015, para 132; Khamtokhu and Aksenchik v Russia [GC] App Nos 60367/08 and 961/11 (ECtHR, 24 January 2017), para 85; Alexandru Enache v Romania App No 16986/12 (ECtHR, 3 October 2017), para 72. 63 Vinter and Others v UK [GC] ECHR 2013 (extracts), para 105. 64 Achour v France [GC] ECHR 2006-IV, para 44; Maktouf and Damjanović v Bosnia and ­Herzegovina [GC] ECHR 2013 (extracts), para 75; Del Río Prada v Spain [GC] ECHR 2013, para 116.

Positive Obligations and Coercion  67 respect for private life or freedom of expression. In such a case, the interference with that person’s right must have a legal basis, it must pursue a legitimate aim and it may not be disproportionate to that legitimate aim. Sometimes it is the legislature’s choice to criminalise a certain conduct that is considered ­disproportionate per se.65 Where there would be a conflict between two human rights, the one requiring criminal law protection and the other limiting the state’s possibility to interfere, a fair balance between the competing rights must be found. All this is plain application of the principles of the Convention. Furthermore, it is true that, exceptionally, the Court has stated that only a given criminal policy could be considered compatible with the Convention, with the consequence that the application of a different domestic policy would be held to violate the Convention. This was the case where the policy was considered to be incompatible with the applicant’s human dignity. In particular, with respect to imprisonment,66 the Court at one point noted that while punishment remained one of the aims of imprisonment, European penal policy had evolved towards according increasing importance to the rehabilitative aim of imprisonment.67 Later it went even further and considered that ‘the emphasis on rehabilitation and reintegration has become a mandatory factor that the member States [of the Council of Europe] need to take into account in designing their penal policies’.68 However, it is difficult to imagine that the Court would consider that a given sentencing policy is per se incompatible with the Convention. To sum up, it is generally within the power of the states to design their penal policy, including their policy with respect to criminalisation and sentencing. The Convention sets certain limits, but generally leaves a wide margin of appreciation to the competent authorities. To some extent, when fundamental values or essential aspects of human dignity are at stake, the Convention ‘pushes’ the states to adopt measures in the criminal law area. However, the obligation following in this respect from the Convention is to be read in light of a particular context: the need to provide for a deterrent effect against certain acts. Punishment of the offender is not as such a positive obligation imposed by the Convention.

65 See, eg, Dudgeon v UK Series A No 45 (1981), para 60. For other examples, see Pinto de Albuquerque (n 2) 68–69. 66 See R Spano, ‘Deprivation of Liberty and Human Dignity in the Case-Law of the European Court of Human Rights’ (2016) Bergen Journal of Criminal Law and Criminal Justice 150. 67 Dickson v UK [GC] ECHR 2007-V, para 75; Vinter and Others (n 63) para 115; Harakchiev and Tolumov v Bulgaria ECHR 2014 (extracts), para 264; Murray v The Netherlands [GC] App No 10511/10 (ECtHR, 26 April 2016), para 101. 68 Khoroshenko (n 62) para 121. See also Hutchinson v UK [GC] App No 57592/08 (ECtHR, 17 January 2017), para 43.

68

Part II

Perspectives on Victims’ Protection and Redress

70

4 Retribution through Reparations? Evaluating the European Court of Human Rights’ Jurisprudence on Gross Human Rights Violations from a Victim’s Perspective ALINA BALTA

A

s pointed out in other chapters in this collection,1 the European Court of Human Rights (hereinafter ECtHR or ‘the Court’) appears to place certain obligations on states in its case law, to protect the European Convention on Human Rights (ECHR) rights by recourse to criminal law. These obligations are either of a substantive nature, eg, obligations of states to criminalise certain human rights violations, or of a procedural nature, referring, for instance, to the obligations for states to prosecute and punish perpetrators.2 Amid criticism that the Court, by placing such obligations on states, may be guilty of ‘coercive overreach’,3 this chapter looks at these obligations from a victim’s perspective and aims to scrutinise their implications for victims at the ECtHR. To be precise, the chapter will explore whether the provision of such obligations may contribute to a sense of reparative justice for the victims within the ECtHR’s jurisdiction.

1 See various chapters within this volume, including those by Laurens Lavrysen and Natasa Mavronicola (chs 2 and 9 respectively). 2 For an elaboration on substantive and procedural obligations, see Lavrysen, ch 2 in this volume. 3 L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce?’ in L Zedner and J Roberts (eds), Principles and Values in Criminal Law and Criminal Justice (Oxford, Oxford University Press, 2012).

72  Alina Balta The chapter is structured as follows: in section I, it will be conceptualised that, at a theoretical level, these obligations can be viewed as a form of reparations for victims. Then, in section II, the chapter will investigate, from a victimological perspective, their implications for victims. In section III, the chapter will zoom in on the ECtHR’s reparations system and selected case law in order to investigate the context and content of the ECtHR’s provision of such obligations. In section IV, the Court’s rationale for including coercive obligations in its case law will be discussed, bearing in mind the implications for victims, discussed in section II. Conclusions with regard to the overall impact on victims are put forward at the end of the chapter. It is worth clarifying from the outset that this chapter is concerned with gross human rights violations, as defined for the purposes of the van Boven/Bassiouni Basic Principles on Reparations.4 Therein, the word ‘gross’ qualifies the term ‘violations’ and indicates the serious character of the violations, while at the same time it is also related to the type of human rights that are being violated, ie, civil and political rights. The serious character of the violation is to be considered having in mind crimes comparable to international crimes, such as genocide, crimes against humanity and war crimes.5 Against this background, the implications of Court-ordered obligations are considered from the perspective of the victims of these crimes.6 I.  RETRIBUTIVE RESPONSES AS REPARATIONS

Increasingly, human rights courts such as the ECtHR and the IACtHR have started to make recourse to criminal law in their case law and include obligations for states to criminalise certain human rights violations or obligations to prosecute and punish perpetrators.7 In the literature, it is asserted that these 4 Basic Principles and Guidelines on the Right to a Remedy and Reparation for Victims of Gross Violations of International Human Rights Law and Serious Violations of International Humanitarian Law, UNGA Res 60/147 (16 December 2005). 5 For more explanation of what gross human rights violations are considered to be for the purposes of the van Boven/Bassiouni Basic Principles on Reparations, see T van Boven, ‘The United Nations Basic Principles and Guidelines on the Right to a Remedy and Reparation for Victims of Gross Violations of International Human Rights Law and Serious Violations of International Humanitarian Law’ (2010) United Nations Audiovisual Library of International Law 2. See also Anja Seibert-Fohr, who explains that international human rights law protection of serious human rights violations is occasionally viewed as a legal basis to complement as well as fill in the gaps of international criminal law. She asserts that international human rights protection covers not only the international crimes outlined above – for which international criminal law already provides ­sanctions – but also serious human rights violations in general. See A Seibert-Fohr, Prosecuting ­Serious Human Rights Violations (Oxford, Oxford University Press, 2009) 3. 6 The chapter is a byproduct of research carried out for the author’s doctoral thesis, investigating the reparations regimes of the International Criminal Court (ICC), the Extraordinary Chambers in the Court of Cambodia (ECCC), the ECtHR and the Inter-American Court of Human Rights (IACtHR), and their contribution to reparative justice for victims of international crimes and gross human rights violations. 7 See eg, Benzer and Others v Turkey App No 23502/06 (ECtHR, 12 Nov 2013); Abuyeva and Others v Russia App No 27065/05 (ECtHR, 2 December 2010). For an analysis of the case law at

Retribution through Reparations?  73 courts started to turn into quasi-criminal jurisdictions, against a background of concerns regarding the accountability of states.8 The benefits of this approach include potential impact on local processes of justice, memory and judicial reform.9 Against this background, this section will investigate the potential implications of obligations such as those the ECtHR is placing on states in the context of reparations for victims of gross human rights violations. Across the chapter, these obligations will be referred to as ‘retributive responses’, as they appear to be infused with the retributive justice doctrine centred on prosecution and punishment.10 Furthermore, the reason for considering these obligations through the lens of reparations for victims of gross human rights violations is that, according to victim-based instruments, such obligations may amount to reparations.11 Finally, such measures, as awarded by the courts investigated in my doctoral project, fall within their reparations mandates, either directly or indirectly.12 Under international law, it is already established that, as a result of gross human rights violations, victims may be entitled to reparations.13 The Basic

the IACtHR, see FF Basch, ‘The Doctrine of the Inter-American Court of Human Rights Regarding States’ Duty to Punish Human Rights Violations and its Dangers’ (2007) 23 American University International Law Review 195. 8 AV Huneeus, ‘International Criminal Law by Other Means: The Quasi-criminal Jurisdiction of the Human Rights Courts’ (2013) 107 American Journal of International Law 1. 9 ibid 4. 10 Retributive justice has been related to the institution of criminal punishment. See DHJ Hermann, ‘Restorative Justice and Retributive Justice: An Opportunity for Cooperation or an Occasion for Conflict in the Search for Justice’ (2017) 16 Seattle Journal for Social Justice 71, 72. See also M Wenzel, TG Okimoto, NT Feather and MJ Platow, ‘Retributive and Restorative Justice’ (2007) 32 Law and Human Behavior 375. This article mentions that retributive justice essentially refers to the repair of justice through the unilateral imposition of punishment. The authors also explain that philosophers distinguish between consequentialist and retributivist justifications for retributive justice, which correspond to the motives of behaviour control (reducing the re-occurrence of the offence through incapacitation, deterrence or rehabilitation) and justice restoration, respectively; see also JVH Holtermann, ‘Can I Be Brought before the ICC? Deterrence of Mass Atrocities between Jus in Bello and Jus ad Bellum’, iCourts Working Paper Series, 154/2019, 7. 11 Basic Principles (2005), Updated Set of principles for the protection and promotion of human rights through action to combat impunity, ECOSOC E/CN.4/2005/102/Add.1 (8 February 2005). 12 See V Colandrea, ‘On the Power of the European Court of Human Rights to Order Specific Non-monetary Measures: Some Remarks in Light of the Assanidze, Broniowki and Sejdovic Cases’ (2007) 7 Human Rights Law Review 410; Huneeus (n 8) 2. The ICC, for instance, provides for the right to reparations under art 75 of the Rome Statute, which includes satisfaction and guarantees of non-repetition. On the other hand, the ECtHR’s reparations regime is, in theory, limited to just satisfaction under art 41 ECHR; however, as the research will show, it does provide for measures under art 46 ECHR, which can be considered as amounting to reparations. 13 Victims’ right to reparation is well established, inter alia, in public international law, human rights law and international criminal law. See Case Concerning the Factory at Chorzów (Germany v Poland) (Merits) PCIJ Rep Series A No 17; Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (adopted and opened for signature, ratification and accession by General Assembly Resolution 39/46 of 10 December 1984, entry into force 26 June 1987, in accordance with art 27(1)), http://www.ohchr.org/EN/ProfessionalInterest/Pages/CAT.aspx; Rome Statute of the International Criminal Court (adopted 17 July 1998, entered into force 1 July 2002) A/CONF.183/9 (hereinafter ‘Rome Statute’).

74  Alina Balta Principles and Guidelines on the Right to a Remedy and Reparation for Victims of Gross Violations of International Human Rights Law and Serious Violations of International Humanitarian Law (hereinafter ‘the van Boven/Bassiouni Principles’) is the best-known and most comprehensive instrument with regard to the right to a remedy and reparation for victims of gross human rights violations at the international level.14 In accordance with domestic and international law and taking into account the individual circumstances, victims of the above specified crimes should, as appropriate and proportional to the gravity of the violation and/or of the harm suffered and circumstances of each case, be provided with full and effective reparations.15 They can include restitution, compensation, rehabilitation, satisfaction and guarantees of non-repetition. Restitution aims to restore the victim’s status to the original situation before the violation took place. Compensation can be provided to any economically assessable damage, as appropriate and proportional to the gravity of the violation, which took place. This measure aims to compensate not only for economic loss, but also for moral, physical or mental harm. Rehabilitation refers to the set of measures that provide medical, psychological, legal and social services. Satisfaction includes, but is not limited to, effective measures aimed at the cessation of continuing violations, verification of the facts, and full and public disclosure of the truth under certain conditions, searching for the whereabouts of the disappeared, assistance in the recovery, identification and reburial of the bodies in accordance with the expressed or presumed wish of the victims or the cultural practices of the families and communities, inclusion of an accurate account of the violations that occurred in international human rights law and international humanitarian law training and in educational material at all levels, judicial and administrative sanctions against persons liable for the violations etc. Guarantees of non-repetition aim to further contribute to prevention and may include measures to ensure effective civilian control of military and security forces, to strengthen the independence of the judiciary, to promote mechanisms for preventing social conflicts and their resolution, reviewing and reforming laws contributing to or allowing gross violations of international human rights law and serious violations of international humanitarian law etc. Finally, as the van Boven/Bassiouni Basic Principles on Reparations contend, adequate, effective and prompt reparation, by means of or through a combination of the five measures, is intended to promote justice by redressing gross violations of international human rights law or serious violations of international humanitarian law.16 A close examination of the measures above reveals that the measures of ­satisfaction carry a symbolic value and may include those involving criminal law, such as truth-finding (by means of an investigation, for instance) and

14 Basic

Principles (n 4). para 18. 16 ibid para 15. 15 ibid

Retribution through Reparations?  75 judicial and administrative sanctions against liable persons. By the same token, ­guarantees of non-repetition refer to measures to ensure that the crimes will not re-occur and, as such, may include the review and reform of laws or measures to strengthen the independence of the judiciary. Therefore, it is asserted that for the victims of gross human rights violations, placing obligations on states to criminalise certain behaviours, to punish perpetrators or to run investigations may have a reparative value and may infuse victims with a sense of justice. Against this background, the next section will assess such retributive responses’ implications for victims. II.  VICTIMS AND RETRIBUTIVE RESPONSES

As pointed out above, retributive responses consisting of prosecuting offenders and imposing criminal sanctions may have a far-reaching impact, fostering local processes of justice, memory, and judicial reform, and pairing victim-centred remedies with retributive justice.17 In addition, they are framed as being part of potential justice provided by means of reparations and assumed to mend the harm following victimisation.18 A non-exhaustive review of studies researching this topic has indeed highlighted that employing retributive responses has important implications for victims. These implications include, but are not limited to, alleviation of harm, restoration of social standing and worth of victims, and restoration of shared values. Of course, as argued elsewhere, victims’ perceptions of justice, in whichever form, depend on personal circumstances, which may inform different views about the essence of justice after rule breaking and therefore how it ought to be restored.19 However, a closer examination of retributive responses’ implications for victims is warranted. As mentioned above, there is an understanding that providing victims with reparations may mend the harm incurred as a result of a crime.20 This can indeed be the case, as illustrated by the following example. If a person has been the victim of a theft, by means of restitution (ie, the offender returns the object to the victims), the harm created by the loss of the object is restored – that is, the material harm is restored. However, it is only natural that the victim may have also suffered mental harm or a form of trauma as a result of the victimising experience, which may call for other reparative interventions, for instance, rehabilitation. In situations involving serious crimes such as murder, or worse, 17 Huneeus (n 8) 4. 18 Basic Principles (n 4) paras 15–17. 19 A Balta, M Bax and R Letschert, ‘Trial and (Potential) Error: Conflicting Visions on Reparations within the International Criminal Court System’ (2018) 29 International Criminal Justice Review 221; Wenzel et al (n 10) 3. 20 See also T van Boven, ‘Victim-Oriented Perspectives: Rights and Realities’ in T Bonacker and C Safferling (eds), Victims of International Crimes: An Interdisciplinary Discourse (The Hague, Springer, 2013) 25.

76  Alina Balta in situations of mass victimisation, which generally involve a large number of victims and are characterised by immense harm and trauma,21 repairing the harm acquires new dimensions. Measures such as rehabilitation and compensation, if available, may mend the material harm visited upon the victims. However, most of the time, victims are left to grapple with a seemingly unbridgeable moral harm.22 To compound the issue further, gross human rights violations and the massive harm they involve are by their very nature irreparable and any remedy is bound to fail to fully repair the harm inflicted on the victims, especially when the violations have been committed on a massive scale.23 Most of the time, harm extends to the societal level too.24 As has been contended, mass crimes produce a tear in the fabric of societies,25 wreaking havoc in terms of (legal) infrastructure and eroding relationships built on shared standards and trust between individuals, amongst other things.26 To this end, retributive responses may start to play a key role in tackling the moral harm. As Wenzel et al highlight, victims’ justice in the aftermath of victimisation demands that the offender be punished, above and beyond a possible restitution or compensation to the victim.27 Similarly, empirical studies surveying victims’ expectations of justice following mass victimisation have repeatedly confirmed that justice for victims necessarily includes the 21 See A Pemberton et al, who explain how mass victimisation is generally characterised by a large number of victims, as well as harm which can only be described in terms of its enormity. Harm is both material and psychological/moral, the latter being characterised by intense psychological trauma that may impact the population for generations to come. See A Pemberton, RM Letschert, A-M de Brouwer and RH Haveman, ‘Coherence in International Criminal Justice: A Victimological Perspective’ (2015) 15 International Criminal Law Review 339. 22 In line with Pemberton et al (ibid), moral harm is harm that is not material and is characterised by intense psychological trauma. At the same time, as philosopher Margaret Urban Walker posited, repairing moral harm entails the task of restoring or stabilising moral relationships. These are trustbased relations anchored in expectations of one another that require us to take responsibility for what we do or fail to do, as well as enabling us to call others to account for what they do or fail to do. See MU Walker, Moral Repair: Reconstructing Moral Relations after Wrongdoing (Cambridge, Cambridge University Press, 2006) 23. Before the ECtHR, harm that is not financial is defined as non-pecuniary harm, and refers to mental or physical suffering. In its practice, as will be seen in this chapter, the Court generally provides awards for victims in relation to pecuniary and non-pecuniary harm. See ECtHR, ‘Just Satisfaction Claims’ (19 September 2016), para 13. Dinah Shelton writes that the ECtHR awards for non-pecuniary harm are difficult to comprehend other than as subjective judgments about the moral worth of the victim and the wrongdoer. See D Shelton, Remedies in International Human Rights Law, 2nd edn (Oxford, Oxford University Press, 2005) 345. 23 UN Sub-Commission on the Promotion and Protection of Human Rights, ‘Study Concerning the Right to Restitution, Compensation and Rehabilitation for Victims of Gross Violations of Human Rights and Fundamental Freedoms: Final Report/Submitted by Theo van Boven, Special Rapporteur’ E/CN.4/Sub.2/1993/8, para 131. 24 Pemberton et al (n 21) 366. 25 L May, Crimes against Humanity: A Normative Account (Cambridge, Cambridge University Press, 2009) 244. 26 Walker (n 22) 24. See also LE Fletcher and HM Weinstein, ‘Violence and Social Repair: Rethinking the Contribution of Justice to Reconciliation’ (2002) 24 Human Rights Quarterly 573, 623, explaining that one of the consequences of mass violence is that the social fabric of a society is torn apart. 27 Wenzel et al (n 10) 6.

Retribution through Reparations?  77 punishment of perpetrators.28 The imposition of punishment may have at least three functions in this respect – namely, just deserts, deterrence and the establishment of truth – and is important for victims inasmuch as it plays a part in the undoing of the moral harm. Victims demand the punishment of the offender, which appears necessary in order for justice to be re-established,29 especially when the punishment is proportional to the severity of wrongdoing30 and is conducive to finding the truth as to why the crimes took place.31 Viewed in this way, punishment is seen as a precondition for the restoration of justice, which may indeed enable the process of mending the moral harm caused to victims. At the same time, it may contribute to reassuring the victims that a system of law concerned with punishing the perpetrators for their wrongdoing is in place32 and may ultimately avoid the spiralling effect of impunity.33

28 P Vinck, P Pham, S Baldo and R Shigekane, ‘A Population-Based Survey on Attitudes about Peace, Justice, and Social Reconstruction in Eastern Democratic Republic of Congo: Living with Fear’ (2008), 44, https://hhi.harvard.edu/publications/living-fear-population-based-survey-attitudesabout-peace-justice-and-social; P Pham, P Vinck, M Balthazard, S Hean and E Stover, ‘A PopulationBased Survey on Attitudes about Social Reconstruction and the Extraordinary Chambers in the Courts of Cambodia: So We Will Never Forget’ (2009), 29, https://escholarship.org/content/ qt74x685xm/qt74x685xm.pdf; P Vinck and P Pham, ‘A Population-Based Survey on Attitudes about Accountability and Social Reconstruction in the Central African Republic: Building Peace, Seeking Justice’ (2010), 27, https://hhi.harvard.edu/publications/building-peace-seeking-justicepopulation-based-survey-attitudes-about-accountability. 29 For instance, civil party Antonya Tiulong, indirect victim before the ECCC, stated: ‘A sentence must be handed down and I wish that the Tribunal – and I beg the Tribunal to hand down the most fairest possible judgment which will be commensurate with the crimes committed by the accused.’ ECCC, ‘Transcript of Proceedings – Duch Trial (Day 60)’, https://eccc.gov.kh/en/document/court/ transcript-proceedings-duch-trial-trial-day-60. See also the study surveying victims’ perceptions of the ICC, explaining that victims generally prioritised convictions over monetary compensation: Human Rights Center of UC Berkley School of Law, ‘The Victims’ Court? A Study of 622 Victim Participants at the International Criminal Court’ (2015), 68, https://www.law.berkeley.edu/ wp-content/uploads/2015/04/VP_report_2015_final_full2.pdf. 30 See Hermann (n 10) 88; A Duff, Punishment, Communication, and Community (New York, Oxford University Press, 2001). 31 Finding the truth is indeed one recurrent request of victims. Similarly, studies focusing on victims before the ECtHR echoed these expectations. See F van der Vet, ‘Seeking Life, Finding Justice: Russian NGO Litigation and Chechen Disappearances before the European Court of Human Rights’ (2012) 13 Human Rights Review 303, 320–21. However, the idea of finding the truth about crimes, especially by a court, ties into the debate about the truth-telling functions of international tribunals; for an interesting summary of the debate, see B Sander and JVH Holtermann, ‘Symposium: Doing Justice to Truth in International Criminal Courts and Tribunals’ (2017), http://humanityjournal.org/blog/ symposium-doing-justice-to-truth-in-international-criminal-courts-and-tribunals. See also T Hunt, who mentions that there exists a fundamental chasm between the legal and historical professions and truth they generate: T Hunt, ‘Whose Truth? Objective Truth and a Challenge for History’ in WA Schabas and S Darcy (eds), Truth Commissions and Courts (The Hague, Springer, 2004) 193. 32 See C Stahn, who elaborates on the expressivist theories of punishment, positing that the merit of punishment is in the message that the denunciation of the crime conveys to the victims, amongst other things: C Stahn, A Critical Introduction to International Criminal Law (Cambridge, Cambridge University Press, 2018) 381–411. 33 For instance, K Sikkink demonstrates that whereas in countries such as Brazil, the lack of punishment for past crimes during the dictatorship explains the high level of impunity and violations today in this country, other countries in the same region that prosecuted human rights

78  Alina Balta Equally importantly, as Hegel framed it, the punishment may annul the moral wrong entailed in a criminal act and may attempt to restore the moral balance between the victim and the offender.34 This argument ties into the idea that imposing a punishment and thus declaring the violation as wrongful, as something that should not have happened,35 has implications for the social standing and worth of the victim. In this sense, a failure to punish indicates indifference towards the victim and even disdain.36 As explained, this contention is not only put forward by moral philosophers, but it is also supported by empirical evidence which proves that victims are not motivated to punish only by instrumental motives such as deterrence, just deserts or rehabilitation, but also by retribution which is driven by social standing concerns.37 As explained by Murphy and Hampton: ‘One reason we so deeply resent moral injuries done to us is not simply that they hurt us in some tangible or sensible way; it is because such injuries are also messages – symbolic communications … Intentional wrongdoing insults us and attempts (sometimes successfully) to degrade us – and thus it involves a kind of injury that is not merely tangible and sensible.’38 The victimisation humiliates the victims and places them outside the sphere of the perpetrator’s moral obligations.39 In the most egregious cases of gross human rights violations, the victim’s degradation is a goal in itself.40 Failing to employ retributive responses and thus to re-emphasise victims’ membership of the moral sphere carries the risk of sentencing the victim to so-called ethical loneliness, a process referring to the experience of having been abandoned by humanity, compounded by the experience of not being heard.41 Viewed in this way, retributive responses may have a positive effect in the rebalancing of the scales (in terms of social standing or worth) for one party relative to the other.42 crimes do relatively well today and have learned their lessons from the past: K Sikkink, The Justice Cascade: How Human Rights Prosecutions are Changing World Politics (London, WW Norton and Company, 2011). See also a critical account on impunity, arguing that too much focus on impunity ultimately results in a limited response to human rights violations: K Engle, Z Miller and DM Davis, Anti-impunity and the Human Rights Agenda (Cambridge, Cambridge University Press, 2016). 34 GWF Hegel, Philosophy of Right (Wotton-under-Edge, Clarendon Press, 1952) 248. 35 See Wenzel et al (n 10) 6, who assert that ‘retributive justice essentially refers to the repair of justice through imposition of punishment’. 36 K Bilz, ‘The Puzzle of Delegated Revenge’ (2007) 87 Boston University Law Review 1059, 1088. 37 ibid. 38 JG Murphy and J Hampton, Forgiveness and Mercy (Cambridge, Cambridge University Press, 1988) 25. 39 Pemberton et al (n 21) 352. See also S Clayton and S Opotow, ‘Justice and Identity: Changing Perceptions of What is Fair’ (2003) 7 Personality and Social Psychology Review 298. 40 Testimonies of victims express all over again that victims are treated inhumanly or subhumanly. See, for instance, Primo Levi, a survivor of the Holocaust, who states that victims were treated as animals in concentration camps: P Levi, If This is a Man (London, Abacus Books, 1971). See also D Livingstone Smith, Less Than Human: Why We Demean, Enslave, and Exterminate Others (New York, St Martin’s Press, 2012). 41 J Stauffer, Ethical Loneliness: The Injustice of Not Being Heard (New York, Columbia University Press, 2015) 1. 42 M Wenzel and TG Okimoto, ‘On the Relationship between Justice and Forgiveness: Are All Forms of Justice Made Equal?’ (2014) 53 British Journal of Social Psychology 466.

Retribution through Reparations?  79 Finally, retributive responses have the potential to restore the values that underlie the laws that have been broken. To be more precise, crimes are a violation of values that are expected to be shared between the offender and the victim due to their shared membership in a relevant community.43 For instance, the idea that a person may peacefully take a morning walk in a quiet village in southern France presupposes the commonly shared value of ‘safety’, which can be quickly shattered when crime strikes.44 As Durkheim put it, crime is contrary to what he called common conscience, which is presumably shared sentiments that constitute and derive from the social cohesion in a collective.45 Therefore, by violating these values, it can be implied that the offender questions them, challenges their validity or undermines the consensus that lends validity to them.46 A starting point in the restoration of values may be an apology by the offender, acknowledging his wrongful behaviour and thus (re)pledging allegiance to the value. However, this situation again puts the victim in a lower-status situation (see above), finding herself in a situation of dependency to an apology by the offender.47 Conversely, employing retributive responses such as punishing the perpetrator may convey a more powerful message. This is all the more important in the case of mass victimisation, when the state usually plays an important role as perpetrator or accomplice to the crimes.48 As Kahan asserted, ‘what a community chooses to punish and how severely, tells us what (or whom) it values and how much’.49 As such, the punishment incorporates a reaction to the crimes committed and conveys the message that engagement in the forbidden conduct results in punishment,50 with the implication that victims’ trust (value) in society and the state may be restored.51 In cases of gross human rights violations, the situation is more complex, inasmuch as attempting to restore a value must necessarily include both a backward-looking function that involves a reaction to the atrocities already committed and a forward-looking function, which contributes to the perceived legitimacy of government and legal authorities, and the rule of law in a material sense.52

43 Wenzel et al (n 10) 7. 44 See the narrative of Susan Brison in Aftermath Violence and the Remaking of a Self (Princeton, Princeton University Press, 2003). 45 E Durkheim, The Division of Labor in Society, G Simpson (trans) (New York, Free Press, 1964). 46 Wenzel et al (n 10) 8. 47 Moreover, the value of apology should not be overestimated. Studies have shown that the victims frequently challenge the sincerity of such apologies. See, for instance, E Stover, M Balthazard and K Alexa Koenig, ‘Confronting Duch: Civil Party Participation in Case 001 at the Extraordinary Chambers in the Courts of Cambodia’ (2011) 93 International Review of the Red Cross 503. 48 D Luban, ‘A Theory of Crimes against Humanity’ (2004) 29 Yale Journal of International Law 116. 49 DM Kahan, ‘Social Meaning and the Economic Analysis of Crime’ (1998) 27 Journal of Legal Studies, 615. 50 Luban (n 48) 128; see also Pemberton et al (n 21) 344–45. 51 For a critical view on punishment as value restoration, see Wenzel and Okimoto (n 42) 466. 52 See a discussion on the backward and forward functions of international criminal justice in Pemberton et al (n 21) 344–45.

80  Alina Balta Therefore, I have contended above that employing retributive responses may have several implications for victims, including but not limited to the alleviation of harm, the restoration of social standing and worth of victims, and the restoration of shared values. The actual realisation of these implications may indeed amount to satisfaction and guarantees of non-repetition, and thus, overall, afford a measure of reparative justice. For instance, the alleviation of harm and restoration of social standing and worth of victims may amount to satisfaction. Indeed, satisfaction measures aim ‘to restore the dignity of the victims’53 and to acknowledge their victimisation,54 and are conducive to the establishment of truth, justice and accountability.55 Similarly, the restoration of shared values may amount to guarantees of non-repetition, to the extent that it paves the way for prevention and for the assurance that such crimes will not re-occur in the future.56 At the same time, it should also be acknowledged that retributive responses should not be considered the only means to achieve reparative justice and, indeed, in isolation, they may not be a completely adequate and sufficient response for the victims to achieve personal, public and socially shared validation and vindication.57 However, for the time being, the implications of retributive responses remain at a theoretical level, which will be scrutinised by looking in the following sections into the case law of the ECtHR. Consequently, the next section will first introduce the legal framework on reparations at the ECtHR and will then look at the relevant case law in order to understand how and in what form retributive responses emerge in this context. III.  REPARATIONS AT THE ECTHR

A.  Legal Framework As early as 1959, the ECtHR came into existence with the mandate to, inter alia, deliver legally binding and enforceable judgments on reparations.58 Individuals who fulfil the admissibility criteria set out by Article 35 ECHR can file 53 C Evans, The Right to Reparation in International Law for Victims of Armed Conflict (Cambridge, Cambridge University Press, 2012) 67. 54 ibid 71. 55 JB Jeangène Vilmer, Repairing the Irreparable: Reparations for Victims at the International Criminal Courts (Paris, PUF, 2009). 56 Van Boven argued that: ‘There exists a definite link between effective remedies to which the victim(s) is (are) entitled, remedies aimed at the prevention of the recurrence of similar violations and the issue of the follow-up given by the State party.’ UN Sub-Commission on the Promotion and Protection of Human Rights (1998), para 55. 57 Walker (n 22) 16. See R Letschert and T van Boven, ‘Providing Reparation in Situations of Mass Victimisation: Key Challenges Involved’ in R Letschert, RH Haveman, A de Brouwer and A Pemberton (eds), Victimological Approaches to International Crimes (Antwerp, Intersentia, 2011), discussing the challenges involved in providing reparative justice for victims. 58 ‘The Execution of Judgments of the European Court of Human Rights’ (Council of Europe, January 2008) 7.

Retribution through Reparations?  81 complaints directly with the ECtHR (before the entry into force of Protocol 11 in 1998, the complaints were first addressed to the European Commission of Human Rights),59 alleging violations of human rights. The right of individuals to an effective remedy is set out in Article 13 ECHR: ‘Everyone whose rights and freedoms as set forth in this Convention are violated shall have an effective remedy before a national authority notwithstanding that the violation has been committed by persons acting in an official capacity.’60 This is usually interpreted in conjunction with Article 41 ECHR, which, on just satisfaction, establishes that: ‘If the Court finds that there has been a violation of the Convention or the Protocols thereto, and if the internal law of the High Contracting Party concerned allows only partial reparation to be made, the Court shall, if n ­ ecessary, afford just satisfaction to the injured party.’61 Although the ECtHR was empowered to deliver just satisfaction since its inception, the Court was rather silent on the matter until it gained the status of a full-time Court, with the entry into force of Protocol 11.62 It was in Aksoy v Turkey in 1996 that the ECtHR’s jurisprudence moved the meaning of ‘effective remedy’ from symbolic satisfaction through the judgment itself, to include compensation as well as ‘a thorough and effective investigation capable of leading to the effective identification and punishment of those responsible and including effective access for the complainant to the investigatory procedure’.63 It is worth noting that although the Convention does not include provisions on gross human rights violations or international crimes as such, through the number of cases filed with the European system relating to serious violations of human rights, including of the right to life and the prohibition of torture, the ECtHR has dealt with such crimes and correspondingly has evolved its

59 Protocol No 11 to the Convention for the Protection of Human Rights and Fundamental Freedoms, Restructuring the Control Machinery Established Thereby (Council of Europe, 11 May 1994). 60 European Convention on Human Rights, as amended by Protocol Nos 11 and 14, supplemented by Protocol Nos 1, 4, 6, 7, 12, 13 and 16 (Council of Europe, November 1950), art 13. 61 ibid. 62 At its inception, the Court struggled to maintain a narrow legal authority; its limited influence was an artefact of its very small case load during the first 15 years of its operation and the reality that key member states of the ECHR – notably France and the UK – were unwilling to accept the Court’s jurisdiction out of fear that it would meddle in the decolonisation struggles of the period. See M Madsen, ‘The Challenging Authority of the European Court of Human Rights: From Cold War Legal Diplomacy to the Brighton Declaration and Backlash’ (2016) 79 Law and Contemporary Problems 141, 142. 63 Aksoy v Turkey App No 21987/93 (ECtHR, 18 December 1996), para 98. It was in McCann v UK that the Court articulated for the first time the state’s duty to investigate: ‘The obligation to protect the right to life … requires by implication that there should be some form of effective official investigation when individuals have been killed as a result of the use of force by, inter alios, agents of the State.’ However, it did not award just satisfaction, as it held that it did not consider it appropriate to do so, since damages were requested in regard to three terrorist suspects, who were killed due to their intention to plant a bomb in Gibraltar. McCann v UK App No 18984/91 (ECtHR, 27 Sept 1995), paras 161 and 219.

82  Alina Balta jurisprudence in the area of reparations.64 To be precise, the Court has witnessed a gradual inflow of cases on gross human rights violations, prompted by the enlargement of the Council of Europe and the entry into force of Protocol 11.65 Consequently, the most well-known situations involving gross human rights violations on the territory of Europe in the past few decades have been brought before the ECtHR mainly by means of individual applications.66 They have been analysed in this chapter and include, but are not limited to, the situation in south-east Turkey,67 the two Chechen wars,68 as well as other cases involving human rights violations which can be considered gross due to the serious character of the violations, as discussed above.69 The Cyprus–Turkey conflict is also an emblematic example of gross human rights violations on the territory of Europe and complaints have been brought before the Court both by means of an inter-state complaint and individual complaints.70 B.  Just Satisfaction: Requirements With the expansion of case law on just satisfaction, the Court has issued a detailed explanation in regard to its practice on just satisfaction claims.71 As the Court clarified, under Article 41 ECHR, just satisfaction is limited to awards in the form of pecuniary damage, non-pecuniary damage, and costs and expenses. For just satisfaction to be awarded, a clear causal link must be established between the damage claimed and the violation alleged by the applicant. The purpose of the award in respect of damage is to compensate the applicant for the actual harmful consequences of a violation and it is not intended to punish

64 Evans (n 53) 61. 65 P Leach, ‘What is Justice? Reflections of a Practitioner at the European Court of Human Rights’ (2013) 4 European Human Rights Law Review 392, 393; O Ichim, Just Satisfaction under the ­European Convention on Human Rights (Cambridge, Cambridge University Press, 2014) 234. 66 See, for instance, MT Kamminga, ‘Is the European Convention on Human Rights Sufficiently Equipped to Cope with Gross and Systematic Violations?’ (1994) 12 Netherlands Quarterly of Human Rights 153, 154, elaborating upon the limited number of cases brought before the Court in Strasbourg through inter-state applications. 67 P Sardaro, ‘Jus Non Dicere for Allegations of Serious Violations of Human Rights: Questionable Trends in the Recent Case Law of the Strasbourg Court’ (2003) 6 European Human Rights Law Review 601. 68 K Koroteev, ‘Legal Remedies for Human Rights Violations in the Armed Conflict in Chechnya: The Approach of the European Court of Human Rights in Context’ (2010) 21(1) International Humanitarian Legal Studies 275. 69 As also employed by Sardaro (n 67) 601. 70 A Solomou, ‘Cyprus v Turkey’ (2015) 109 American Journal of International Law 393; The Court’s jurisprudence on inter-state cases is very limited, as the states appear to be reluctant to denounce other states’ unlawful behaviour. For an overview of inter-state cases at the ECtHR level, see ‘Background Paper for Seminar Opening of the Judicial Year: International and National Courts Confronting Large-Scale Violations of Human Rights – Genocide, Crimes against Humanity and War Crimes’ (ECtHR, January 2016). 71 ECtHR (n 22).

Retribution through Reparations?  83 the Contracting Party responsible.72 Furthermore, as the Court held, the rationale of pecuniary damage awards is to achieve restitutio in integrum, which is to place the applicant, as far as possible, in the position in which he or she would have been had the violation found not taken place. This may involve compensation for both loss actually suffered (damnum emergens) and loss, or diminished gain, to be expected in the future (lucrum cessans).73 Depending on the (scarcity of) facts at its disposal, the Court may decide on an ‘equitable basis’ to award less than the full amount of the loss requested by the applicants. In regard to non-pecuniary damage, the Court’s award is intended to provide compensation for non-material harm, such as mental or physical suffering. However, due to the nature of non-pecuniary damage, it does not lend itself to precise calculation, therefore here too, if the existence of such damage is established, the Court may make a monetary award on an equitable basis, having regard to the standards that emerge from its case law.74 C.  Case Law Selection Bearing in mind the context of this research, the case law of the Court was selected, taking into account certain criteria: the sort of reparation provided; the nature and seriousness of victimisation; the number of applicants; link to a bigger conflict; type of document (just satisfaction or, alternatively, merits and just satisfaction when they were decided together); key search terms – ‘armed conflict’, ‘genocide’, ‘international crimes’, ‘war crimes’ and ‘crimes against humanity’; and language (English). Approximately 300 cases were selected, which were then analysed to ensure they fitted the scope of this research. Following a second scrutiny of this batch, 72 cases appeared to be relevant and were subjected to coding using qualitative data software Atlas.ti.75 Not all 72 cases that were studied closely provide for reparations that might amount to retributive responses, but the analysis of all 72 was necessary to first identify the Court’s general approach to reparations and then zoom in on cases that provide retributive responses. In addition, the list of cases does not by any means aim to be exhaustive, but to generally expose the practice of the Court with regard to retributive responses in the context of reparations. The 72 cases concern six countries, namely Russia, Turkey, Romania, Croatia, the UK and Armenia; however, the cases indicative of retributive responses concern only Russia, Turkey and Romania. Following this exposition of the cases and the obligations



72 ibid

para 9. para 10. 74 ibid paras 13–15. 75 In addition, the selection includes case law issued by the ECtHR until 1 December 2018. 73 ibid

84  Alina Balta for states provided therein, the goal is to ultimately connect the results with the theoretical illustration given in section II. D. Results The analysis reveals that the case law involving gross human rights violations decided upon by the Court revolves around some common denominators. In a large number of cases, the Court offered reparations in cases on large-scale enforced disappearances/missing persons,76 large-scale unlawful killings by actors of the state,77 war crimes78 and large-scale human rights violations set against a backdrop of inter-state issues.79 Furthermore, the research identified that outside of the just satisfaction framework whereby the Court establishes clear-cut obligations for states in respect of the payment of pecuniary and non-pecuniary damages, the judgments may establish further obligations, which fall under the category with potential reparative value for the victims. Although the Court, according to Article 41, asserts that it will make only awards for restitutio in integrum and compensation, as will be seen, in ‘exceptional cases’80 of gross human rights violations, such as those under current investigation, it takes a proactive stance and recommends measures that move its reparatory regime beyond just satisfaction. These measures – individual and/or general – are awarded as obligations for states under Article 46 on the execution of judgments, and might amount to measures of satisfaction and guarantees of non-repetition, as defined by the van Boven/Bassiouni Basic Principles on Reparations and explained in section I above.81 However, insofar as the legal basis for these measures is Article 46 and not Article 41, legally speaking, these measures cannot be considered a 76 The situation is prevalent in the cases against Russia, involving applicants residing in the Chechen Republic. See Kaykharova and Others v Russia App Nos 11554/07, 7862/08, 56745/08 and 61274/09 (ECtHR, 1 August 2013); Abakarova v Russia App No 16664/07 (ECtHR, 15 October 2015); Dzhabrailov and Others v Russia App Nos 8620/09, 11674/09, 16488/09, 21133/09, 36354/09, 47770/09, 54728/09, 25511/10 and 32791/10 (ECtHR, 7 July 2014); Gakayeva and Others v Russia App Nos 51534/08, 4401/10, 25518/10, 28779/10, 33175/10, 47393/10, 54753/10, 58131/10, 62207/10 and 73784/10 (ECtHR, 10 October 2013). See also Cyprus v Turkey (Just Satisfaction) [GC] App No 25781/94 (ECtHR, 12 May 2014); Varnava and Others v Turkey [GC] App Nos 16064/90, 16065/90, 16066/90, 16068/90, 16069/90, 16070/90, 16071/90, 16072/90 and 16073/90 (ECtHR, 18 September 2009). 77 For instance, see Association ‘21 December 1989’ and Others v Romania App No 33810/07 (ECtHR, 24 May 2011). 78 For instance, see Jelic v Croatia App No 57856/11 (ECtHR, 12 June 2014). 79 For instance, see Chiragov and Others v Armenia App No 13216/05, (ECtHR, 12 December 2017); Cyprus v Turkey (n 76). 80 ‘The Execution of Judgments of the European Court of Human Rights’ (n 58) 52. 81 For a detailed elaboration of the Court’s practice and different legal bases for providing these measures, see L-A Sicilianos, ‘The Involvement of the European Court of Human Rights in the Implementation of its Judgments: Recent Developments under Article 46 ECHR’ (2014) 32 Netherlands Quarterly of Human Rights 235.

Retribution through Reparations?  85 form of reparation as such; rather, they are elements of the states’ obligation to abide by final judgments, under the Committee of Ministers’ supervision,82 which includes the responsibility to adopt, when appropriate, individual and/or general measures.83 First and foremost, in the cases concerning deprivations of life, followed by situations where the respondent state fails to investigate the circumstances around the crimes (eg, breach of the procedural limb of Article 2), the Court stresses its concern with regard to the continued impunity of perpetrators of serious human rights violations.84 In this context, the Court reiterates a twofold obligation for the national authorities: to punish those responsible and to officially establish the truth.85 As regards punishment, the Court holds that the object and purpose of the Convention as an instrument for the protection of individual human beings require that its provisions be interpreted and applied so as to make its safeguards practical and effective, and that impunity for those responsible for gross human rights violations would not be compatible with the state’s obligations under the Convention.86 Although mindful of the fact that it is up to the respondent state, under the supervision of Committee of Ministers, to take measures to tackle the situation, the Court nevertheless recommends a variety of individual and general measures that the state institutions should take in this regard.87 The starting point concerns an effective criminal investigation with a view to identifying, prosecuting and punishing those responsible.88 To fulfil this obligation, the states should carry out a new, independent investigation, in full compliance with the requirements of the ECHR;89 however, it may also entail the obligation to ensure that the legislative and administrative framework set up to protect rights is properly implemented.90 In this connection, the Court notes that among the main purposes91 of imposing criminal sanctions

82 ECHR, art 46. 83 ‘The Conscience of Europe: 50 Years of the European Court of Human Rights’ (ECtHR, 10 October 2010) 88. 84 Abakarova v Russia (n 76) paras 111–12. 85 Benzer and Others v Turkey (n 7) para 197; Aslakhanova and Others v Russia App Nos 2944/06 and 8300/07, 50184/07, 332/08, 42509/10 (ECtHR, 18 December 2012), para 214. 86 Abakarova v Russia (n 76) para 112. 87 ibid paras 112–14. 88 Benzer and Others v Turkey (n 7) para 219; Bitiyeva and Others v Russia App No 36156/04 (ECtHR, 23 April 2009), para 87; Bopayeva and Others v Russia App No 40799/06 (ECtHR, 7 November 2013), para 83; Esmukhambetov and Others v Russia, App No 23445/03 (ECtHR, 29 March 2011), para 107; Jelic v Croatia App No 57856/11 (ECtHR, 12 June 2014), para 56. 89 Abuyeva and Others v Russia (n 7) para 243; Aslakhanova and Others v Russia (n 85) para 228; Benzer and Others v Turkey (n 7) para 219; Elena Apostol and Others v Romania App Nos 24093/14 and 16 (ECtHR, 23 May 2016) para 32; Association ‘21 December 1989’ and Others v Romania (n 77) para 194. 90 Bopayeva and Others v Russia (n 88) para 83. 91 Moreover, when deciding on the adequacy of the national investigation, the Court held that an amnesty is generally incompatible with the duty incumbent on the states to investigate acts of torture (see Ould Dah v France App No 13113/03 (ECtHR, 17 March 2009)) and to combat impunity for international crimes (see Association ‘21 December 1989’ and Others v Romania (n 77) para 144).

86  Alina Balta are retribution as a form of justice for victims and general deterrence aimed at the prevention of new violations and upholding the rule of law.92 It also stresses that this obligation is a necessary measure to guarantee the ­non-repetition of such crimes.93 In addition, one important dimension of tackling impunity are the reparative measures aimed at the cessation of continuing violations, followed by investigations with the aim of establishing the truth about crimes.94 This approach appears to have led to the emergence of a limited form of right to truth for victims, a right that seems to have recently taken on a public dimension.95 To this end, the research has identified that the Court may require the respondent state to be respectful of the suffering of victims and to assist them in obtaining information and uncovering relevant facts.96 The Court also held that the silence of the authorities of the respondent state in the face of the real concerns of relatives, relating to anxiety caused by the death or disappearance of relatives, can be categorised as inhuman treatment.97 In situations of recurring victimisation, the Court suggested setting up a mechanism equipped with all the necessary resources and expertise to deal with disappearances in the area of the state concerned,98 including the location and exhumation of presumed burial sites, as well as the collection, storage and identification of remains.99 Interestingly enough, in the cases characterised by a pervasive failure of national authorities to carry out investigations, the Court stressed that the importance of establishing the truth and punishing perpetrators has implications at both the personal and societal levels.100 Within satisfaction measures, the Court often also referred to the reparative value of an official declaration or a judicial decision, which represents a public vindication of the wrong suffered by the applicant.101 Other measures recommended by the Court include recourse to non-judicial means of collecting information and establishing the truth about the events, public acknowledgement, as well as condemnation of serious violations of rights in the course of security operations etc.102

92 Jelic v Croatia (n 88) para 90. 93 Esmukhambetov and Others v Russia (n 88) para 107. 94 For instance, Abakarova v Russia (n 76) para 111; Benzer and Others v Turkey (n 7) para 197; Jelic v Croatia (n 88) paras 94–95; Association ‘21 December 1989’ and Others v Romania (n 77) para 144. 95 See ‘Background Paper’ (n 70) 10. 96 Aslakhanova and Others v Russia (n 85) para 215. 97 ibid. 98 ibid para 225. 99 ibid para 226. 100 Jelic v Croatia (n 88) para 94; Association ‘21 December 1989’ and Others v Romania (n 77) para 142. 101 For instance, Chiragov and Others v Armenia (n 79) para 57; Cyprus v Turkey (n 76) para 56. 102 See Abakarova v Russia (n 76) para 112.

Retribution through Reparations?  87 Furthermore, although not labelled as such, the Court recommends to states to take reparative measures, which amount to guarantees of non-repetition of similar crimes in the future.103 It suggests setting up non-judicial mechanisms focused on drawing lessons from the past, raising awareness of the applicable legal and operational standards, deterring new violations of a similar nature, assessing the adequacy of the national legal instruments pertaining to large-scale security operations and the mechanisms governing military–civilian cooperation in such situations, greater dissemination of information and better training for both military and security personnel in order to ensure strict compliance with the relevant legal standards, including human rights and international humanitarian law.104 Finally, in cases where the integrity of the national judiciary was at stake, the Court has taken the further step of laying down characteristics of an impartial and independent tribunal in order to strengthen the independence of the judiciary.105 IV. DISCUSSION

In section II, I outlined, from a victim’s perspective, three implications of retributive responses that consist of obligations placed on states to, amongst other things, criminalise certain human rights violations and prosecute or punish offenders. I contended that employing retributive responses contributes to the alleviation of harm, the restoration of social standing and worth of victims, and the restoration of shared values. As posited in this research, retributive responses are seen as a precondition for justice restoration, which may indeed enable the process of mending the moral harm caused to victims. Furthermore, they have the potential to re-emphasise the victims’ membership of the moral sphere and attempt to restore the moral balance between the victim and the offender. Similarly, retributive responses may contribute to value restoration, inasmuch as the punishment incorporates a reaction to the crimes committed, and conveys the message that engagement in the forbidden conduct is punished, with the implication that victims’ trust (value) in society and state may be restored. In situations of gross human rights violations, all the issues are further compounded and thus retributive responses appear to be all the more important.

103 These measures may also be taken in the context of pilot judgment procedures, whereby ­shortcomings in the legal order – the systemic problem – that are the cause of the violation and that affect a whole class of individuals are identified. In the judgment in the pilot case, the Court gives advice to the government, in terms of recommended measures on how to solve the systemic problem. See E Fribergh, ‘Pilot Judgments from the Court’s Perspective’ (2008) 87–88, https://rm.coe.int/ applying-and-supervising-the-echr-towards-stronger-implementation-of-t/1680695ac3. 104 See Abakarova v Russia (n 76) paras 112 and 114. 105 Elena Apostol and Others v Romania (n 89) para 33.

88  Alina Balta In section III, a scrutiny of the relevant case law of the ECtHR highlighted that the Court appears to be departing from its clear-cut legal framework on just satisfaction so as to place certain obligations on states, which amount to satisfaction and guarantees of non-repetition measures, in the sense of the van Boven/Bassiouni Basic Principles on Reparations. The Court, although mindful of its limited role with regard to the execution of judgments, recommends a variety of individual and general measures that the state institutions should take to tackle the situation exposed in the judgments. These individual and general measures draw on criminal law and amount to retributive responses, as described in this chapter. However, it is worth clarifying that these observations emerged from the analysis of the jurisprudence concerning gross human rights violations, as this was the corpus of cases under current investigation. Against this background, the question remains whether the retributive ­obligations the ECtHR is placing on states may have positive implications for the victims involved and may afford any measure of reparative justice to them. In what follows, I explain why the answer to this question is riddled with challenges. Of the cases under current investigation, the instances when the Court placed such obligations on states are rather limited (approximately 14 per cent); however, they are indicative of the Court’s tendency to become more proactive in the area of reparations in certain cases characterised by gross human rights violations.106 As the analysis shows, the Court’s approach to award such measures is motivated by concerns regarding prolonged impunity at the state level.107 The case law is rife with examples to support this idea. For instance, in Abakarova v Russia, the Court stated that the failure to carry out a criminal investigation to establish, inter alia, the factual circumstances leading up to the events ‘gives rise to particular concerns about the continued impunity of the perpetrators of a serious human rights violation’.108 Similarly, in Benzer and Others v Turkey, the Court stressed that investigatory steps should be taken at the national level in order to ‘prevent impunity and they should include the carrying out of an effective criminal investigation’.109 In Bitiyeva and X v Russia, the Court, referring to the failure of the state to carry out an investigation to find out the status of a detention centre where the applicants were detained, held that: This situation fosters impunity for all kinds of abuses and is absolutely incompatible with the responsibility of the authorities to account for individuals under their 106 A Mowbray, ‘An Examination of the European Court of Human Rights’ Indication of Remedial Measures’ (2017) Human Rights Law Review 451. 107 Of course, the Court is also motivated by pragmatic considerations relating to its mounting challenges of its increasing case law, which has skyrocketed, as well as increasing criticism from its Member States who have sought ways to limit its powers. See MR Madsen, ‘From Cold War Instrument to Supreme European Court: The European Court of Human Rights at the Crossroads of International and National Law and Politics’ (2007) 32 Law and Social Inquiry 137. 108 Abakarova v Russia (n 76) para 111. 109 Benzer and Others v Turkey (n 7) para 219.

Retribution through Reparations?  89 control. Once brought to the attention of the competent bodies, it should have prompted urgent and comprehensive steps in order to identify and bring to justice those responsible, to provide redress for the victims and to ensure that no such ­situation would arise in the future.110

Although not explicitly mentioned, the Court’s approach in these cases suggests the Court’s employment of the doctrine of evolutionary interpretation, according to which the ECHR is a living instrument that must be adapted to the realities of the present day.111 Indeed, when placing such obligations on states, the Court appears to respond to challenges in the landscape of human rights protection in Europe112 as well as concerns for the plight of victims.113 In addition, the award of these measures by the Court might serve the symbolic function of countering the denial of the authorities with regard to these crimes,114 establish the human rights violations115 and attempt to catalyse change at the national level. Its counterpart, the IACtHR has already developed a large jurisprudence on reparations and impunity, in an attempt to grapple with the gross human rights violations on the Latin American continent.116 Consequently, the Court’s approach, observed from a victim’s perspective, appears to be positive. Acknowledging that a criminal act took place and that the state is responsible (by way of omission, due to failure to investigate, or by way of commission, when the state authorities themselves engaged in or condoned the criminal activities) and, furthermore, providing for measures to be taken to tackle the situation, ties into the argument on harm restoration. Indeed, the symbolic value of having the victimisation acknowledged at the

110 Bitiyeva and X v Russia (n 88) para 118. 111 Tyrer v UK App No 5856/72 (ECtHR, 25 April 1978) para 31. 112 A former president of the Court stressed that, in managing its case law, the Court should be realistic regarding its ambitions to discipline national systems, especially those generating systemic gross human rights violations, and should develop an acute sense for effectivity and priorities. See L Wildhaber, ‘Rethinking the European Court of Human Rights’ in J Christoffersen and MR Madsen (eds), The European Court of Human Rights between Law and Politics (Oxford, Oxford University Press, 2014), 214. In addition, the power of the Court to award these measures was enabled by the Committee of Ministers after the adoption of Protocol 14, which aimed to improve the functioning of the system amid concerns relating to the Court’s excessive case load. See Explanatory Report to Protocol No. 14 to the Convention for the Protection of Human Rights and Fundamental Freedoms (ECtHR, 13 May 2004) paras 2 and 7. 113 As indicated in some selected cases, eg, Mocanu and Others v Romania App Nos 10865/09, 45886/07 and 32431/08 (ECtHR, 17 September 2014); Aslakhanova and Others v Russia (n 85); Dogan and Others v Turkey (Just Satisfaction) App Nos 8803-8811/02, 8813/02 and 8815-8819/02 (ECtHR, 13 July 2006). 114 Van der Vet (n 31) 319. 115 Koroteev (n 68) 303. 116 K de Feyter, S Parmentier, M Bossuyt and P Lemmens, Out of the Ashes: Reparation for Victims of Gross and Systematic Human Rights Violations (Antwerp, Intersentia, 2005); Huneeus (n 8) 5. See also Fernando Felipe Basch’s critical view on the IACtHR’s provision of a duty to punish, arguing that it promotes the violation of an individual’s right to equal treatment and to be presumed innocent, as well as of a defendant’s right to defence in a fair trial, ultimately infringing the very objectives of the Inter-American system of human rights protection: Basch (n 7) 227–28.

90  Alina Balta regional level, by the ECtHR, as well as its awards on just satisfaction are significant because they deliver a degree of justice for the victims (perhaps mending at least the material harm) and provide a measure of accountability.117 This is all the more relevant for victims whose needs for accountability have not been met at the national level118 due to a lack of or prolonged investigations into crimes at the national level (eg, in the Chechen and south-east Turkey cases). In addition, the finding of state responsibility119 for crimes is highly significant as well as indicative of the Court’s potential to restore the worth and social standing of victims. Indeed, acknowledging through a judgment that certain crimes took place and designating the perpetrator may indeed amount to satisfaction for victims.120 Similarly, having these obligations come from a regional court conveys that international oversight is directed at the region and its authorities, and might contribute to the restoration of a belief in a value shared at the international level.121 As Daly put it, the judgment may make amends for the victim’s loss of the presumptions of security, for example, by some effort to re-assure the victim that his or her rights are now respected.122 Furthermore, the message the ECtHR puts forward is that a criminal act has indeed taken place and that action has to be taken in this regard, and this is significant. If the obligations the Court is placing on states are indeed implemented, as exemplified, such as with the establishment in Russia of a national body to supervise the cases at the national level,123 this gives rise to hope that similar crimes will not re-occur. 117 P Leach, ‘The Chechen Conflict: Analysing the Oversight of the European Court of Human Rights’ (2008) European Human Rights Law Review 17. 118 Interviews with victims before the Court have captured their needs for accountability and truth; see ‘Annual Report’ (European Human Rights Advocacy Centre, 2010) 15; van der Vet (n 31) 321. 119 As Leach contends, state responsibility has been established in a very high proportion of Chechen cases before the ECtHR, despite all the evidential difficulties: Leach (n 65) 6. To clarify the concept, as regards the Chechen cases and some of the south-east Turkey cases under current investigation, the ECHR found Russia and Turkey responsible for the violation of, amongst others, art 2 in the substantive aspect, inasmuch as sufficient evidence was provided to establish that the state’s security forces had carried out the acts in violation of the right to life. At the same time, the Court found the states responsible for the violation of the procedural aspect of art 2, in relation to the failure of the states’ authorities to carry out investigations to ascribe individual criminal responsibility for the acts in violation of the right to life. 120 C McCarthy, ‘Victim Redress and International Criminal Justice: Competing Paradigms, or Compatible Forms of Justice?’ (2012) 10 Journal of International Criminal Justice 351, 366. See also Kirill Koroteev, who asserts that the judgments in the Chechen cases helped to render justice to individual applicants by establishing what had happened to their relatives and by qualifying those instances under the ECHR: Koroteev (n 68) 303. 121 Leach (n 117) 17. 122 K Daly, ‘Revisiting the Relationship between Retributive and Restorative Justice’ in H Strang and J Braithwaite (eds), Restorative Justice: From Philosophy to Practice (Aldershot, Dartmouth Press, 2000) 22. 123 See the Committee of Ministers on the establishment of a Special Investigating Unit in Russia, Action of the security forces in the Chechen Republic of the Russian Federation: general measures to comply with the judgments of the European Court of Human Rights, Update of the Memorandum CM/Inf/DH(2008)33, Memorandum prepared by the Department for the Execution of Judgments of the European Court of Human Rights (DG-HL), https://search.coe.int/cm/Pages/result_details.aspx ?ObjectID=09000016805aec57.

Retribution through Reparations?  91 However, more can be done at the national level to ensure the actual i­mplementation and follow-up of Court-ordered obligations, particularly of those that fall outside the just satisfaction framework. Research shows that whereas states tend to pay just satisfaction for pecuniary and non-pecuniary damages,124 the implementation of obligations relating to the re-opening of criminal investigations or punishment of perpetrators is not always forthcoming.125 As Koroteev put it, referring to the rich case law coming out of the armed conflict in Chechnya, ‘failure to investigate the serious violations of the ECHR found by the Court and prosecute those responsible maintains impunity for grave crimes in Chechnya and denies victims their right to know the truth about what happened to their loved ones’.126 Similarly, a study looking into the legacy of cases in south-east Turkey highlighted that in light of recent uncovering of burial sites and human remains, expectations arose that prosecutions will follow, 20 years after the crimes took place. However, due to statutes of limitations at the national level, the reality is that attempts to bring perpetrators to justice in Turkey are extremely limited.127 As regards the Romanian cases analysed in the current study, concerning the violent suppression of demonstrations during the events of December 1989, resulting in hundreds of deaths and injuries, the Court has yet again issued a case in late 2018 reiterating the failure of investigations at the national level.128 Some positive developments in the implementation of these obligations at the state level can be noted. For instance, in Russia, a number of investigations have been re-opened, a Special Investigative Unit for Russian actions in Chechenia has been set up, and an Action Plan was adopted, to train the authorities, raise awareness on human rights violations and provide victims with access to the files.129 In Turkey, measures have been taken to improve the rights protection of people in custody, train members of the security forces, raise awareness with regard to human rights, as well as provide remedies for victims at the national level.130 Notwithstanding these developments, the Committee of Ministers remains concerned at the longstanding lack of progress of the investigations in Russia (in 258 cases of enforced disappearances coming out of 124 Leach (n 65) 5. 125 Leach (n 117) 16–17. Conversely, in a famous instance, Poland did amend its legislation following a Court order in this regard. See Broniowski v Poland [GC] App No 31443/96 (ECtHR, 22 June 2004). 126 Koroteev (n 68) 301. 127 Leach (n 65) 2. 128 Case of Picu and Others v Romania and 22 Other Applications App No 74269/16 (ECtHR, 30 October 2018). 129 See the Committee of Ministers’ examination regarding the Chechen situation, which surveyed the implementation of measures awarded in some cases under current investigation: Khashiyev and Akayeva v Russia App No 57942/00 (Committee of Ministers, Department for the Execution of Judgments of the ECHR), https://hudoc.exec.coe.int/eng#{%22EXECDocumentType Collection%22:[%22CEC%22],%22EXECAppno%22:[%2257942/00%22],%22EXECIdentif ier%22:[%22004-9%22]}. 130 See the Committee of Ministers’ examination regarding the situation in south-east Turkey, looking at the implementation of measures in some cases under current investigation: Aksoy v Turkey App

92  Alina Balta Chechenia), as well as the ineffectiveness of measures in Turkey in preventing new, similar violations.131 To this end, the failure of states to rigorously implement the obligations the Court is timidly starting to place on them has, without doubt, negative consequences for the victims of these crimes.132 The victims’ right to know the truth remains denied, while frustrated expectations regarding domestic responses continue to grow. V. CONCLUSION

As this chapter has aimed to show, the Court’s intentions to proactively foster changes to enhance the domestic system’s ability to respond to gross human rights violations is commendable. However, the issue is not that retributive responses do not have a positive impact on the victims; the problem is the lack, or inadequacy, of implementation of these retributive responses. They carry symbolic value, in the sense that they acknowledge the harm and victimisation, may contribute to the process of harm restoration and the rebalancing of the worth and social standing of victims, and may also restore the belief in shared values in a society. I welcome this proactive approach of the Court to step outside of the traditional legal framework on just satisfaction and attempt to tackle impunity by ordering retributive responses. Yet, at the same time, it may be wise for the Court system to further find ways to enhance the power of Committee of Ministers in order to pressure non-abiding states to actually carry out those measures. Only through this might a discussion of their reparative impact on the victims actually commence.

No 21987/93 (Committee of Ministers, Department for the Execution of Judgments of the ECHR), https://hudoc.exec.coe.int/eng#{%22EXECDocumentTypeCollection%22:[%22CEC%22], %22EXECAppno%22:[%2221987/93%22],%22EXECIdentifier%22:[%22004-37124%22]}. 131 See Khashiyev and Akayeva v Russia (n 129); Aksoy v Turkey (n 130). 132 See also Hannah Russell, who points out deficiencies in the implementation of measures in the south-east Turkey cases and Russia (in relation to the Chechen conflict): H Russell, The Use of Force and Article 2 of the ECHR in Light of European Conflicts (Oxford, Hart Publishing, 2017).

5 Shaping Coercive Obligations through Vulnerability The Example of the ECtHR CORINA HERI

I. INTRODUCTION

T

he idea that states’ positive obligations under the European Convention on Human Rights (ECHR) can require them to protect individuals from violations of their rights, even when these are committed by third parties, is well established.1 In some cases, this means that the European Court of Human Rights (ECtHR) dictates certain aspects of how rights-harmful acts must be deterred and policed; since the 1980s,2 the Court has repeatedly found that states are required to criminalise rape, forced labour, human trafficking and other acts.3 In other words, while human rights are often cast as a check on state measures aimed at protecting law and order, this is not the extent of their relationship with criminal justice. Instead, in addition to limiting coercive measures, human rights must also be understood as, under certain circumstances, requiring states to introduce coercive measures or to shape these in a certain way.4 In the words

1 See, on this topic, L Lavrysen, Human Rights in a Positive State: Rethinking the Relationship between Positive and Negative Obligations under the European Convention on Human Rights (Cambridge, Intersentia 2016). 2 The case in question was X and Y v The Netherlands (1985) Series A No 91. 3 Söderman v Sweden ECHR 2013, concurring opinion of Judge Pinto de Albuquerque, references and emphasis removed. 4 L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce?’ in L Zedner and JV Roberts (eds), Principles and Values in Criminal Law and Criminal Justice: Essays in Honour of Andrew Ashworth (Oxford, Oxford University Press, 2012) 135.

94  Corina Heri of Liora Lazarus, there are ‘coercive duties on the state to criminalize, prevent, police, and prosecute harmful acts’.5 For scholars of the ECHR, the fact that human rights place positive obligations on states is a familiar one.6 One example of coercive obligations arises under Article 3 ECHR, the prohibition of torture and inhuman or degrading treatment, which imposes far-reaching obligations on states to deter and investigate alleged ill-treatment. In some instances, states’ positive obligations also dictate the choice of means to be used to deter a given act, requiring its criminalisation. One prominent example is the case law on the state obligation to enact and apply criminal law provisions punishing rape, as discussed below. However, as a complement to this obligation, the ECHR also protects the human rights of alleged perpetrators – thus, for example, the right to a fair trial in Article 6 ECHR, including the presumption of innocence contained in its paragraph 2. The present chapter will interrogate these ‘coercive’ positive obligations – and the alleged ‘ambiguity’ that arises from the fact that human rights both constrain and require coercion7 – from the perspective of vulnerability, a concept that, it will be argued here, informs and shapes states’ coercive obligations. It should be noted that the approach to ‘coercive obligations’ used in this chapter is a broad one, placing its main focus not on a precise definition of coercion, but on a discussion of how vulnerability relates to various forms of overlapping positive obligations to deploy the coercive legal machinery of the state to prevent and effectively deter human rights violations by third parties. The ­chapter will therefore err on the side of over-inclusivity and, like Lazarus, consider coercive obligations to ‘criminalise, prevent, police, and prosecute’.8 In other words, the focus will not be only on obligations to criminalise a given conduct, but also on other aspects of ECHR-mandated state coercion of individuals, for example, the obligation to effectively apply the criminal law to prevent a given act of ill-treatment from being perpetrated. The chapter will focus especially on three examples: the protection of children, victims of sexual and domestic violence, and members of the LGBTIQA+ community. It will do so largely within the ambit of Article 3 ECHR, for a number of reasons: a significant portion of the relevant cases fall under this provision; this has not been extensively investigated;9 it is useful to highlight the particularities of this fundamental and absolute human rights norm; and

5 ibid 136. 6 Lavrysen (n 1). 7 Lazarus (n 4) 137. 8 ibid 136. 9 cf the literature discussed below about vulnerability in the Court’s case law, which touches on the coercive obligations discussed herein, but does not focus on them. Compare also Laurens Lavrysen, Alina Balta and Natasa Mavronicola, chs 2, 4 and 9 respectively in this volume, as well as A Seibert-Fohr, Prosecuting Serious Human Rights Violations (Oxford, Oxford University Press, 2009); N Zimmermann, ‘Legislating for the Vulnerable? Special Duties under the European Convention on Human Rights’ (2015) 25 Swiss Review of International and European Law 539.

Shaping Coercive Obligations through Vulnerability  95 a database on the relevant case law was available to the author.10 However, the chapter will to some extent also consider the case law on Article 8 (the right to respect for private and family life), especially because these two provisions display a certain overlap in the sense that ill-treatment that does not reach the threshold of severity of ill-treatment necessary to constitute a violation of Article 3 may nonetheless constitute a violation of Article 8, which protects physical, psychological and sexual integrity.11 II.  VULNERABILITY IN THE COURT’S ARTICLE 3 CASE LAW

Before assessing the impact that vulnerability has on coercive obligations under the ECHR, a clarification of the concept, which has to date been used prolifically by the Court in its case law (with over 500 references to vulnerability in the merits of the Article 3 case law until September 2018) is in order. Despite its frequent invocation, the Court does not define vulnerability or the criteria for applying it; instead, it usually simply establishes that applicants in a particular case are vulnerable (or not). However, a finding of vulnerability can have farreaching legal ramifications. Generally speaking, vulnerability-related reasoning serves to interpret rights in a way that render them effective in practice for their bearers because it allows for a more context-sensitive assessment of certain factors that shape a particular applicant’s experience.12 Other scholars have noted and discussed the Court’s use of vulnerability.13 Lourdes Peroni and Alexandra Timmer14 have submitted that the ECtHR uses vulnerability to identify flaws in human rights law, including hidden assumptions and biases, and thereby uses it to create a case law that better reflects the principle of equality and responds to the reality that some groups experience human rights violations ‘more routinely and acutely’ than others.15 Alexandra Timmer understands vulnerability as a tool that allows for ‘context-sensitive judgments’ and prioritising the interests of the vulnerable over the state margin of appreciation.16 10 This database is exhaustive in that it covers every instance in which the Court relied on the concept of vulnerability under art 3 between its inception and 12 September 2018. It covers both official languages and contains 512 cases. This database was the basis for C Heri, ‘The Rights of the Vulnerable under Article 3 ECHR: Promoting Dignity, Equality and Autonomy by Reconceptualizing the Human Rights Subject’, PhD thesis, University of Zurich, 2017. 11 Söderman v Sweden (n 3) para 80. 12 R Estupiñan-Silva, ‘La vulnérabilité dans la jurisprudence de la Cour inter-américaine des droits de l’homme: esquisse d’une typologie’ in L Bourgorgue-Larsen (ed), La vulnérabilité saisie par les juges en Europe (Paris, Pedone, 2014) 89–113. 13 cf, for example, Samantha Besson, ‘La vulnerabilité et la structure des droits de l’homme: l’exemple de la jurisprudence de la cour européenne des droits de l’homme’ in Bourgorgue-Larsen (n 12). 14 L Peroni and A Timmer, ‘Vulnerable Groups: The Promise of an Emerging Concept in European Human Rights Convention Law’ (2013) 11 International Journal of Constitutional Law 1056, 1061. 15 ibid 1061, 1075–76 and 1085. 16 A Timmer, ‘A Quiet Revolution: Vulnerability in the European Court of Human Rights’ in MA Fineman and A Grear (eds), Vulnerability: Reflections on a New Ethical Foundation for Law and Politics (Farnham, Ashgate, 2013) 162–64.

96  Corina Heri This requires some explanation in the present context. It is well established that, when it comes to Article 3 ECHR, the Court’s assessment is contextsensitive, given that the minimum level of severity that must be reached in order for ill-treatment to fall under the provision is relative and ‘depends on all the circumstances of the case’.17 However, vulnerability ‘acts as a magnifying glass’, making ill-treatment look bigger.18 In other words, the Court more readily finds that the threshold of severity has been met when it comes to vulnerable applicants.19 Elsewhere,20 I have argued that it is helpful to understand vulnerability as portrayed by Martha Albertson Fineman, namely as one end of a spectrum and as the opposite of resilience to harms.21 Fineman sees vulnerability as a universal part of the human condition, which is experienced differently by different individuals.22 Based on Fineman’s argumentation, other scholars have described vulnerability as a foundation of human rights, as a platform for advancing social justice and as a way for clarifying state obligations to ensure substantive equality and protect human agency.23 However, the ECtHR’s approach to vulnerability is quite different from that of Fineman. The Court does not consider vulnerability to be a universal condition or a spectrum; instead, it understands vulnerability as a rather binary characteristic of specific individuals and groups, and often links it to an experience of dependency, marginalisation or victimisation.24 Such an approach does not always prove to be empowering – for example, by negating the agency of the individual concerned – and it can also negatively affect the rights protection of those who are not considered vulnerable by reducing receptiveness to that non-vulnerable person’s claims.25 However, the Court’s use of vulnerability can have a number of positive effects for applicants who are considered vulnerable. This includes – and for present purposes this is most important – tailoring the extent of states’ positive obligations.26 As Nesa Zimmermann has written, Member States’ vulnerability-related legislative duties contain positive obligations to criminalise certain acts committed against vulnerable individuals.27 Her analysis, which covers various Convention

17 Ireland v UK (1978) Series A No 25, para 162. 18 Peroni and Timmer (n 14) 1061 and 1091. 19 ibid 1079; Heri (n 10). 20 Heri (n 10). 21 MA Fineman, ‘Equality, Autonomy, and the Vulnerable Subject in Law and Politics’ in Fineman and Grear (n 16) 20 and 22–24. 22 ibid. Timmer and Peroni (n 14) and Timmer (n 16) have also made this argument. 23 CY Furusho, ‘Uncovering the Human Rights of the Vulnerable Subject and Correlated State Duties under Liberalism’ (2016) 5 UCL Journal of Law and Jurisprudence 175, 177–78. 24 As reflected in Timmer (16) and Besson (n 13). 25 As argued in Heri (n 10). 26 cf D Xenos, ‘The Human Rights of the Vulnerable’ (2004) 13 International Journal of Human Rights 97. 27 Zimmermann (n 9) 539–62.

Shaping Coercive Obligations through Vulnerability  97 articles, found that vulnerability reasoning has led to a clear obligation under the prohibition of slavery and forced labour in Article 4 ECHR to enact criminal law provisions against the exploitation of domestic workers and against human trafficking. She furthermore argued that such obligations are also emerging – albeit less consistently – under other Convention provisions, including Article 3.28 This, as outlined above, is the focus of the present contribution. In many cases, vulnerability reasoning as employed under Article 3 provides a sort of minimum content to states’ positive obligations; however, this standard is still emerging.29 Often also referring to states’ obligation under Article 1 ECHR to secure Convention rights ‘to everyone within their jurisdiction’, the Court has repeatedly held that while states are not required to take measures to prevent every risk of ill-treatment,30 they should at least effectively protect vulnerable persons and take reasonable steps to prevent known risks of ill-treatment from materialising.31 The following will shed some more light on this by discussing examples from the case law. III.  THREE EXAMPLES OF THE COERCIVE POWER OF VULNERABILITY

Like vulnerability reasoning itself, coercive obligations in the ECHR can, generally speaking, be traced to the requirement of giving practical and effective protection to individuals’ rights.32 This was the justification for the coercive obligation found in the groundbreaking judgment in X and Y v The Netherlands, a case brought in respect of an intellectually disabled 16-year-old victim of rape who was precluded from initiating criminal proceedings against the perpetrator.33 Before the Court, the applicant claimed, inter alia, that her rights under Articles 3 and 8 ECHR had been violated. The Court examined Article 8 first, noting that recourse to the criminal law is not the only way to ensure compliance with the Convention, but finding that civil remedies were not enough to protect persons such as the applicant from this kind of wrongdoing.34 The Court therefore held that: ‘Effective deterrence is indispensable in this area and it can be achieved only by criminal-law provisions.’35 Ultimately, the Court found that the respondent state had violated its obligations under Article 8 and thus it was not necessary to examine Article 3.36 This case is relevant because it represents the starting point of the Court’s modern-day jurisprudence on coercive obligations and because, if brought today,

28 ibid

561. argued in Heri (n 10). 30 O’Keeffe v Ireland ECHR 2014, para 144. 31 ibid. 32 X and Y v The Netherlands (n 2) para 30. 33 ibid paras 12 and 27. 34 ibid para 27. 35 ibid. 36 As noted in O’Keeffe v Ireland (n 30) para 147. 29 As

98  Corina Heri it would likely be considered in a rather different way – namely as an Article 3 case, wherein that provision consumes the less severe Article 8 claim, and as one involving the concept of vulnerability.37 Later cases on the interplay between vulnerability and coercive duties have largely been similar to X and Y, both in the fact pattern (concerning grievous violence inflicted by private parties) and in the coercive obligations found by the Court. The following will describe some of the case law that followed X and Y and will focus specifically on three constellations of vulnerability-related cases: those concerning children, those concerning victims of sexual and domestic violence, and those concerning people who face discrimination because of their sexual orientation or gender identity. A. Children Under Article 3 ECHR, the Court regularly finds the deployment of the criminal law to be necessary in order to prevent and address the ill-treatment of children. The majority of relevant cases concern the protection of children from sexual abuse, but others concerning domestic abuse, corporal punishment and other grave acts follow a similar approach. The reason often cited for granting special protection to children is their vulnerability, which has been used to support the finding that Article 3 ECHR requires coercion in relation to wilful harms perpetrated against minors as old as 17.38 Under Article 3, the ECtHR often requires states to deploy a criminal law response with regard to harms intentionally inflicted on minors by adults. However, the same is not necessarily true for every violation of Article 3 that does not fall under this specific constellation or for every violation of a child’s ECHR rights. When it comes to harms against minors that fall under other Convention provisions, even when those harms are of a sexual nature, the Court may consider that the existence of civil law remedies suffices to show that the respondent state has established an adequate legal framework to protect against such acts. This was the case in Söderman v Sweden, which concerned an Article 8 ECHR complaint related to the covert filming of an undressing teenager by her stepfather.39 The discussion below will consider three groups of cases concerning children in which the Court has established that the state does have coercive obligations: those concerning the criminalisation and investigation of rape and sexual violence; those concerning other types of violence inflicted by private persons; and, to complete the picture, the ways in which the ECHR protects vulnerable minors from the coercive power of the state. 37 cf ibid paras 188–95; M and M v Croatia App No 10161/13 (ECtHR, 9 September 2015), para 143. 38 Ateşoğlu v Turkey App No 53645/10 (ECtHR, 20 January 2015), paras 11–12. 39 Söderman v Sweden (n 3).

Shaping Coercive Obligations through Vulnerability  99 i.  Sexual Abuse of Children In 1997, the Court found that rape could constitute torture and since then, it has repeatedly acknowledged that, if not always constituting torture, sexual violence reaches the threshold of severity required to constitute less severe violations of Article 3, namely inhuman or degrading treatment.40 The Court has since recognised that ‘[s]exual abuse is unquestionably an abhorrent form of wrongdoing, with debilitating effects on its victims’, and, in relation to the sexual abuse of children, adopted the standard language that ‘[c]hildren and other vulnerable individuals are entitled to state protection, in the form of effective deterrence, from such ill-treatment’.41 The Court also regularly holds that, under Article 3 taken together with Article 1 ECHR, states must take measures to ensure that individuals in their jurisdiction do not suffer ill-treatment, including where administered by third parties, and should ‘provide effective protection, in particular, of children and other vulnerable persons and include reasonable steps to prevent ill-treatment of which the authorities had or ought to have had knowledge’.42 These passages indicate how vulnerability informs coercive obligations: vulnerable individuals are particularly entitled to effective deterrence of and protection from attacks on their Article 3 rights. In addition, authorities’ knowledge of ill-treatment against which they must take reasonable preventive steps can be imputed (‘ought to have had knowledge’), and known vulnerability might contribute to imputing knowledge of a risk of ill-treatment.43 They also indicate the existence of different overlapping obligations: to provide for effective deterrence, to effectively protect and to take reasonable steps to prevent ill-treatment. The following cases show what is meant by these three obligations, as well as setting out in more detail what is required of the state. The first obligation, the obligation to provide for effective deterrence by creating and applying provisions of the criminal law applicable to acts of ­ill-treatment, was concretised in MC v Bulgaria, which concerned the alleged rape of a 14-year-old girl.44 Because she had not been subjected to or threatened with violence, her alleged rapists were not prosecuted. Regarding Article 8 ECHR, the Court held that while states have a margin of appreciation in their choice of means, in order to deter grave acts, including rape, they must enact effective criminal law provisions and, in particular, effectively protect children 40 See Aydin v Turkey ECHR 1997-IV, paras 80–88; CAS and CS v Romania App No 26692/05 (ECtHR 20 March 2012); C McGlynn, ‘Rape, Torture and the European Convention on Human Rights’ (2009) 58 International and Comparative Law Quarterly 565. 41 Stuart v UK App No 41903/98 (ECtHR, 6 July 1999), para 1; M and Others v Italy and Bulgaria App No 40020/03 (ECtHR, 31 July 2012), para 105. 42 Z and Others v UK ECHR 2001-V, para 73. 43 See O’Keeffe v Ireland (n 30), where the Court found that the state, given the vulnerability of children and the nature of the institutional context, ‘must be considered to have been aware of the sexual abuse of children by adults’ (paras 144–46, 168). 44 MC v Bulgaria ECHR 2003-XII (extracts).

100  Corina Heri and other vulnerable individuals.45 This obligation is reiterated later in the judgment (now concerning both Articles 3 and 8 ECHR) as requiring states ‘to enact criminal-law provisions effectively punishing rape and to apply them in practice through effective investigation and prosecution’.46 In MC, the Court followed this finding by reiterating the wide state margin of appreciation in the choice of means (a somewhat unfortunate conflation of principles and a result of its taking Articles 3 and 8 together),47 and that local, cultural and traditional considerations may be taken into account.48 The Court in MC then draws heavily on comparative analysis,49 on the basis of which the Court established that rape is constituted by the absence of consent and not by the use of force.50 It also drew on scientific evidence that victims of sexual abuse, especially young girls, often do not physically resist such abuse because of the psychological factors at play, as well as their fear of provoking (further) violence.51 As a result, the Court held that a rigid approach to prosecuting sexual offences, including by demanding proof of physical resistance in every case, risks impunity and jeopardises the effective protection of victims’ rights. Member States therefore have a positive obligation under Articles 3 and 8 to penalise and effectively prosecute all non-consensual sexual acts, including absent physical resistance.52 State authorities must assess all of the circumstances via an investigation based around the question of consent, instead of requiring ‘direct’ proof of rape or evidence of physical resistance; in the concrete case, they should have attached more weight to the applicant’s vulnerability and the special psychological factors at play.53 The Court did not express an opinion concerning the guilt of the alleged perpetrators – something that it cannot do, because it is not a domestic criminal court – but did find that the state had fallen short of its positive obligation ‘to establish and apply effectively a criminal-law system punishing all forms of rape and sexual abuse’.54 Judge Françoise Tulkens, concurring in MC, argued that while recourse to the criminal law was understandable for offences as grave as rape, more generally speaking, criminalisation is not the only tool available to ensure respect

45 ibid para 150. 46 ibid para 153. 47 Article 3 is an absolute right; although the meaning of this absoluteness may be more straightforward with regard to negative than to positive obligations, as discussed in more detail below, the idea of a ‘wide margin of appreciation’ in the choice of means to prevent ill-treatment falling under art 3 is difficult to reconcile with the remainder of the case law discussed herein and seems more relevant to the art 8 claim concurrently in this judgment. 48 MC v Bulgaria (n 44) para 154. 49 ibid paras 88–100. 50 ibid paras 155 and 159. 51 ibid para 164. 52 ibid para 166. 53 ibid paras 181–83. 54 ibid para 185.

Shaping Coercive Obligations through Vulnerability  101 for private life.55 Nevertheless, once a state has made the choice to deploy the criminal law vis-a-vis a given act, it must be ‘fully and rigorously applied’.56 Meanwhile, the majority in MC went beyond tolerance of criminalisation and actually required it. And while Judge Tulkens’ call for caution is reasonable in relation to ensuring compliance with the ECHR as a whole – which covers a range of rights, some of which can effectively and sufficiently protected through non-criminal avenues – it can also be argued that this may not always suffice, given the grievous nature of certain violations, especially under Article 3 ECHR, which ‘enshrines one of the most fundamental values of democratic society’.57 This position – and counter-arguments to it58 – will be discussed more thoroughly below. The second obligation under discussion here is the obligation to effectively protect by meaningfully applying the existing criminal law. In 2012, the Court applied this obligation when it examined the alleged rape of a seven-year-old Romanian boy. At the time, rape of juvenile males was not criminally punishable as rape in Romania, but did fall under other criminalised offences.59 This, together with the fact that the law had since changed, satisfied the Court in respect of the duty to criminalise and it did not consider the criminalisation issue further; instead, it noted the subsidiarity of its role, stating that it cannot judge the criminal responsibility of alleged perpetrators.60 Despite this, it expressed concern about the domestic handling of the case, including the failure to weigh conflicting evidence and apply a context-sensitive assessment.61 It found that the authorities had not sufficiently considered the vulnerability of the minor applicant and the special psychological factors at play, which could have explained his hesitation to report and describe the abuse.62 In such cases, Articles 3 and 8 ECHR require respect for the best interests of the child; the Court accordingly regretted the failure to offer the victim counselling or psychological support.63 All of this together, the Court held, left ‘the criminal proceedings in the case devoid of meaning’, and it found a violation of both Articles 3 and 8 ECHR.64 The third obligation requires the taking of steps to prevent ill-treatment of which the state knew, or should have known. In 2014, the Grand Chamber again drew on vulnerability in O’Keeffe v Ireland, which concerned an adult 55 ibid, concurring opinion of Judge Tulkens, para 2. See also F Tulkens, ‘The Paradoxical Relationship between Criminal Law and Human Rights’ (2011) 9 Journal of International Criminal Justice 577. 56 ibid para 3. 57 O’Keeffe v Ireland (n 30) paras 144 and 146. 58 Arguing against this position, see Lazarus (n 4). 59 CAS and CS v Romania (n 40) para 76. 60 ibid paras 77–78. 61 ibid para 78. 62 ibid para 81. 63 ibid para 82. 64 ibid para 83.

102  Corina Heri applicant who, as a child, had been sexually abused in a school setting. It noted the general vulnerability of young children in institutional settings and found that the state ‘must be considered to have been aware of the sexual abuse of children by adults’.65 Thus, it focused not on the vulnerability of a particular child or school setting, but on the fact that sexual abuse of children is a reality known to the state that requires the taking of criminal law measures. This case therefore draws on the state’s ‘empirical social experience’66 and takes a different approach than cases wherein the state’s imputed knowledge of ill-treatment came from concrete indications concerning a particular victim or threat.67 In O’Keeffe, while the Court found that states’ positive obligations should not create an excessive burden for the authorities given ‘the unpredictability of human conduct and operational choices which must be made’, it reiterated the obligations detailed above.68 It relied on X and Y and argued that ‘effective measures of deterrence against grave acts … can only be achieved by the existence of effective criminal-law provisions backed up by law-enforcement machinery’.69 It also noted that ‘the existence of useful detection and reporting mechanisms are fundamental to the effective implementation of the relevant criminal laws’.70 These cases show that the three types of obligations under discussion are interlinked. Effective protection relates to effective deterrence, which requires enforcement of the law. Likewise, the reasonable steps required for the prevention of ill-treatment will often involve the criminal law, thereby relating to the effective protection obligation. Beyond the above cases, the Court has continued to shape state obligations. It has since reiterated that in order to ensure effective deterrence regarding the rape of minors, states must take measures that aim at ensuring respect for victims’ human dignity and that protect the best interests of the child; this means sufficiently investigating the circumstances surrounding an alleged rape and attaching sufficient weight to minors’ vulnerability and special psychology.71 These measures should include preventive measures and, when faced with vulnerable victims, the Court must display particular vigilance and provide increased protection, given that vulnerable minor victims’ ‘capacity or willingness to pursue a complaint’ may be limited.72 Particular attention is also required

65 ibid paras 144–46, 168. 66 D Xenos, The Positive Obligations of the State under the European Convention on Human Rights (Abingdon, Routledge, 2012) 82; Lavrysen (n 1) 133–34 on the ‘knowledge condition’. 67 Compare DMD v Romania App No 23022/13 (ECtHR, 3 October 2017), para 51; Lavrysen (2016) (n 1) 3.1 on the ‘knowledge condition’. 68 O’Keeffe v Ireland (n 30) para 144. 69 ibid para 148. 70 ibid. 71 MGC v Romania App No 61495/11 (ECtHR, 15 March 2016), para 73. 72 IC v Romania App No 36934/08 (ECtHR, 24 May 2016), para 51.

Shaping Coercive Obligations through Vulnerability  103 when it comes to alleged rape in the family context, which may impair victims’ ability to report and describe their rape.73 When it comes to the intersections74 of young age and intellectual disability in the context of sexual violence, the Court has been particularly exacting. It requires states not only to penalise and effectively prosecute all non-consensual sexual acts, regardless of whether there has been physical resistance,75 but also to consider victims’ personal circumstances – specifically their age and disability – and the circumstances of the case in order to avoid misinterpretations of their failure to come forward and to adequately analyse the validity of consent.76 Intellectual disability puts applicants ‘in a heightened state of vulnerability’, which requires particular diligence in analysing their statements.77 ii.  Non-sexual Violence Beyond the context of sexual violence, the Court has referred to vulnerability reasoning in cases concerning child victims of violent acts to support, as above, three separate obligations.78 These obligations require, first, effective deterrence against such acts. Second, the state has a positive obligation to conduct an adequate official investigation into allegations of ill-treatment. Third, where the state knew or should have known about ill-treatment, children have the right to effective state protection. This arises where ongoing violence or abuse comes to the authorities’ attention.79 One context of particular relevance here is the protection of vulnerable minors from corporal punishment. In 1998, the Court considered the complaints of a minor who had been subjected to corporal punishment by his stepfather. Finding that the beating was severe enough to reach Article 3’s threshold of severity, it applied the state’s obligation to take measures designed to prevent ill-treatment, even where administered by private individuals, noting that children and other vulnerable individuals are particularly entitled to protection through effective deterrence.80 The Court found a violation of Article 3 because the domestic criminal law provided inadequate protection, given that it accepted ‘reasonable chastisement’ of a child as a valid defence against a charge of assault.81 In later

73 GU v Turkey App No 16143/10 (ECtHR, 18 October 2016), para 72, translation from the ­original French by the author. 74 K Crenshaw, ‘Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory and Antiracist Politics’ (1989) University of Chicago Legal Forum 139. 75 ibid para 52. 76 IC v Romania (n 72) para 56. 77 ibid. 78 NP and NI v Bulgaria App No 72226/11 (ECtHR, 3 May 2016), para 76; M and M v Croatia (n 37); DMD v Romania (n 67) para 51. 79 M and M v Croatia (n 37); DMD v Romania (n 67) para 51. 80 A v UK ECHR 1998-VI, para 22. 81 ibid.

104  Corina Heri cases, the Court reiterated this, holding that respect for children’s dignity means that domestic courts cannot accept any justifications for corporal punishment that reaches Article 3’s threshold of severity, because this is incompatible with the provision’s absolute nature.82 Another context of relevance is that of domestic violence. As concerns adult victims, this will be considered in more depth below. However, as the Court has noted, domestic violence may also affect children.83 Vulnerability-related considerations apply here when children are themselves the direct victims of violence or where they suffer indirectly because their caregivers are ­victimised.84 States must ‘strive to expressly and comprehensively protect children’s dignity’ through an adequate legal framework which protects them by deterring, preventing and investigating domestic violence.85 However, it is not yet entirely clear what children’s vulnerability means in this context. Below, this chapter will discuss the vulnerability of adult victims of domestic violence. Given that children are already vulnerable given their minor age, their exposure to domestic violence should render them doubly vulnerable, and this intersection of dependence and victimisation should arguably lead to particularly stringent obligations. Nevertheless, the case law on this is not entirely clear, given that the vulnerability-based coercive obligations vis-a-vis child victims of domestic violence are often not differentiated from those vis-a-vis their affected parents or caregivers. Lastly, violence outside the home – for example, that inflicted by strangers or even other children – can also fall under Article 3. Thus, in 2011, the Court considered injuries allegedly inflicted on a teenage boy of Roma origin by his schoolmates. The Court reiterated its previous findings concerning states’ positive obligations vis-a-vis vulnerable minors.86 However, it found that the state cannot be expected to be aware of the risk of accidental injuries (the applicant had, on one occasion, been hurt when another pupil swung his head back and hit him on the nose), which therefore cannot trigger the state’s positive ­obligations.87 In another instance, the applicant had been hit on the head with an iron bar outside his school by unspecified perpetrators. The allegations of recurring violence had been partially accidental (the nose injury) and partially too vague and unspecific to trigger the state’s obligation to take concrete steps to protect the applicant (the iron bar incident), as they did not sufficiently detail the time, place and circumstances of the ill-treatment.88 This case indicates where the limits of the obligation to prevent ill-treatment lie, even when the risk of harm pertains to vulnerable children: states cannot be expected to prevent all violence everywhere, especially not when it is accidental, and their obligation

82 DMD

v Romania (n 67) para 51. v Turkey ECHR 2009, para 132. 84 NP and NI v Bulgaria (n 78) para 79. 85 DMD v Romania (n 67) para 51. 86 Đurđević v Croatia ECHR 2011 (extracts), para 102. 87 ibid para 113. 88 ibid paras 116–17. 83 Opuz

Shaping Coercive Obligations through Vulnerability  105 to take measures against known ill-treatment applies only if they have sufficient ­information to be able to intervene, even when it comes to allegations of recurring violence. It may also make a difference whether the uncertainty in question concerns whether a harm took place or its specifics (ie, information that would have allowed the state to anticipate it). Even a harm that in fact occurred may not violate the state’s positive obligations to take reasonable steps to prevent ­ill-treatment where it was not foreseeable. iii.  Abuse by the State and Protection from the Criminal Law As mentioned earlier, other authors have referred to the ‘ambiguity’ of a Convention that protects both the perpetrators and the victims of a crime.89 In this regard, it should be noted that the same ‘ambiguity’ applies to vulnerability reasoning, in the sense that vulnerability and its impact on coercive action cuts both ways: it not only protects people by requiring coercive action, but also protects them from (abuses of) coercive power. In this vein, the Court has heard a number of cases brought by minors who were ill-treated during police interrogation. The Court has repeatedly found that the state has an obligation to dissuade and punish such acts by meaningfully applying the criminal law. For example, in the 2006 Okkali v Turkey judgment, the Court held, concerning the imposition of minimum and suspended sentences on police officers found guilty of ill-treating a 12-year-old boy, that the domestic criminal system was ‘far from being rigorous’ and did not have the dissuasive effect required to effectively prevent the ill-treatment.90 It concluded that the domestic criminal proceedings had not afforded appropriate redress and that no civil action could have remedied the violation of Article 3 at stake.91 This finding was picked up again in 2015, concerning alleged ill-treatment of a 17-year-old applicant during police interrogation. The perpetrators were found guilty of criminal charges, but their sentences were later suspended.92 Examining these proceedings, the Court recalled the particular vulnerability of children to violence and their entitlement to state protection and effective deterrence. It then found that not only had the authorities failed to give extra consideration to the applicant’s vulnerability, but that the perpetrators’ impunity raised doubts about the domestic judicial machinery’s ability ‘to produce a sufficiently deterrent effect to protect anybody at all, minor or otherwise’.93 So, in other words, the Court’s understanding of the vulnerability of children also protects the targets of coercive measures. This not only applies to minors (adults, too, are entitled to protection and care given their vulnerability – due to



89 Lazarus

(n 4) 137. v Turkey ECHR 2006-XII, para 78. 91 ibid paras 70 and 78. 92 Ateşoğlu v Turkey (n 38) paras 11–12. 93 ibid para 27. 90 Okkali

106  Corina Heri their dependency on the state – when detained).94 However, it means that states have special obligations whenever a vulnerable minor is detained or affected by policing operations and must protect them more strongly than they must an adult in a comparable situation. For example, when minors suffer violence perpetrated by their co-detainees, proper regard for the detained minors’ vulnerability means that a formal complaint should not be required as a precondition for penal measures.95 When vulnerable minors are detained in the migration context, their extreme vulnerability takes precedence over their migration status, their detention conditions must be child-appropriate, and detaining them may be considered categorically impermissible outside of a certain limited timeframe.96 Lastly, when children may potentially be affected by coercive measures taken against their caregivers, the state must take this into account in planning and executing police operations.97 In addition, it should be noted that, in the context of criminal proceedings, other Convention provisions also mandate particular protections for minors. When minors are accused of a crime, the resulting criminal proceedings must respect their best interests, and while it is not incompatible with the ECHR for states to prosecute children as young as 10, their trial and sentencing must be modified to reflect their age and vulnerability.98 Children charged with an offence must be treated in a manner that takes full account of their age, maturity, and intellectual and emotional capacities.99 This includes taking steps to minimise the child’s sense of intimidation and inhibition, and ensuring that he or she broadly understands the nature of the investigation, what is at stake, the potential punishment and his or her right to remain silent.100 B.  Victims of Sexual and Domestic Violence The above cases show that, in the context of Article 3, the vulnerability of ­children requires recourse to the criminal law when it comes to the sexual abuse and rape of children. The question now is whether the same obligation applies when adults are the victims of sexual or domestic violence. The short answer is that while adult victims of domestic violence are considered vulnerable,101 this is less clear for adult victims of sexual violence and that the requirements in this regard are in any event not identical to those vis-a-vis minors. 94 On this, cf, for example, Slimani v France ECHR 2004-IX (extracts), para 27. 95 AŞ v Turkey App No 58271/10 (ECtHR, 13 September 2016), para 73. 96 cf AM and Others v France App No 24587/12 (ECtHR, 12 July 2016), with further references, including to Mubilanzila Mayeka and Kaniki Mitunga v Belgium ECHR 2006-XI, para 55. 97 Slavov and Others v Bulgaria App No 58500/10 (ECtHR, 10 November 2015), paras 82–83; Govedarski v Bulgaria App No 34957/12 (ECtHR, 16 February 2016), para 62. 98 V v UK ECHR 1999-IX, para 64. 99 Blokhin v Russia ECHR 2016, para 195. 100 ibid. 101 Opuz v Turkey (n 83) para 160.

Shaping Coercive Obligations through Vulnerability  107 The below discussion concerns ill-treatment inflicted by private parties, but it should be made clear at the outset that physical and sexual violence inflicted by state actors, especially on victims who are in the custody of the state and therefore vulnerable, also falls under Article 3.102 In the case of allegations of ill-treatment in custody, the state has an obligation to protect those in its care; given that the events lie largely within the exclusive knowledge of the authorities, strong presumptions of fact arise where detainees suffer injuries, and the burden of proof is then on the state to satisfactorily and convincingly explain these injuries.103 When it comes to sexual violence, all of the vulnerability-related Article 3 cases to date, save one,104 have intersected with minor age (as discussed above), with intellectual disability105 or with detention (whereby the principles outlined in the previous paragraph apply).106 Concerning adults with intellectual or cognitive disabilities, the Court acknowledges that their ability to complain of ill-treatment may be impaired, and requires states to adequately respond to and investigate alleged sexual abuse or risk producing ‘a background of impunity’ that may violate Article 3.107 Deficiencies regarding the effective investigation and prosecution of ill-treatment are therefore aggravated by the vulnerability of disabled adult applicants.108 What, then, of victims of sexual violence who are not minors, or ­disabled, or in detention – are they vulnerable? The only case in which the Court applied the vulnerability concept to this type of situation was truly grievous: the a­ pplicant had been kidnapped, threatened with sale into prostitution, raped, had attempted to defenestrate herself and was ultimately hospitalised for psychiatric reasons. In that case, the Court held that the manifest lack of diligence by the domestic authorities and the excessive length of proceedings (including an eight-year preliminary investigation) had violated Article 3, given the negative effects on the applicant, who was ‘clearly psychologically very vulnerable as a result of the attack’.109 The finding of vulnerability here seems to be linked to the severe psychological repercussions of the applicant’s ordeal, and while this chapter would in no way suggest that sexual assault is ever not traumatic, it is unclear whether the Court would apply the concept of vulnerability in cases with less severe psychological effects. Given that vulnerability has not been used in other non-intersecting sexual assault cases, it is possible that the Court may consider the victims of such cases adequately able to come forward about their

102 cf Maslova and Nalbandov v Russia App No 839/02 (ECtHR, 24 January 2008); Bouyid v Belgium ECHR 2015. 103 Bouyid v Belgium (n 102) para 83. 104 SZ v Bulgaria App No 29263/12 (ECtHR, 3 March 2015). 105 cf W v Slovenia App No 24125/06 (ECtHR, 23 January 2014). 106 cf Aydin v Turkey (n 40). 107 IC v Romania (n 72) para 55. 108 W v Slovenia (n 105) para 70. 109 SZ v Bulgaria (n 104) para 39.

108  Corina Heri experiences and endure the resulting proceedings, even if these are protracted. This approach must be flagged as problematic, as will be further discussed below, because of the very nature of sexual violence, which – as the Court has acknowledged – ‘leaves deep psychological scars on the victim which do not respond to the passage of time as quickly as other forms of physical and mental violence’.110 Though the experiences and responses of each victim of sexual violence are different, and the Court can and does find Convention violations where victims of sexual violence are not considered vulnerable, being considered vulnerable works in applicants’ favour.111 Therefore, declining to consider the vulnerability of whole groups of applicants raises an equality problem, a problem that is particularly acute where there are considerable grounds for a finding of vulnerability, as there may be here. Turning to the issue of domestic violence, the leading vulnerability-related case is Opuz v Turkey. This case concerned a gruesome pattern of domestic violence inflicted on the applicant by her husband over the course of several years, during which time she was severely injured multiple times; ultimately, her husband shot and killed her mother.112 While the Court does not consider women vulnerable as a group,113 it found in this case that the applicant was vulnerable, given the violence and threats she had suffered, her fear of further violence and her social situation as a woman in south-east Turkey.114 Based on this vulnerability, the Court found that the domestic legal framework had failed to ‘provide for specific administrative and policing measures designed to protect vulnerable persons against domestic violence’.115 It held that while it was not its role to replace the national authorities and determine which measures would have been appropriate, the legislative framework should have – given the severity and continuing threat of violence – allowed the authorities to pursue criminal investigations even after the applicant and her mother had withdrawn their criminal complaints.116 The required due diligence had not been exercised; the applicant’s husband had perpetrated his repeated violent attacks ‘without hindrance and with impunity’.117 As a result, the Court held that the state’s response ‘was manifestly inadequate to the gravity of the offences’ and the judicial decisions taken showed ‘a lack of efficacy and a certain degree of tolerance, and had no noticeable preventive or deterrent effect’.118 Accordingly, it found a violation of Article 3.119



110 Aydin

v Turkey (n 40) para 83. Heri (n 10); Peroni and Timmer (n 14); Timmer (n 16) and the case law cited herein. 112 Opuz v Turkey (n 83). 113 cf Valiulienė v Lithuania App No 33234/07 (ECtHR, 26 March 2013), para 69. 114 Opuz v Turkey (n 83) para 160. 115 ibid para 171. 116 ibid paras 149 and 168. 117 ibid paras 169–70. 118 ibid para 170. 119 ibid para 176. 111 See

Shaping Coercive Obligations through Vulnerability  109 Since Opuz, the Court has held on various occasions that states must effectively protect victims of domestic violence, given their vulnerability and the violence against them, by providing for effective deterrence and launching an official investigation.120 It has also reiterated that states have positive obligations to establish and effectively apply a system to punish all domestic violence and to provide victims with sufficient safeguards.121 Findings of domestic violence victims’ vulnerability have been based on the fact that these victims often do not report violence,122 their fear of repeated violence123 and their experiences, which cause in them ‘feelings of fear, vulnerability and insecurity’.124 The Court has emphasised the need to take into account victims’ precariousness and vulnerability, which can be ‘particular, moral, physical and material’.125 Specifically for the obligation to investigate, the threshold is quite low here: precisely because of victims’ vulnerability, the authorities must investigate as soon as there is a credible assertion of repeated domestic violence or harassment.126 When it comes to the effectiveness of the criminal investigation, this ‘does not necessarily require a conviction, but effective implementation of the law, particularly criminal’, and in particular it requires prompt state action to ensure that domestic law provides more than theoretical protection.127 There are a number of limits on coercive obligations in this context. First, as recently as 2008, the Court held in the context of Article 8 ECHR that, regardless of the vulnerability of victims of domestic violence, states must not necessarily turn to the criminal law and provide for state-assisted prosecution of perpetrators, but have a margin of appreciation in the choice of means.128 In other words, vulnerability-based reasoning does not require the criminalisation of all acts of physical violence (ie, those that fall below the threshold of Article 3), even against vulnerable victims. Second, vulnerability reasoning has not yet been applied to victims of sexual violence without some intersection with another ground for vulnerability; this is problematic, given that many of the factors that render victims of domestic violence vulnerable (their lack of willingness to report and the psychological effects of their ordeals) are also at play here.

120 EM v Romania App No 43994/05 (ECtHR, 30 October 2012), para 58; TM and CM v Republic of Moldova App No 26608/11 (ECtHR, 28 January 2014), para 60; Rumor v Italy App No 72964/10 (ECtHR, 25 May 2014), paras 58 and 60; MG v Turkey App No 646/10 (ECtHR, 22 March 2016), para 76. 121 Bălşan v Romania App No 49645/09 (ECtHR, 23 May 2017), para 57. 122 TM and CM v Republic of Moldova (n 120) para 60. 123 Rumor v Italy (n 120) para 60. 124 MG v Turkey (n 120) para 105, translation by the author. 125 ibid para 76. 126 Irina Smirnova v Ukraine App No 1870/05 (ECtHR, 13 October 2016), para 79. 127 Talpis v Italy App No 41237/14 (ECtHR, 2 March 2017), paras 103–06. 128 Bevacqua and S v Bulgaria App No 71127/01 (ECtHR, 12 June 2008), para 82.

110  Corina Heri C.  Members of the LGBTIQA+ Community The last group of cases to be examined here concerns Article 3 cases brought by members of the LGBTQIA+ community. Like the above, this jurisprudence to some extent highlights the limits of the Court’s approach to vulnerability and coercion. There is still very little vulnerability-related Article 3 case law about criminalising acts of violence or hate speech based on victims’ sexual orientation or gender identity. The most well-known case concerning such violence is Identoba and Others v Georgia, wherein the Court considered participants in an LGBTQIA+ rights march to be part of a ‘vulnerable community’ based on a ‘history of public hostility’ towards this group in Georgia.129 The imputed acts in this case were already criminalised under domestic law;130 in its judgment, the Court found that the authorities had failed to adequately protect the applicants from violence, to apply the appropriate provisions of the domestic criminal code and to take reasonable steps to unmask discriminatory motives for the acts in question.131 To understand how the Court deals with cases where there is no existing criminalisation of the acts concerned, it is helpful to turn to the factually rather similar MC and AC v Romania case, where there was no law on the books that criminalised hate crimes based on sexual orientation. There, the Court invoked the need to particularly protect vulnerable individuals132 and held that the domestic authorities’ failure to initiate criminal proceedings against the perpetrators of homophobic violence had violated the Convention. It came to this conclusion because, it argued, the acts in question – which involved a violent attack accompanied by homophobic hate speech – could have been ‘assigned a legal classification that would have allowed the proper administration of justice’.133 The authorities are required to be ‘rigorous’ here because otherwise crimes based on prejudice are treated ‘on an equal footing with cases involving no such overtones, and the resultant indifference would be tantamount to official acquiescence to, or even connivance with, hate crimes’.134 In other words, here, while not requiring the existence of a specific criminal offence, the Court nonetheless required criminalisation of the hate-based violence in question. Apart from these cases, case law concerning the criminalisation of haterelated acts, whether framed as relevant to vulnerability or not, is scarce.135 Much more plentiful, by contrast, is jurisprudence concerning the decriminalisation of homosexual acts. This serves to show that there are two sides to the



129 Identoba

and Others v Georgia App No 73235/12 (ECtHR, 12 May 2015), para 72. paras 72 and 77. 131 ibid paras 74, 75 and 78. 132 MC and AC v Romania App No 12060/12 (ECtHR, 12 April 2016), para 114. 133 ibid para 124. 134 ibid. 135 See also Stephanos Stavros, ch 6 in this volume. 130 ibid

Shaping Coercive Obligations through Vulnerability  111 coin when it comes to the Court’s review of criminalisation: it determines not only whether compliance with the Convention requires criminalisation of a given conduct, but also whether it precludes criminalisation.136 The main and earliest case of note in this regard is the 1981 judgment in Dudgeon v UK, which concerned legislation that classified homosexual relations between consenting adult men as a criminal offence.137 The Court held that there had been a violation of Article 8 ECHR because the restriction, given its breadth and absolute character, had been disproportionate to its aims. Incidentally, vulnerability also entered into play in Dudgeon, namely as the justification for the coercive measure in place: the respondent state argued that the law aimed to protect ‘vulnerable members of society, such as the young, against the consequences of homosexual practices’.138 The Court held that there was insufficient proof of harm to these vulnerable people and thus no sufficient ‘pressing social need’ for the law. Of course, the test under Article 3 is different – there is no ‘pressing social need’ test and no legitimate aim analysis. Instead, coercive obligations are required to provide effective deterrence against acts that the Court describes as particularly grave, with this gravity being intensified but not per se constituted by the vulnerability of those affected. IV.  VULNERABILITY AND THE RISK OF COERCIVE OVERREACH

As highlighted above, states have a number of different positive obligations under Article 3, from an obligation to criminalise certain acts, to the obligation to provide for an effective investigation, to the obligation to protect individuals from harm under certain circumstances. While they do not require states to prevent every risk of ill-treatment from materialising or to criminalise every human rights violation, these obligations together mean that the Court has quite extensive coercive reach. The requirements specific to vulnerable individuals are one part of this. Under Article 3, vulnerability has a number of effects, including providing the rationale for criminalising certain harms, allowing the Court to impute knowledge of ill-treatment to the state, making the need for effective protection particularly acute and permitting the accommodation of inconsistencies or procedural issues in claims from applicants who face particular hurdles in accessing protection and redress. However, there are various constraints to this jurisprudence. For one thing, vulnerability-related positive (and negative) ­obligations also protect those affected by state coercion. Moreover, the decision not to criminalise a given act can be subject to scrutiny just like the decision to criminalise another. In addition, vulnerability does not necessarily mean that a

136 cf

Söderman v Sweden (n 3), concurring opinion of Judge Pinto de Albuquerque, para 18. v UK (1981) Series A No 45. 138 ibid para 47. 137 Dudgeon

112  Corina Heri given act must be criminalised. Lastly, vulnerability does not apply to all victims of ill-treatment. Liora Lazarus has noted the alleged ambiguity that arises from the fact that human rights both constrain and require coercion139 or, in other words, that they protect both the suspected and convicted perpetrators of a given proscribed act, and their victims. Writing on the Opuz case, she explored the Court’s finding that ‘in domestic violence cases perpetrators’ rights cannot supersede victims’ rights to life and to physical and mental integrity’140 and considered it to be ‘striking that the countervailing rights and interests of those subject to protective measures received so little regard here’.141 She also refers to an example of a case before the UK House of Lords, wherein a 15-year-old boy was convicted of the crime of ‘rape of a child under 13’ after having sexual intercourse with a 12-year-old girl, whom he believed to be older and who expressed her willingness to have sexual intercourse with him.142 The boy argued before the House of Lords that he should have been sentenced instead for the crime of ‘child sex offences committed by children or young persons’, given the circumstances of the case, the stigma of being convicted of rape and the effects of this on his private life as protected by Article 8 ECHR. Baroness Hale wrote, in response to this argument, that the applicant, who was ‘vulnerable by reason of her age if nothing else’, deserved protection and respect. The domestic law tried to provide this, which she did not consider to mean ‘a lack of respect for the private life of the penetrating male’.143 This frames the issue as one of two competing rights under Article 8; the resulting balancing exercise, according to Lazarus, was not sufficiently stringent, but ‘rested on broad brush rhetorical assertions of the protective duties of the state, and included almost no analysis of the content of the competing right’.144 To better understand what this means for the present context, it should be noted that a balancing exercise is applied where the equally positioned ECHR rights of two individuals conflict. In conducting this exercise, to ensure consistency, the Court must reach the same outcome regardless of the perspective from which a case is brought.145 The relevant case law, for example, balances the right to freedom of expression of one party against the right to respect for private life and reputation of another who is affected by that expression; in doing so, ‘the outcome of the application should not, in principle, vary’ depending on which of the parties concerned lodges the claim before the Court.146 This is due to the fact that ‘as a matter of principle these rights deserve equal respect’.147

139 Lazarus

(n 4) 137. v Turkey (n 83) para 147. 141 Lazarus (n 4). 142 R v G [2008] UKHL 37, as discussed in Lazarus (n 4). 143 ibid para 54, as discussed in Lazarus (n 4). 144 Lazarus (n 4). 145 Annen v Germany App No 3690/10 (ECtHR, 26 November 2015), para 56. 146 ibid. 147 ibid. 140 Opuz

Shaping Coercive Obligations through Vulnerability  113 However, this is not necessarily true when other Convention rights conflict with Article 3 ECHR. In this regard, the Court has recognised that ‘every human being has an absolute, inalienable right not to be subjected to torture or to inhuman or degrading treatment under any circumstances, even the most difficult’.148 Article 3 does not allow for exceptions, justification of interference or balancing of interests, irrespective of the conduct of those concerned and the nature of the relevant offence.149 Thus, it is not possible to balance this provision against the rights of others, or at least not in the way outlined above.150 That said, there is necessarily a degree of discretion and interpretation involved in the positive obligations under Article 3, given that these are often duties to take reasonable measures. Indeed, even the negative obligation under Article 3 contains a certain ambiguity in this regard, given that its protection is absolute, but its minimum level of severity is relative.151 When it comes to positive obligations, what is reasonable can vary according to the circumstances, and these obligations are thus less conducive to the establishment of an absolute standard than the negative enjoinment not to ill-treat. The absoluteness of Article 3 cannot demand that the state must prevent every instance of ill-treatment from materialising, or that every act of ill-treatment meeting the threshold of Article 3 should be met with a criminal law response, or that every perpetrator must be convicted and punished, or that a specific provision of the criminal code must necessarily be applied. Nevertheless, these variables aside, there remain absolute and ‘non-displaceable’ obligations under Article 3, both in the negative and positive senses.152 Does this mean that persons who violate others’ Article 3 rights lose their right to the protection of their own Convention rights, save perhaps for their own equally most fundamental rights under Articles 2 and 3? Certainly not, even though some of their rights, such as their rights to physical liberty, reputation or family life, may be limited to some degree in the context of criminal proceedings or the execution of a sentence. What this means here is that the state must take measures to ensure that the rights of victims are safeguarded along with those of perpetrators. However, it would be incongruous to argue that (if attenuated by all the appropriate safeguards) the use of the criminal law, where oriented at criminalising violations of fundamental human rights, per se violates perpetrators’ rights.

148 Gäfgen v Germany ECHR 2010, para 107. 149 ibid para 107. 150 cf, for a much more nuanced study of this, S Smet, ‘The “Absolute” Prohibition of Torture and Inhuman or Degrading Treatment in Article 3 ECHR: Truly a Question of Scope Only?’ in E Brems and J Gerards (eds), Shaping Rights in the ECHR (Cambridge, Cambridge University Press, 2013). 151 N Mavronicola, ‘Is the Prohibition against Torture and Cruel, Inhuman and Degrading Treatment Absolute in International Human Rights Law? A Reply to Steven Greer’ (2017) 17 Human Rights Law Review 479. 152 For a more developed if not quite identical argument, see ibid 488–90.

114  Corina Heri What, then, of the provocative example before the House of Lords employed by Liora Lazarus? Lazarus frames this case as a question of two conflicting claims under Article 8 – that of the 15-year-old perpetrator and that of the 12-year-old victim – and argues that that of the older child unjustifiably received less weight. The present chapter cannot determine whether the conviction and sentencing of this particular boy was justified. However, it does not need to do so, because it is concerned with obligations under the ECHR as interpreted by the ECtHR, which has never demanded that facts similar to those in this case must be prosecuted as rape and not another criminal offence. Indeed, regarding sexual assault, even where there is an obligation to apply the criminal law, states have quite a bit of leeway when it comes to the precise qualification of a given act under domestic criminal law.153 Lazarus herself also does not seem to be arguing that recourse to the criminal law in and of itself represented an erroneous choice of means in this case. The question instead concerns the fairness of convicting the boy of rape instead of a lesser offence. The Court would surely leave this choice up to states, while likely emphasising that both parties in this case are vulnerable minors and determining whether the domestic proceedings adequately responded to their vulnerability. However, the framing of this situation both by Lazarus and Baroness Hale presents the criminal law as an act of balancing between the interests of perpetrators and victims. This insinuates that the prosecution of rape – including here statutory rape – and other forms of sexual assault should be seen from the perspective of the perpetrator just as much as from the perspective of the victim. It could also be seen as meaning that the prosecution of these acts only serves the purpose of punishment or retribution, instead of also pursuing various other purposes, including, for example, prevention, reformation and restoration. Again, Lazarus’ argument seems not to be whether this is an issue for criminal law, but that the criminal law was applied too harshly. Focusing on this example distracts from the fact that coercive obligations, at least under Article 3 ECHR, primarily mean that for certain egregious human rights violations, and especially where the victims are at a disadvantage because of their vulnerability, the Court has found that the only way to give human rights the ‘bite’ they need in order to be effective in practice is to turn to the criminal law, officially investigate potential violations and take measures to protect those who may be at risk.154 In effect, obligations to investigate will lose much of their salience if there is no criminal law basis on which to conduct an official investigation, and this in turn means that there is little the state can do to protect people against ill-treatment.155 This does not mean that the criminal law alone serves to deter violations of Article 3 or to guarantee non-repetition. However, as the Court has 153 CAS and CS v Romania (n 40) para 73. 154 At least in the Court’s case law. For counter-arguments, see K Engle, Z Miller and DM Davis, Anti-impunity and the Human Rights Agenda (Cambridge, Cambridge University Press, 2017). 155 As shown by Opuz v Turkey (n 83).

Shaping Coercive Obligations through Vulnerability  115 found, non-criminal measures such as civil remedies do not always suffice, and it is here that the criminal law comes into play. Vulnerability, in this context, requires a degree of attention and diligence from the domestic authorities. If, generally speaking, the concept serves as a tool that allows for ‘context-sensitive judgments’ and prioritising the interests of the vulnerable,156 here it requires particular sensitivity for a particular aspect of the context, namely the victim’s experience. For the present purposes, this is described as a victim-oriented perspective. This is one that makes it possible to obtain redress because it is responsive to the context and lived reality of an applicant, and can capture inequalities and imbalances of power.157 It is sensitive to how trauma affects victims and their statements, so that there can be a realistic possibility of a criminal investigation and of bringing perpetrators to justice. This does not mean that it requires that every alleged perpetrator of sexual assault or other ill-treatment must be found guilty or receive a particular sentence, or that victims may dictate how a given procedure must be managed. It does mean that if individuals are to have a meaningful right to be protected from ill-treatment, it is necessary that when any entity – be it the state authorities or a private person – violates this most fundamental right through grievous acts such as rape or violence, it is appropriate to deploy the criminal law. If the problem is not the application of criminal law, but the fact that human rights mandate the application of the criminal law (more concretely, that the ECtHR does so), then we must ask ourselves the following question: what authority could the Court claim to have (and what impact would Convention rights have) without any possibility to require deployment of the criminal law, at least against some of the most grievous wrongs under the Convention? Again, it should be noted that these coercive obligations are generally deployed in cases concerning heinous human rights violations, supported by national and international law, and do not require that anyone be found guilty of a given crime or be given a specific sentence;158 indeed, and importantly, the Court’s case law does not require that a criminal wrong must have occurred in order to find a violation of Article 3. These coercive obligations also do not invalidate the safeguards that protect suspected and convicted perpetrators. Instead, they provide a certain minimum content of positive obligations, without which the underlying human rights could not be rendered practical and effective. Vulnerability, in this context, serves as a lens that focuses attention on those victims who are least likely to be treated respectfully, believed, heard or granted justice. This allows for a more nuanced and targeted understanding of coercive positive obligations, as opposed, for example, to the blanket obligation to ensure the criminalisation

156 Timmer (n 16) 162–64. 157 Using this term differently, see T van Boven, ‘Victim-Oriented Perspectives: Rights and R ­ ealities’ in T Bonacker and C Safferling (eds), Victims of International Crimes: An Interdisciplinary Discourse (Asser, Springer, 2013). 158 See Lavrysen, ch 2 in this volume.

116  Corina Heri of all violations of an individual’s physical or psychological integrity. In this sense – and because it also serves to protect vulnerable applicants from coercive power – it may actually temper the risk of ‘coercive overreach’. The fact that the ECHR both requires and constrains coercion may seem ‘ambiguous’ to some extent. However, this ‘ambiguity’ is central to human rights law in the sense that states are both the main guarantors and the main violators of human rights. The above cases represent just one snapshot of how the ECHR protects individuals from the coercive power of the state, but indeed much of the mechanism of the Convention is aimed at doing just this – from the rights to a fair trial and an effective remedy, to the presumption of innocence, to the protection of people in the custody of the state, who are vulnerable even if they are adults,159 to the large range of other guarantees that constrain the state’s coercive power. V. CONCLUSION

Like vulnerability reasoning itself, coercive obligations serve to make Convention rights and freedoms practical and effective, instead of theoretical and illusory. It is not always clear where the various obligations to deter and/or criminalise, investigate and prevent violations of Article 3 begin and end, for there is considerable overlap between them, but it is clear that there are, under certain circumstances, coercive obligations under Article 3 ECHR. These obligations are informed by the vulnerability of certain applicants, for example, children. States have an obligation to criminalise ill-treatment against vulnerable people, along with other positive obligations. In these cases, states must take a victim-oriented perspective, engage with alleged victims of ill-treatment in a way that makes it possible to determine whether there has been ill-treatment and, if so, to take appropriate measures. It neither dictates the concrete modalities of the domestic criminal law nor requires that a given perpetrator be found guilty or receive a particular sentence; instead, the Court examines whether the domestic law is arranged in a way that promises to provide effective deterrence against ill-treatment, which in some cases has been found to require the deployment of the criminal law. However, vulnerability does not always automatically entail coercive obligations. In some cases, the Court has considered laws arguably promulgated to protect vulnerable individuals to be incompatible with the Convention. In other, less grievous cases, the Court has deferred to the states’ choice of means to ensure compliance with the Convention. What emerges is that vulnerability-based reasoning can lead to an obligation to criminalise, but it may actually temper the risk of ‘coercive overreach’ by targeting specific situations and circumstances, and not requiring a criminal law response to all violations of Article 3.

159 On

this, cf Slimani v France (n 94) para 27.

6 Criminal Law Responses to Hate Speech Towards a Systematic Approach in Strasbourg? STEPHANOS STAVROS*

I. INTRODUCTION

H

ate speech is on the increase everywhere in our continent. So are the racism and intolerance to which it is linked.1 Various factors have contributed to the rise of these phenomena, including social tensions related to growing inequality, conflicts in and around Europe,2 populist appeals to bias and technological change in the way in which information is shared and views circulate. Hate speech is the product of prejudice, which may derive from ideology, but may also be rather superficial (not less stigmatising for being so). It can be linked to denial of rights and hate crime. Therefore, the fight against it must be seen as part of the struggle against racism, discrimination and intolerance. The victims of hate speech are many. Usually, they will be targeted by virtue of belonging to groups traditionally discriminated against on grounds of ‘race’,3

* I am grateful to Lawrence Early, Robert Spano and the two editors for their comments; any errors are my own and the views expressed do not bind the organisation I work for. This chapter takes into account developments up to 31 October 2019. 1 See recent reports by Council of Europe (CoE) human rights monitoring bodies and EU agencies: https://rm.coe.int/annual-report-2018/168094d6fe and http://fra.europa.eu/en/speech/2017/ freedom-expression-and-combat-hate-speech-europe. 2 Migration to and across Europe is usually related to conflicts and unequal economic opportunities. 3 Although all humans belong to the same species, some are discriminated against because of perceived ‘racial’ differences.

118  Stephanos Stavros national/ethnic origin, colour, citizenship, religion or belief, language, gender or sexual orientation. However, another personal characteristic, such as age, disability or status, may also become the triggering factor.4 The victims’ need for protection has always been recognised as a legitimate concern by the European Court of Human Rights (hereinafter ECtHR or ‘the Court’). The Court has often referred in this connection to the work of the European Commission against Racism and Intolerance (ECRI), the Council of Europe (CoE)’s specialised monitoring mechanism on racism, ‘racial’ discrimination,5 xenophobia, anti-semitism and intolerance, which is discussed in section II below.6 However, until recently, the Court appeared reluctant to accept that states had an obligation – under the European Convention on Human Rights (hereinafter ECHR or ‘the Convention’) – to prosecute hate speech (as opposed to hate-motivated violence). The emphasis of the Court’s case law was on freedom of expression under Article 10 of the Convention (see section III below). The hate-speech victims’ interests were taken into account only in the context of the second paragraph of Article 10, which allows – in some circumstances – for limitations thereof (or when the Court would apply, in exceptional cases,7 Article 17, which prohibits the abuse of any Convention right). The aim of this chapter is to examine the implications of four recent ECtHR judgments finding a violation of either Article 88 (right to respect for private and family life) or Article 9 (freedom of thought, conscience and religion) of the Convention, as a result of the authorities’ failure to mobilise the criminal law to counter some forms of hate speech (see section IV below). These judgments are welcomed, in section V below, as a natural development of the Convention case law on positive obligations. Following this, two questions are discussed: in what kind of cases would/ should a Convention-based obligation to prosecute hate speech arise (section VI)? And how would/should this obligation be discharged (section VII)? Finally, conclusions are drawn in section VIII.

4 All these concepts are used by the European Commission against Racism and Intolerance (ECRI) in its General Policy Recommendation (GPR) No 15 on combating hate speech. In my view, gender identity is an aspect of gender, while descent is a status issue. 5 Discrimination based on ‘race’, colour, national/ethnic origin, citizenship, religion, language, sexual orientation or gender identity. 6 ECRI was set up in 1993. It organises a monitoring visit to each CoE Member State every five years, followed by dialogue with its authorities and the adoption of a report. The latter would contain a description of the situation in the country, ECRI’s assessment thereof and its recommendations. ECRI also issues GPRs addressed to all CoE governments. 7 Perinçek v Switzerland [GC] App No 27510/08 (ECtHR, 15 October 2015), ECHR 2015-VI, para 114. 8 On its own or with art 14, which prohibits discrimination in the enjoyment of Convention rights.

Criminal Law Responses to Hate Speech  119 II.  THE COE AND HATE SPEECH

The CoE, whose tasks include ensuring respect for human rights in all its 47 Member States, is a complex organisation. Political bodies adopt standards, often developed by committees of government experts. Various organs monitor states’ compliance therewith; many of them, including the ECtHR, function independently of national governments. Finally, international civil servants take the initiative of organising different kinds of activities to assist in the implementation of these standards. While all the parts of the organisation work towards common goals, occasionally their strategies are not identical. The CoE takes pride in the number of international treaties it has produced. However, soft law – which may take the form of recommendations by political or monitoring bodies – is also an important device that is often resorted to by the organisation in its efforts to enhance human rights protection across the continent. One area in which the CoE experiments with different tools is hate speech.9 Its Committee of Ministers (CM) has produced Recommendation No R(97)20.10 Its secretariat has organised a youth campaign against such speech online. Other parts of the organisation seek to raise awareness among additional target groups, including parliamentarians11 and journalists, or organise expert reviews of the compatibility with the ECHR of ‘laws and practices that place restrictions on freedom of expression’.12 Of course, no discussion concerning the fight against hate speech in Europe may ignore the contribution made by ECRI. For ECRI, hate speech is the advocacy, promotion or incitement, in any form, of the denigration, hatred or vilification of a person or group of persons on one of the grounds mentioned in the introductory paragraphs of this chapter. It is also any harassment, insult, negative stereotyping, stigmatisation or threat in respect of such a person or group of persons and the justification of all the preceding types of expression. This definition is contained in ECRI’s General Policy Recommendation (GPR) No 15 into which it has distilled the experience it has gathered during 25 years of monitoring European states’ responses to this phenomenon. GPR No 15 reminds us that states need different tools to tackle hate speech online and offline. These would include criminal, civil or administrative law, self-regulation, prevention policies and more speech – in terms of counter-­narratives as well as positive messages. The GPR’s aim is to offer

9 https://www.coe.int/en/web/no-hate-campaign/coe-work-on-hate-speech. 10 Drafted by a committee of government experts whose work is continued by the Steering Committee on Media and Information Society (CDMSI). While the latter has taken many normative initiatives, none appears to target hate speech specifically; see https://www.coe.int/en/web/ freedom-expression/hate-speech. 11 www.assembly.coe.int/nw/Page-EN.asp?LID=NoHate. 12 Proposed by the part of the secretariat that provides support to the CDMSI; see https://www. coe.int/en/web/freedom-expression/hate-speech.

120  Stephanos Stavros national authorities an à la carte menu from which they may choose the most appropriate response in each set of circumstances. Criminal law could be such a response: [A]gainst the use, in a public context, of hate speech which is intended or can reasonably be expected to incite acts of violence, intimidation, hostility or discrimination against those targeted by it … provided that no other, less restrictive, measure would be effective and the right to freedom of expression and opinion is respected.13

One may find examples of situations where criminal sanctions would have been ECRI’s preferred option by looking at its state-monitoring reports.14 Some of the GPR’s responses to hate speech clearly involve direct state interference with freedom of expression. This makes it ‘controversial’, as some regard any such legislation as quintessentially illiberal. Others would argue that not enough is done to protect vulnerable groups from expressions of hatred.15 Comparisons are drawn between the European and the US approach. Attitudes harden when the discussion focuses on the criminal law response. Some members of vulnerable groups want it because of its symbolic value, the powerful message that it carries. They also consider it preferable – or even necessary – for practical reasons, as in this case the main resource-mobilisation burden is borne by the state. Others find it particularly objectionable, making reference to the very function of the criminal law and the need for restraint in terms of its mobilisation.16 All relevant arguments have been rehearsed – at one time or another – in front of the ECtHR, the final arbiter in such matters in Europe. It is well known that the Court, when interpreting the Convention, often relies on the common European ground,17 other CoE standards and international standards. In cases concerning many forms of discrimination as well as hate-motivated violence and hate speech, the Court often cites ECRI’s findings, assessments and ­recommendations – even in the main part of its judgments, entitled ‘The Law’.18 Some of these judgments also refer, among the background material 13 Article 10. 14 See, inter alia, ECRI Report on the Netherlands CRI(2013)39, June 2013, § 21. 15 In ECRI’s practice, ‘vulnerable’ groups include non-nationals, migrants and persons with a migration background, the Roma and other historical minorities, followers of minority religions, including Muslims and Jews, as well as the LGBTI community/communities. 16 This paragraph refers to I Hare and J Weinstein (eds), Extreme Speech and Democracy (Oxford, Oxford University Press, 2009); M Herz and P Molnar (eds), The Content and Context of Hate Speech: Rethinking Regulation and Responses (Cambridge, Cambridge University Press, 2012); A Brown, Hate Speech Law: A Philosophical Examination (New York, Routledge, 2015); and E Heinze, Hate Speech and Democratic Citizenship (Oxford, Oxford University Press, 2016). Regarding the ultima ratio principle constraining the criminal law, see Nina Peršak, ch 7 in this volume. 17 If such is the conclusion of its comparative analysis of Member States’ legislation and practice. 18 See, eg DH and Others v Czech Republic [GC] App No 57325/00 (ECtHR, 13 November 2017), paras 184, 192, 200 and 205; Féret v Belgium App No 15615/07 (ECtHR, 16 July 2009), para 74; Aksu v Turkey [GC] App Nos 4149/04 and 41029/04 (ECtHR, 15 March 2012), para 85; Fedorchenko and Lozenko v Ukraine, App No 387/03 (ECtHR, 20 September 2012), para 68; and Mariya Alekhina and Others v Russia App No 38004/12 (ECtHR, 17 July 2018), para 222.

Criminal Law Responses to Hate Speech  121 (under ‘The Facts’), to CM Recommendation No R(97)20.19 Accordingly, no one can criticise the CoE for not having taken a – broadly speaking – systematic approach to hate speech.20 This is true, despite the fact that the Court has so far refrained from attempting to define the latter in a precise manner – since it has not explicitly or implicitly questioned ECRI’s definition, which also coheres with that in Recommendation No R(97)20. III.  CRIMINAL LAW MOBILISATION TO COUNTER HATE SPEECH IN FREEDOM OF EXPRESSION CASES BEFORE THE ECtHR

Until recently, all applications concerning the mobilisation of criminal law to counter hate speech were lodged in Strasbourg under Article 10 of the Convention by individuals whose freedom of expression had been limited at the national level. The Court adopted its usual careful stance in these cases. On the one hand, it called on several occasions, in obiter, for restraint in the use of criminal law.21 On the other hand, it never excluded that states might find it necessary ‘to sanction or even prevent’ (including by means of the criminal law) forms of expression that spread, incited, promoted or justified hatred based on intolerance.22 This would be done in the interests of ‘tolerance’ and the need to respect ‘the equal dignity of all human beings’; these are – together with freedom of expression – among the foundations of a democratic, pluralistic society.23 It must be stressed that, in Article 10 cases, the Court’s task was to decide whether it was permissible for the country concerned to curtail hate speech. This has facilitated the recognition of a margin of appreciation in favour of national authorities, which are the first to be entrusted with the task of striking a balance between the demands of all the above-mentioned values. In such a context, it is unsurprising that the ECtHR has found that Article 10 was not violated in cases involving criminal convictions for xenophobic insults or threats, group defamation, Holocaust denial, or incitement to ethnic and/or religious discrimination or hatred. Usually,24 the Court would reach this

19 See, eg, Féret (n 18) para 44; and Savva Terentyev v Russia App No 10692/09 (ECtHR, 28 August 2018), paras 36 and 37. 20 cf https://www.coe.int/en/web/no-hate-campaign/coe-work-on-hate-speech and https://www. coe.int/en/web/freedom-expression/hate-speech. 21 eg, Féret (n 18) para 80; and Balsytė-Lideikienė v Lithuania App No 72596/01 (ECtHR, 4 November 2008), paras 83 and 84. 22 Gündüz v Turkey App No 35071/97 (ECtHR, 4 December 2003), ECHR 2003-XI, para 40, a case involving criminal sanctions. 23 ibid paras 37 and 40. 24 eg, Sürek v Turkey (No 1) [GC] App No 26682/95 (ECtHR, 8 July 1999), ECHR 1999-IV; Soulas and Others v France, App No 15948/03 (ECtHR, 10 July 2008); Féret (n 18); Vejdeland and Others v Sweden App No 1813/07 (ECtHR, 9 February 2012); and Le Pen v France (dec) App No 18788/09 (ECtHR, 20 April 2010).

122  Stephanos Stavros conclusion by examining, under the second paragraph of the above-mentioned provision, whether the interference with freedom of expression was ‘necessary in a democratic society’. Relatively few cases decided in Strasbourg turned on the question whether the particular form of speech was in the first place protected by Article 10;25 despite the fact that Article 17 of the Convention makes it clear that none of its provisions may be interpreted as implying any right to engage in any activity aimed at the destruction of any of the freedoms set forth therein. Even fewer cases26 involved consideration of other relevant international conventions, ie, the International Convention on the Elimination of All Forms of Racial Discrimination (ICERD) and the International Covenant on Civil  and  Political Rights (ICCPR), which require, respectively, the ­criminalisation of: [A]ll dissemination of ideas based on racial superiority or hatred, incitement to racial discrimination, as well as all acts of violence or incitement to such acts against any race or group of persons of another colour or ethnic origin, and also the provision of any assistance to racist activities, including the financing thereof.27

And that: Any advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence shall be prohibited by law.28

These are among the most widely ratified UN instruments29 and may serve as inspiration30 even when the conditions of Article 31(3)(c) of the Vienna Convention on the Law of Treaties are not met. In other words, the Court may

25 For an exception, see Jersild v Denmark App No 15890/89 (ECtHR, 23 September 1994), Series A No 298, para 35, with further references; Norwood v UK (dec) App No 23131/03 (ECtHR, 16 November 2004), ECHR 2004-XI; M’Bala M’Bala v France (dec) App No 25239/13 (ECtHR, 20 October 2015), ECHR 2015-VIII; and Belkacem v Belgium (dec) App No 34367/14 (ECtHR, 27 June 2017). 26 See Jersild (n 25). 27 ICERD, art 4(a). 28 ICCPR, art 20(2); see also J Temperman, Religious Hatred and International Law: The Prohibition of Incitement to Violence or Discrimination (Cambridge, Cambridge University Press, 2016). 29 A total of 182 ratifications for ICERD and 173 for the ICCPR, compared to 196 for the Convention on the Rights of the Child, 189 for CEDAW and 181 for the Convention on the Rights of Persons with Disabilities. 30 K Dzehtsiarou and V Lukashevich, ‘Informed Decision-Making: The Comparative Endeavours of the Strasbourg Court’ (2012) 30 Netherlands Quarterly of Human Rights 272; E Brems, ‘Should Pluriform Human Rights Become One? Exploring the Benefits of Human Rights Integration’ (2014) 4 European Journal of Human Rights 447; P Mahoney and R Kondak, ‘Common Ground: A Starting Point or Destination for Comparative-Law Analysis by the ECtHR?’ in M Andenas and D Fairgrieve (eds), Courts and Comparative Law (Oxford, Oxford University Press, 2015); LA Sicilianos, ‘Précédent et le Dialogue des Juges: L’Exemple de la Cour Européennne des Droits de l’Homme (“CEDH”)’ in N Aloupi and C Kleiner (eds), Le Précédent en Droit International (Paris, Pedone, 2016); and A Nussberger, ‘The ECtHR’s Use of Decisions of International Courts and Quasi-Judicial Bodies’, in A Müller (ed), Judicial Dialogue and Human Rights (Cambridge, Cambridge University Press, 2017).

Criminal Law Responses to Hate Speech  123 heed the message conveyed by ICERD and the ICCPR even if it does not have to do so. The reasons for this seem obvious: [T]he international law background to the legal question before it [b]eing made up of a set of rules and principles that are accepted by the vast majority of States … reflect[s] a reality that the Court cannot disregard when it is called upon to clarify the scope of a Convention provision that more conventional means of interpretation have not enabled it to establish with a sufficient degree of certainty.

The passage is from Opuz v Turkey,31 a case of relevance to our topic, since it concerned gender-based violence and the use, inter alia, of criminal law to further the aims of the UN Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW). Nevertheless, the Court has always privileged ‘internal’ CoE sources32 when looking for support for its interpretational choices in its judgments concerning criminal law mobilisation against hate speech.33 The only notable exception is Jersild v Denmark, which refers to ICERD – without, however, drawing any conclusions from this that affect the outcome of the case.34 It is striking how differently the Court treats ‘external sources’ in the above cases, especially since the judgments in question concern related fields: gender-based discrimination, on the one hand, and racism and intolerance, on the other. One possible explanation is that, in the sensitive area of criminal law responses to hate speech, the Court wants to stress the need for a distinctly European voice.35 To recapitulate, the Court has adopted a balanced approach to Article 10 cases concerning criminal law mobilisation against hate speech. It has resisted the temptation of applying Article 17 too widely, has preferred internal to external sources when looking for support for its interpretational choices and has even called, in obiter, for restraint in the use of criminal law. At the same time, it has recognised a margin of appreciation for national authorities in the area, which has paved the way for many judgments concluding that there has not been a violation.

31 Opuz v Turkey App No 33401/02 (ECtHR, 9 June 2009), para 184; it also refers to ‘the common … domestic law standards of European states’. 32 See n 30; see also K Dzehtsiarou, ‘What is Law for the European Court of Human Rights?’ (2017) 49 Georgetown Journal of International Law 89; and L Glas, ‘The European Court of Human Rights’ Use of Non-binding and Standard-Setting Council of Europe Documents’ (2017) 17 Human Rights Law Review 97. 33 See the cases in nn 18, 19 and 26. 34 The Court found a violation of art 10 because of a lack of proportionality between the means employed (the applicant’s conviction and sentence) and the aim pursued (protecting the reputation or rights of others); it reached this conclusion although it had accepted, in para 30 of Jersild (n 25), that ICERD’s object and purpose were of great weight in determining whether the applicant’s conviction was necessary within the meaning of the second paragraph of this provision. 35 Several CoE states have entered reservations to arts 4 of ICERD and 20 of the ICCPR; cf ECRI’s GPR No 15, art 2.

124  Stephanos Stavros IV.  THE COURT RECOGNISES A CONVENTION-BASED OBLIGATION TO PROSECUTE HATE SPEECH

It is only recently that the Court seems to have opted for a bolder stance by considering that the Convention does not simply allow states to use criminal law to combat hate speech (as distinct from hate-motivated violence). In certain circumstances, it would require them to do so. This has happened in a string of cases decided in the past five years. Two of these cases were brought under Article 8 of the Convention, which guarantees the right to respect for private life, and involved hate speech during anti-Roma rallies in Hungary. The first, RB,36 concerned a demonstration by a ‘right-wing’ political party in a Hungarian municipality, as well as marches organised in this connection by two ‘right-wing’ paramilitary groups in its Roma neighbourhood. During these marches, someone, carrying an axe (or a whip), allegedly subjected the applicant, who was Roma and was watching the march with her daughter from her garden, to verbal abuse: ‘Go inside, you damned dirty gypsies!’ He also threatened that he would build a house in the Roma neighbourhood ‘out of their blood’. The Court, in April 2016, considered that his behaviour could raise issues under two provisions of the Hungarian Criminal Code (CC): violence against a member of a national ethnic, racial or religious group under Article 174/B (as per the applicant’s complaint) and incitement against a member of a group under its Article 269 (a point raised by the Court proprio motu). However, the police considered that the use of force could not be established and decided to investigate whether harassment had occurred. They concluded that this had not been the case, as the various utterances had not been directed at the applicant personally. The Court found a violation of Article 8, inter alia, because the offence the police had chosen to investigate did not contain any element alluding to racist motives. Király and Dömötör,37 decided in January 2017, involved another anti-Roma ‘right-wing’ demonstration in another Hungarian municipality during which the speakers called upon participants ‘to fight back’ and ‘sweep the rubbish out of the country’; they also referred to an ‘ongoing ethnic conflict’ and ‘the use of all necessary means of protection’. Following the speeches, the demonstrators – some of whom belonged to ‘far-right’ paramilitary groups and were equipped with sticks and whips – marched down a Roma neighbourhood, where the police had set up a security cordon. There were threats – ‘Roma, you will die’, ‘We will burn your house down’ and ‘We will come back when the police are gone’ – and ‘obscene insults’. The demonstrators encouraged the police not to protect the Roma; there were instances when the cordon was dismantled and when 36 RB v Hungary App No 64602/12 (ECtHR, 12 April 2016). 37 Király and Dömötör v Hungary App No 10851/13 (ECtHR, 17 January 2017). For an analysis of RB and Király and Dömötör, see A Nieuwenhuis, ‘A Positive Obligation under the ECHR to Ban Hate Speech?’ [2009] PL 326.

Criminal Law Responses to Hate Speech  125 things were thrown into gardens. The Court again found a violation of Article 8. It took issue with the narrowing down of the scope of the national proceedings: the authorities had considered that although the statements made by the speakers during the demonstration had been hateful and abusive, they had not incited to violence; as a result, no offence could have been committed under Article 269 CC. The Court also criticised the ineffectiveness of the investigation into the incidents during the march. No effort had been made to identify many of those involved and, of the few who were questioned, only one was charged and convicted. Most importantly, the Court stressed that the investigation was limited to actual acts of violence under Article 174/B CC. That the Court took issue with the above-mentioned threats and insults being excluded is important for the purposes of this chapter. The same reasons led to the finding of violation of the right to freedom of religion under Article 9 in the 2015 Karaahmed v Bulgaria judgment.38 This case concerned the authorities’ failure to protect those who had gathered for prayer in a mosque from an anti-Muslim, anti-Turkish demonstration.39 Although national law criminalised religious hate speech, the altogether-inadequate investigation focused ‘on physical acts of violence’, as opposed to the threats against the worshippers. The latter were quite serious. At a certain point, a demonstrator had started cutting a fez with a pocketknife, saying: ‘Can you hear me? We shall now show you what will happen to each one of you!’ Threats had also been made against the Muslim Roma applicant of the Alković v Montenegro case,40 in which the Court delivered its judgment in December 2017. The words ‘move out or you’ll bitterly regret it’ were scrawled on his wall, together with a cross, on Bayram (a Muslim holiday). This was a culmination of a series of incidents, which also involved the shooting of bullets into the outside walls of his house. The applicant had complained to the authorities, informing them that he suspected his neighbours, whom he had overheard making threats concerning him and who seemed to have an issue with both his ethnic origin and his religion. Montenegrin criminal law provided for the offences of jeopardising someone’s security, incitement to ethnic and religious hatred and racial discrimination. However, the police had remained largely inactive until the offences became time-barred,41 and the Court found a breach of Article 8 in conjunction with Article 14, which protects all persons within

38 Karaahmed v Bulgaria App No 30587/13 (ECtHR, 24 February 2015); see S Stavros, ‘A Duty to Prosecute Hate Speech under the ECHR?’, Oxford Human Rights Hub, 9 April 2015, http://ohrh. law.ox.ac.uk/a-duty-to-prosecute-hate-speech-under-the-european-convention-on-human-rights. 39 The Court considered that there were positive obligations under art 9 and that the national authorities had failed to strike a proper balance when taking ‘steps … to ensure the effective and peaceful exercise rights of the demonstrators and the rights of the applicant’. The authorities were also criticised by the ECtHR for their ‘subsequent failure properly to respond to’ the impugned events. 40 Alković v Montenegro App No 66895/10 (ECtHR, 5 December 2017). 41 Moreover, the prosecution had tried to play down the shooting incident.

126  Stephanos Stavros the jurisdiction of a State Party from discrimination in the enjoyment of their Convention rights.42 Hate speech, in the form of racist insults and threats, was at the heart of all four cases.43 The Court made this clear by taking issue with the kind of offences that the Hungarian authorities had chosen to prosecute and with the facts excluded from the scope of the investigation in Karaahmed. Although the Montenegrin case also involved a shooting incident (against the applicant’s home), the Court stressed that Alković had not been physically harmed and focused on how the authorities should have ensured respect for his right to psychological (as distinct from physical) integrity.44 The emphasis was on the discriminatory messages addressed to him, as evidenced by the finding of violation. As a result, the key question in all judgments was whether the applicants were effectively protected against hate speech. In order to be able to examine it, the Court first had to find a basis in one of the Convention provisions guaranteeing substantive rights. Having found a jurisdictional basis45 in Article 8 or 9,46 the Court could have answered the question in many different ways: would the Convention provision providing the jurisdictional basis be complied with if the state concerned ‘maintain[ed] and appl[ied] in practice an adequate legal framework’?47 Or did this provision require a specific response – in the form, for example, of criminal law? If so, was the Court prepared to go as far as to demand the criminalisation of certain forms of hate speech? Or would it limit itself in reviewing whether existing criminal law remedies were used? And what would the Court’s rationale be for insisting on the mobilisation of the criminal justice apparatus? It is clear that the only issue that the Court examined in all four cases was whether criminal law could have ensured respect for the applicants’ rights under Articles 8, 9 and 14. However, was this symptomatic of a casuistic approach, largely driven by the way in which the applicants’ complaints had been ­formulated?48 Some might argue, in this connection, that the judgments should be narrowly interpreted: they simply require states to ‘maintain and

42 Alković and the two Hungarian cases were decided by different Court sections. Moreover, discrimination was part of Alković’s claim under Montenegrin law. 43 Three art 3 cases against Georgia – Members of the Gldani Congregation of Jehovah’s Witnesses and Others v Georgia App No 71156/01 (ECtHR, 3 May 2007); Begheluri and Others v Georgia App No 28490/02 (ECtHR, 7 October 2014); and Identoba and Others v Georgia App No 73235/12 (ECtHR, 12 May 2015) – do not provide a clear precedent. In some instances, the authorities were directly involved in the ill-treatment. Moreover (as pointed out in respect of the last case in RB (n 36) para 50), the Court, when examining the authorities’ obligation to protect the Georgian applicants, did not make a clear distinction between their physical and moral integrity; cf n 62. 44 Alković (n 40) paras 69 and 72. 45 In terms of compatibility ratione materiae. 46 cf section V below. 47 See n 49. 48 eg RB (n 36) para 53.

Criminal Law Responses to Hate Speech  127 apply in practice an adequate legal framework’49 affording protection to alleged victims of hate speech. Any remedy (including, for example, a civil law injunction or action for damages) that gave such victims the possibility of having their interests under the above-mentioned provisions balanced against the interests of other individuals would fit the bill. In the circumstances of the cases, this remedy happened to be a criminal law one. The four judgments, seen in this light, would not go beyond what was decided in Aksu v Turkey,50 a case involving civil law proceedings, in which a Roma applicant had complained about failed claims for damages he had brought against those responsible for two publications allegedly casting his community in a negative light: a dictionary reproducing several expressions that were disparaging of the Roma community and an academic study on the Roma lifestyle, parts of which were considered less than flattering by the applicant. The Court had accepted in Aksu that the applicant had, in principle, a right under Article 8 to a remedy to air his complaints. However, given the content of the publications and the national margin of appreciation, the Court considered that the national courts’ failure to award the applicant damages did not amount to a violation of Article 8. Those supporting a narrow interpretation of the four judgments could draw an argument from the margin of appreciation, which the Court is, in principle, prepared to recognise in favour of national authorities in cases involving the balancing of Convention rights, especially those concerning positive obligations.51 However, there is evidence that the Court wanted the four judgments to be given a broader reading. This is to be found in the cases cited as precedent, which all concern a state duty to adopt adequate positive measures ‘in the sphere of criminal-law protection’.52 Moreover, a criminal law remedy is what RB expressly required in respect of the insults and threats to which the applicant had been subjected.53 It follows that, in the circumstances of each of the four cases, the authorities were under a Convention-based obligation to respond to hate speech by mobilising pre-existing criminal law provisions. In particular, they should have used, as a basis for their investigation, criminal law provisions that would have allowed them to inquire into the racist character of the offences.

49 As per para 45 of Sandra Janković v Croatia App No 38478/05 (ECtHR, 5 March 2009), a case concerning positive obligations under art 8 to protect the applicant’s physical integrity. 50 See n 18. Panayotova and Others v Bulgaria (dec) App No 12509/13 (ECtHR, 7 May 2019), which concerned a brochure that ‘clearly sought to portray Roma … as exceptionally prone to crime and depravity’, largely follows Aksu (n 18). 51 Eweida and Others v UK App Nos 48420/10, 36516/10, 51671/10 and 59842/10 (ECtHR, 15 January 2013), para 106; and Von Hannover v Germany (No 2) [GC] App Nos 40660/08 and 60641/08 (ECtHR, 7 February 2012), ECHR 2012-I, para 106. 52 RB (n 36) para 83; see section V below. 53 RB (n 36) paras 84 and 85.

128  Stephanos Stavros It is significant that the Court reached this conclusion without relying on the national margin of appreciation, despite some ritual references thereto.54 Instead, the Court took into account the ‘fundamental values of democratic societies’ and ‘the increasingly high standard required in the area of human rights and fundamental freedoms’ that calls for ‘greater firmness’ when these are attacked.55 This, combined with the need to avoid giving the impression that the authorities approved of the intolerance,56 seems to be the main rationale for the recognition, in the four judgments, of a Convention-based obligation to prosecute hate speech. V.  PROSECUTING HATE SPEECH AND OTHER POSITIVE OBLIGATIONS UNDER THE CONVENTION

To those familiar with the case law on positive obligations, such recognition might not seem a surprising development. The idea that criminal law should be used to make sure that private individuals respect certain Convention rights is not new.57 Articles 2 and 3 of the Convention, for example, require a criminal law remedy against ‘serious’ or ‘grave acts’ by, inter alios, private individuals; usually, these would involve violence.58 In this connection, the Court will inquire into the existence of adequate statutory provisions and into whether these were backed up by effective law enforcement.59 Whenever there is suspicion of a racist or homophobic/transphobic motive for these acts, the Convention would require this to be properly taken into account in the investigation; it should also be taken into account in the prosecution of the relevant offence (if there is enough evidence) and, if guilt is proven, in the penalty to be imposed on the offender.60

54 Király and Dömötör (n 37) para 82. RB (n 36) para 81 refers to the ‘wide’ margin of appreciation enjoyed by national authorities in the field of positive obligations under art 8 (a departure from previous case law, which recognised a wide margin only in special areas, including rape – MC v Bulgaria App No 39272/98 (ECtHR, 4 December 2003), ECHR 2003-XII, para 154 – and police decisions in ‘operational matters’ – Karaahmed (n 38) para 105). 55 RB (n 36), para 84; see also Sandra Janković (n 49) para 47. 56 Király and Dömötör (n 37) para 80. 57 The idea is critically discussed in A Seibert-Fohr, Prosecuting Serious Human Rights Violations (Oxford, Oxford University Press, 2009); and L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce?’ in L Zender and JV Roberts (eds), Principles and Values in Criminal Law and Criminal Justice: Essays in Honour of A Ashworth (Oxford, Oxford University Press, 2012). 58 eg, Opuz (n 31) para 128; and O’Keeffe v Ireland [GC] App No 35810/09 (ECtHR, 28 January 2014), ECHR 2014-I, para 148. 59 eg, O’Keeffe (n 58) para 148. 60 Nachova and Others [GC] App Nos 43577/98 and 43579/98 (ECtHR, 6 July 2005), para 160, ECHR 2005-VII; Balázs v Hungary App No 15529/12 (ECtHR, 20 October 2015), para 75; and Identoba and Others (n 43) para 75.

Criminal Law Responses to Hate Speech  129 Moreover, there are numerous judgments on positive obligations under Article 8 of the Convention (which protects, inter alia, the right to ‘personal’61 – and even ‘moral’62 or ‘psychological’63 – integrity) going as far back as 1985. In such cases, the Court will first examine whether the acts complained of have interfered with an interest coming within the scope of application of Article 8. In Aksu, for example, the Court, in order to examine whether the applicant could claim protection against the two publications under the Convention, had first to assert that ethnic identity (in this case, the Roma one) was ‘an element embraced’ by this provision.64 Of course, the same holds true for sexual orientation and gender identity.65 The Court will then examine what kind of protection Article 8 requires in the circumstances, as there are different ways of ensuring respect for it. In practice, the Convention does not always require a criminal law response to acts interfering with interests falling within the scope of application of Article 8.66 Aksu again provides an example. However, it is significant that the Court had never excluded the use of criminal law in Article 8 cases concerning the right to one’s honour and reputation, including the right to be free from insults and defamation.67 Moreover, there had been cases – dating from long before RB, Király and Dömötör and Alković – in which the Court went as far as to require a criminal law remedy for the protection of personal integrity from third parties, whenever ‘fundamental values’ and ‘essential aspects of private life’ were at stake.68 In my view, there are several links between the racist-violence cases, in which – as seen above – the ECtHR requires criminal law responses, and RB, Király and Dömötör and Alković. Hate speech often lays the ground for racist and homophobic/transphobic violence. although this is not the only reason why hate speech is criminalised.69 Moreover, the method used by the Court to assess whether the failure to take account of the alleged racist homophobic/transphobic motive amounts to a breach of Article 2 or 3 lends itself par excellence to the examination of the adequacy of the authorities’ response to hate speech.70 In both instances, the ECtHR will consider whether the authorities had sufficient indications that a hate-motivated offence might have been committed and will assess the manner in which they reacted thereto, as well as their general willingness to engage with hate-crime allegations. 61 Sandra Janković (n 49) para 46. 62 Distinguished from ‘physical’ integrity (X and Y v The Netherlands App No 8978/80 (ECtHR, 26 March 1985), Series A No 91, para 22). 63 RB (n 36) para 101. 64 Aksu (n 18) para 58. 65 Dudgeon v UK App No 7525/76 (ECtHR, 22 October 1981), Series A No 45; and YY v Turkey App No 14793/08 (ECtHR, 10 March 2015), ECHR 2015-I. 66 X and Y (n 62) para 24; see also Judge Tulkens’ concurring opinion in MC (n 54). 67 eg, Mihaiu v Romania App No 42512/02 (ECtHR, 4 November 2008). 68 X and Y (n 62) § 27. 69 ECRI’s GPR No 15, Explanatory Memorandum, §§ 173 and 28–30. 70 The question of the method to be applied is different from that concerning the standard to be used to assess the effectiveness of the authorities’ response (see section VII below).

130  Stephanos Stavros Parallels may be also drawn with the Article 8 cases in which the Court has recognised an obligation to mobilise criminal law to protect some interests falling within the scope of application of this provision. Hate speech is incompatible with a democratic society’s fundamental values71 and forms of it affect victims’ sense of identity, as well as their self-worth and self-confidence.72 To recapitulate, Article 8 covers ethnic and gender identity and sexual orientation. Moreover, the Convention requires criminal law to be used to protect against most instances of racist and homophobic/transphobic violence. It also requires, in certain circumstances, criminal law to be used to protect some Article 8 interests. Against such a background, it is not surprising that the Court has also decided that Article 8 requires, in certain circumstances, a criminal law remedy against hate speech. In the light of what has been explained in sections I and II above about the links between hate speech, vulnerability, stigmatisation and discrimination, the four judgments examined are – in my view – a positive development. Their terse reasoning notwithstanding, they cover significant ground towards achieving three objectives. First, they reinforce the idea that the CoE mechanisms function as a system. What the Court has essentially done in these cases is to throw its weight behind what has always been a key pillar of ECRI’s policy: using criminal law to combat hate speech is not excluded; and in some circumstances, it constitutes the most appropriate response thereto. Second, the four judgments bring the ECHR closer to ICERD (and, to a certain degree, the ICCPR).73 One must not forget that it is by reference to ICERD’s object and purpose that the Court sought to justify its seminal statement in Jersild, its first hate-speech case, concerning the centrality of the fight against racism for the CoE human rights edifice. Third, the four judgments send a signal – which is clearer than, but not significantly different from, the one timidly emitted in the earlier Article 10 cases – about the fundamental premises of judicial policy at the European level. While freedom of expression is unquestionably a key value, human rights law cannot ignore the most vulnerable,74 who often fall victim to hate speech. That the needs of the vulnerable cannot be neglected has already been accepted by the ECtHR in a number of judgments concerning many different types of cases.75

71 See section III above. 72 Aksu (n 18) para 58. Whether these would qualify as ‘essential aspects of private life’ is a separate question, which so far the Court has not addressed expressly. 73 See n 107, which addresses the issue of whether the Convention may be interpreted in a manner that overrides a Member State’s reservation under art 4 ICERD or art 20 ICCPR. 74 Sometimes the Court seems to use the term ‘vulnerable’ as shorthand for complex problems that are known to exist; see Redfearn v UK App No 47335/06 (ECtHR, 6 November 2012), para 44. cf D Xenos, ‘The Human Rights of the Vulnerable’ (2009) 13 International Journal of Human Rights 591, starting with a rather developed vulnerability theory. 75 L Peroni and A Timmer, ‘Vulnerable Groups: The Promise of an Emerging Concept in ECHR Law’ (2013) 11 International Journal of Constitutional Law 1056; and S Besson, ‘La Vulnérabilité et la Structure des Droits de l’Homme: L’exemple de la Jurisprudence de la CEDH’ in L BurgorgueLarsen (ed), La Vulnérabilité Saisie par les Juges en Europe (Paris, Pedone, 2014).

Criminal Law Responses to Hate Speech  131 Several of them, as Corina Heri points out, use the vulnerability concept to shape some of the coercive obligations the Court has discovered in the Convention.76 Arguably, what the ECtHR has done in the four judgments analysed in this chapter is to take the next logical step in its exploration of this concept. This would be so because many people consider that the criminalisation of particular forms of hate speech is necessary at both a symbolic level and a practical level: it de-stigmatises certain groups and, in some situations, it may prove instrumental in empowering victims to pursue, before the domestic courts, their rights under Articles 8, 9 and 14 of the Convention.77 The final reason why the four judgments are a positive development is that they achieve all the above without upsetting the delicate balance that the Court has struck so far in the process of interpreting, in an admittedly dynamic manner, key Convention provisions. As a matter of fact, no surprise seems to have been sprung on states in RB, Király and Dömötör or Alković (let alone in Karaahmed, the Article 9 case);78 they had received clear warning that the interpretation of Article 8 was moving in this direction in Aksu.79 Arguably, it is also wiser to inject additional content into what has become one of the most open-ended articles of the Convention80 than to try to find safeguards against hate speech in Article 3 (which would have been another option, according to some).81 The label ‘inhuman and degrading treatment’ has been traditionally reserved for the most serious human rights violations. Many believe that the powerful message carried by this provision should not be weakened, as would have

76 See Corina Heri, ch 5 in this volume, with further bibliographical references. 77 Vulnerability can affect access to justice in many ways; see CAS and CS v Romania App No 26692/05 (ECtHR, 20 March 2012), para 81. Some victims prefer a criminal law response to hate speech because, in this case, the main resource-mobilisation burden is borne by the state (see section II above). It is, of course, true that the Convention does not always require state-assisted prosecution, as opposed to prosecution by the victim; see Bevacqua and S v Bulgaria App No 71127/01 (ECtHR, 12 June 2008), para 82; and Sandra Janković (n 49) para 50 (art 8 cases to be distinguished from Opuz (n 31) paras 168 and 171, an art 3 case). However, the Court has also accepted that art 6 ‘might require the domestic courts to assist the most vulnerable party to the proceedings in collecting evidence’ (Trapeznikova v Russia App No 21539/02 (ECtHR, 11 December 2008), para 100). This could well be the case even when the prosecution is conducted by the victim. 78 cf n 81. It is interesting to compare the composition of the Fourth Section chambers that decided, on the one hand, Karaahmed (n 38), and, on the other, Identoba and Others (n 43), as well as RB (n 36) and Király and Dömötör (n 37). 79 cf G Letsas, A Theory of Interpretation of the ECHR (Oxford, Oxford University Press, 2007) 72–74, on the application of the principle of legal certainty in the interpretation of human rights treaties. 80 M Burbergs, ‘How the Right to Respect for Private and Family Life, Home and Correspondence Became the Nursery in Which New Rights are Born: Article 8 ECHR’ in E Brems and J Gerards (eds), Shaping Rights in the ECHR: The Role of the ECtHR in Determining the Scope of Human Rights (Cambridge, Cambridge University Press, 2013); and A Sajó, ‘The Social in the Private: A Genealogy of “Private Life”’ in J Casadevall, R Raimondi, E Fribergh et al (eds), Liber Amicorum D Spielmann (Oisterwijk, Wolf Legal Publishers, 2015). 81 See Judge Wojtyczek’s dissenting opinions in RB (n 36) and Király and Dömötör (n 37); cf Identoba and Others (n 43).

132  Stephanos Stavros happened had the Court applied Article 3 to cases involving threats made or minor physical injuries inflicted by private individuals.82 VI.  IN WHAT KIND OF CASES WOULD/SHOULD THE OBLIGATION TO PROSECUTE HATE SPEECH ARISE?

To recapitulate, there is clear support for the line taken by the Court in RB, Király and Dömötör, Karaahmed and Alković in previous case law about positive obligations under Articles 2, 3 and 8. Tracing the origins of the recognition of the obligation to prosecute hate speech in this body of case law also prepares the ground for answering, in a contextualised manner, our next question: in what kind of cases would/should this obligation arise? As already seen, there are few theoretical considerations in the judgments under examination that could provide guidance in this respect. However, the four cases have a common characteristic: they concern hate speech during racist or intolerant demonstrations or marches that could at any time degenerate into violence83 or hate speech in the context of violent interpersonal relations. The point is expressly made in Király and Dömötör.84 It is also implicit in RB,85 which refers to an ‘attempted assault’, and in the choice of incidents on which the Court decided to focus its examination in Alković (although stressing that they were not isolated).86 The parallels with US Supreme-Court (SCOTUS) case law on speech that is not protected by the First Amendment are obvious.87 In Király and Dömötör, there was incitement to ‘immediate and likely illegal action’, which would have satisfied the Brandenburg v Ohio test;88 the danger for the applicant was both ‘clear’ and ‘present’. In the two Hungarian cases and in Karaahmed, the incitement in question was combined with ‘true threats’.89 Such threats were also scrawled on Alković’s wall. Finally, ‘fighting words’ were used in RB.90 82 cf Sandra Janković (n 49) para 47. Bouyid v Belgium [GC] App No 23380/09 (ECtHR, 28 September 2015), ECHR 2015-V can be distinguished as it concerned police violence. 83 In Karaahmed (n 38), there were violent incidents; as already observed, part of the problem was that the investigation had focused on them. 84 Király and Dömötör (n 37) paras 78 and 80. 85 RB (n 36) para 80. 86 Alković (n 40), paras 69 and 72. 87 According to SCOTUS case law, the First Amendment to the US Constitution prohibits the authorities from abridging freedom of speech. There are, however, several categories of ‘unprotected speech’, including obscenity, fighting words, defamation, child pornography, incitement to imminent lawless action and true threats. 88 Brandenburg v Ohio 395 US 444 (1969). 89 Statements meant to frighten or intimidate one or more specified persons into believing that they would be seriously harmed by the speaker or someone acting at his or her behest; see Virginia v Black 538 US 343 (2003). 90 Words that, by their very utterance, inflict injury – to use part of the Chaplinsky v New Hampshire 315 US 568 (1942) test. In this respect, the ECtHR distinguishes RB (n 36) from Alković (n 40) para 69.

Criminal Law Responses to Hate Speech  133 This parallelism is probably key for the understanding of our four judgments, three of which concern Hungary, a country whose legal culture has a clear pro-US tint on free-speech issues. This is evidenced in the judgment in Király and Dömötör,91 which refers to the Hungarian Supreme Court’s understanding of the incitement-against-a-group offence. US precedent is also cited in two more judgments issued against Hungary in 2012 and 2013 involving far-right associations and objectionable symbols.92 In general, while many ECtHR judges consider that the Convention allows for significant hate-speech limitations of freedom of expression,93 others’ views on this matter seem similar to those expressed by the majority in the above-mentioned SCOTUS cases. In such a context, the common ground required for concluding on a violation of Article 8 (and, to a lesser extent, Articles 9 and 14)94 might well have been found in US free-speech case law. Of course, the latter does not require the mobilisation of the criminal law in hatespeech cases. However, it might have been felt that some ECtHR judges would not have been prepared to accept the recognition of such an obligation beyond the very limited category of cases in which the SCOTUS allows for such prosecutions, and the finding of violation might have been fashioned accordingly. However, this might not be the only possible reading of the four judgments. In Király and Dömötör, the Court considered that in order to determine the scope of the positive obligation under Article 8 to use the criminal justice system to provide protection against hate speech, it should look for inspiration in its Article 10 case law.95 Since the latter allows for many convictions for hate speech that would not have met the various constitutionality tests applied across the Atlantic, the protection provided by Article 8 would be rather broad if victims were entitled under this provision to all speech restrictions that were not in breach of Article 10.96 Drawing its lines with reference to the SCOTUS or alternatively turning for guidance to its own Article 10 case law therefore seem to be, for the moment, the two main options for the ECtHR when deciding how wide-ranging the obligation to prosecute hate speech should be.97 In my view, it would be natural 91 Király and Dömötör (n 37) para 78. 92 Vona v Hungary App No 35943/10 (ECtHR, 9 July 2013), ECHR 2013-IV, para 66; and Fáber v Hungary App No 40721/08 (ECtHR, 24 July 2012), para 18. 93 See section III above. 94 See n 78. 95 Király and Dömötör (n 37) paras 73–74. 96 An approach rejected by the Committee that decided Panayotova and Others (n 50). 97 Nieuwenhuis (n 37) proposes that the ban under art 8 should target, inter alia, speech that would qualify as an abuse of rights under art 17. However, this is a provision with heavy historical baggage (see Perinçek (n 7) para 97 and its references to the Spanish Constitutional Court’s judgment of 7 November 2007 discussing the ‘militant democracy’ concept) whose application has always caused controversy (H Cannie and D Voorhoof, ‘The Abuse Clause and Freedom of Expression in the ECHR: An Added Value for Democracy and Human Rights Protection?’ (2011) 29 Netherlands Quarterly of Human Rights 54). Therefore, it might prove to be of limited help (cf Perinçek (n 7) para 114, M’Bala M’Bala (n 25) and Belkacem (n 25) concerning the exceptional cases to which art 17 would apply).

134  Stephanos Stavros that the Court should prefer the latter and seek to speak with a truly European voice.98 As will be shown immediately below, this should not necessarily result in overbroad restrictions on freedom of expression. One of the main features of the Court’s Article 10 case law is its ‘highly context-specific’ approach99 (inter alia, the statements at the heart of each case should be ‘fairly construed’ and ‘seen in their immediate or wider context’).100 The situational vulnerability of the applicant (ie, vulnerability that is not attributed only to personal characteristics, but results from a combination of the latter with external factors)101 should be seen as an essential part of this context. Taking situational vulnerability into account would involve asking whether, in the circumstances of a case, it would be realistic to expect the hate-speech victim to bear the burden of mobilising the criminal law apparatus on his or her own.102 Making the obligation to prosecute largely dependent on the applicant’s situational vulnerability would be in line with standard case law to the effect that the Convention ‘is intended to guarantee not theoretical or illusory but practical and effective rights’.103 It would also let the Court take into account the overall situation in the Member State concerned by the complaint, with regard to discrimination, hate crime and the potential existence of an intolerant climate. This already happened in the two Hungarian judgments, which refer to ‘the general hostile attitude against the Roma community’ in the municipality where the march had taken place and to ‘large-scale, coordinated intimidation’.104 Other factors relating to the national situation that the Court could usefully take into account would be the following: how effective would non-criminal law responses to hate speech be in the circumstances of the case? (This would be in line with ECRI’s approach. Although there are situations in which the criminal law constitutes the most appropriate response to hate speech, it should be used only when ‘no other, less restrictive, measure would be effective’.)105 How ‘mature’, ie, resistant to hateful messages, is local society?106 And how strong

98 cf Judge Keller’s dissenting opinion in Fáber (n 92). 99 Perinçek (n 7) para 208. 100 Király and Dömötör (n 37) para 73. 101 cf Opuz (n 31) para 160 (referring to victims of domestic violence and the situation of women, in general, in south-east Turkey); Akdivar and Others v Turkey App No 21893/93 (ECtHR, 16 September 1996), paras 73 and 105, Reports 1996-IV; and Identoba and Others (n 43) para 94 (citing Bączkowski and Others v Poland App No 1543/06 (ECtHR, 3 May 2007), para 64). See also, a contrario, Karaahmed (n 38) para 76. 102 Alković was not required by the Court to pursue a private prosecution; cf n 77. 103 eg Opuz (n 31) para 165; and, a contrario, Dilipak v Turkey App No 29680/05 (ECtHR, 15 September 2015), para 63, referring to the ‘dominant position’ of state authorities, which ‘requires them to show restraint in their recourse to criminal law, especially if they have other means of replying’ at their disposal. 104 RB (n 36) para 88; and Király and Dömötör (n 37), para 77. Panayotova and Others (n 50) seems to have been declared inadmissible because the applicants had not relied on the brochure’s ‘pernicious effect on them specifically’. 105 ECRI’s GPR No 15, art 10; see also Dilipak (n 103) para 63. 106 RB (n 36) para 84, in fine, seems to be pointing in this direction.

Criminal Law Responses to Hate Speech  135 are the national institutions (other than the criminal law machinery) that are entrusted with the fight against discrimination and intolerance?107 Of course, these are all factors that could also determine whether a criminal law response is allowed under Article 10 of the Convention. Moreover, these are factors that would have an incidence on the breadth of the margin of appreciation to be recognised in favour of national authorities in each case.108 There are many reasons why the Court might consider enriching its current approach under both Article 8 and Article 10 by relying on the situational vulnerability of the individual victim and the general national/local situation. One seems to stand out. The ‘highly context-specific’ character of such an approach would make it fully compatible with the true Strasbourg tradition. VII.  DISCHARGING THE OBLIGATION TO PROSECUTE HATE SPEECH

Having addressed the question of how far the obligation to prosecute hate speech would apply, it is now proposed to turn to a related one: how would this obligation be discharged? The answer might seem straightforward: in order to assess whether the national authorities complied with their obligation to prosecute hate speech in the three Article 8 cases, the Court applied the standards that it had developed under Articles 2 and 3 concerning the effective investigation and prosecution of hate-motivated violence. Few would doubt that this was the normal course of action, given the specific context of the cases: there was an imminent risk that hate speech could degenerate into violence. However, the above-mentioned case law may provide less solid guidance should the Court decide to apply the obligation to prosecute more widely. So much transpires from Blumberga v Latvia, which was about Convention requirements concerning the protection to be afforded by criminal law to property rights.109 In cases of ‘less serious crimes, such as those involving property’, the obligations under the Convention are ‘less exacting than with regard to more serious ones, such as violent crimes, and in particular those which would fall within the scope of Articles 2 and 3 of the Convention’.110 In such cases, the Court will consider that ‘the State will only fail to fulfill its positive

107 Some of these factors would have also affected the Member State’s decision to enter or maintain its reservation under art 4 ICERD or art 20 ICCPR. 108 In a series of recent cases against the Russian Federation (Dmitriyevskiy v Russia App No 42168/06 (ECtHR, 3 October 2017), Stomakhin v Russia App No 52273/07 (ECtHR, 9 May 2018), Mariya Alekhina and Others (n 18) and Savva Terentyev (n 19)) the Court has allowed national authorities a rather narrow margin of appreciation, something that may be attributed to its highly context-specific approach. 109 L Lavrysen, Human Rights in a Positive State: Rethinking the Relationship between Positive and Negative Obligations under the ECHR (Cambridge, Intersentia, 2016) 66. 110 Blumberga v Latvia App No 70930/01 (ECtHR, 14 October 2008), para 67.

136  Stephanos Stavros obligations … where flagrant and serious deficiencies in the criminal investigation and prosecution can be identified’.111 It cannot be excluded that a similar approach may be followed in hate-speech cases not involving an imminent risk of violence. The question of whether the criminal law was effectively applied in a case is, of course, separate from the question of whether the statute book contains offences that take the discrimination dimension of hate speech adequately into account. This is a real issue, as there are some European states – such as Hungary – where the legislator has traditionally adopted a strongly pro-freedom of expression stance. As a result, their criminal codes/laws lack many of the usual hate-speech offences. The ECtHR has so far seemed reluctant to tackle the ‘adequate legal framework’ question head-on. This is not surprising. There are several judgments in which it has tried to draw its findings as closely as possible on the facts of the case – focusing, for example, on the application of the law as opposed to the adequacy of the provisions themselves.112 Moreover, there are many – including Judge Bošjnak in Király and Dömötör – who consider that the Court has a tendency to conclude too readily that ‘the criminal-law mechanisms … appear to provide an apt legal basis for launching a criminal investigation into alleged bias-motives’.113 It will be recalled that one of the problems of the case was that racist insults could only be prosecuted as violence against a member of a national, ethnic or racial group or incitement against a group. However, can one realistically expect the Court to be able to avoid the ‘adequate legal framework’ issue indefinitely? Probably not, especially if it decides to apply the obligation to prosecute hate speech in cases not involving an imminent risk of violence. This, of course, raises the question of what would constitute an adequate legal framework for these purposes. ECRI’s GPR No 7 (on national legislation to combat racism and racial discrimination) offers a complete list of hate-speech offences: public incitement to violence, hatred or discrimination; public insults and defamation; threats; the public expression of a supremacist ideology; the public denial of genocide or crimes against humanity or war crimes; and offences related to racist material.114 The list is admittedly extensive and, as ECRI itself has recognised,115 there is always a risk that hate-speech legislation might be abused.116 111 ibid. 112 eg, Kokkinakis v Greece App No 14307/88 (ECtHR, 25 May 1993), Series A No 260-A. 113 Judge Bošjnak’s concurring opinion in Király and Dömötör (n 37). 114 All of them defined in GPR No 7, art 18. GPR No 15 sheds further light on the negationism offence. 115 GPR No 15, art 10(c); see also, inter alia, ECRI Report on Turkey (CRI(2011)5, December 2010), §§ 25 and 26; ECRI Report on the Russian Federation (CRI(2013)40, June 2013), § 23; and ECRI Report on Azerbaijan (CRI(2016)17, March 2016), § 34. 116 See also, inter alia, Dicle (No 2) v Turkey App No 46733/99 (ECtHR, 11 April 2006); and Stomakhin (n 108) (discussed by D Voorhoof, ‘European Court of Human Rights: Stomakhin v Russia’ (2018), http://hdl.handle.net/1854/LU-8571016).

Criminal Law Responses to Hate Speech  137 However, using the ECRI list as a standard will again show that the CoE human rights monitoring mechanisms operate as a system. This is necessary to protect the authority of the Court’s findings against unwarranted attacks using as a pretext the allegedly selective use of CoE sources.117 It might also be necessary on a practical level; the CM plays a role in all monitoring processes of the CoE (including those of ECRI) and would, in principle, intervene if the system’s internal coherence risks being jeopardised.118 Moreover, many of the concepts figuring on ECRI’s list have been at the core of key ECtHR judgments.119 In addition, as already seen,120 ECRI’s GPR No 15 makes important distinctions between hate speech that should be responded to by criminal law and hate speech that should be responded to by other means (civil/administrative-law, self-regulation, prevention or counter/positive speech).121 The criminal law response is circumscribed by several safeguards. Some of them have to do with the way in which the relevant offences are defined (eg, ‘clearly’ and ‘in a manner that permits their application to keep pace with technological developments’). Others concern the operational choices to be made when deciding how to respond to particular incidents (for example, criminal law should only be mobilised against hate speech ‘in a public context’, when ‘no other less restrictive measure would be effective’ and prosecutions should be brought ‘on a non-discriminatory basis’ and should not be used ‘to suppress criticism of official policies, political opposition or religious beliefs’). Arguably, all the above go a long way towards curbing the risk of overreach.122 However, the most conclusive argument in favour of a comprehensive statutory definition of hate speech has already been made. Criminal law should send a powerful message showing that it does not neglect the weakest members of society. In Europe, it can afford to do so because of the scrutiny that the ECtHR will exercise over its application in individual cases, which may usefully focus on victims’ situational vulnerability123 and the general situation in the country concerned.

117 As the Court often relies on ECRI material, including other GPRs (see section III above). 118 The CM always discusses ECRI’s GPRs and reports, even though it cannot formally adopt them (as it would those of other monitoring bodies). 119 In addition to the case law reviewed in section III above, there are numerous judgments concerning insults, defamation and even threats (eg Vona (n 92)). 120 See section II above. 121 S Stavros, ‘ECRI’s New GPR on Combating Hate Speech’ in JP Jacqué, F Benoît-Rohmer, P Grigoriou and MD Marouda (eds), Liber Amicorum S Perrakis (Athens, I Sideris, 2017). 122 N Mavronicola, ‘Taking Life and Liberty Seriously: Reconsidering Criminal Liability under Article 2 of the ECHR’ (2017) 80(6) MLR 1026; and Lazarus (n 57). 123 As transpires from the analysis above, victims’ vulnerability becomes a consideration only when reviewing operational choices concerning the appropriate response to specific hate-speech incidents. It cannot serve as the basis for foreseeable criminal law provisions.

138  Stephanos Stavros VIII. CONCLUSIONS

There is no doubt, following the publication of the judgments in RB, Király and Dömötör, Karaahmed and Alković, that the Convention does not only allow states to prosecute hate speech; in certain circumstances, it actually requires them to do so. This obligation is mostly derived from Article 8, which protects the right to respect for one’s private life. The judgments recognising such an obligation build on solid precedent concerning the concept of positive obligations to protect physical and ‘moral’ integrity. They are a welcome development: they show that the CoE human rights mechanisms operate as a system; they align the Convention with key UN instruments; and they send a powerful message concerning the weight attached by the Court to the need to protect the most vulnerable members of society. For the moment, there are no reasons for concern regarding the implications of these judgments for the enjoyment of other human rights, notably freedom of expression. Of course, the Court will inevitably be called upon to make, sooner or later, some important judicial policy choices in this, as in every other, area of case law. Some of them will relate to its willingness to review the very existence of adequate criminal law provisions, as opposed to their application, and the role the national margin of appreciation will be allowed to play in general in hate speech applications, as well as with the extent to which the ECtHR will rely on US or other sources of inspiration to carry on developing its framework for the examination of such issues. Some uncertainty is, of course, unavoidable when elaborating the content of open-ended provisions such as Article 8. However, what is certain is that the Convention has always been and continues to be a living instrument interpreted in a way that allows the Court to address the real concerns of real people in Europe today. Few will question the fact that the spread of hate speech is one of the most burning contemporary issues affecting the very existence of our ‘democratic societies’.

Part III

Critical Reflections: Theory, Impact, Limitations

140

7 Positive Obligations in View of the Principle of Criminal Law as a Last Resort NINA PERŠAK*

I. INTRODUCTION

T

he positive obligations doctrine requires states to actively engage in protecting people’s human rights and not only to refrain from interfering with them (illegally). Moreover, it extends the state’s obligations into securing people’s rights against other private actors, not merely against the state itself. Through this doctrine, European human rights law is said to have acquired its offensive or ‘sword’ function, in contrast to its merely defensive or ‘shield’ function.1 Over the years, the doctrine has been responsible for influencing changes in contracting parties’ criminal law; on the procedural plane, they have had to adopt measures for effective prosecution of crimes, on the adjudication plane, they have to choose sanctions that deter similar crimes from occurring in the future and, on the legislative plane, they have to appropriately criminalise harmful conduct. In line with this doctrine, the European Court of Human Rights (hereinafter ECtHR or ‘the Court’) has been progressively assessing the content of domestic laws and demanding the criminalisation of harmful acts to prevent harms and protect potential victims.2 National criminal law systems have thus been increasingly influenced by ECtHR jurisprudence, ‘which imposes standards and * The research behind this chapter has been partly supported by the Institute for Advanced Study of the Central European University (under the Senior Core Fellowship award) and the CEU Foundation of Budapest. The views expressed in the chapter are the author’s own. 1 F Tulkens, ‘The Paradoxical Relationship between Criminal Law and Human Rights’ (2011) 9 Journal of International Criminal Justice 577. 2 M Eriksson, ‘The Prevention of Human Trafficking: Regulating Domestic Criminal Legislation through the European Convention on Human Rights’ (2013) 82 Nordic Journal of International Law 339.

142  Nina Peršak constraints on the legislators’ discretion in criminal matters’.3 In light of this development, some law scholars have legitimately expressed concerns over the Court’s interference in this area of law and over potentially expansive criminalisation that does not bode well with the traditional criminal law principle of ultima ratio, ie, of using criminal law (criminalisation) as a last resort.4 It should be noted that this concern does not relate only to the ECtHR’s case law, but that it has also been voiced with respect to the EU’s engagement in the area of criminal policy, with the Lisbon Treaty expanding the EU’s competence in this field, giving it clear criminalisation powers.5 Other scholars have more generally raised questions in relation to the ECtHR extending the Convention rights beyond what could have been foreseen by the drafters of the Convention.6 The aim of this chapter is to examine the positive obligations doctrine through the eyes of this traditional criminal law principle, to flesh out some relevant issues and to critically examine certain problematic concepts (or rather their use) relating to the doctrine’s engagement in promoting criminalisation. First, the principle of criminal law as a last resort (ultima ratio) and its origin and rationale are presented. Next, we turn to the doctrine of positive obligations affecting criminal law and appraise the ECtHR’s judicial activity from the viewpoint of the Continental (civil law) criminal law theory in relation to the mandate that criminal law be used as ultima ratio, discussing certain empirical and normative challenges that arise from it. The chapter concludes with a 3 AM Maugeri, ‘Fundamental Rights in the European Legal Order, Both as a Limit on Punitive Power and as a Source of Positive Obligations to Criminalise’ (2013) 4 New Journal of European Criminal Law 374, 389. 4 Some notice that the ECtHR, rather than using criminal law as the ultimate remedy, ‘regards the effective protection of human rights and fundamental freedoms as increasingly obliging the contracting states to enact criminal law provisions’ (JW Ouwerkerk, ‘Criminalisation as a Last Resort: A National Principle under the Pressure of Europeanisation?’ (2012) 3 New Journal of E ­ uropean Criminal Law 228, 241). Others warn that: ‘Human rights law and human rights bodies are not arenas of criminal law-making or criminal justice’ (N Mavronicola, ‘Submission to the Human Rights Committee on its Draft General Comment on Article 6 of the International Covenant on Civil and Political Rights – Right to Life’ (5 October 2017) 6). 5 The Lisbon Treaty has provided new opportunities for the development of the EU criminal law legislation, ‘an explicit legal basis for the EU to adopt criminal law directives to ensure the effective implementation of EU policies, which have been subject to harmonisation measures’. See European Commission, ‘Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, towards an EU Criminal Policy: Ensuring the Effective Implementation of EU Policies through Criminal Law’ COM (2011) 573 final, 20 September 2011, 2; N Peršak, ‘EU Criminal Law and its Legitimation: In Search for a Substantive Principle of Criminalisation’ (2018) 26 European Journal of Crime, Criminal Law and Criminal Justice 20. 6 For example, A Mowbray, ‘Between the Will of the Contracting Parties and the Needs of Today: Extending the Scope of Convention Rights and Freedoms Beyond What Could Have Been Foreseen by the Drafters of the ECHR’ in E Brems and J Gerards (eds), Shaping Rights and the ECHR: The Role of the European Court of Human Rights in Determining the Scope of Human Rights (Cambridge, Cambridge University Press, 2015). For a more in-depth examination of positive obligations under the ECHR, see L Lavrysen, Human Rights in a Positive State: Rethinking the Relationship between Positive and Negative Obligations under the European Convention on Human Rights (Cambridge, Intersentia, 2016).

Positive Obligations in View of the Ultima Ratio Principle  143 reflection on the ways in which the Court could contribute to the mitigation of concerns regarding the use of the positive obligations doctrine to criminalise and to the respect of the ultima ratio principle. II.  THE PRINCIPLE OF CRIMINAL LAW AS A LAST RESORT

Among the principles of criminal law in Continental criminal legal theory (of German as well as French legal circles), we find principles such as the principle of legality, the principle of ultima ratio, the principle of culpability and the principle of proportionality.7 One can find the principle of criminal law as the last resort also invoked, albeit less frequently, in Anglo-American criminal theory, usually by more liberally minded scholars who warn against overcriminalisation or advocate reticence and minimalism in criminalisation.8 However, the ultima ratio principle is probably most discussed in German criminal law doctrine and those criminal law systems that have been influenced by it. While ultima ratio is not limited to criminal law, but, according to Wendt, represents a basic concept in many fields of law, its ‘most popular application … can probably be found in criminal law where this theorem of modern criminal law theory is a kind of common property … which is treated in nearly every [German criminal law] textbook or commentary’.9 The principle of ultima ratio or criminal law as the last resort (ultimum remedium) presents a limit on the use of criminal law and is, as such, a principle against criminalisation. This principle is thus predominantly a ‘negative’ and ‘formal’10 criminalisation principle, as it compels the state to desist from criminalising. Even if the conduct is harmful to others, the ultima ratio requirement may put a brake on its criminalisation if there are less intrusive, less repressive alternatives available.11

7 N Peršak, Criminalising Harmful Conduct: The Harm Principle, its Limitations and Continental Counterparts (New York, Springer, 2007). 8 See, for example, AP Simester and GR Sullivan, Criminal Law Theory and Doctrine, 2nd edn (Oxford, Hart Publishing, 2003) 21; A Ashworth, Principles of Criminal Law, 3rd edn (Oxford, Oxford University Press, 1999); D Husak, ‘The Criminal Law as Last Resort’ (2004) 24 Oxford Journal of Legal Studies 207. 9 R Wendt, ‘The Principle of “Ultima Ratio” and/or the Principle of Proportionality’ (2013) 3 Oñati Socio-Legal Series 81, 85. See also, eg, HH Jescheck, ‘Einleitung’ in Leipziger Kommentar zum StGB (Berlin, De Gruyter, 1992); G Jakobs, Strafrecht allgemeiner Teil: die Grundlagen und die Zurechnungslehre, 2nd edn (Berlin, De Gruyter, 1993); C Roxin, Strafrecht Allgemeiner Teil Band I: Grundlagen. Der Aufbau der Verbrechenslehre, 3rd edn (MMunich, Beck, 1997); W Wohlers, ‘Strafrecht als ultima ratio – tragender Grundsatz eines rechtsstaatlichen Strafrechts oder Prinzip ohne eigenen Aussagegehalt’ in A von Hirsch, K Seelmann and W Wohlers (eds), Mediating Principles (Baden-Baden, Nomos, 2006). 10 As opposed to, eg, the harm principle, which should be considered a substantive and positive principle of criminalisation, as it offers guidance on the content (or substance) of that which ought to be criminalised. See Peršak (n 7) 22. 11 Peršak (n 7).

144  Nina Peršak The content of the principle more or less follows from its name. The state ought to use the tool of criminalisation only as a last resort when there is no other way of addressing the undesirable conduct. It is often mentioned in connection with the so-called ‘fragmentary character’12 of criminal law (ie, that criminal law is protecting only some legal goods) and the ‘subsidiarity’13 of criminal law (ie, that the task of the criminal law is merely to provide subsidiary protection for the interests that the law recognises as warranting protection).14 The criminal law should therefore be employed only when other means have been exhausted. While Jareborg deliberately refrains from discussing what should be the prima ratio or secunda ratio and so forth, he stresses that: [T]here must be a basic presumption that the State should not interfere at all. If interference is necessary, then aid, support, care, insurance and licence arrangements should take precedence over coercive measures. If coercive measures are necessary, they need not consist in sanctions. If sanctions are necessary, private law sanctions might be preferable to administrative sanctions.15

And, as it would logically follow, if public law sanctions are necessary, administrative sanctions might be preferable to criminal law sanctions. The ultima ratio principle is first and foremost a ‘principle of legislative ethics’,16 a principle that can sometimes be found in the criminal codes,17 even though it might better fit in the constitution, as it (primarily) addresses the legislator.18 Some maintain that the principle speaks not only to the legislator but also to any official or body within the criminal justice system, playing a role in deliberations on the type of sanction to be used, deciding between alternative modes of punishment and so forth. Tulkens here distinguishes between the 12 N Jareborg, ‘Criminalization as Last Resort (Ultima Ratio)’ (2005) 2 Ohio State Journal of Criminal Law 521. 13 ‘der fragmentarische und akzessorische Character der Strafrechts’ (HH Jescheck, Lehrbuch des Strafrechts (Berlin, Duncker & Humblot, 1988) 46). 14 Jareborg (n 12) 532. 15 ibid 524. 16 ibid 521. If we want it to have any independent normative function, Jareborg argues, the ultima ratio principle should be a meta-principle – more precisely, a ‘metaprinciple summarising sufficient penal value reasons for criminalisation’. He proposes a ‘positive’ conception of the said principle, ie, as a principle providing reasons for (not against) criminalisation. However, his article finishes with a warning that the scope of such a conception ‘should be, however, fairly limited’ (at 534) and mentions (only) the humanity and utility principles as the possible limiting factors. Yet, as the humanity principle itself undoubtedly encompasses (that is, ought to encompass) the ultima ratio principle as it is commonly understood (ie, in the ‘negative’ conception, providing reasons against criminalisation), it could be argued that the ‘positive’ conception of the ultima ratio principle would be, at the end of the day, limited (also) by its ‘negative’ conception. 17 This used to be the case in Slovenia, for example, with the former art 2 of the Criminal Code stipulating that ‘it shall be legitimate to lay down criminal offences and to prescribe sentences only when, and to the extent that, the protection of human life and other basic values cannot otherwise be assured’. 18 Bengoetxea similarly states that ‘ultima ratio is a constitutional and constitutive legal principle which is primarily and primordially addressed to the legislator, not to the judiciary’ (J Bengoetxea, ‘Ultima Ratio and the Judicial Application of Law (2013) 3 Oñati Socio-Legal Series 107, 111).

Positive Obligations in View of the Ultima Ratio Principle  145 subsidiarity of criminal law (alternative forms of intervention should be considered first) and subsidiarity within criminal law (less severe penalties should be considered first).19 The latter: [C]an also be assessed in terms of ‘proportionality’, as the European Court of Human Rights construes the term today; that is to say, its application is justified only where there are no other less harmful means of ensuring the protection of a right guaranteed by the Convention.20

However, the majority of scholars tend to see the last resort principle as belonging primarily to the theory of criminalisation, not the theory of sentencing, and thus as primarily addressing the legislator, although acknowledging that there is a relationship between the two. Bengoetxea thus argues that ‘ultima ratio is conceptually and methodologically prior to proportionality’; while the former addresses the question of whether criminal law should be used at all, proportionality deals with the question of how much criminal law should be used (once the justification for resorting to criminal law has been established).21 While the proportionality principle – in its narrow, most functional sense, ie, proportionality between the sanction prescribed or used for the offence committed – reflects the idea of retributive justice, which is concerned with the appropriate or just allocation of blame and punishment for the undesirable conduct, the last resort principle reflects the wider social or civilisational idea of restraint in meting out any sort of punishment and therefore a general restraint in proscribing human conduct. The ultima ratio principle is thus a preference for non-intervention or, at least, non-criminal law intervention, an ‘obligation of restraint’22 or principle of reticence in criminalisation. In order to appreciate the principle’s current meaning, it is important to understand how such a principle came about in the first place. The origins of the ultima ratio principle have been linked to the Age of Enlightenment and its realisation of the repressiveness of, and hence necessary constraints on, criminal law. Its inception has been traced to the ultima ratio regum or ‘the final argument of kings’, which was inscribed on the royal cannons of the Sun King, Louis XIV of France, conveying the idea that military force should be employed only when other means failed.23 This notion was then transferred to the legal field, acquiring its current meaning. The principle is thus predicated on the 19 A somewhat comparable distinction is drawn by Melander, who differentiates between inner and outer dimensions of ultima ratio. Its outer dimension reflects the essence of the traditional understanding of this principle, positing criminal law in relation to other less intrusive legislative means and legal domains. The inner dimension, on the other hand, ‘does not make the criminal justice system proportionate to other legislative means but operates to guarantee proportionality within criminal law or the criminal justice system’ (S Melander, ‘Ultima Ratio in European Criminal Law’ (2013) 3 Oñati Socio-Legal Series 42, 50). 20 Tulkens (n 1) 582. 21 Bengoetxea (n 18) 111. 22 W Hassemer, Einführung in die Grundlagen des Strafsrechts, 2nd edn (Munich, Beck, 1990) 316. 23 Bengoetxea (n 18) 121; Wendt (n 9) 84.

146  Nina Peršak idea that criminal law is in itself ‘an evil’,24 which should be employed only where ‘the other legal responses prove inadequate’.25 It logically follows from the presumption in dubio pro libertate (in favour of liberty). Libertas, in this context, refers to non-criminalisation. Hanack argues that, like a judge who may not convict a person without proof of her guilt, the legislator should not criminalise a particular kind of conduct without evidence that criminalisation is necessary.26 In dubio pro reo for the judge should translate into in dubio pro libertate or in dubio contra delictum27 for the legislator. Feinberg endorses this presumption in favour of liberty and formulates it as being a requirement that: [W]henever a legislator is faced with a choice between imposing a legal duty on citizens or leaving them at liberty, other things being equal, he should leave individuals free to make their own choices. Liberty should be the norm; coercion always needs some special justification.28

The aforementioned presumption is ‘a reason against criminalisation’29 and, as such, a logical outgrowth of the ‘principle of liberty’30 as well as of the ultima ratio principle. The task of this presumption and these principles is to make the task of criminalising difficult. Moreover, the mentioned presumption allocates the burden of persuasion, resting it on those who wish to criminalise rather than on those who wish to live free of such restrictions.31 Many consider the ultima ratio principle as deriving from the principle of the rule of law or from the ‘constitutional framework of the rule-of-law state’.32 According to Jareborg, Rechtsstaat limits the state from using criminal punishment at will, the basic reason for this being the fact that ‘punishment involves hard treatment, inflicting harm that is often serious’:33 My arguments rely on the assumption that punishment is a State’s most intrusive means of enforcement in cases of illegal conduct … Loss of liberty is in itself grossly intrusive, and as a punishment it constitutes a clear expression of severe societal censure. In addition, incarceration often involves stigmatization and humiliation. If one disregards very short prison sentences, going to prison often affects family life and working life in ways that can never be repaired. Legislators try to make

24 J Bentham, An Introduction to the Principles of Morals and Legislation (Oxford, Clarendon Press, 1996). 25 R Koering-Joulin and JF Seuvic, ‘Droits fondamentaux et droit criminel’ (1998) L’Actualite juridique. Droit administratif 106. 26 EW Hanack, in Jareborg (n 12) 531. 27 T Vormbaum, in ibid. 28 J Feinberg, Harm to Others: The Moral Limits of the Criminal Law (New York, Oxford University Press, 1984) 9. 29 J Schonsheck, On Criminalization: An Essay in the Philosophy of the Criminal Law (Boston, Kluwer, 1994) 15. 30 JS Mill, On Liberty and Other Essays (Oxford, Oxford University Press, 1991 [1859]). 31 Schonsheck (n 29) 63. 32 P Minkkinen, ‘“If Taken in Earnest”: Criminal Law Doctrine and the Last Resort’ (2006) 45 Howard Journal 521. See also Jescheck (n 13); Roxin (n 9); Bengoetxea (n 18). 33 Jareborg (n 12) 522.

Positive Obligations in View of the Ultima Ratio Principle  147 punishment less detrimental by using less intrusive alternatives to imprisonment, but from the perspective of criminalization in such cases the threat of punishment still concerns imprisonment, and it is criminalization we are talking about.34

Criminal law inflicts punishment, which apart from its hard treatment (loss of liberty) includes stigmatisation or censure.35 According to Husak, criminal law should thus be ‘a last resort because the right not to be subjected to punishment – understood to include both hard treatment and censure – is extremely valuable and therefore very difficult to override’.36 For Stratenwerth, the punishment is, as a rule, the severest intrusion into the personal rights of an individual and therefore to be used only where other measures, especially private law and administrative law measures, fail.37 However, in some cases, such severe intrusion is not only warranted but may also be required in light of the doctrine of positive obligations, to which we will turn next. III.  POSITIVE OBLIGATIONS AFFECTING CRIMINAL LAW

The main rationale behind the European Convention on Human Rights and Fundamental Freedoms (hereinafter ECHR or ‘the Convention’) when it was passed in 1950 was to protect individuals against the unlimited and excessive use of power by the state. Therefore, in most cases, the rights and freedoms written therein were primarily for the states to respect and refrain from interfering in the individual’s exercise thereof (ie, negative obligations). Yet, in relation to some other rights and freedoms, the text of the ECHR itself requires states to take action, such as to protect the individual’s right to life (Article 2) or to deliver a fair trial within a reasonable time by an independent and impartial tribunal (Article 6(1)). However, while some positive obligations have been explicitly imposed upon states by the text of the ECHR, some other, ‘implied’ positive obligations have been read into the Convention text later.38 Yet, even in the case of explicit positive obligations, the Court has shown substantial judicial creativity over the years. The development of the theory of positive obligations has led to the ‘horizontal effect’ of human rights jurisprudence, where the responsibility for the violation of one’s human rights by another private individual could be attributed to the state ‘where the state has made it possible or probable, 34 ibid 526. 35 A von Hirsch, Censure and Sanctions (Oxford, Clarendon Press, 1993). 36 Husak (n 8) 233. The mentioned right refers of course to the right in the moral philosophical sense, not to a recognised human right. 37 Jareborg (n 12) 525. 38 AR Mowbray, The Development of Positive Obligations under the European Convention on Human Rights by the European Court of Human Rights (Oxford, Hart Publishing, 2004) 222. See also D Russell, ‘Supplementing the European Convention on Human Rights: Legislating for Positive Obligations’ (2010) 61 Northern Ireland Legal Quarterly 281.

148  Nina Peršak either through simple negligence or through benign tolerance’.39 The Belgian Linguistics Case40 was the first case in which the Court recognised that ‘positive obligations’ were inherent in the text of the Convention.41 The case of Marckx v Belgium,42 which concerned a violation of Article 8 (and Article 14 in conjunction with Article 8) in view of the lack of appropriate administrative measures that would establish a legal bond between an unmarried mother and her child from the moment and mere fact of birth, paved the way for the idea that the state’s obligations, which are primarily negative, may be accompanied by a positive obligation to guarantee effective respect for the rights and freedoms secured. In this case, the Court thus first applied positive obligations as ‘inherent’ in the effective protection of human rights,43 concretely, ‘inherent in an effective “respect” for family life’,44 and held that the state was under an obligation to assist the integration of children into the family, including the children of unmarried women. Positive obligations in the area of criminal law represent the case of criminal law being called into play to protect human rights.45 Mowbray observes that: One of the most prevalent types of positive obligation is the duty upon states to take reasonable measures to protect individuals from infringement of their Convention rights by other private persons. At its most basic level this positive obligation may be satisfied by the respondent state having adequate domestic legal provisions criminalising the conduct which threatens another’s Convention rights.46

In X and Y v The Netherlands,47 which is seen as a leading judgment in this respect, the applicants (a father and daughter) claimed a violation of Article 8 ECHR (respect for private life), as they were unable to instigate criminal proceedings against the rapist of the daughter who was a mentally disabled young girl and thus was not capable of filing a complaint herself (yet the law required that the victim herself personally lodges a complaint). Therein the Court stressed that ‘there are different ways of ensuring the respect for private life, and the nature of the state’s obligation will depend on the particular aspect of private

39 Tulkens (n 1) 583. 40 Belgian Linguistics Case Series A No 5 (1968). 41 Since then, ‘the European Court has constantly broadened this category [positive obligations] with the addition of new elements, to the point where virtually all the standard-setting provisions of the Convention now have a dual aspect in terms of their requirements, one negative and the other positive’. JF Akandji-Kombe, Positive Obligations under the European Convention on Human Rights: A Guide to the Implementation of the European Convention on Human Rights (Strasbourg, Council of Europe, 2007) 5–6. 42 Marckx v Belgium Series A No 31 (1979). 43 D Xenos, The Positive Obligations of the State under the European Convention of Human Rights (Abingdon, Routledge, 2012) 24. 44 Marckx v Belgium Series A No 31 (1979), para 31. 45 Tulkens (n 1) 582. 46 Mowbray (n 38) 225. 47 X and Y v The Netherlands Series A No 91 (1985).

Positive Obligations in View of the Ultima Ratio Principle  149 life that is at issue’.48 However, while recognising that ‘[r]ecourse to the criminal law is not necessarily the only answer’,49 the Court added that in this case: [T]he protection afforded by the civil law in the case of wrongdoing of the kind inflicted on Miss Y is insufficient. This is a case where fundamental values and essential aspects of private life are at stake. Effective deterrence is indispensable in this area [meaning the area of sexual abuse] and it can be achieved only by criminal-law provisions; indeed, it is by such provisions that the matter is normally regulated.50

Similarly, for example, in the case of Osman v UK, the states had to adopt ‘effective criminal law provisions to deter the commission of offences against the person backed up by law-enforcement machinery for the prevention, suppression and sanctioning of breaches of such provisions’.51 The same or almost the same phrasing appears in a number of other cases, for example, in Paul and Audrey Edwards v UK, where the Court held that taking appropriate steps to safeguard the lives of those within the state’s jurisdiction under Article 2 ‘involves a primary duty on the State to secure the right to life by putting in place effective criminal-law provisions to deter the commission of offences against the person’52 and MS v Croatia (No 2), where the phrase was repeated in connection with Article 3 ECHR.53 In Cestaro v Italy, the Court held that ‘[f]or an investigation to be effective in practice it is a prerequisite that the State has enacted criminallaw provisions penalising practices that are contrary to Article 3’,54 referring to the Gäfgen55 case. It added that ‘the absence of criminal legislation capable of preventing and effectively punishing the perpetrators of acts contrary to Article 3 can prevent the authorities from prosecuting violations of that fundamental value of democratic societies, assessing their gravity, imposing adequate penalties and precluding the implementation of any measure likely to weaken the penalty excessively, undermining its preventive and dissuasive effect’.56 Most57 of the actual cases where the ECtHR mandated criminalisation or the use of criminal measures would not be considered problematic, as they refer

48 ibid para 24. 49 ibid. 50 ibid para 27. 51 Osman v UK ECHR 1998-VIII, para 115. 52 Paul and Audrey Edwards v UK ECHR 2002-II, para 54. 53 MS v Croatia App No 75450/12 (ECtHR, 19 February 2015), para 74. 54 Cestaro v Italy App No 6884/11 (ECtHR, 7 April 2015), para 209. 55 Gäfgen v Germany App No 22978/05 (ECtHR, 1 June 2010), para 117. 56 Cestaro v Italy App No 6884/11 (ECtHR, 7 April 2015), para 209. 57 However, instances where criminalisation is required for an offence which has not been committed intentionally are likely or likelier to raise some concerns and doubts as to the EU-wide consensus or as to the ‘normalcy’ of such acts constituting criminal offences than criminalisation of harms committed with intent. In Sinim v Turkey App No 9441/10 (ECtHR, 6 June 2017), para 63, for example, the responsible parties’ ‘voluntary and reckless disregard of their legal duties under the relevant legislation’ (concretely, of the relevant rules on the transportation of dangerous goods) was enough to demand criminalisation, ie, ‘a criminal-law reaction to ensure effective deterrence against similar threats to the right to life in the future’, notwithstanding the lack of intent.

150  Nina Peršak to situations or conduct that is usually or ‘normally’58 regulated by the criminal law, ie, mostly found in criminal statutes already. However, the non-specification of sufficiently clear and detailed criteria for when criminal law should be used, a lack of sufficiently articulated good reasons justifying the imposing of the obligation to criminalise on states, and a reliance on certain vague formulas and phrases do raise concerns from the viewpoint of criminal law theory and human rights. The three elements that can be extracted from the case of X and Y v The Netherlands as justifying criminal legislation in that case were severity of the offence (wrongdoing against fundamental values), normality (the fact that these kinds of infringements ‘normally’ constitute criminal offences) and the assumption that effective deterrence requires criminal law provisions. ‘Effective deterrence’ is the reoccurring keyword here and the most problematic of the three elements mentioned, as (or when) its interpretation seems to pull in the opposite direction than the principle of ultima ratio does. Effective deterrence is a goal in the name of which the Court applies and stretches positive obligations of the state to include the criminalisation of certain conduct, and it represents another realisation of that longstanding justification behind the idea of positive obligations, namely that the Convention rights must be effective. The ‘effective interpretation’ of Convention rights follows from the need to ensure an effective protection of rights. As first formulated in the case of Artico v Italy: ‘The Convention is intended to guarantee not rights that are theoretical or illusory but rights that are practical and effective.’59 With this in mind, there are two groups of issues that need addressing here, one being empirical and the other theoretical or normative. We shall now inspect both of these in turn. A.  The Empirical Issue of Effectiveness If ‘effective deterrence’ is used as a reason why the Court feels legitimised to require the state to amend (adopt new offences or modify) its criminal law, then the first question that arises is of course: will criminalisation ‘effectively deter’?60 Will it be effective for (the goal of) deterrence of some undesirable social conduct/crime (x)? The sub-questions that follow from this are: what deters? Is there any evidence to support the claim that criminalisation will deter x? And are decision-makers (in this case ECtHR judges) really interested in the question of actual deterrence/effectiveness or is the idea that criminal law deters rather an 58 cf X and Y v The Netherlands App No 8978/80 (ECtHR, 26 March 1985); Siliadin v France App No 73316/01 (ECtHR, 26 July 2005), para 112: ‘in accordance with contemporary norms and trends in this field’. 59 Artico v Italy Series A No 37 (1980), para 33. 60 To ‘effectively deter’ probably just means to ‘deter’, as to ‘ineffectively deter’ is not to deter at all.

Positive Obligations in View of the Ultima Ratio Principle  151 assumption or legal fiction that is simply taken for granted? In other words, how interested are ECtHR judges in the criminological knowledge61 on these issues? Rogers asserts that: Few criminal lawyers (as opposed to theorists or criminologists) give serious thought to whether certain types of crimes, such as those committed under duress, can be effectively deterred, because the issue lacks legal relevance, but this may change if positive obligations arguments are made more frequently.62

While one can easily agree with the first part of his claim, the second, optimistic claim or expectation of the future may raise some scepticism. In the last 15 years, at least in the Court’s jurisprudence, we would be hard-pressed to find any concern with providing the evidence for the effective deterrence-based claims. While individual judges may entertain such thoughts – a notable example being the former Belgian ECtHR judge Françoise Tulkens63 – the Strasbourg case law appears to remain unperturbed by such considerations. However, warning against the erosion of the principle of subsidiarity of the criminal law in view of the modern positive obligations doctrine, Tulkens argues that under this ‘new form of instrumentalism’, the ‘legitimacy of the end pursued once again justifies the choice of means, no matter how harmful, as long as they are the most effective’, adding that the latter ‘would still need to be demonstrated on a case-by-case basis’.64 Should the Court not be really interested in ‘what works’ in terms of prevention, then it is probably not unfair to say that the ‘effective deterrence’ mantra is simply a new phrase, perhaps a more objective and neutral phrase in the lexicon of ‘symbolism’ or ‘expressiveness’ of criminal law. Appeals to the latter tend to justify criminalisation purely on the grounds that it is a declaration or expression of condemnation, a symbolic statement to the outside world that some types of behaviour will not be tolerated. Those scholars and judges who see 61 Such as knowledge on the prevalence of certain crimes and measuring crime, victim’s needs, the difference between certainty and severity of punishment in terms of influencing deterrence and, more generally, on if and when deterrence works (see, for example, A von Hirsch, AE Bottoms, E Burney and PO Wikstrom, Criminal Deterrence and Sentencing Severity (Oxford, Hart Publishing, 1999); M Tonry (ed), Why Punish? How Much? (Oxford, Oxford University Press, 2010); DS Nagin, FT Cullen and C Lero Jonson (eds), Deterrence, Choice, and Crime: Contemporary Perspectives (New York, Routledge, 2018). 62 J Rogers, ‘Applying the Doctrine of Positive Obligations in the European Convention on Human Rights to Domestic Substantive Criminal Law in Domestic Proceedings’ [2003] Criminal Law Review 690, 707. 63 Incidentally, she holds a Master’s degree in criminology, which perhaps contributed to her being more attuned to such empirical questions. In her concurring opinion to the judgment in the case of MC v Bulgaria App No 39272/98 (ECtHR, 4 December 2003), para 2, for example, she stressed: ‘As to the assumption that criminal remedies are, in any event, the most effective in terms of deterrence, the observations set out in the Report on Decriminalisation by the European Committee on Crime Problems clearly show that the effectiveness of general deterrence based on the criminal law depends on various factors and that such an approach “is not the only way of preventing undesirable behaviour”.’ 64 Tulkens (n 1) 594.

152  Nina Peršak the main purpose of criminal law in its expression of the societal sentiments of resentment and disapproval do not usually concern themselves with questions of whether a criminal law prohibition actually prevents or deters such conduct, as this seems irrelevant for the justification of criminal law in their view. Yet there are some problems with this approach. While it is undoubtedly the case that any passing of criminal laws is also an expressive act of ‘making a statement’, ie, denouncing certain conduct as reprehensible and blameworthy, enacting such laws (or demanding criminalisation) solely for their expressive function, ie, without regard for the criminal law’s actual, real-life consequences and effects of punishment, is not responsible and therefore not justified.65 In Sunstein’s words, ‘good expressivists are also consequentialists’.66 While it is difficult to establish from afar to what extent this guides the judgments of the ECtHR, it is clear that the judgments themselves refer frequently to deterrence, which allows us to infer that, at least rhetorically, the Court (primarily) subscribes to the preventive (rather than to the retributivist or to the expressive)67 theory of punishment, ie, that the sole or main objective of the criminal law is to prevent and deter crime. The fact that it remains unclear under which precise circumstances the ECtHR finds that non-criminal law measures are not sufficient or effective is problematic under the ultima ratio principle with its inherent requirement that a choice to criminalise be extensively justified. Such considerations ought to be evident in ECtHR judgments, ‘which should examine not only the question of whether the domestic law allows criminal-law measures but also the question whether resorting to the criminal law is desirable in the first place’.68 Van Kempen emphasises an even more heightened need for persuasive argumentation in this case, maintaining that ‘since a positive obligation more intrusively violates freedom and possibly even fundamental rights, the urgency to justify a duty increases and the reasoning must be more powerful’.69 Consequently, positive obligations to use criminal law require strong argumentation and the burden of proof lies with the Court.70

65 Sunstein (CR Sunstein, ‘On the Expressive Function of Law’ (1996) 144 University of Pennsylvania Law Review 2021) similarly argues that ‘the expressive function of law makes most sense in connection with efforts to change norms and that if legal statements produce bad consequences, they should not be enacted even if they seem reasonable or noble’ (at 2025). In short, ‘legal “statements” producing bad consequences should not be endorsed’ (at 2052). 66 ibid 2047. 67 The Court has on occasion invoked retribution as a rationale as well. For example, in Jelić v Croatia App No 57856/11 (ECtHR, 12 June 2014), para 90, with regard to war crimes in Croatia, the Court noted that ‘among the main purposes of imposing criminal sanctions are retribution as a form of justice for victims and general deterrence aimed at prevention of new violations and upholding the rule of law’. 68 Ouwerkerk (n 4) 240. 69 PHPHMC van Kempen, Repressie door mensenrechten. Over positieve verplichtingen tot aanwending van strafrecht ter bescherming van mensenrechten (Nijmegen, Wolf Legal Publishers, 2008) 66 (author’s translation). 70 ibid.

Positive Obligations in View of the Ultima Ratio Principle  153 B.  Theoretical-Normative Issues In addition to more empirically focused issues relating to the positive obligations doctrine that has repercussions in the domain of criminal law, there is also a series of more theoretical or normative issues that should be considered. i.  The Meaning of Effectiveness Explaining the Court’s approach to human trafficking, Eriksson observes that if the domestic law tackling human trafficking is similar to that of the Palermo Protocol to the United Nations Convention against Transnational Organised Crime, the domestic law is considered effective, but that the Court has indicated that other legal formulations may also reach the required level of effectiveness. She argues that the methodology of the Court in delineating state obligations is ‘flawed’ in the sense that the Court demands effective laws, but does not clarify what this ‘effectiveness’ encompasses: The casuistic style of the Court negates its increasingly outspoken goal of developing the rules of the Convention for all Member States. The rather broad margin of appreciation of states in formulating domestic criminal laws conflicts with the demands of ‘effectiveness’ in protection.71

The first theoretical issue is thus conceptual and pertains to the meaning of ‘effectiveness’. What does effectiveness mean in the Court’s jurisprudence?72 What is the conception of effectiveness under which the Court operates? Effectiveness as a concept presupposes that the right means or measure is chosen for the desired goal – ‘an appropriate means-ends relationship’.73 But what is the desired goal? First, if ‘effective’ is something that completely prevents the same or similar conduct in the future, it is (a) impossible to make such claims without empirical evidence and (b) unreasonable to believe, as some crime will always persist.74 Second, if we took ‘effective’ to mean merely something that reduces 71 Eriksson (n 2) 339. 72 In the Council of Europe’s own document on the effectiveness of the Convention, effectiveness is understood as decreasing the backlog of pending applications (ie, increasing efficiency), the Member States’ reinforcing of the principle of subsidiarity ‘by more effectively embedding Convention standards in their domestic legal order’, getting Member States to sign and ratify protocols, and by tasking the Committee of Ministers with reinforcing and improving the means at their disposal to accelerate the implementation of the Court’s judgments, to take firmer measures in cases of repetitive non-compliance, and to grant more money to the Court in order to enable it to clear the backlog (Council of Europe, Committee on Legal Affairs and Human Rights, The Effectiveness of the European Convention on Human Rights: The Brighton Declaration and Beyond, Report (Strasbourg, Council of Europe, 2014)). Nowhere in the 19-page-long document on the effectiveness of the Convention are positive obligations mentioned. 73 L Lavrysen, ‘No “Significant Flaws” in the Regulatory Framework: E.S. v. Sweden and the Lowering of Standards in the Positive Obligations Case-law of the European Court of Human Rights’ (2013) 7 Human Rights and International Legal Discourse 145, 162. 74 E Durkheim, Les Règles de la Méthode Sociologique (Paris, Les Presses universitaires de France, 1895); Husak (n 8).

154  Nina Peršak the occurrence of similar events in the future, we would still need empirical evidence to back it up. However, this is a more plausible and defensible position regarding crime than the first version. Reducing means acknowledging that the crime will still occur only with less frequency. Third, if ‘effective’ is what the victim wishes and is satisfied with,75 then victimological knowledge should be sought on this. While it clearly would not be justified for justice to depend on the wishes of the concrete applicant, since justice delivered needs to be general, objective or averaged, there might be some room for an average or typical experience of this type of victim to be taken into consideration. However, drawing on victimological insights, the Court would be able to observe that punishment is not always (or regularly) what victims want or need,76 nor what they are entitled to. Seeing it as such would easily lead to ‘coercive overreach’77 or ‘widening of the net’ of social control.78 ii.  The Expressivity of Invoking Criminal Law If, however, the meaning cannot be determined or established (or there is no interest in clarifying these matters, suggesting that the invoking of ‘effectiveness’ is a symbolic rhetoric), then the second theoretical question is as follows: if symbolic/expressive function is invoked (even if couched in the rhetoric of ‘effectiveness’), is it a sufficient reason to criminalise certain conduct in view of the ultima ratio principle? If the expressed condemnation is all that should be required as an objective from criminal law, then it might easily be argued that criminal law should be employed as a first rather than as a last resort.79 This is thus essentially a question of criminal law justifications and ultimately of the scope of criminal law: how much criminal law do we want? If we still subscribe to the so-called minimalist approach to criminal law,80 that is, the approach that sees the criminal law as an ‘evil’, as the biggest interference with the individual’s liberty and therefore to be avoided if possible, then

75 Some indeed interpret the development of the positive obligations doctrine in this victimoriented fashion, seeing positive obligations under ECHR as creating ‘unavoidable duties to uphold victims’ rights’ (C de Than, ‘Positive Obligations under the European Convention on Human Rights: Towards the Human Rights of Victims and Vulnerable Witnesses?’ (2003) 67 Journal of Criminal Law 165, 182. 76 A recent US research has, for example, revealed that victims overwhelmingly favour rehabilitation over punishment (Alliance for Safety and Justice, Crime Survivors Speak (ASJ, 2016), https:// shadowproof.com/wp-content/uploads/2016/08/Crime-Survivors-Speak-Report.pdf). 77 L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce?’ in L Zedner and J Roberts (eds), Principles and Values in Criminal Law and Criminal Justice: Essays in Honour of Andrew Ashworth (Oxford, Oxford University Press, 2012) 149. 78 S Cohen, Visions of Social Control (Cambridge, Polity Press, 1985). 79 However, it must be borne in mind that ‘criminal code’ is not meant to be a comprehensive societal moral code, but rather the societal minimum of morality or, in Jellinek’s words, ‘the ethical minimum’ in a given society. See G Jellinek, Die Sozialethische Bedeutung von Recht, Unrecht und Strafe (Berlin, Verlag von O Härin, 1908) 45. 80 Ashworth (n 8).

Positive Obligations in View of the Ultima Ratio Principle  155 one wonders whether ‘effectiveness’ could ever be seen as a sole justification for inflicting criminal punishment or criminalisation, which precedes it. Jareborg insists that: To the extent that we live in a social/liberal/conservative-democratic Rechtsstaat, we are not any longer allowed to use criminal punishment just because we want it, or because a god is said to want it, or because we have always done it, or because it seems to be a natural or an effective means to some end. (Emphasis added)81

This suggests that under the rule of law, if taken seriously, criminal law should not be used only because it might be effective to achieve some end. In the ECtHR jurisprudence, any interference with human rights must be necessary in a democratic society and proportionate to the legitimate aim pursued. If freedom/liberty is taken as the most elemental human value and right – and many would agree that various ECHR human rights are merely the realisation of this core human right to freedom (libertas) and autonomy – then the Court should be careful in requiring the Contracting Parties to resort to criminalisation without being certain (and able to support it with evidence) that this interference with human rights is really necessary in a democratic society and is proportionate to the aim.82 iii.  The Risk of Circularity of Reasoning The Court has emphasised on several occasions that it does not aim to replace the national authorities and that the latter are free to choose among the possible measures in order to secure compliance with the ECHR, but that the protection of Convention rights has to be effective (‘not theoretical or illusory, but practical and effective’). The Court requires an effective protection of fundamental rights, which includes the deterrent effect on possible future human rights violations. If ‘effective’ means deterrent, as it seems to in this context, and if deterrence requires criminal law prohibition (as it most likely seems to, since the very fact that the term ‘deterrence’ is invoked suggests that it is a matter for criminal law),83 this assumes that some state measure was not (or could not be) effective if it had not been a proper criminal law measure. We have already queried above whether the Court has any evidence upon which it could base such conclusions. However, statements such as ‘[f]or an investigation to be effective in practice it is a prerequisite that the State has enacted

81 Jareborg (n 12) 522. 82 It seems that the Court forgets that criminalisation is by default in itself an interference with the individual’s core human right to freedom. 83 One rarely finds reference to deterrence in administrative law or civil law. While the concept of ‘punitive damages’ in American civil law could be understood as having a deterrent function, this is a criminal law element introduced into civil law rather than civil law proper. They are ‘a way of punishing the defendant in a civil lawsuit’ (https://legal-dictionary.thefreedictionary.com/ punitive+damages, emphasis added).

156  Nina Peršak criminal-law provisions penalising practices that are contrary to Article 3’84 also seem circular. Leaving aside the fact that probably no one would problematise seeing Article 3 violations (such as torture) as criminal,85 such formulations run the risk of committing a logical fallacy of petitio principii (begging the question), where a premise is assumed to be true without warrant or that which is yet to be proved is implicitly already taken for granted. The premise that an investigation cannot be effective without prior criminalisation (as if civil, disciplinary or administrative matters somehow cannot be investigated) represents – in the above quoted sentence – the justification for criminalisation (a criminal law measure), yet it is something that itself first needs to be proven. The Court has also emphasised that effective investigation is ‘not an obligation of result, but one of means’;86 however, criminalisation is also a ‘result’ of sorts, so the above-mentioned explicit argument by the Court that criminalisation (a criminal law measure) should have been used by the state in order not to violate the applicant’s human right seems somewhat inconsistent with this obligation-of-means requirement. iv.  Practical Reasons? Lastly, we should also not forget a more sociologically informed practical reason behind the development of positive obligations. Reading the Report of the Evaluation Group to the Committee of Ministers on the European Court of Human Rights from 27 September 2001, Mowbray has pointed out that: [T]he Court’s desire to minimise the use of time-consuming and expensive factfinding missions to examine specific complaints had been achieved, in part, by imposing positive obligations to conduct effective investigations into killings/illtreatment-disappearances on states. Likewise, the full-time Court’s establishment of an obligation upon states to provide an effective domestic remedy to deal with complaints of unreasonable delays in civil and criminal proceedings (Kudla) was partly inspired by the wish to reduce the overwhelming number of cases being lodged at Strasbourg alleging a breach of the reasonable time guarantee enshrined in Article 6(1).87

Although rarely so openly stated, this rationale behind positive obligations throws an interesting additional light on the ‘effectiveness’ debate – one that has more to do with the Court and its workload than with any deterrence or other legal goals. Many concepts that seem and sound highly theoretical and normative may thus, on a closer or more socio-legal inspection, turn out to have (partially or wholly) non-normative, pragmatic roots, which can have unspoken implications for the ways in which such concepts further develop.

84 Gäfgen

v Germany App No 22978/05 (ECtHR, 1 June 2010), para 117. Natasa Mavronicola, ch 9 in this volume. 86 Grimailovs v Latvia App No 6087/03 (ECtHR, 25 June 2013). 87 Mowbray (n 38) 222. 85 See

Positive Obligations in View of the Ultima Ratio Principle  157 IV. CONCLUSION

Positive obligations doctrine has led to the expansion of human rights, which can in addition to many positive consequences also lead to some adverse results that can be: (a) detrimental for criminal law in view of the ultima ratio principle being effectively negated; and (b) detrimental for human rights law. For the latter, they may be disadvantageous partly for the same reason, as the ultima ratio principle protects the liberty of individuals, and partly for the reason that the over-expansion of human rights – like any over-extension of a concept that should act as a critical limiting tool on state power – can lead to the watering down of the concept of human rights and to the latter concept losing its critical bite when everything can be conceived as a human rights issue or subsumed under human rights.88 Shue makes a persuasive point, arguing that the division between positive and negative duties as well as the concept of every single right having a single correlating duty is misleading, as, according to his view, every basic right has three types of correlating duties (ie, duties to avoid depriving, duties to protect against deprivation and duties to aid the deprived)89 and that therefore all basic rights involve some positive obligations, which can explain the development of Convention’s positive obligations.90 However, the question is whether the Convention is equipped to cover all those duties and whether the Court has the competence to extend the Convention rights in such a way. A positive obligation to criminalise would represent the ‘duty to protect’,91 which in itself is beyond reproach; nevertheless, whether the protection can and should be achieved by criminalisation is far from obvious. The ‘dynamic interpretation’ of the Convention leaves enough room for manoeuvre to the Court for it to continue developing positive obligations with respect to the ‘legitimate needs

88 H Hannum, ‘Back to the Future: A Proposal for a New Human Rights Strategy for the 21st Century’ (Lecture at the Central European University), 1 March 2018 at Central European University, Budapest. A more parsimonious approach to human rights is also advocated by others who are concerned with the disorganised proliferation of rights or ‘rights inflation’ that may lead to the devaluing of the notion of human rights as a whole. See, eg, A Buchanan, Justice Legitimacy and Self-Determination (Oxford, Oxford University Press, 2003); J Tasioulas, ‘Towards a Philosophy of Human Rights’ (2012) 65 Current Legal Problems 1. 89 Some claim that the distinction between positive and negative obligations is in fact passé, finding more current a tripartite typology of obligations, namely, the obligation to respect, the obligation to protect and the obligation to fulfil human rights. The latter two categories largely overlap with ‘positive obligations’. See DJ Harris, M O’Boyle, E. Bates and CM Buckley, Law of the European Convention on Human Rights (Oxford, Oxford University Press, 2014). Moreover, others argue that it has become increasingly difficult today to draw a strict line between action and inaction and to be able to properly distinguish between the two (Lavrysen (n 73)) or that the negative-positive dichotomy is a false one, since all rights have correlative obligations or duties (B Dickson, ‘Positive Obligations and the European Court of Human Rights’ (2010) 61 Northern Ireland Legal Quarterly 203). 90 In Mowbray (n 38) 224–25. 91 H Shue, Basic Rights, 2nd edn (Princeton, Princeton University Press, 1996).

158  Nina Peršak of the applicants and itself assessed against the background of contemporary European societies’.92 However, as argued above, every criminalisation or state intrusion into the individual’s liberties via criminal law requires good reasons. Similar to the fact that making room for victims’ rights in the criminal process does not mean that the rights of the accused should be diminished, the positive obligations imposed on the Contracting Parties under the ECHR should not lead to the automatic interference with the individual’s liberty represented by criminalisation. While positive obligations such as requiring effective investigations do not, when accompanied by appropriate legal safeguards, encroach upon the individual’s rights and freedoms in a long-term intrusive way, positive obligations that require criminalisation are much more complex and inevitably widen the net of social control, thereby reducing the space of individual freedom with long-lasting consequences. This makes them different from other types of positive obligations and calls for a certain restraint or une main tremblante in the ECtHR’s adjudication of this sort. In other words, notwithstanding the theoretical and factual possibility of developing human rights in such a way that positive obligations leading to the extension of criminal law follow from this development, the question is whether such a development is also desirable.93 Bearing in mind the principle of subsidiarity and the doctrine of the margin of appreciation, added by Protocol 15 also as a new recital into the Preamble to the Convention, any attempts at supranational judicial criminal law-making (albeit indirectly, ie, through finding violations of the positive obligation to criminalise) should be seriously measured, if not completely avoided. The positive obligations doctrine has undoubtedly done a lot of good. It has taught the Contracting States that in this day and age, it is simply not sufficient to only abstain from interference with human rights, but that their protection should be ensured with more proactive measures as well. The ECtHR’s case law with regard to positive obligations has also ‘contributed to the partial erosion of the generational gap between Convention rights and later generations of international human rights’, which can be seen as a considerable achievement.94 It has helped refine the substance of the domestic laws, criticising, for example, their lack of clarity.95 The doctrine of positive obligations has thus certainly expanded our rights and made them more tangible through the detailing of the state’s duties, which should be rightly recognised. However (and this relates only to those positive obligations that lead to increased repression),96 acknowledging such accomplishments should not prevent scholars from warning against the commensurate treatment of judicial action that creates positive obligations



92 Mowbray

(n 38) 229. Kempen (n 69) 66. 94 Mowbray (n 38) 230–31. 95 Eriksson (n 2) 368. 96 Many positive obligations obviously do not result in coercion or increased repression. 93 Van

Positive Obligations in View of the Ultima Ratio Principle  159 concerning effective investigations, adjudications and other measures, on the one hand, and judicial action creating obligations that directly demand criminalisation, on the other, as they are different. While the former is easier to defend, the latter requires some further justification to be provided. If the states are allowed to limit certain human rights of individuals only when such limitations are not only lawful (in accordance with the law) and pursuing a recognised legitimate aim but also necessary in a democratic ­society, the Court would honour the Convention by abiding by the same rules – that is, demonstrating that criminalisation (when lacking) is a measure that is absolutely necessary in a democratic society for the protection of a legitimate aim (human rights), when applying or creating positive obligations under Articles 8–11 ECHR. It follows from this that the Court should abide by the ultima ratio principle, which is in its essence very similar to the requirement of necessity, and therefore mandate that criminalisation in the context of positive obligations (as well as more broadly) can only be used as a measure of last resort. While still fairly contained at this point, the doors of obligations to criminalise are kept ajar and can perhaps be more widely opened with a slightly stronger wind arising from more coercive, non-liberal dispositions or even from compassionate responses to the victim-applicant’s suffering and strong reactions to otherwise atypical events with devastating, out-of-the-ordinary consequences. The need for a thorough legal analysis and presentation of compelling good reasons for judicial choices that would reduce doctrinal opacity and charges of potential unjust ‘widening of the net’ is therefore pronounced. To end on a more positive note, one should not forget that the ECtHR is, through its concept of ‘criminal charge’ (Article 6 ECHR) as autonomously conceived, also in a very good position to quash certain attempts by Contracting Parties to disguise – for example, by invoking the justification of ultima ratio – their criminalisation and consequent punishment of a certain conduct through the use of administrative or regulatory law instead of criminal law proper. By revealing the disingenuous national bids to (ab)use the principle of last resort in order to camouflage the real criminal charge and criminal sanction as non-criminal ones, for which criminal law procedural guarantees under Article 6 need not apply, the Court can also in this way importantly contribute to the respect of the principle of last resort.

160

8 Sowing a ‘Culture of Conviction’ What Shall Domestic Criminal Justice Systems Reap from Coercive Human Rights? MATTIA PINTO*

They that sow the wind, shall reap the whirlwind (Hosea 8:7)

I. INTRODUCTION

I

n general, we tend to think of human rights bodies as moderating and restraining state penal policies. These institutions appear to have ‘humanised’ criminal justice by upholding and strengthening procedural and substantive rights to protect the individual from the state’s arbitrary use of criminal mechanisms.1 However, human rights bodies have also worked to expand penality.2 In recent years, they have increasingly interpreted their mandate in monitoring compliance with human rights conventions as comprising

* I am very grateful for the comments and discussion with Conor Gearty, Peter Ramsay, Diletta Machesi, Natasa Mavronicola, Laurens Lavrysen, Corina Heri, Barrie Sanders and Virginia Mantouvalou. I also benefited from thoughtful discussion and criticism at the ICON-S Annual Conference in Santiago (July 2019). 1 Substantive rights include, for instance, the provisions that crimes and punishment must be established by law, the prohibition of inhuman and degrading treatment or punishment, and the prohibition upon unlawful or arbitrary detention. Procedural rights include the right to a fair trial, the presumption of innocence and the need to prove the guilt of the accused beyond reasonable doubt. 2 In this chapter, I use ‘penality’ to refer to the entire penal sphere, including its laws, sanctions, institutions, practices, discourses and representations. See, eg, D Garland, Punishment and Welfare: A History of Penal Strategies (Aldershot, Gower, 1985) x; D Garland, ‘Penality and the Penal State’ (2013) 51 Criminology 475, 476.

162  Mattia Pinto the imposition of state obligations in criminal matters.3 In particular, the European Court of Human Rights (ECtHR) has developed ample case law demanding that states criminalise, prosecute and punish serious human rights violations and avoid acts of ‘criminal clemency’ (decriminalisation and amnesty laws).4 The result is the emergence of a jurisprudence of ‘coercive human rights law’, whereby rights and freedoms simultaneously humanise and enhance state penal practices.5 Some attention has been devoted in the literature to the evolution towards coercive human rights in human rights bodies’ case law, and its immediate effects on the theory of both human rights and criminal law.6 Yet, there is little to no analysis on how this jurisprudence affects domestic criminal justice systems. Aiming to fill this gap, this chapter assesses some actual and potential implications that the ECtHR’s duties to deploy the criminal law (might) have on domestic legal systems. It shows that the Court tends to present criminal accountability as indispensable to protect human rights. This approach may foster a ‘culture of conviction’ at the domestic level whereby punishment is seen as the end to pursue, whatever the cost.7 While the ECtHR’s jurisprudence currently refers to the duty to punish as an obligation of means, increased concern with the efficiency of criminal justice systems in preventing crime is leading the Court to consider whether adequate punishment has been imposed. Such an invocation of penal measures may have adverse implications in practice at the domestic level. Through the case study of the Modern Slavery Act (MSA), adopted in England and Wales in 2015, this chapter illustrates how domestic authorities may appropriate the language of coercive human rights to foster and legitimise penal expansion. The MSA appears to be heading towards a more repressive criminal justice system, with growing limitations on due process rights (for both

3 See generally M Pinto, ‘Awakening the Leviathan through Human Rights Law: How Human Rights Bodies Trigger the Application of Criminal Law’ (2018) 34 Utrecht Journal of International and European Law 161; A Seibert-Fohr, Prosecuting Serious Human Rights Violations (Oxford, Oxford University Press, 2009). 4 See, eg, K Kamber, Prosecuting Human Rights Offences: Rethinking the Sword Function of Human Rights Law (Leiden, Brill, 2017). 5 F Tulkens, ‘The Paradoxical Relationship between Criminal Law and Human Rights’ (2011) 9 Journal of International Criminal Justice 577, 538. 6 See, eg, Pinto (n 3); Tulkens (n 5); F Tulkens, ‘Human Rights as the Good and the Bad Conscience of Criminal Law’ in S Snacken and E Dumortier (eds), Resisting Punitiveness in Europe (Abingdon, Routledge, 2012); M Sorochinsky, ‘Prosecuting Torturers, Protecting Child Molesters: Toward a Power Balance Model of Criminal Process for International Human Rights Law’ (2009) 31 Michigan Journal of International Law 157; Seibert-Fohr (n 3); L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce?’ in L Zedner and JV Roberts (eds), Principles and Values in Criminal Law and Criminal Justice (Oxford, Oxford University Press, 2012); Kamber (n 4). 7 The idea that coercive human rights may encourage a ‘culture of conviction’ was first exposed, in the context of the Inter-American Court of Human Rights, by F Mégret and JP Calderón, ‘The Move towards a Victim-Centred Concept of Criminal Law and the “Criminalization” of Inter-American Human Rights Law: A Case of Human Rights Law Devouring itself’ in Y Haeck, O Ruiz-Chiriboga and C Burbano-Herrera (eds), The Inter-American Court of Human Rights Theory and Practice, Present and Future (Cambridge, Intersentia, 2015) 438.

Sowing a ‘Culture of Conviction’  163 defendants and victims of crime), harsher punishments and wider powers of arrest and detention. Conversely, criminal justice reform initiatives, directed at reducing unnecessary criminalisation and implementing alternatives to prison, are totally neglected. The discussion in this chapter is divided up as follows. Section II presents the current trends of penal expansion in Europe. Drawing on a number of influential texts on the ‘carceral turn’ in contemporary society,8 it connects the growing punitive tendencies in European states with the widespread assumption that criminal law can help to address pressing social problems. Section III outlines the ECtHR’s case law on coercive human rights and considers how the Court has laid the groundwork for promoting a ‘culture of conviction’ at the domestic level. Section IV uses the MSA as a case study to investigate how domestic criminal justice systems may react to coercive human rights. Ultimately, it is argued that the ECtHR’s jurisprudence has encouraged an expansion of penality in the UK under a benevolent guise that masks the oppressive effects of extensive powers to criminalise and punish. II.  THE RISE OF THE PENAL STATE IN EUROPE

Several studies have observed a rise of punitiveness in the Western world since the 1980s.9 This phenomenon, characterised by a decline of rehabilitative ideals, a growth in incarceration rates, expanded application of criminal sanctions and increased police control, concerns not only the US, but also, to a lesser extent, certain European countries.10 The ever-expanding punitive outlook of contemporary democracies has raised great concern among progressive activists and scholars, who have long believed that reliance on criminal law and punishment should be as limited as possible. The history of criminal law is indeed largely a history of abuses and cruel practices.11 Criminalisation and punishment constitute the most severe forms of state interference with the individual’s private sphere, and penal power, when increased, is prone to abuse and injustice.12

8 On the ‘carceral turn’ in US and Western European societies, see generally E Bernstein, ‘Carceral Politics as Gender Justice? The “Traffic in Women” and Neoliberal Circuits of Crime, Sex, and Rights’ (2012) 41 Theory and Society 233, 236. 9 See, eg, D Garland, The Culture of Control: Crime and Social Order in Contemporary Society (Oxford, Clarendon Press, 2001); J Pratt, The New Punitiveness: Trends, Theories, Perspectives (Cullompton, Willan, 2005); J Simon, Governing through Crime: How the War on Crime Transformed American Democracy and Created a Culture of Fear (Oxford, Oxford University Press, 2007); L Wacquant, Punishing the Poor: The Neoliberal Government of Social Insecurity, Englishlanguage edn (Durham, NC, Duke University Press, 2009). 10 S Snacken, ‘Resisting Punitiveness in Europe?’ (2010) 14 Theoretical Criminology 273. 11 N Jareborg, ‘What Kind of Criminal Law Do We Want?’ in Annika Snare (ed), Beware of Punishment: On the Utility and Futility of Criminal Law, vol 14 (Oslo, Pax Forlag, 1995) 18. 12 PH van Kempen (ed), Criminal Law and Human Rights (Farnham, Ashgate, 2014) xi.

164  Mattia Pinto In the influential book The Culture of Control, David Garland critically examined these trends, by presenting a historical-cultural account of strategies of crime control since the 1970s in the US and the UK.13 He illustrated how a new ‘culture of control’ has risen and expanded in these two countries. This ‘culture of control’ is characterised by two patterns of action: a strategy of prevention and partnership between state and non-state agencies, and a strategy of ‘punitive segregation’ aimed at harsher punishment and increased incapacitation.14 Garland singled out several examples of this phenomenon, including mandatory sentencing laws, more proactive and preventive policing efforts, harsher penalties and emphasis on ‘the victim’ as justification for more severe responses to crime, increased concern with prevention, harm reduction and risk management, community notification laws and offenders’ registers, private policing, zero-tolerance policies and Anti-social Behaviour Orders.15 Garland also suggested that, despite considerable variation from place to place and some exceptions, other Western industrialised nations appeared to be heading down the same path.16 This is also the position taken by the French sociologist Loïc Wacquant. For him, ‘the explosive growth of the scope and intensity of punishment’ also affected Western Europe.17 In Punishing the Poor, he reconnected this phenomenon to the global expansion of neoliberalism, which has been imported into the European continent from the US and has transformed the welfare state into a penal one.18 According to him, the ‘European road’ to punitiveness (with French, Dutch, Italian and other variants) is characterised by social and penal policies that seek to regulate those on the margins of society.19 Welfare services and punitive interventions are combined for the purpose of supervising, controlling, criminalising and punishing ‘the poor’. Wacquant observed that in Europe, the ‘penalization of poverty’ is mainly driven by the police and the courts rather than prisons.20 Elements of the European punitive tendency are the upsurge in police intervention, the escalation of penalties for repeated offences, the subordination of social services to judicial authorities, and the extension of powers to arrest and detain.21 The punitive trends described by Garland and Wacquant sped up in the wake of the attacks of 11 September 2001.22 On both sides of the Atlantic, 13 Garland (n 9). 14 ibid 139–46. 15 ibid chs 6–7. 16 ibid 91. 17 Wacquant (n 9) xvi. 18 ibid 20–21. 19 ibid 24. 20 ibid. 21 ibid pt 4. 22 E Dumortier et al, ‘The Rise of the Penal State: What Can Human Rights Do about it?’ in Sonja Snacken and Els Dumortier (eds), Resisting Punitiveness in Europe (Abingdon, Routledge, 2012) 109–10.

Sowing a ‘Culture of Conviction’  165 several countries adopted numerous counter-terrorism measures that bolstered their penal systems. Western democracies passed laws aimed at facilitating arrests and detentions and at reducing defendants’ rights, whilst increasing governmental powers to interfere with individuals’ privacy and private life.23 However, the rise of the penal state has involved not only tangible developments within the criminal justice system, but also more punitive attitudes at the level of social and political discourse. This phenomenon involves a growing sensibility about revenge by crime victims,24 renewed attention towards ‘popular’ demands of punishment,25 diffusion of images of insecurity by the media26 and political initiatives promoting ‘law and order’ thinking.27 The normal citizen is transformed into a victim of crime, whose vulnerability requires a greater resort to criminalisation and penal severity.28 The underlying assumption is that criminal sanctions are an effective tool for reducing crime, providing justice and solving complex social issues. It is a form of ‘penal optimism’,29 summed up by the assertion that ‘prison works’.30 This punitive fervour attracts not only ‘law and order’ politicians, but at times also liberal and progressive movements (victim advocates, feminist activists and ecological groups), which view the penal system as a tool of social reform and a source of protection for vulnerable individuals.31 The rise of the penal state has not involved every European country in the same way. As Sonja Snacken reports, some states ‘do not easily fit this

23 See generally G Agamben, State of Exception (Chicago, University of Chicago Press, 2005) 3; B Evans, Liberal Terror (Cambridge, Polity Press, 2013). 24 I Aertsen, ‘Punitivity from a Victim’s Perspective’ in S Snacken and E Dumortier (eds), Resisting Punitiveness in Europe (Abingdon, Routledge, 2012). 25 See generally J Pratt, Penal Populism (Abingdon, Routledge, 2007); A Bottoms, ‘The Philosophy and Politics of Punishment and Sentencing’ in CMV Clarkson and R Morgan (eds), The Politics of Sentencing Reform (Oxford, Clarendon Press, 1995). 26 Simon (n 9) 135. 27 R Reiner, Law and Order: An Honest Citizen’s Guide to Crime and Control (Cambridge, Polity Press, 2007). 28 P Ramsay, The Insecurity State: Vulnerable Autonomy and the Right to Security in the Criminal Law (Oxford, Oxford University Press, 2012) 228; Simon (n 9) 76. 29 CM Webster and AN Doob, ‘Penal Optimism: Understanding American Mass Imprisonment from a Canadian Perspective’ in KR Reitz (ed), American Exceptionalism in Crime and Punishment (Oxford, Oxford University Press, 2017). 30 The phrase was coined by former British Home Secretary Michael Howard in a speech at the British Conservative Party conference on 6 October 1993 (‘Let us be clear. Prison works. It ensures that we are protected from murderers, muggers and rapists, and it makes many who are tempted to commit crime think twice’). See C Brown, ‘Howard Seeks to Placate “Angry Majority”: Home Secretary Tells Party’, The Independent, 7 October 1993, www.independent.co.uk/news/ uk/howard-seeks-to-placate-angry-majority-home-secretary-tells-party-that-balance-in-criminaljustice-1509088.html. 31 AJ Do Amaral, ‘The Penal Policy of Human Rights’ (2013) 1 Zeitschrift für Internationale Strafrechtsdogmatik 61, quoting ML Karam, Recuperar o Desejo Da Liberdade e Conter o Poder Punitivo, vol 1 (1st edn, Lumen Juris 2009) 4; H Aviram, ‘Progressive Punitivism: Notes on the Use of Punitive Social Control to Advance Social Justice Ends’ (2019) UC Hastings Research Paper No 364, www.ssrn.com/abstract=3404276.

166  Mattia Pinto bleak vision of punitiveness’.32 While in certain countries imprisonment rates have grown steadily (the UK, Spain, the Netherlands and France), in others the prison population has remained stable (the Scandinavian countries and Germany) or has even decreased (the Eastern European countries).33 In this regard, it has been observed that levels of punitiveness are to be correlated with welfare investment and political economy,34 but also with different democratic political structures.35 Moreover, for some authors, if Europe has never achieved the same level of punitiveness as the US, this is due to the different role that human rights law has played in the two contexts.36 The European Convention on Human Rights (ECHR) is often mentioned as one of the barriers against an uncontrolled expansion of penality.37 According to Denis Salas, for instance, the monitoring role of the ECtHR is a key factor in explaining the more humane and moderate penal and sentencing policies on the European continent.38 It is probably true that ECtHR case law has somewhat constrained the penal actions of European states. Examples that are often reported are the abolition of the death penalty, the protection of prisoners’ rights, the prohibition of indefinite detention, the decriminalisation of homosexuality and the protection of freedom of expression against undue criminalisation.39 Nonetheless, the jurisprudence of the ECtHR is also ambivalent. Not only has the Court at times hesitated to counter expansive national penal policies, but in some other cases it has backed or even promoted the introduction of new criminal offences, more policing powers, the institution of criminal proceedings and harsher punishments. III.  ECTHR CASE LAW ON COERCIVE HUMAN RIGHTS

A.  State Obligations in Criminal Matters The ECtHR has developed a robust body of case law requiring states to deploy their domestic criminal law to counter human rights violations and protect

32 Snacken (n 10) 274. 33 ibid. 34 N Lacey, The Prisoners’ Dilemma: Political Economy and Punishment in Contemporary Democracies (Cambridge University Press, 2008). 35 T Lappi-Seppälä, ‘Explaining National Differences in the Use of Imprisonment’ in S Snacken and E Dumortier (eds), Resisting Punitiveness in Europe (Abingdon, Routledge, 2012). 36 Dumortier et al (n 22) 113. 37 Snacken (n 10) 283. 38 D Salas, La Volonté de Punir : Essai Sur Le Populisme Pénal (Paris, Hachette Littératures, 2008) 124–40; mentioned in Dumortier et al (n 22) 113. 39 Dumortier et al (n 22) 114–20. On freedom of expression, see also Stephanos Stavros, ch 6 in this volume.

Sowing a ‘Culture of Conviction’  167 the rights and freedoms enshrined in the ECHR.40 As Laurens Lavrysen has shown in his chapter in this volume,41 the duty to have recourse to criminal law solutions generally has two strands: first, the Court demands that states criminalise conduct that harms individuals’ liberties; and, second, it requires that effective criminal law mechanisms (investigation, prosecution and punishment) are put in place and pursued effectively as enforcement tools of human rights provisions. The ECtHR has provided for the duty to criminalise in relation to Articles 2, 3, 4 and 8 ECHR.42 Given the scope of this chapter, I shall focus here only on Article 4. In Siliadin v France, for instance, the Court censured French law for not providing sufficient and effective criminal law protection against ‘servitude’.43 The case involved a Togolese girl who had worked for four years as a domestic servant for a married couple without remuneration or freedom of movement. The ECtHR found that Article 4 ECHR ‘must be seen as requiring the penalisation and effective prosecution’ of ‘any act aimed at maintaining a person’ in a situation of slavery, servitude, forced or compulsory labour.44 Similarly, in CN v UK, the Court found that the government’s failure to properly investigate the complaints of domestic servitude made by the applicant, a woman who had travelled to the UK from Uganda, was due to the absence of specific legislation criminalising such a form of wrongdoing.45 In Rantsev v Cyprus and Russia, the victim, a Russian woman, had been trafficked from Russia to Cyprus for reasons of sexual exploitation and the European judges extended the protection offered by Article 4 to trafficking in human beings.46 As a result, states are required to adopt measures intended to prevent trafficking, protect victims, and criminalise and punish traffickers.47 The ECtHR also orders states to enforce their criminal law.48 The Court has recently clarified that in cases of serious violence such as assaults for racial reasons or police ill-treatment, civil compensation or administrative proceedings

40 Kamber (n 4); S Manacorda, ‘“Dovere Di Punire”? Gli Obblighi di Tutela Penale Nell’era della Internazionalizzazione del Diritto’ in M Meccarelli, P Palchetti and C Sotis (eds), Il lato oscuro dei diritti umani: Esigenze emancipatorie e logiche di dominio nella tutela giuridica dell’individuo (Madrid, Universidad Carlos III de Madrid 2014); Lazarus (n 6). 41 See Laurens Lavrysen, ch 2 in this volume. 42 X and Y v The Netherlands Series A No 91 (1986) 8 EHRR 235, para 27; MC v Bulgaria (2005) 40 EHRR 20, paras 150 and 174; Kiliç v Turkey (2001) 33 EHRR 58, para 62; Cestaro v Italy App No 6884/11 (ECtHR, 7 May 2015), para 246; CN and V v France App No 67724/09 (ECtHR, 11 October 2012), para 105. 43 Siliadin v France (2006) 43 EHRR 16, para 89. 44 ibid para 112. 45 CN v UK (2013) 56 EHRR 24, para 81. 46 Rantsev v Cyprus and Russia (2010) 51 EHRR 1, para 282. See also Chowdury and Others v Greece App No 21884/14 (ECtHR, 30 March 2017), para 86. 47 Rantsev (n 46) para 285. 48 Öneryildiz v Turkey (2005) 41 EHRR 20, para 94; Ali and Ayşe Duran v Turkey App No 42942/02 (ECtHR, 8 May 2008), para 61.

168  Mattia Pinto are not enough and prosecution is required.49 The European judges have tried to attenuate this provision by clarifying that: ‘The obligation to conduct an effective investigation is an obligation not of result but of means.’50 However, this qualification is rendered immaterial when the judges hold that the institution of criminal proceedings cannot be detached from their ‘issue naturelle’, namely the punishment of those responsible.51 In the Court’s opinion, ‘not doing enough to prosecute and punish those responsible’ and confining the state’s reaction to the mere payment of compensation makes any investigation ineffective in practice.52 In Ali and Ayşe Duran v Turkey, the European judges explained that: ‘While there is no absolute obligation for all prosecutions to result in conviction or in a particular sentence, the national courts should not under any circumstances be prepared to allow life-endangering offences and grave attacks on physical and moral integrity to go unpunished.’53 Finally, the ECtHR insists that those responsible for serious human rights violations should serve a severe sentence.54 In its view, the duty to resort to criminal law would lose much of its meaning if perpetrators were punished with too lenient a sanction.55 Where the ECtHR is confronted with situations of human rights abuses of a certain gravity (eg, torture, killings, trafficking and forced labour, as well as domestic and sexual violence), criminal justice appears as an indispensable element of human rights protection and promotion. The underlying rationales are mostly deterrence, prevention and adherence to, or restoration of, the rule of law.56 For the Court, criminal law safeguards society as a whole by ending impunity and providing general human rights protection.57 The ECtHR also has a tendency to require criminal accountability in the interests of individual victims. In Jelić v Croatia, for instance, the European judges noted that ‘among the main purposes of imposing criminal sanctions [is] retribution as a form of justice for victims’.58 Even if the Court has yet to recognise the victim’s right to have criminal proceedings instituted against third parties,59 criminal law mechanisms 49 Movsesyan v Armenia App No 27524/09 (ECtHR, 16 November 2017), para 74; Jeronovičs v Latvia App No 44898/10 (ECtHR [GC] 05 July 2016), para 105; OR and LR v Moldova App No 24129/11 (ECtHR, 30 October 2018), para 77. 50 Mustafa Tunç and Fecire Tunç v Turkey App No 24014/05 (ECtHR [GC], 14 April 2015), para 173. 51 Nasr and Ghali v Italy App No 44883/09 (ECtHR, 23 February 2016), para 272. 52 Jeronovičs (n 49) para 105; Gäfgen v Germany (2011) 52 EHRR 1, para 119. 53 Ali and Ayşe Duran v Turkey (n 48) para 61. See also Öneryildiz (n 48) para 96; Jelić v Croatia (2015) 61 EHRR 43, para 77. 54 OR and LR v Moldova (n 49) para 79 (the ECtHR considered that five years’ imprisonment was a too lenient sentence in a case of police ill-treatment). 55 Nikolova and Velichkova v Bulgaria (2009) 48 EHRR 40, para 62 (the ECtHR held that punishing police ill-treatment resulting in death with suspended terms of imprisonment amounted to a violation of Article 2 ECHR). 56 Opuz v Turkey (2010) 50 EHRR 28, para 128. 57 OR and LR v Moldova (n 49) paras 58–59. 58 Jelić (n 53) para 90; El-Masri v FYROM (2013) 57 EHRR 25, para 192. 59 Movsesyan v Armenia App No 27524/09 (ECtHR, 16 November 2017), para 62. See also Mazepa and Others v Russia (2019) 68 EHRR 11, para 69.

Sowing a ‘Culture of Conviction’  169 are to be available to those affected by serious human rights violations.60 As Krešimir Kamber has persuasively contended, the victim has a ‘right-claim to the application of criminal law mechanisms’ without any flaws or culpable failures in the process.61 This tendency is clearly visible in Al Nashiri v Romania and Abu Zubaydah v Lithuania, where the ECtHR held that ‘the notion of an “effective remedy”’ for the victim entails criminal proceedings ‘leading to the identification and punishment of those responsible’.62 B.  Encouraging a ‘Culture of Conviction’ Throughout its case law, the ECtHR has maintained that the criminalisation, prosecution and punishment of certain abuses are required to secure human rights. In so doing, the Court shows a great confidence in the deployment of penal powers and gives criminal law a positive outlook, as a painkiller to redress social and moral harms.63 The gravity of the harm caused and the seriousness of the abuses committed also contribute to presenting criminal law as the obvious and self-evident instrument to respond to serious human rights violations. Critical arguments concerning the risks of overcriminalisation are, on the other hand, dismissed.64 The European judges appear to regard criminal law solutions as the principal and most effective way to prevent grave ECHR breaches and give redress to victims. Currently, the ECtHR’s jurisprudence refers to the duty to punish and prosecute as obligations of means. However, increased concern with the effectiveness of the criminal justice system in preventing crime and impunity is leading the Court to ultimately consider whether adequate punishment has been, or could have been, imposed.65 When criminal justice systems are criticised, it is usually because they are not effective enough in securing convictions and punishments. In Nasr and Ghali v Italy, for example, Italian investigators and judges had identified and convicted those responsible for the ‘extraordinary rendition’ of the Egyptian imam Abu Omar.66 The ECtHR also acknowledged that the investigations had been ‘effective et profonde’.67 However, Italy was found to be in violation of the Convention because it had created a situation

60 Pulfer v Albania App No 31959/13 (ECtHR, 20 November 2018), para 87. 61 Kamber (n 4) 507. 62 Al Nashiri v Romania (2019) 68 EHRR 3, para 706; Abu Zubaydah v Lithuania App No 46454/11 (ECtHR, 31 May 2018), para 673. 63 Similarly, in the context of the IACtHR, see DR Pastor, ‘La Deriva Neopunitivista y el Desprestigio Actual de los Derechos Humanos’ [2006] Jura Gentium, www.juragentium.org/topics/ latina/es/pastor.htm. 64 Tulkens (n 6) 157. 65 Similarly, see Sorochinsky (n 6) 186. 66 Nasr and Ghali (n 51). 67 ibid para 272.

170  Mattia Pinto of impunity by not punishing some of the convicted persons.68 Clearly, the Court’s concern here is not with the adequacy of the process, but rather with the efficiency of results. Only the imposition of an adequate (ie, non-lenient) punishment on those deemed responsible is considered as an effective remedy.69 On the other hand, absence of criminal accountability is regarded as a human rights violation per se.70 The assumption that a general atmosphere of impunity contributes to the commission of further crimes has led Judge Ksenija Turković to assert that passivity in prosecuting ‘may be equated with complicity’.71 For this reason, the Court appears to be willing to accommodate various limitations to the principle of fair trial,72 statutes of limitations and acts of criminal clemency when they are perceived to foster impunity.73 The insistence on coercive human rights also has the effect of pulling nonpunitive remedies into the criminal justice orbit. While the Court at times demands states to combine criminal law with a wide range of other social and legislative measures, the latter nonetheless tends to be conceived within a criminal law paradigm.74 By the same token, victims’ protection is mostly to be ensured by means of criminal proceedings and even restorative measures are incorporated into the existing punitive framework. As stated in Rantsev v Cyprus and Russia, states have positive obligations to put in place an appropriate legislative and administrative framework, to ensure the enjoyment of human rights, to take protective measures on behalf of victims, to investigate human rights breaches and to cooperate with each other in cases of cross-border violations.75 Although the preventive and protective measures that, according to the ECtHR, should accompany criminal law mechanisms are not strictly ‘penal’ in nature, it is primarily for criminal justice institutions (police, prosecutors and criminal courts) to implement and enforce them. As a result, the focus shifts from protection and compensation to prosecution and punishment. In August v UK, for example, the applicant, a minor, complained that his rights under Article 8 ECHR had been violated because he had not been awarded compensation as a victim of prostitution.76 The Court rejected his claim, arguing that the conviction of his abuser was a sufficient response to the violation.77

68 ibid. Italy never sought the extradition of the CIA agents who had been convicted and the Italian President of the Republic pardoned three of them. 69 Atalay v Turkey App No 1249/04 (ECtHR, 18 September 2008), para 46. 70 El-Masri (n 58) para 192. 71 K Turković, ‘International and National Courts Confronting Large-Scale Violations of Human Rights’ (2016), www.echr.coe.int/Documents/Speech_20160129_Turkovic_JY_ENG.pdf (speech at a seminar for the opening of the ECHR’s judicial year in 2016). 72 See Kelly Pitcher, ch 13 in this volume. 73 Opuz v Turkey (n 56); GIEM Srl and Others v Italy App Nos 1828/06, 34163/07, 19029/11 (ECtHR [GC], 28 June 2018), para 260; Marguš v Croatia (2016) 62 EHRR 17, para 139. 74 See, eg, Rantsev (n 46) para 284. See also Vladislava Stoyanova, ch 10 in this volume. 75 Rantsev (n 46) paras 286–89. 76 August v UK (2003) 36 EHRR CD115. 77 ibid.

Sowing a ‘Culture of Conviction’  171 Commenting on the assertion that the Court has become a ‘machine that encourages criminal inflation’,78 former ECtHR judge Françoise Tulkens contends that ‘while it is true … that the Court sometimes gives in to the questionable charm of more severe criminalization, the reality is more complex’.79 In fact, the Court has not ceased to set restrictions on the use of penal powers by European states. However, the latter function co-exists with the other presented here, aimed at holding perpetrators of human rights violations criminally accountable.80 Coercive human rights are the result of a certain doublethink in the Court’s jurisprudence. While the ECtHR still advances its moderating influence in some areas of criminal law, it has started bolstering and requiring penality in others. In so doing, the ECtHR sends an ambiguous message to domestic criminal justice systems, which are required to enforce coercive human rights obligations while guaranteeing due process rights or limiting their penal tendency in some other contexts.81 It is likely that this mixed message leads to undesirable outcomes. As presented above, some European states are already expanding their criminal justice systems with troublesome results. Coercive human rights obligations may, in turn, foster a further ‘culture of conviction’ whereby punishment is considered as the end to pursue, whatever the cost.82 As Liora Lazarus has noted, ‘case law is one thing, it is another matter how states and governments translate the messages they are hearing from the ECtHR … and how they are incorporating these messages into the rhetoric of public protection’.83 The danger, far from being hypothetical, is that domestic authorities interpret the decisions of the Court as enabling them to increase the level of punitiveness. In other words, domestic authorities may take advantage of the jurisprudence on coercive human rights to enhance penal measures to the detriment of individuals’ freedoms.84 This expansion, at the same time, is easily welcomed into the system because it is justified by the language and rhetoric of rights.85 Growing policing powers, the introduction of new criminal offences and wide-ranging incarceration become acceptable and even desirable if presented as necessary for the protection of rights. As such, the Court’s human rights case law is used as a form of leverage for justifying the expansion of the penal state.

78 Y Cartuyvels, ‘Droit Pénal et Droits de l’homme, un Retournement?’ in Michel van de Kerchove et al (eds), Les droits de l’Homme : épée ou bouclier du droit pénal? (Brussels, Presses de l’Université Saint-Louis Series, 2007) 43. 79 Tulkens (n 6) 157. 80 Dumortier et al (n 22) 128; Tulkens (n 5) 578; Lazarus (n 6) 137. 81 Sorochinsky (n 6) 163–64. 82 Similarly, see K Engle, ‘A Genealogy of the Criminal Turn in Human Rights’ in K Engle, Z Miller and DM Davis (eds), Anti-impunity and the Human Rights Agenda (Cambridge, Cambridge University Press, 2016) 37; Mégret and Calderón (n 7) 438. 83 Lazarus (n 6) 141. 84 PH van Kempen, ‘Four Concepts of Security: A Human Rights Perspective’ (2013) 13 Human Rights Law Review 1, 19. 85 Sorochinsky (n 6) 208.

172  Mattia Pinto IV.  THE MODERN SLAVERY ACT 2015

In this section, the MSA is taken as a case study to illustrate how the jurisprudence of the ECtHR on coercive human rights is translated at the domestic level. The Act and its application are analysed to show that domestic authorities may appropriate the language of human rights to foster and legitimise penal expansion. Contrary to the opinions of certain authors,86 it is contended that a rights-based approach to complex social problems (eg, human trafficking and labour exploitation) – especially when influenced by the ECtHR case law – is not in opposition, but rather is conducive, to the deployment of more far-reaching criminal law. The MSA is an anti-trafficking legislation passed in England and Wales in 2015. It aims to address what is commonly called ‘modern slavery’, namely a wide range of exploitative practices, including slavery, forced labour and trafficking in human beings. It does so by strongly relying on criminal law measures, such as heavy penalties and provisions to facilitate investigation, prosecution and conviction. The underlying justification for the Act is that ‘modern slavery’ is a serious human rights violation that requires prompt criminal accountability. The increasingly intertwined relationship between ‘human rights problems’ and ‘criminal law solutions’ was underscored by Theresa May in an article she wrote in The Telegraph one year after the enactment of the MSA: These crimes must be stopped … This is the great human rights issue of our time, and as Prime Minister I am determined that we will make it a national and international mission to rid our world of this barbaric evil.87

The case law of the ECtHR had a prominent influence on the adoption of the MSA.88 In 2012, the Court rendered the judgment of CN v UK, involving a Ugandan national who had been forced into working as a live-in carer with no significant payment and confiscated identity and travel documents.89 Recalling a previous decision in Siliadin v France and Rantsev v Cyprus and Russia,90 the

86 C Leigh, ‘Anti-trafficking Campaigns, Sex Workers and the Roots of Damage’ in P Kotiswaran and S Okyere (eds), State and the Law: Beyond Trafficking and Slavery Short Course, vol 3 (London, Beyond Trafficking and Slavery, 2015); Danielle Worden, ‘Sex Trafficking: Towards a Human Rights Paradigm’ (2018) 22 International Journal of Human Rights 709. 87 T May, ‘My Government Will Lead the Way in Defeating Modern Slavery’, The Telegraph, 31 July 2016, www.telegraph.co.uk/news/2016/07/30/we-will-lead-the-way-in-defeating-modern-slavery. 88 Home Office, ‘Modern Slavery Bill: European Convention on Human Rights: United Nations Convention on the Rights of the Child: Memorandum by the Home Office’ (2014), paras 3, 8 and 40, https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/ file/318618/ECHRmemoModernSlavery.pdf; see also V Mantouvalou, ‘The UK Modern Slavery Act 2015 Three Years on’ (2018) 81 Modern Law Review 1017, 1020; I Turner, ‘Positive Obligations and Article 4 of the European Convention on Human Rights: A Defence of the UK’s Human Rights Act 1998’ (2014) 18 International Journal of Human Rights 94, 105; J Haynes, ‘The Modern Slavery Act (2015): A Legislative Commentary’ (2016) 37 Statute Law Review 33, 37. 89 CN (n 45). 90 Siliadin (n 43); Rantsev (n 46).

Sowing a ‘Culture of Conviction’  173 ECtHR held that the UK legislative framework in place at the relevant time had been inadequate to afford practical and effective protection against violations of Article 4 ECHR.91 The Court also censured as ineffective the investigation by the British authorities into the applicant’s complaint of domestic servitude.92 The UK had already introduced the various offences of ‘slavery, servitude, and forced or compulsory labour’ in section 71 of the UK Coroners and Justice Act 2009. However, the ECtHR case law on Article 4 ECHR, together with difficulties in enforcement and the lack of clarity of section 71, fuelled interest in enacting a more comprehensive act,93 which would ‘equip the United Kingdom to fight modern slavery’.94 The MSA consolidated and simplified existing offences and increased sentences. The maximum penalty for slavery, servitude, forced or compulsory labour and human trafficking is now life imprisonment.95 In addition, the Act provides for a confiscation regime for individuals convicted of ‘modern slavery’ offences,96 introduces Slavery and Trafficking Prevention and Risk Orders to restrict the activity of those who pose a risk of causing harm,97 and establishes an Independent Anti-Slavery Commissioner.98 Finally, the Act lays down various provisions for the protection of victims, including a new defence for victims who commit an offence,99 as well as a section on transparency in supply chains.100 The stated purpose of the MSA is to strengthen criminalisation initiatives, facilitate the work of law-enforcement officers and increase prosecution rates.101 The approach adopted builds upon the British government’s strategy to counter organised crime and terrorism.102 The regulation of trafficking and exploitation relies heavily on law enforcement and crime control, while the causes of trafficking and the structural factors that create vulnerability remain unaddressed.103 Arguably, these exploitative factors are even promoted by the government, through immigration policies of exclusion, such as preventing asylum-seekers from working and tying overseas domestic workers to

91 CN (n 45) para 76. 92 ibid paras 79–81. 93 Home Office (n 88) para 8; Haynes (n 88) 5; C Barclay and S Foster, ‘The Modern Slavery Act 2015: Good Intentions and Sending out the Right Message’ (2017) 22 Coventry Law Journal 1, 2. 94 See Centre for Social Justice, ‘It Happens Here: Equipping the United Kingdom to Fight Modern Slavery’ (2013), https://www.centreforsocialjustice.org.uk/library/happens-equippingunited-kingdom-fight-modern-slavery. 95 MSA, s 5(1)(a). 96 ibid s 7. 97 ibid pt 2. 98 ibid s 41. 99 ibid s 45. 100 ibid s 54. 101 Mantouvalou (n 88) 1021. 102 J Fudge, ‘The Dangerous Appeal of the Modern Slavery Paradigm’ in Kotiswaran and Okyere (n 86) 21. 103 J Collins, ‘Exploitation of Persons and the Limits of the Criminal Law’ [2017] Criminal Law Review 169.

174  Mattia Pinto their employers.104 The assumption is that ‘modern slavery’ can be ‘solved’ by targeting individual players rather than the social processes that normalise exploitation.105 This assumption, among other reasons,106 finds justification in the ECtHR case law, which has transformed criminal law into the primary instrument of human rights protection. Criminal law is also used as a communication tool to send a clear message to abusers that their conduct is not tolerated. In Theresa May’s words: This landmark legislation sends the strongest possible signal to criminals that if you are involved in this vile trade you will be arrested, you will be prosecuted and you will be locked up.107

The focus on crime reduction, deterrence and punishment is in line with, and implements, the UK’s international obligations under the UN Trafficking Protocol, the Council of Europe Convention on Action against Trafficking in Human Beings and the EU Anti-trafficking Directive.108 However, it is also compatible with, if not required by, the UK’s obligations under the ECHR.109 The jurisprudence of the ECtHR supports rather than challenges the use of a criminal justice framework to address ‘modern slavery’. Section 1(2) of the MSA is indicative in this respect, when it clarifies that the offence of slavery, servitude, forced and compulsory labour has to be construed in accordance with Article 4 ECHR.110 Three examples illustrate the limitations of an act that uses adherence to international obligations as ways to justify extensive penality. A. Sentencing First, in terms of penalties, the MSA fully endorses the ‘prison works’ rhetoric. Theresa May’s words are again illustrative: Offences committed in connection with modern slavery are some of the most serious that can be committed, so the Bill extends the maximum available sentence to life 104 K Roberts, ‘The Modern Slavery Bill: Migrant Domestic Workers Fall through the Gaps’ in Kotiswaran and Okyere (n 86) 14–19. 105 Fudge (n 102) 21–23. 106 A focus on individual criminal responsibility may also be explained by the increasing tendency of contemporary democracies to use criminal law as a relatively easy and symbolic tool to address social problems, thereby avoiding more reasoned and comprehensive legal and political actions. See section II above. 107 Home Office, ‘Historic Law to End Modern Slavery Passed’, www.gov.uk/government/news/ historic-law-to-end-modern-slavery-passed. 108 These instruments mainly promote a criminal law approach to human trafficking. See, eg, JC Hathaway, ‘The Human Rights Quagmire of “Human Trafficking”’ (2008) 49 Virginia Journal of International Law 1, 26; M Eriksson, ‘The Prevention of Human Trafficking: Regulating Domestic Criminal Legislation through the European Convention on Human Rights’ (2013) 82 Nordic Journal of International Law 339. 109 A Weatherburn, ‘Using an Integrated Human Rights-Based Approach to Address Modern Slavery: The UK Experience’ (2016) 2 European Human Rights Law Review 184, 188. 110 MSA, s 1(2).

Sowing a ‘Culture of Conviction’  175 imprisonment. That will ensure that the worst perpetrators can receive the lengthy custodial sentences that they deserve. Tough sentences will also act as a powerful deterrent to others.111

Although life in prison is yet to be used, the existing case law has shown a tendency to impose higher custodial sentences.112 In R v Zielienski (David), the offender had conspired with his family to deceive a group of Polish nationals to travel to the UK, on the promise of a well-paid job.113 They were instead housed in appalling conditions and their wages were confiscated. The Court of Appeal increased a sentence of four years for trafficking offences to seven years, confirming that ‘modern slavery’ offences are to be treated with severity.114 First, Lord Justice Davis explained that sentences in cases of ‘modern slavery’ must make clear that ‘every vulnerable victim of exploitation will be protected by the criminal law’.115 Endorsing the judge of first instance, Lord Justice Davis accepted that there was a ‘clear, obvious and continuing need to punish’ traffickers and ‘deter others who would seek to profit from the oppression of others’.116 Then, on this basis, he had no doubt that Zielienski’s sentence was ‘unduly l­ enient’ and had to be ‘significantly increased’.117 The trend towards higher penalties is even more visible in R v Iyamu (Josephine).118 Iyamu had been convicted for having arranged the travel of five individuals to Europe with a view to exploitation. With one exception, all victims had been made aware that they were to work as prostitutes once in Germany and Italy. They had been subject to juju rituals, trafficked to Europe and forced to sell their bodies to repay their ‘debts’. Iyamu was initially sentenced to 14 years’ imprisonment. However, the Court of Appeal found the latter penalty ‘unduly lenient’ and increased it to 18 years’ imprisonment.119 Without wishing to assess whether the latter penalty is appropriate, it is important to focus here on the language used by the Court and its emphasis on the need for a ‘severe sentence’.120 The Court of Appeal described Iyamu’s lack of remorse and continued denial of responsibility for what occurred121 and concluded: Overall, therefore, the facts and circumstances of this case compel the conclusion that this sentence, in our judgment, ought to be increased. Offending of this gravity 111 Hansard HC col 172 (8 July 2014). 112 Mantouvalou (n 88) 1026. 113 R v Zielinski (David) [2017] EWCA Crim 758. 114 ibid para 28. 115 ibid, quoting Attorney General’s Reference Nos 2, 3, 4 and 5 of 2013 (R v William Connors and Others) [2013] EWCA Crim 324 para 10. 116 Zielinski (n 113) para 21, quoting the decision of the Crown Court at Bedford, which sentenced Zielinski in the first instance. 117 ibid para 27. 118 R v Iyamu (Josephine) [2018] EWCA Crim 2166. 119 ibid paras 50–52. 120 ibid para 51. 121 ibid para 46.

176  Mattia Pinto requires condign sentencing. It also requires that persons be deterred from engaging in offending of this kind … Those who are within the reach of the jurisdiction of the courts of England and Wales should expect severe sentencing in this context. Those who engage in vicious, ruthless and heartless human trafficking of this particular kind are playing for high stakes. They need to know that the courts will show no mercy when their criminality is exposed. (Emphasis added)122

B.  Prosecution of Victims Second, a framework heavily focused on crime encourages a rigid dichotomy between victims and perpetrators. However, it is not uncommon for people who have been exploited to commit offences or become involved in trafficking themselves.123 A criminal record for victims of ‘modern slavery’ is highly detrimental because it would limit their access to lawful employment, have an impact on their immigration status and render contact with state authorities less likely in the future.124 In other words, it makes victims more vulnerable to further exploitation. A series of international and European instruments have acknowledged this problem and have encouraged the adoption of non-punishment clauses for victims of trafficking and labour exploitation.125 In place of a non-punishment clause, section 45 of the MSA provides for a statutory defence for individuals compelled to commit certain criminal offences as a result of their exploitation.126 The defence does not apply to various crimes, such as sexual offences, offences involving serious violence, robbery or burglary.127 Despite good intentions, section 45 is problematic.128 First, it is designed not primarily to protect victims of ‘modern slavery’, but to transform them into reliable witnesses who can facilitate the work of law-enforcement officers.129 Its stated purpose is to encourage victims to come forward and give evidence

122 ibid para 51. 123 S Rodríguez-López, ‘(De)Constructing Stereotypes: Media Representations, Social Perceptions, and Legal Responses to Human Trafficking’ (2018) 4 Journal of Human Trafficking 61, 68. 124 R v EK [2018] EWCA Crim 2961, para 2. 125 See, eg, Council of Europe Convention on Action against Trafficking in Human Being, adopted on 1 February 2008, CETS No 197, art 26; Directive 2011/36/EU, 5 April 2011, art 8. According to Piotrowicz and Sorrentino, the obligation to protect victims of trafficking in Rantsev ((n 46) para 284) requires the enactment of non-punishment clauses (RW Piotrowicz and L Sorrentino, ‘Human Trafficking and the Emergence of the Non-punishment Principle’ (2016) 16 Human Rights Law Review 669, 681). 126 MSA, s 45. 127 ibid s 45(7). 128 For a critique on how the defence is formulated, see SSM Edwards, ‘Coercion and Compulsion: Re-imagining Crimes and Defences’ [2016] 12 Criminal Law Review 876; K Laird, ‘Evaluating the Relationship between Section 45 of the Modern Slavery Act 2015 and the Defence of Duress: An Opportunity Missed?’ [2016] Criminal Law Review 395. 129 On victims of human trafficking as witnesses, see T Ward and S Fouladvand, ‘Human Trafficking, Victims’ Rights and Fair Trials’ (2018) 82 Journal of Criminal Law 138.

Sowing a ‘Culture of Conviction’  177 without fear of being convicted.130 Second, a criminal law approach to trafficking criminalises victims who are compelled to commit offences more often than not, even when a defence is provided. Trafficked victims often do not selfidentify as having been trafficked and may distrust the authorities and even their lawyer.131 As a result, in the UK victims continue to be prosecuted for offences they were forced to commit.132 The following examples may help illustrate the problem here. The first case is R v Joseph (Verna Sermanfure).133 The human rights non-governmental organisation (NGO) Anti-Slavery International, acting as intervener, invited the Court of Appeal to reconsider the approach the UK courts had taken prior to the enactment of the MSA in respect of prosecution of trafficked victims who offended. In five out of the six conjoined appeals, the interested parties had been granted the status of victims of trafficking after their conviction. Only in two cases the Court of Appeal allowed the appeals and quashed the convictions.134 The judges’ observations demonstrate the difficulty of reconciling the need to protect victims of trafficking and the duty to prosecute ‘modern slavery’ offences in line with the UK’s international obligations.135 According to the Court of Appeal, ‘it is in the public interest to prosecute’ when the crime committed is ‘grave’, regardless of whether the defendant is a victim of human trafficking.136 The Court, by refusing or dismissing four out of six appeals, arguably favoured the criminalisation of victims rather than protecting them.137 Another case is R v GS.138 GS was a Jamaican national who was sentenced in 2007 to seven years’ imprisonment for a drug-related offence. She was also a victim of human trafficking and her status was confirmed by the Home Office. In 2017, she applied for an extension of time for leave to appeal against conviction, claiming it had been an abuse of process for her trial to proceed. The Court of Appeal acknowledged the material change concerning the treatment of victims between 2007 and after the MSA.139 Nonetheless, Lord Justice Gross refused the application, arguing that victims of trafficking are not exempted

130 Home Office, ‘Modern Slavery Act 2015 Review: Fourth Interim Report’ (2019), s 4.1.2, https:// www.gov.uk/government/publications/modern-slavery-act-review-2015-fourth-interim-report. 131 Laird (n 128); Edwards (n 128) 895. 132 Prison Reform Trust, ‘Still No Way out: Foreign National Women and Trafficked Women in the Criminal Justice System’ (Prison Reform Trust and Hibiscus Initiatives, 2018), para 2.2, www.prisonreformtrust.org.uk/Portals/0/Documents/Still%20No%20Way%20Out%20full%20report.pdf; see also S Bird and P Southwell, ‘Does the New “Slavery” Defence Offer Victims of Trafficking Any Greater Protection?’ (2015) 9 Archbold Review 7; S Mennim, ‘Case Comment: The Wrong Side of the Line? Trafficking Victims Compelled to Commit Offences and Prosecutorial Discretion’ (2019) 83 Journal of Criminal Law 111. 133 R v Joseph (Verna Sermanfure) [2017] EWCA Crim 36. 134 ibid paras 136 and 160. 135 ibid para 28. 136 ibid para 31. 137 See ibid paras 68, 100, 125 and 145. 138 R v GS [2018] EWCA Crim 1824. 139 ibid para 64.

178  Mattia Pinto from prosecution should the latter be in the public interest and that this was fully in line with the UK’s international obligations.140 He said: ‘[A]s repeatedly made clear, where crimes have been committed by [victims of human trafficking], even arising from their own trafficking, there is no blanket immunity.’141 The prosecutor, the judge continued, had a wide discretion to decide whether to proceed against a victim of trafficking.142 According to Lord Justice Gross, GS’s conviction was not unsafe, given that she had committed a ‘serious offence’ whose ‘gravity should not be minimised’.143 From the cases above, it is clear that the statutory defence in section 45 of the MSA does not afford victims any significant protection;144 in fact, it prioritises prosecution over protection.145 Nonetheless, the British government can take advantage of the rhetoric surrounding the new defence to assert its position as a human rights champion.146 C.  Risks and Preventive Orders Third, the MSA introduces civil preventive orders to restrict the activity of individuals who pose a risk of committing ‘modern slavery’ offences. First, courts can issue Prevention Orders against individuals convicted under the MSA or not guilty by reason of insanity who are still at risk of offending.147 Second, they can issue Risk Orders on the basis of a generic risk that a person will exploit another one.148 The issuance of these orders is highly discretionary even when the required criteria are met. Should an order be granted, the person concerned will be prohibited from doing anything described in it.149 Apart from time limits, there are no specific constraints on the severity of orders, which can have a substantial impact on the recipient. Failure to comply, without reasonable excuse, with any of the prohibitions or requirements imposed by such an order

140 ibid para 76. 141 ibid para 1. 142 ibid para 76. 143 ibid para 78–80. 144 For another case along the lines of Verna Sermanfure and GS, see R v EK [2018] EWCA Crim 2961. 145 The major interest in conviction rather than the welfare of victims also emerges from the fact that the government rejected proposals to introduce an adequate civil remedy for those affected by ‘modern slavery’. In A and B v CICA and Secretary State for Justice [2018] EWCA Civ 1534, Lord Justice Gross observed in passing that: ‘It is by no means inevitable’ that Article 4 ECHR ‘leaps to the provision of state compensation’ in addition to the requirement to prevent trafficking, prosecute offenders and protect victims (at para 66). 146 Home Office (n 107). 147 MSA, s 14. 148 ibid s 21. 149 ibid ss 17(1) and 24(1).

Sowing a ‘Culture of Conviction’  179 is considered a criminal offence.150 Until March 2019, courts in England and Wales adopted 119 Prevention Orders and 46 Risk Orders.151 In two cases, the recipients breached one of the conditions in their orders and were sentenced to three and a half years and two years and four months in prison, respectively.152 While Prevention Orders are contingent upon individuals having already committed a ‘modern slavery’ offence, Risk Orders are to be seen as pre-emptive steps based on suspicious events.153 Their introduction finds validation in the jurisprudence of the ECtHR, which requires states to take preventive measures to protect individuals from harm by others.154 The British government explicitly invoked these positive duties in a memorandum prepared by the Home Office in June 2014 to address the compatibility of the then Modern Slavery Bill with the ECHR.155 However, the asserted necessity of intervening preventively is also used as leverage to expand the state’s power to coerce and control.156 Prevention and Risk Orders may impose harsh coercive measures that are not dissimilar to actual punishment before a harm occurs. The content of the prohibitions is also decided following civil rules of evidence and not the more rigorous criminal ones.157 A custodial sentence may also be imposed if the conditions of the orders are breached. With the excuse of providing more security, these orders ultimately afford law enforcement a high degree of flexibility in controlling and limiting individuals’ liberties. In so doing, they reinforce the punitive stance of the MSA. D.  Supporting a Culture of Conviction The MSA plays a crucial role in the UK’s strategy to combat crime. The National Audit Office’s report Reducing Modern Slavery, published in December 2017, devotes a great deal of space to the increase in recorded crimes, the number of 150 ibid s 30. 151 HM Government, ‘2019 UK Annual Report on Modern Slavery’ (2019), para 2.63, https:// assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/840059/ Modern_Slavery_Report_2019.pdf. 152 HM Government, ‘2018 UK Annual Report on Modern Slavery’ (2018), para 2.76, https://assets. publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/749346/ 2018_UK_Annual_Report_on_Modern_Slavery.pdf. 153 Worden (n 86) 716. 154 Osman v UK (2000) 29 EHRR 245, para 115; Branko Tomašić and Others v Croatia App No 46598/06 (ECtHR, 19 January 2009), para 50. On the relationship between preventive orders and human rights, see JW Nickel, ‘Restraining Orders, Liberty, and Due Process’ in A Ashworth, L Zedner and P Tomlin (eds), Prevention and the Limits of the Criminal Law (Oxford, Oxford University Press, 2013); Ramsay (n 28) ch 6. 155 Home Office (n 88) para 28, citing Stubbings and Others v UK (1997) 23 EHRR 213. 156 For a critical account of the ‘turn to preventive measures’ in the name of public security in some Western states, see A Ashworth and L Zedner, Preventive Justice (Oxford, Oxford University Press, 2014). See also Liora Lazarus, ch 12 in this volume. 157 Ashworth and Zedner (n 156) 88.

180  Mattia Pinto investigations, and the rates of prosecutions and convictions.158 The focus is on efficiency and crime reduction. This approach has been endorsed by commentators, who now assess the success (or lack thereof) of the British policy on human trafficking by relying on the number of criminal trials that have been held.159 They decry the (relatively) small number of prosecutions and convictions as they were the main way to assess the UK’s effort to fight ‘modern slavery’.160 However, through such an insistence on criminal accountability, the MSA ultimately reveals a tendency towards those punitive trends that characterise contemporary democracies and have attracted criticism among activists and scholars. The MSA shows uncritical reliance on criminal sanctions, a leaning towards long custodial sentences and an extension of preventive police powers. It also combines welfare services and punitive strategies to regulate and exclude those on the margins, such as migrants, sex workers and people from the poorer classes. A great emphasis on prevention, harm reduction and victims’ protection is used as a justification for larger criminal justice institutions and increased power to state officials. Finally, the MSA promotes partnerships between corporations, NGOs and individual citizens on the one hand, and the state on the other hand, ‘in order to help reduce criminal opportunities and enhance crime control’.161 This punitive tendency is worrying because the UK is already heading towards a more repressive penality, with large-scale incarceration, severe punishment and proactive policing strategies.162 Some elements of this penal expansion are to be linked to the jurisprudence on coercive human rights. Of course, beyond the requirements of human rights law, there are various interests and rationales that play a role in the criminalisation of human trafficking. The ECtHR can hardly be considered the main reason behind the UK’s approach to ‘modern slavery’ and its narrow focus on criminal justice responses. However, the contribution of human rights in facilitating criminal justice responses cannot be overlooked. The ECtHR’s insistence on effective criminal enforcement motivates policing and prosecutorial responses rather than efforts to understand and deal with situations of systemic injustice, in which global inequality, immigration controls and exploitative labour conditions all play a part.163 Coercive human rights obligations support and

158 National Audit Office, ‘Reducing Modern Slavery’ (Home Office 2017) HC 630 Session 2017–2019. 159 Mantouvalou (n 88) 1026. In her article, Virginia Montouvalou argues that, given the low level of prosecutions, the MSA has failed to meet its purpose. Although she welcomes the criminalisation of human trafficking, she acknowledges that the reason why the British policy on ‘modern slavery’ is not a success is also because the government still creates structures of exploitation, while pretending to be very serious about tackling the particular wrong. 160 See also National Audit Office (n 158) para 4 (observing that prosecution and conviction rates of ‘modern slavery’ crimes are still very low). 161 Ward and Fouladvand (n 129), mentioning Garland (n 9) 126. 162 See, eg, Garland (n 9). 163 J Christman, ‘Human Rights and Global Wrongs’ in Diana Tietjens Meyers (ed), Poverty, Agency, and Human Rights (Oxford, Oxford University Press, 2014) 323.

Sowing a ‘Culture of Conviction’  181 strengthen the MSA’s crime control machinery, whilst deflecting attention, energy and resources from initiatives in the opposite direction. The ECtHR promotes an individualistic approach whereby social issues can be solved by tackling single acts of abuse and violence. As a matter of fact, this attitude tends to disregard alternative, non-criminal responses to ‘modern slavery’, such as inclusive immigration policies or labour regulation.164 By the same token, the Act marginalises calls for criminal justice reforms aimed at reducing custodial sentences or decriminalising less serious infractions. Even accepting that human rightsdriven criminalisation has not directly increased the net of social control, it has nonetheless normalised criminal law as the default answer to human trafficking, whilst limiting any room for change. While the ECtHR might not be the direct cause of a statute centred on ‘crime’ rather than on victim protection and structural causes of exploitation, its jurisprudence has sustained this approach by providing a human rights rationale for crime-control measures. The confidence shown by the ECtHR in calling for criminalisation, prosecution and punishment has given criminal law a new legitimacy as the ‘sharp edge of the sword’ of human rights.165 It has also favoured the illusion that criminal management is the answer to problems that are commonly acknowledged as complex and multifaceted. Coercive human rights, when transposed to the domestic level, confer power to the state in pursuing punitive agendas by invoking better protection for victims and compliance with international obligations. At the same time, they neutralise critique about the current conditions of imprisonment or the current policing of borders.166 Harsh prison sentences, greater power to prosecute and convict, and extensive pre-emptive measures become acceptable and even desirable when advanced in the name of human rights.167 The ‘culture of conviction’ spreads out and gathers strength. V. CONCLUSION

It is often argued that the ECtHR is partly responsible for constraining European states’ penal policies and supporting a more ‘humane’ criminal law. While this may in fact be true, the Court has also contributed to sustaining certain punitive 164 Similarly, see A Gruber, AJ Cohen and K Mogulescu, ‘Penal Welfare and the New Human Trafficking Intervention Courts’ (2016) 68 Florida Law Review 1333, 1379 and 1380; C Costello, ‘Migrants and Forced Labour: A Labour Law Response’ in Alan Bogg et al (eds), Autonomy of Labour Law (Oxford, Hart Publishing, 2015) 191. 165 RK Woetzel, ‘International Criminal Law and Human Rights: The Sharp Edge of the Sword’ (1968) 62 Proceedings of the American Society of International Law 117. 166 E Stanley, ‘Human Rights as a Protective Force’ in L Weber, E Fishwick and M Marmo (eds), The Routledge International Handbook of Criminology and Human Rights (Abingdon, Routledge, 2017) 508. 167 D Otto, ‘Decoding Crisis in International Law: A Queer Feminist Perspective’ in B Stark (ed), International Law and its Discontents: Confronting Crises (Cambridge, Cambridge University Press) 121.

182  Mattia Pinto impulses through its coercive human rights obligations. When the ECtHR orders states to resort to criminal law, human rights protection is advanced to bolster penality. However, the European judges have generally failed to ponder the potential pitfalls of their jurisprudence. They have shown a sort of ‘criminal fetishism’, which overlooks the risks associated with penal expansion and underestimates the fact that criminal law always poses a threat to the individual’s freedoms.168 Even in cases of abuses of a certain gravity, uncritical reliance on criminal law instruments is problematic if it disregards crucial considerations on the opportunity and implications of deploying penal measures. As the case of the MSA has illustrated, national reforms led by coercive human rights turn out to extend and give legitimacy to forms of penal control, while failing to tackle the root causes of abuses and protect victims. In conclusion, coercive human rights obligations appear to have a dual function: they require new penal intervention, but they also legitimise punitiveness by covering it up with a humane veil. While activists and scholars have generally condemned the expansion of penality led by ‘law and order’ rhetoric, such expansion is instead demanded when criminal law is used in the name of human rights. The same individuals who criticise harsh prison conditions and overcriminalisation in the context of counter-terrorism or ‘tough on crime’ policies are often advocates of extensive penal control to stop human trafficking, protect the right to life and prohibit torture.169 By endorsing criminal accountability as the best instrument to protect human rights, the ECtHR has given strength to this form of cognitive dissonance, which, depending on the case, sees penality as problematic or benevolent. But it has also offered European states a new excuse to have recourse to ever-more expansive criminal law with little criticism and in a way that seems consistent with progressive thinking.

168 Similarly, in the context of international criminal law, see D Zolo, ‘The Lords of Peace: From the Holy Alliance to the New International Criminal Tribunals’ in H Barry (ed), Global Democracy: Key Debates (London, Routledge, 1999) 81–82. 169 For some examples of how various international institutions and NGOs support the expansions of penality in some areas, whilst criticising criminalisation and punishment in others, see A Sattar, Criminal Punishment and Human Rights: Convenient Morality (Abingdon, Routledge, 2019) 194–200.

9 Coercive Overreach, Dilution and Diversion Potential Dangers of Aligning Human Rights Protection with Criminal Law (Enforcement) NATASA MAVRONICOLA

I. INTRODUCTION The object and purpose of the Convention as an instrument for the protection of individual human beings require that its provisions be interpreted and applied so as to make its safeguards practical and effective. Any interpretation of the rights and freedoms guaranteed has to be consistent with the general spirit of the ­Convention.1

P

rotecting people from the grave wrongs proscribed by human rights law is central to the human rights project and to the European Convention on Human Rights (ECHR) in particular.2 This has been reflected in the extensive elaboration of positive obligations demanding State action by the European Court of Human Rights (ECtHR).3 At the same time, the ECtHR has reiterated the centrality of the value of human freedom throughout its rich

1 Svinarenko and Slyadnev v Russia App Nos 32541/08 and 43441/08 (ECtHR, 17 July 2014), para 118. 2 See D Harris, M O’Boyle, E Bates and C Buckley, Law of the European Convention on Human Rights, 3rd edn (Oxford, Oxford University Press, 2014) 203. 3 The European Court of Human Rights is hereinafter referred to as ECtHR, ‘the Strasbourg Court’ or ‘the Court’.

184  Natasa Mavronicola jurisprudence on the Convention,4 and this has underpinned the far-reaching safeguards that the Court has cemented in the context of criminal justice and beyond, constraining the coercive force of the State. The premise of this chapter is that the ECHR should be interpreted coherently, in line with its spirit and purpose, and to make its safeguards practical and effective. In light of this starting point, it examines three key dangers arising from the coercive ‘sting’5 of positive duties imposed on states through ECtHR doctrine, with particular focus on Articles 2 and 3 ECHR. The dangers highlighted are those of coercive overreach, dilution and diversion in respect of the human rights standards at play. The chapter proceeds as follows. Section II provides an illustrative account of duties of criminalisation and criminal redress, focusing particularly on Articles 2 and 3 ECHR. Section III highlights how the coercive duties emerging from these rights tend towards coercive overreach – that is, the excessive, or otherwise inapposite, demand for criminalisation and punishment. This is followed in section IV by the argument that the coercive slant of positive obligations entails a danger of weakening the (rightly) expansive understanding of the wrongs at issue and the stringent (negative) duties imposed on the state under these rights. Section V highlights the ways in which the coercive orientation of positive obligations under core rights can divert the Court from other practical and effective tools of protection, and concludes in favour of a protective and preventive re-orientation of the doctrine. II.  DUTIES TO MOBILISE THE CRIMINAL LAW IN THE PROTECTION OF LIFE AND PERSONAL INTEGRITY UNDER ARTICLES 2 AND 3 ECHR

Convention rights give rise to a range of positive duties applicable in different contexts,6 which under Articles 2, 3 and 4 tend to encompass: general, or framework, duties; operational duties; and investigative duties. Duties of criminalisation and criminal redress chiefly emanate from and are shaped through the general duties to set up and adequately implement a legal framework that protects persons from unlawful takings of life or proscribed

4 See, for example, the frequently reiterated idea that ‘the very essence of the Convention is respect for human dignity and human freedom’: Goodwin v UK (2002) 35 EHRR 18, para 90; Pretty v UK (2002) 35 EHRR 1, para 65. 5 L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce?’ in L Zedner and J Roberts (eds), Principles and Values in Criminal Law and Criminal Justice (Oxford, Oxford University Press, 2012). 6 See the analysis in A Mowbray, The Development of Positive Obligations under the European Convention on Human Rights by the European Court of Human Rights (Oxford, Hart Publishing, 2004). See also the interesting normative treatment in L Lavrysen, Human Rights in a Positive State: Rethinking the Relationship between Positive and Negative Obligations under the European Convention on Human Rights (Cambridge, Intersentia, 2016).

Coercive Overreach, Dilution and Diversion  185 ill-treatment,7 and the duties to investigate allegations of unlawful takings of life or proscribed ill-treatment and to provide the requisite procedures for accountability and redress (often referred to as procedural duties).8 The notion of redress is broader than that of remedy; the former more clearly accommodates recourse to both civil and criminal means of redress. Indeed, the state’s duties of redress are often decoupled from the right to a remedy under Article 13 ECHR and seen to stem from the rights themselves, particularly insofar as they are tied to recourse to the criminal law rather than compensation.9 The ECtHR has elaborated on such duties under Articles 2 and 3 ECHR on various occasions. In relation to Article 2, it has stated that: [The framework] obligation requires the State to secure the right to life by putting in place effective criminal law provisions to deter the commission of offences against the person backed up by law enforcement machinery for the prevention, suppression and punishment of breaches of such provisions.10

In relation to Article 3, it has also stipulated that: Article 3 requires States to put in place effective criminal-law provisions to deter the commission of offences against personal integrity, backed up by law-enforcement machinery for the prevention, suppression and punishment of breaches of such provisions … and this requirement also extends to ill-treatment administered by private individuals …11

The significance and potency of duties of redress comes into particularly sharp relief in the Court’s elaboration of relevant investigative duties, often set out in broader terms as ‘procedural’ duties. According to the ECtHR, the investigation into a lethal use of force must be ‘effective in the sense that it is capable of leading to a determination of whether the force used in such cases was or was not justified in the circumstances … and to the identification and punishment of those responsible’.12 In the context of Article 3 too, the language is similar: Where an individual has an arguable claim that he has been ill-treated by agents of the State, the notion of an ‘effective remedy’ entails, in addition to the payment of compensation where appropriate, a thorough and effective investigation capable of leading to the identification and punishment of those responsible.13 7 On this, see L Lavrysen, ‘Protection by the Law: The Positive Obligation to Develop a Legal Framework to Adequately Protect ECHR Rights’ in Y Haeck and E Brems (eds), Human Rights and Civil Liberties in the 21st Century (Dordrecht, Springer, 2014). 8 See, further, K Kamber, Prosecuting Human Rights Offences: Rethinking the Sword Function of Human Rights Law (Leiden, Brill, 2017) chs 2 and 4. See also Laurens Lavrysen, ch 2 in this volume on duties to criminalise and duties to punish. 9 See, for instance, Perevedentsevy v Russia App No 39583/05 (ECtHR, 24 April 2014), paras 74–126. This is noted in A Seibert-Fohr, Prosecuting Serious Human Rights Violations (Oxford, Oxford University Press, 2009) 126–27. 10 Rantsev v Cyprus and Russia (2010) 51 EHRR 1, para 218. See also Osman v UK (2000) 29 EHRR 245, para 115. 11 Beganović v Croatia App No 46423/06 (ECtHR, 25 June 2009), para 71. 12 McShane v UK (2002) 35 EHRR 23, para 96. 13 El-Masri v FYROM (2013) 57 EHRR 25, para 255, with reference to Article 13 ECHR. See, too, O’Keeffe v Ireland (2014) 59 EHRR 15, para 172.

186  Natasa Mavronicola The ECtHR has indicated that the procedural duty reaches into prosecutorial decision-making,14 as well as judicial proceedings: While there is no absolute obligation for all prosecutions to result in conviction or in a particular sentence, the national courts should not under any circumstances be prepared to allow life-endangering offences and grave attacks on physical and moral integrity to go unpunished.15

Increasingly, this has brought the Court’s reasoning on investigative duties full circle back to the requirement of criminalisation – as the Court reasoned in Cestaro v Italy, for example: ‘For an investigation to be effective in practice it is a prerequisite that the State has enacted criminal-law provisions penalising practices that are contrary to Article 3.’16 The Court has taken it upon itself to ‘intervene in cases of manifest disproportion between the gravity of the act and the punishment imposed’.17 It has found, in several cases, that the penalty imposed on persons found to have committed torture or inhuman or degrading treatment or to have taken life unlawfully was inadequate.18 In some instances, the Court has made such finding in connection with determining whether the applicant(s) maintain victim status following domestic proceedings and redress.19 It has taken a particularly punitive stance in relation to the sanction meted out to persons who have been found to have engaged in what it calls ‘wilful’ ill-treatment. It has stipulated that ‘in cases of wilful ill-treatment a violation of Articles 2 or 3 cannot be remedied exclusively through an award of compensation to the victim’.20 The above account is only a cross-section of a vast domain of coercive human rights doctrine emerging from the ECtHR, much of which is covered in more detail in other chapters in this volume, notably by Laurens Lavrysen in Chapter 2. Following this illustrative account of key elements of the doctrine, we may proceed to consider the fundamental dangers emanating from it.

14 This is within limits. See, for example, the art 2 judgments of Da Silva v UK (2016) 63 EHRR 12, paras 259–82; Kolevi v Bulgaria (2014) 59 EHRR 23, paras 191–215. But see the art 3 (and 8) judgment in MC v Bulgaria (2005) 40 EHRR 20, paras 148–87. 15 Ali and Ayşe Duran v. Turkey App No 42942/02 (ECtHR, 8 April 2008), paras 61–62. See also Okkalı v Turkey App No 52067/99 (ECtHR, 17 October 2006), paras 65–66; MC v Bulgaria (2005) 40 EHRR 20, para 131. 16 Cestaro v Italy App No 6884/11 (ECtHR, 7 April 2015), para 209 (citations omitted). See also Gäfgen v Germany (2011) 52 EHRR 1, para 117. 17 Nikolova v Bulgaria (2009) 48 EHRR 40, para 62. See the analysis in Lavrysen, ch 2 in this volume. 18 See, for example, the art 2 and 3 case of Ali and Ayşe Duran v Turkey App No 42942/02 (ECtHR, 8 April 2008), paras 59–73; and the art 3 cases of Gäfgen (n 16), paras 119–30; Sidiropoulos and Papakostas v Greece App No 33349/10 (ECtHR, 25 January 2018), paras 83–100. 19 See, for example, Nikolova (n 17) paras 47–64; Gäfgen (n 16), paras 119–30. 20 Aleksakhin v Ukraine App No 31939/06 (ECtHR, 19 July 2012), para 60. In relation to art 2, see also the Court’s approach in Tarariyeva v Russia App No 4353/03 (ECtHR, 14 December 2006), para 75.

Coercive Overreach, Dilution and Diversion  187 III.  THE DANGER OF COERCIVE OVERREACH

Liora Lazarus, writing on the ‘coercive sting’ of obligations often cast as ‘protective’ of individual rights, warned of the danger of ‘coercive overreach’ through human rights.21 She suggested that such overreach can arise within judicial doctrine and its immediate implications, as well as in the ‘rhetorical assertion of coercive duties’22 within a ‘broader politics of security’.23 In this section, I focus on probing the potential for coercive overreach within the doctrine itself.24 This would involve demanding the penalisation of acts or omissions which might, as a matter of principle or policy, not necessarily warrant penal sanction. A danger of coercive overreach within the Court’s doctrine carries significance not only for the individuals who stand to face detriment from such overreach, but also for the integrity25 of the Convention itself. This is because deploying the criminal law may well – in relation to killings or ill-treatment – entail the deprivation of liberty, and respect for liberty or ‘human freedom’ is the ‘very essence’ of the Convention.26 How might such coercive overreach arise? Substantively, the criminal law is meant to capture particularly blameworthy wrongs.27 Defences – whether partial or total – to such prima facie wrongs reflect the circumstances in which it is considered that people should not be exposed to the punitive force of the criminal law because their (in)actions are justified or excused and their blameworthiness is accordingly extinguished or diminished.28 The criminal law’s punitive force may be withheld where someone lacked an intention to cause harm, acted on an honest error or misjudged the proportionality of their response to a threat. In such circumstances, their action may nonetheless amount to a civil wrong, such as a tort. While the sanctity of life, for example, is of foundational significance to the criminal law,29 such nuance in relation to culpability may well apply in the context of homicide offences or offences against personal integrity. Approaches to the delimitation of defences vary across jurisdictions,30 although they often include necessity and self-defence

21 Lazarus (n 5) 136, 147. 22 ibid 149. 23 ibid 141. 24 Mattia Pinto and Liora Lazarus, chs 8 and 12 in this volume respectively, explore the broader implications of the doctrine within particular cultures of penality and securitisation. 25 R Dworkin, Law’s Empire (Oxford, Hart Publishing, 1998 [1986]) chs 2 and 3. 26 See, for instance, Pretty (n 4), para 65. 27 A Ashworth, ‘Conceptions of Overcriminalization’ (2008) 5 Ohio State Journal of Criminal Law 407, 408–09. 28 See the nuanced account of defences in criminal law in J Gardner, Offences and Defences (Oxford, Oxford University Press, 2007) ch 4. 29 See the collection of essays on puzzles and controversies in this area in J Horder (ed), The Sanctity of Life and the Criminal Law: The Legacy of Glanville Williams (Cambridge, Cambridge University Press, 2013). 30 J Horder (ed), Homicide Law in Comparative Perspective (Oxford, Hart Publishing, 2007) ch 2. Regarding different approaches to the role of self-defence, see A Duff, Answering for Crime: Responsibility and Liability in the Criminal Law (Oxford, Hart Publishing, 2007) chs 9 and 11.

188  Natasa Mavronicola regarding action to avert equivalent harm befalling oneself or another. Defences can range from exculpation to excuse, and encompass different legal and labelling implications.31 For our purposes, what is important to underline is that individual criminal liability may appropriately be shaped through an assessment of subjective blameworthiness,32 and that the legal approach to defences may accord some leeway to the potentially flawed judgement of the individual asserting the defence. This is because it is largely not considered appropriate, for the purposes of establishing individual criminal liability, to assess the actions of a person acting under pressure or in a context of violence through the eyes of a notional ‘suitably informed impartial observer in a calm frame of mind’ who thereby applies an objective test.33 Looking to Article 2 ECHR, the stringent absolute necessity test elaborated by the Court in relation to the negative obligation not to employ (potentially) lethal force is evidently not modelled on a criminal law standard of liability for homicide.34 The absolute necessity test is, after all, ‘a stricter and more compelling test of necessity’ than the proportionality test ‘normally applicable when determining whether State action is “necessary in a democratic society”’.35 The latter is an objective test insofar as it involves checking the adequacy, and not just the good faith character, of the state’s justifications for interfering with the relevant qualified right, and so – on this basis – must be the former, even more stringent test. While the subjective, potentially flawed or mistaken, perception of the perpetrator may suitably play a role in determining individual criminal liability, this is not – or should not be – the case in determining state responsibility for violating the right to life. Indeed, the state may be found to have violated Article 2 in circumstances of diffuse or systemic failings – which may sometimes not be criminally culpable – in the discharge of a (high-risk) state operation involving the deployment of (potentially) lethal force.36 In respect of Article 3 ECHR, many instances of the violation of the right encompass grossly culpable conduct that would – and should – be considered a grave criminal offence in most jurisdictions. Nonetheless, Article 3 violations can also arise out of circumstances of systemic, structural or diffuse failings that may sometimes not involve particular persons engaging in criminally culpable conduct. Accordingly, the standard for state responsibility for human rights breaches is, and can appropriately be, stricter than standards shaping individual criminal 31 See the discussion in F Leverick, Killing in Self-Defence (Oxford, Oxford University Press, 2006) ch 1; and B Sangero, Self-Defence in Criminal Law (Oxford, Hart Publishing, 2006) ch 1. Note the review of the latter and nuanced points on ‘forfeiture’ in A du Bois-Pedain, ‘Publication Review: SelfDefence in Criminal Law’ (2009) 68 Cambridge Law Journal 227. See also J Horder, ‘Self-Defence, Necessity and Duress: Understanding the Relationship’ (1998) 11 Canadian Journal of Law and Jurisprudence 143. 32 See, for example, Leverick (n 31) 11. 33 S Uniacke, ‘Proportionality and Self-Defense’ (2011) 30 Law & Philosophy 253, 256. 34 McCann v UK (1996) 21 EHRR 97, para 149. 35 Pretty (n 4), para 65. 36 eg, McCann (n 34).

Coercive Overreach, Dilution and Diversion  189 liability, with human rights law demanding more from the state apparatus and holding it more stringently to account than the criminal law does vis-a-vis individual actors. Anja Seibert-Fohr aptly put the issue in the following terms in her monograph on prosecuting serious human rights violations: The argument has been made that domestic criminal law should mirror the defence standards developed by the European Court of Human Rights … But these standards were developed to determine State responsibility for the taking of life and do not mean that States parties must criminalize the acts accordingly. State responsibility should not be confused with individual criminal responsibility.37

Darryl Robinson, too, in his critique of international criminal law, warned against: [S]ubstantive and structural conflation – that is, the assumption that criminal norms must be coextensive with similar norms in human rights or humanitarian law, overlooking the different structure and consequences of these areas of law, and thus neglecting the special principles necessary for blame and punishment of individuals.38

Accordingly, the main danger of coercive overreach within ECtHR doctrine on Articles 2 and 3 ECHR lies in equating circumstances amounting to a breach of (the negative obligations under) Articles 2 and 3 ECHR with circumstances demanding criminalisation and the pursuit of criminal redress. The doctrine gives reason for concern in this regard. Some of the Court’s pronouncements tend to suggest not only that the ECtHR assumes that violations of the prohibitions enshrined in Articles 2 and 3 automatically entail individual criminal liability, but that the Court requires that they do so under domestic law. Consider, again, the principle that in order for an investigation to be effective, it is a ‘prerequisite’ for the state to have put in place criminal law provisions penalising ‘practices that are contrary to Article 3’ (emphasis added).39 We know that there are practices and structures that are contrary to Article 3 even in the absence of any intention to cause suffering, anguish or humiliation,40 or that are part of chronic and/or systemic problems such as 37 Seibert-Fohr (n 9) 147, fn 232, citing B Emmerson, A Ashworth and A MacDonald, Human Rights and Criminal Justice, 2nd edn (London, Sweet & Maxwell, 2007) 750, 756. 38 D Robinson, ‘The Identity Crisis of International Criminal Law’ (2008) 21 Leiden Journal of International Law 925, 929; see also at 946–49. See too V Stoyanova, ‘Article 4 of the ECHR and the Obligation of Criminalising Slavery, Servitude, Forced Labour and Human Trafficking’ (2014) 3 Cambridge Journal of International and Comparative Law 407, 441. But note the positive treatment of the intersection between international criminal law and duties of criminal redress in human rights law in JD Ohlin, ‘The Right to Punishment for International Crimes’ in F Jeßberger and J Geneuss (eds), Why Punish Perpetrators of Mass Atrocities? Purposes of Punishment in International Criminal Law (Cambridge, Cambridge University Press, 2020). 39 Cestaro (n 16), para 209. 40 For example, in assessing ‘degrading treatment’, the Grand Chamber has affirmed that ‘although the question whether the purpose of the treatment was to humiliate or debase the victim is a factor to be taken into account, the absence of any such purpose cannot conclusively rule out a finding of violation of art. 3’ – MSS v Belgium and Greece (2011) 53 EHRR 2, para 220. See also V v UK (2000) 30 EHRR 121, para 71.

190  Natasa Mavronicola those associated with certain places of detention, which may in turn be tied to (problematic) government policies. The diffuse nature of the failings at play does not (and ought not to) absolve the state of liability for the inhumanity and degradation inflicted on the victims in such circumstances. At the same time, even though there may well be several such cases where the culpability of an individual or individuals may reach a level that warrants civil and/or criminal liability, this is by no means guaranteed in all cases. Moreover, we know that excessive use of force by law-enforcement authorities – that is, force that was not indispensable and proportionate in the circumstances – may be found to be inhuman or degrading and therefore in violation of Article 3 ECHR.41 Although such excess may in many or even most instances tend to be characterised by a criminal level of culpability, this might not always be the case: the excessive force used may have stemmed from a genuine mistake, lack of training or equipment, or another factor that might be reasonably considered a basis for an exculpatory defence or mitigation of sentence in a state’s criminal law. In relation to Article 2 ECHR, the potential for coercive overreach can be illustrated by the case of Da Silva v UK, concerning the targeting, as part of a counter-terrorism operation, and subsequent shooting of Jean Charles de Menezes by armed police officers on the London Underground during a police operation in which he was mistaken for a suicide bomber.42 The police operation had involved a number of well-documented failings, which culminated in the ultimate decision taken to shoot to kill.43 The applicant, Jean Charles de Menezes’ cousin, did not complain that her cousin was killed by state agents in violation of the negative obligation under Article 2, but complained solely of the fact that no individual police officer had been prosecuted following the fatal shooting. A key argument of both the applicant and an intervener in alleging that the non-prosecution constituted an Article 2 breach was that the criminal law defence of self-defence in English law did not mirror the test of absolute necessity found in Article 2(2) ECHR44 and that this was a breach of the UK’s positive obligations under Article 2.45 Although the ECtHR in Da Silva ultimately found no violation of Article 2, it did not disavow this conflation of criminal liability and state responsibility for breaches of Article 2; rather, it went to great lengths to show that the test for breach of the negative obligation under Article 2 closely approximates the test for criminal liability for takings of life under English law.46

41 See, for example, Güler and Öngel v Turkey App Nos 29612/05 and 30668/05 (ECtHR, 4 October 2011), paras 25–31. 42 Da Silva (n 14) paras 12–38, 136. 43 ibid paras 52–58, 283–88. 44 ibid paras 152, 186 and 192 (applicant’s arguments) and 224 (the Equality and Human Rights Commission’s arguments). 45 See, for instance, ibid paras 191, 193, 205, 225 and 226. 46 See ibid paras 244–56. But cf Tekin and Arslan v Belgium App No 37795/13 (ECtHR, 5 September 2017).

Coercive Overreach, Dilution and Diversion  191 These are just some examples of how the ‘mirroring’, or ‘conflation’, warned against by Seibert-Fohr and Robinson has seeped into ECtHR doctrine and produced statements of principle which substantively overreach by demanding the criminalisation and punishment of conduct that need not, or should not, in all circumstances be deemed criminally wrongful. Negative obligations under the fundamental human rights enshrined in Articles 2 and 3 ECHR are central to delimiting legitimate state violence; they are shaped so as to stringently limit such violence. In their substantive scope and application to particular circumstances, they are not meant to replicate the less exacting standards appropriate to the determination of individual criminal liability. Viewing state liability for Article 2 or Article 3 breaches on the one hand and individual liability for criminal wrongs on the other as coterminous risks coercive overreach. The capacity of the ECtHR’s doctrine to lead to conflation of state responsibility for violations of Articles 2 and 3 ECHR and individual criminal liability is accordingly a cause for concern in itself, even aside from other significant concerns surrounding the coercive implications of the ‘antiimpunity’ turn in human rights.47 It also stands in tension with the commitment to human freedom that underpins the Convention and that has shaped a number of counter-carceral safeguards in the ECtHR’s case law.48 Yet besides substantive coercive overreach, a danger of institutional coercive overreach also arises in the ‘mirroring’ or ‘conflation’ tendencies within the ECtHR’s coercive duties doctrine. As the Court has put it: Responsibility under the Convention is based on its own provisions which are to be interpreted and applied on the basis of the objectives of the Convention and in light of the relevant principles of international law. The responsibility of a State under the Convention, arising for the acts of its organs, agents and servants, is not to be confused with the domestic legal issues of individual criminal responsibility under examination in the national criminal courts. The Court is not concerned with reaching any findings as to guilt or innocence in that sense.49

What the Court establishes in the excerpt cited is twofold. Firstly, it rightly highlights that a domestic finding of no criminal culpability of individual state agents in allegations relating to, say, inhuman treatment (an Article 3 issue) or the use of lethal force (an Article 2 issue) does not resolve the question of state responsibility for a human rights violation. This is because individual criminal liability in domestic law and state responsibility for a breach of the

47 See the discussion of ‘anti-impunity’ in Laurens Lavrysen and Natasa Mavronicola, ch 1 in this volume. See also K Engle, ‘Anti-impunity and the Turn to Criminal Law in Human Rights’ (2015) 100(5) Cornell Law Review 1069; K Engle, Z Miller and DM Davis (eds), Anti-impunity and the Human Rights Agenda (Cambridge, Cambridge University Press, 2016). 48 On the tensions between liberal commitments and the coercive duties doctrine, see Nina Peršak and Mattia Pinto, chs 7 and 8 in this volume respectively. 49 Avşar v Turkey (2003) 37 EHRR 53, para 284 (citations omitted). See also McCann (n 34) para 173.

192  Natasa Mavronicola ECHR are meant to be legally distinct matters. The ECtHR is tasked with determining the latter. It can accordingly go beyond domestic criminal court findings to establish whether, on the ECtHR’s own principles, the respondent state has committed a breach of the relevant human rights. Secondly, the ECtHR distinguishes the process and authority of establishing criminal liability from its own processes and scope of authority. The Court has averred that it is ‘not a criminal court’ and that ‘in determining whether there has been a breach of Article 2 [or Article 3] of the Convention it is not assessing the criminal responsibility of those directly or indirectly concerned … because that responsibility is distinct from international law responsibility under the Convention’.50 Both findings of individual criminal liability and decisions on punishment must take place through a criminal justice process involving several key safeguards protecting the defendant, many of which are encapsulated in Articles 5 and 6 ECHR and extensive ECtHR jurisprudence.51 Yet where the ECtHR demands that heavier penalties be imposed, as it has done in its ‘manifest disproportion’ case law, it veers dangerously close to assuming the role of a tribunal tasked with administering criminal liability and punishment,52 a role that it has neither the authority nor the institutional capacity to play.53 IV.  THE PROSPECT OF WEAKENING HUMAN RIGHTS STANDARDS

Another grave danger that inheres in the ‘mirroring’ or ‘conflation’ identified above is the prospect of dilution. The danger is that a tendency to view human rights violations through a criminal lens might bring about a narrowing or dilution of the stringency of the obligations that the Convention imposes on the state and might create a danger of undermining practical, effective and entirely appropriate presumptions that have been developed to hold states to account. 50 Dimitrov v Bulgaria App No 77938/11 (ECtHR, 1 July 2014), para 129 (‘the same applies in relation to Article 3’ is stated in para 130 of the judgment). See too Tekin and Arslan (n 46) paras 81 and 109. See also André Nollkaemper, ‘Concurrence between Individual Responsibility and State Responsibility in International Law’ (2003) 52 International & Comparative Law Quarterly 615, 638. 51 See, especially, R Goss, Criminal Fair Trial Rights (Oxford, Hart Publishing, 2014). 52 On the reality of international human rights courts’ ‘quasi-criminal jurisdiction’, see A Huneeus, ‘International Criminal Law by Other Means: The Quasi-criminal Jurisdiction of the Human Rights Courts’ (2013) 107 American Journal of International Law 1. See the specific concerns outlined in respect of the Inter-American Court of Human Rights’ duty to punish doctrine in FF Basch, ‘The Doctrine of the Inter-American Court of Human Rights Regarding States’ Duty to Punish Human Rights Violations and its Dangers’ (2007) 23 American University International Law Review 195, 208–21. 53 For a nuanced take on the ECtHR’s fact-finding potential, see L Helfer, ‘Redesigning the European Court of Human Rights: Embeddedness as a Deep Structural Principle of the European Human Rights Regime’ (2008) 19 European Journal of International Law 125, 142–44; P Leach, C Paraskeva and G Uzelac, International Human Rights and Fact-Finding: An Analysis of the FactFinding Missions Conducted by the European Commission and Court of Human Rights (London, Human Rights & Social Justice Research Institute, 2009) 57–63.

Coercive Overreach, Dilution and Diversion  193 The danger of dilution has to some extent materialised in Article 2 case law.54 Having made criminalisation central to the implications of (some) breaches of the right to life and having come close to equating state liability with individual criminal liability, the ECtHR has come to apply a criminal law-styled standard of culpability to determining the absolute necessity of lethal force. Article 2 involves a strict negative obligation in principle, delineated by an objective test of absolute necessity, and structurally extended so that the assessment of absolute necessity may incorporate considerations pertaining to failings in the planning and control of state operations in violent, volatile or otherwise life-endangering situations. According to the Court in McCann: In keeping with the importance of this provision (art. 2) in a democratic society, the Court must, in making its assessment, subject deprivations of life to the most careful scrutiny, particularly where deliberate lethal force is used, taking into consideration not only the actions of the agents of the State who actually administer the force but also all the surrounding circumstances including such matters as the planning and control of the actions under examination.55

In McCann, this translated into a finding that while the actions of the soldiers who had fired the lethal shots did not in themselves give rise to a violation of Article 2, a variety of diffuse organisational failings entailed that the killing of three individuals suspected of imminent ‘terrorist’ violence did not constitute absolutely necessary force within the meaning of Article 2(2) ECHR.56 As another example, in Güleç v Turkey, the Court highlighted the absence of less lethal tools to quell a riot in finding a breach of Article 2 in the shooting of an unarmed civilian in the context of unrest during a demonstration in Turkey.57 Yet in much of the Court’s case law, after pronouncing that only circumstances of absolute necessity can justify the use of lethal force and that ‘careful scrutiny’ is required on the stringent principles set out above, the ECtHR has proceeded to assess the use of such force through an ‘honest belief’ test: The use of force by agents of the state in pursuit of one of the aims delineated in para. 2 of art. 2 of the Convention may be justified under this provision where it is

54 See the further analysis in N Mavronicola, ‘Taking Life and Liberty Seriously: Reconsidering Criminal Liability under Article 2 of the ECHR’ (2017) 80 Modern Law Review 1026. 55 McCann (n 34) para 150; see also, for example, Boukrourou v France App No 30059/15 (ECtHR, 16 November 2017), para 55; and the related approach in Haász and Szabó v Hungary App Nos 11327/14 and 11613/14 (ECtHR, 13 October 2015), paras 56–57. See the analysis in B Dickson, ‘The Planning and Control of Operations Involving the Use of Lethal Force’ in L Early and A Austin (eds), The Right to Life under Article 2 of the European Convention on Human Rights (Eindhoven, Wolf Legal, 2016); N Melzer, Targeted Killing in International Law (Oxford, Oxford University Press, 2008) 102–17. 56 McCann (n 34) paras 213–14. 57 Güleç v Turkey (1999) 28 EHRR 121, para 71. See further H Russell, ‘Understanding “­Quelling a Riot or Insurrection” under Article 2 of the ECHR’ (2015) European Human Rights Law Review 495.

194  Natasa Mavronicola based on an honest belief which is perceived, for good reasons, to be valid at the time but which subsequently turns out to be mistaken.58

This test of ‘honest belief’ is not only applied to the state agents’ assessment of the danger at issue, but also to their choice of reaction. The ‘honest belief’ lens substantially dilutes the ‘absolute necessity’ assessment. As outlined in the above passage, the ‘honest belief’ test includes what may be considered to be an objective element, embodied in the ‘for good reasons’ criterion. However, the Grand Chamber in Da Silva confirmed that it considers that ‘the existence of “good reasons” should be determined subjectively’.59 The Grand Chamber suggested that ‘the Court has not treated reasonableness as a separate requirement but rather as a relevant factor in determining whether a belief was honestly and genuinely held’ (emphasis added),60 referencing prior case law in which this approach had been effectively adopted if not explicitly affirmed, such as Bubbins v UK.61 In Bubbins, a police officer shot and killed an unarmed man who was mistaken for an intruder in his own home and was wrongly thought to be aiming a weapon from the window of his flat towards police officers surrounding it. The substantive complaint on Article 2 grounds was that there had been a breach of Article 2 both in the actions of the officer who shot and killed the man, and in the overall planning and control of the operation that led to the use of lethal force, which was not absolutely necessary. The Court applied the ‘honest belief’ test, reasoning that ‘it cannot substitute its own assessment of the situation for that of an officer who was required to react in the heat of the moment to avert an honestly perceived danger to his life’62 and concluded that: [T]he use of lethal force in the circumstances of this case, albeit highly regrettable, was not disproportionate and did not exceed what was absolutely necessary to avert what was honestly perceived by Officer B to be a real and immediate risk to his life and the lives of his colleagues.63

Andrew Ashworth has suggested that Bubbins effectively embodies a subjective test,64 while Neil Martin has commented that: ‘Given that there were a number of noticeable errors and questionable decisions made in the conduct of the operation, it would seem that the Court will require an extremely high level of error, ineptitude or bad judgement before it will find a breach of Art. 2’ (emphasis added).65

58 See, for example, Giuliani and Gaggio v Italy (2012) 54 EHRR 10, para 178. 59 Da Silva (n 14) para 247. 60 ibid para 248. 61 Bubbins v UK (2005) 41 EHRR 24. 62 ibid para 139. 63 ibid para 140. 64 A Ashworth, Positive Obligations in Criminal Law (Oxford, Hart Publishing, 2015) 205. 65 N Martin, ‘Bubbins v United Kingdom: Civil Remedies and the Right to Life’ (2006) 69 Modern Law Review 242, 246 (citations omitted).

Coercive Overreach, Dilution and Diversion  195 Attesting further to the problem of dilution, which reaches into the assessment of the planning of relevant operations, is the case of Giuliani v Italy. In Giuliani, concerning the shooting of an anti-globalisation protester by a carabiniero when a jeep carrying three carabinieri was surrounded by violent protesters during the G8 summit in Genoa, the Court reasoned similarly to Bubbins through the ‘honest belief’ test.66 Applying it both to the question of whether circumstances called for force and to the question of whether the force ultimately used by the carabiniero – shooting blindly from the jeep – was strictly proportionate to the risk posed,67 it found no violation of Article 2. The Court also found the organisational deficiencies which resulted in three highly inexperienced carabinieri armed with only lethal weapons at their disposal being surrounded by protesters in the context of a pre-planned, highly securitised event – the G8 summit – not to fall foul of its stringent standards in minimising the likelihood of loss of life.68 In a judgment characterised by ‘national security’-styled deference,69 the Court shrank back from a stringent application of the absolute necessity test, which should be seen holistically to encompass the organisational aspect of the state operation in McCann.70 The Court could approach the question of whether there has been a breach of the negative obligation under Article 2 differently by re-asserting the distinction between state liability and individual (civil or criminal) liability. Absolute necessity represents an objective standard of assessment, and not one of which only an extremely high level of error, ineptitude or bad judgement would fall foul. While the subjective perception of the perpetrator may play a role in determining individual criminal liability, Article 2 makes it clear that the right to life is violated where a person has been killed in circumstances where lethal force used against them was not absolutely necessary. The standard is not whether the person inflicting such force considered it, in good faith, to be (absolutely) necessary.71 Thus, while individual criminal liability may appropriately be delineated through an assessment of (potentially flawed) subjective perception and intent  – an assessment that only a full Article 6-compliant criminal trial can provide – the state’s responsibility for unnecessary takings of life is appropriately to be assessed objectively. Moreover, in state killings, the superior competence and knowledge of the state should substantially burden rather than ‘absolve’ the

66 Giuliani (n 58) para 178. 67 ibid paras 178–95. 68 ibid paras 244–62. 69 See S Skinner, ‘Deference, Proportionality and the Margin of Appreciation in Lethal Force Case Law under Article 2 ECHR’ (2014) European Human Rights Law Review 32. 70 See text to n 55. But note the Court’s more stringent assessment of planning and control under a ‘positive obligation’ heading in, for example, Finogenov v Russia (2015) 61 EHRR 4, paras 237–66. 71 Jonathan Rogers observed that the tests at play are distinct in J Rogers, ‘Culpability in SelfDefence and Crime Prevention’ in GR Sullivan and I Dennis, Seeking Security: Pre-empting the Commission of Criminal Harms (Oxford, Hart Publishing, 2012) 266.

196  Natasa Mavronicola state, and the onus to show that the force used was absolutely necessary should lie with the state, which is likely to be – and indeed under an investigative obligation to take reasonable steps to be – in possession of the relevant facts. This requires the state to show adequate objective reasons establishing both the need to use force and that the force used was strictly proportionate – that is, strictly not excessive – to the risk to life or bodily integrity at issue in the circumstances. The latter element means that (potentially) lethal force is meant to be treated as a last resort and that alternatives, such as retreat, warnings and other non-lethal or ‘less lethal’ means, must be shown to have been made available and, where appropriate, used or considered first.72 On this approach, the state should also be held accountable under the right to life in circumstances in which the planning of a particular operation, capable of resulting in the use of lethal force, was collectively mismanaged so as to unnecessarily create the conditions for (potentially) lethal force to be used, even if (potentially) lethal force was ultimately necessary at the fatal moment. Findings of an Article 2 breach in circumstances of such operational mismanagement entail that the state apparatus is held to account, without necessarily demanding the criminal accountability of any of the individuals involved.73 This is the approach ultimately taken in McCann.74 Such an approach would entail that while individuals are not necessarily held criminally liable for diffuse or systemic errors or collective mismanagement, the state may nonetheless be found responsible for violating the right to life and be compelled to revise its planning and management of high-risk operations accordingly. The issue has not quite materialised in the same way in Article 3 case law. Nonetheless, the prospect of a criminal lens operating to narrow the scope of circumstances in which the state is found to be in violation of the negative obligation under Article 3 ECHR remains a worrying one. A ‘mirroring’ approach would mean that Article 3 ill-treatment is only established in circumstances in which criminally wrongful conduct has occurred. Should this path be followed, it is capable of eroding much of Article 3’s protection in a range of contexts where inhumanity or degradation may be inflicted unintentionally, by virtue of a legal regime, or as a result of structural, systemic or diffuse problems, failings or errors. Such contexts include immigration and asylum, sentencing and imprisonment, and others. Consider, for example, the imposition of a de facto irreducible sentence of life imprisonment,75 degrading prison conditions in inadequately resourced prisons, substantively flawed asylum decisions resulting in violation of the non-refoulement duty, or the inadvertently excessive use of force

72 See Russell (n 57) 499. 73 cf Ashworth (n 64) 205. 74 McCann (n 34) paras 213–14. 75 This is contrary to art 3 according to the Court in Vinter v UK (2016) 63 EHRR 1 and Murray v The Netherlands (2017) 64 EHRR 3.

Coercive Overreach, Dilution and Diversion  197 in the conduct of an arrest by an undertrained police officer misperceiving the threat at hand.76 The Court should continue to recognise violations of Article 3 in such circumstances even if no criminal wrong is necessarily made out. The human right not to be subjected to torture or inhuman or degrading treatment or punishment has been interpreted in a dynamic way to capture acts and omissions whose wrongfulness is relational and significant, indeed significant enough for them to be considered conclusively unlawful as a matter of human rights law,77 even if not always of a nature that is criminally culpable. Moreover, there is cause for concern regarding what a criminal lens might entail for the approach taken to proving violations of rights such as those enshrined in Articles 2 and 3 ECHR. Although the Court often uses the language of ‘beyond reasonable doubt’ in determining violations of these provisions,78 the evidential standard applied in Article 2 and 3 cases generally does not (and ought not to) replicate a criminal law standard.79 Instead, the Court has clarified that ‘such proof may follow from the co-existence of sufficiently strong, clear and concordant inferences or of similar unrebutted presumptions of fact’ and that ‘where allegations are made under Articles 2 and 3 of the Convention, the Court must apply a particularly thorough scrutiny’.80 It has, in particular, adopted the position that where the events at issue ‘lie wholly, or in large part, within the exclusive knowledge of the authorities, as in the case of persons within their control in custody’, a presumption of an Article 3 violation will

76 See, for example, Peers v Greece (2001) 33 EHRR 51; MSS v Belgium and Greece (n 40); and the excessive use of force case law outlined in N Mavronicola, ‘Güler and Öngel v Turkey: Article 3 of the European Convention on Human Rights and Strasbourg’s Discourse on the Justified Use of Force’ (2013) 76 Modern Law Review 370. 77 See, generally, N Mavronicola, Torture, Inhumanity and Degradation under Article 3 of the ECHR: Absolute Rights and Absolute Wrongs (Oxford, Hart Publishing, forthcoming). 78 See, for example, Bouyid v Belgium (2016) 62 EHRR 32, para 82; Güler and Öngel (n 41) para 26; Anguelova v Bulgaria (2004) 38 EHRR 31, para 111. 79 See the related discussion in Nollkaemper (n 50) 627–31. 80 See, for example, Güler and Öngel (n 41) para 26. Indeed, the Court has elaborated on this as follows: ‘[I]n assessing evidence, the Court has adopted the standard of proof “beyond reasonable doubt”. However, it has never been its purpose to borrow the approach of the national legal systems that use that standard. Its role is not to rule on criminal guilt or civil liability but on Contracting States’ responsibility under the Convention. The specificity of its task under Article 19 of the Convention – to ensure the observance by the Contracting States of their engagement to secure the fundamental rights enshrined in the Convention – conditions its approach to the issues of evidence and proof. In the proceedings before the Court, there are no procedural barriers to the admissibility of evidence or pre-determined formulae for its assessment. It adopts the conclusions that are, in its view, supported by the free evaluation of all evidence, including such inferences as may flow from the facts and the parties’ submissions. According to its established case-law, proof may follow from the coexistence of sufficiently strong, clear and concordant inferences or of similar unrebutted presumptions of fact. Moreover, the level of persuasion necessary for reaching a particular conclusion and, in this connection, the distribution of the burden of proof are intrinsically linked to the specificity of the facts, the nature of the allegation made and the Convention right at stake. The Court is also attentive to the seriousness that attaches to a ruling that a Contracting State has violated fundamental rights’ – Nachova and Others v Bulgaria (2006) 42 EHRR 43, para 147.

198  Natasa Mavronicola arise in respect of any injuries sustained during the relevant period.81 On this basis, ‘the burden of proof may be regarded as resting on the authorities to provide a satisfactory and convincing explanation’82 for the injuries in order to be absolved of responsibility. These are entirely appropriate approaches to the evidentiary dimension of establishing state violations of human rights, and are justified not only by the superior or exclusive knowledge of the authorities, but also by the ‘vulnerable position’83 in which individuals confronted by the state’s law enforcement officials or similar find themselves. However, such evidentiary approaches are not usually adopted – and may not be appropriate in many or most circumstances – in the determination of individual criminal liability. The danger therefore arises, out of the ‘mirroring’ of Article 2 and Article 3 breaches and criminal liability, that the looser burden of proof and the burden’s reversal in particular circumstances may be relinquished in favour of a standard that more closely resembles the evidentiary standard employed to determine individual criminal liability for the most serious crimes. Such a change in approach would, if it did occur, impair the process of establishing human rights violations and potentially shrink the circumstances in which the state is accordingly held to account. V.  THE PROBLEM OF DIVERSION

The above account takes seriously the importance of upholding human freedom, closely constraining state violence and effectively protecting persons from grave harm. Guarding against coercive overreach and dilution is required in order to maintain and promote, in a coherent fashion, the ideals and values of a democratic society as demanded by the Convention, whose essence is respect for human dignity and human freedom.84 At the same time, in view of the significance placed on making Convention safeguards practical and effective, it is worth underlining that criminalisation and criminal redress may in many circumstances be neither necessary (coercive overreach viewed from a deterrence/ prevention angle)85 nor sufficient86 to prevent ill-treatment or unlawful takings of life, or indeed more generally to protect (potential) victims of such violations. Therefore, it is important to highlight that the ECtHR’s coercive approach to positive obligations for the protection of core rights can divert the Court from alternative tools of protection and obscure alternative ways of conceptualising what is at stake. What I mean by the latter point is that a focus, via this coercive orientation, on the criminally wrongful dimension of human rights violations can entail that other issues might be missed or implicitly downplayed.

81 Bouyid

(n 78) para 83. v Turkey (2010) 50 EHRR 1, para 39. 83 Bouyid (n 78) para 83. 84 Pretty (n 4) para 65. 85 See the account offered above, as well as Peršak, ch 7 in this volume. 86 See, for example, Vladislava Stoyanova, ch 10 in this volume. 82 Diri

Coercive Overreach, Dilution and Diversion  199 Thus, an application of the right to life to a litany of not-quite-criminallyculpable failings, culminating in the avoidable loss of life, that employs a criminal law lens to find that no breach has been committed not only miscasts the test of absolute necessity but also misses the opportunity to repair – or, rather, call for the repair of – the processes, mechanisms and structures that went awry. An emphasis in the investigation of alleged or potential human rights violations on identifying and punishing those responsible can mean that a range of crucial contributing factors or indeed a whole context conducive to the proliferation of such violations are potentially sidelined and left (sometimes conveniently) unaddressed. A finding of a procedural violation of Article 3 in respect of a state’s failure to punish (adequately) a police officer for brutally beating a Roma person may be perceived as pursuing justice, but it is only partly reparative if it fixes its gaze on a bad apple (the racist, violent officer) and misses (and seemingly implicitly absolves) the rotten orchard (an institutionally racist police force).87 A pronouncement by a human rights court that presents and addresses the failure to protect a victim of domestic violence purely or mainly as a failure to criminalise and punish adequately can obscure the systems and structures at play (from misogynist institutions to patriarchal and heteronormative legal and societal norms) and can therefore skew perceptions of the problem itself as well as how to fix it.88 Another important domain in which the criminal approach, or an emphasis on individual responsibility, might limit our vision is the Court’s stance on the reckoning needed after – sometimes long after – mass atrocity or conflict. The Court has favoured criminal investigations as the means of discharging the procedural duty under Articles 2 and 3, with the emphasis being on the investigation’s capability of leading ‘to the identification and punishment of those responsible’.89 Even where the Court has entertained a wider conception of the duty to investigate, such a duty is still understood as seeking the apportioning of criminal, civil, administrative or disciplinary liability and as ‘leading to … an award of compensation’.90 The Court’s interpretation of the investigative duty under Article 2 or 3 ECHR as requiring only proceedings leading to the apportioning of (individual) legal responsibility has had significant and limiting implications for how ‘dealing with the past’91 may be conceived as a matter of human rights law. In particular, the Court has held, in Janowiec v Russia,92 that the procedural duty relates solely to ‘acts undertaken in the framework of criminal, civil, administrative or disciplinary proceedings which are capable

87 M Punch, ‘Rotten Orchards: “Pestilence”, Police Misconduct and System Failure’ (2003) 13 Policing and Society 171. 88 But see Volodina v Russia App No 41261/17 (ECtHR, 9 July 2019). 89 McShane (n 12) para 96. 90 Aleksakhin (n 20) para 60. 91 See K McEvoy and L Mallinder, ‘Truth, Amnesty and Prosecutions: Models for Dealing with the Past’ (Transitional Justice Institute Research Paper No 14-09, 2014). 92 Janowiec v Russia (2014) 58 EHRR 30.

200  Natasa Mavronicola of leading to the identification and punishment of those responsible or to an award of compensation to the injured party’, to the exclusion of ‘other types of inquiries that may be carried out for other purposes, such as establishing a historical truth’.93 This had concrete implications in Janowiec, in which relatives of victims argued that Russia had not discharged its investigative obligation in respect of the disappearance of their relatives at the hands of Russian forces in 1940. The Grand Chamber found that it lacked jurisdiction ratione temporis because most of the relevant liability-determining proceedings had taken place prior to the ECHR entering into force in Russia and, in the Court’s view, no significant investigative steps or novel information had arisen subsequently so as to establish its temporal jurisdiction over Russia’s investigation.94 The Court’s reasoning has been extensively critiqued.95 Looking beyond this particular judgment or indeed the issue of temporal jurisdiction over historical injustices, the ECtHR’s understanding of the investigative duty as requiring exclusively legal responsibility-apportioning processes arguably operates under a thin notion of justice, at the risk of crowding out, delegitimising, or affirmatively excluding other processes that operate under richer – or ‘thicker’96 – understandings of justice.97 Accordingly, there is cause for recognising that a criminal law (enforcement) focus may address only a fraction of the wrong(s) at issue as well as of the substantial obstacles or shortcomings in protecting persons from the harms at issue. A protective and preventive re-orientation in the doctrine is needed, shaped perhaps by a re-orientation in human rights litigation agendas. A protective and preventive re-orientation would require a recalibration of the Court’s attention towards the wider, potentially non-criminal acts and omissions – encompassing laws, policies, practices and structures – involved in the violation of a fundamental human right, and a consideration of ‘practical and effective’ (to quote the Court)98 tools of protection beyond criminal law (enforcement) in the specification of positive obligations. This would entail, for example, a greater readiness to hold states to account for failing to provide thorough training on the use of force and adequate defensive equipment for all state agents who may be called upon to use force, or failing to plan and prepare for potentially violent state operations robustly.99 But it could also entail demanding 93 ibid para 143. 94 ibid paras 158–59. 95 For a critical take on this, see C Heri, ‘Enforced Disappearance and the European Court of Human Rights’ Ratione Temporis Jurisdiction’ (2014) 12 Journal of International Criminal Justice 751. 96 On thicker conceptions of transitional justice, see K McEvoy, ‘Beyond Legalism: Towards a Thicker Conception of Transitional Justice’ (2007) 34 Journal of Law and Society 411. 97 Many of the contributions in Engle, Miller and Davis (n 47) consider the opportunity costs of anti-impunity in transitional justice. Miles Jackson considers the exclusion of particular transitional justice processes and settlements in a discussion on the permissibility of amnesties under ECtHR doctrine in M Jackson, ‘Amnesties in Strasbourg’ (2018) 38 Oxford Journal of Legal Studies 451. 98 Svinarenko and Slyadnev (n 1) para 118. 99 This has, to an extent, materialised in case law since McCann (n 34). See n 55.

Coercive Overreach, Dilution and Diversion  201 safeguards such as: effective access to justice, including through the alleviation of financial or material barriers and relating not just to criminal proceedings but also to family law, immigration and other proceedings, for victims of domestic violence;100 access to refuges for victims of such violence;101 or the operation of firewalls between mechanisms of protection from abuse on the one hand and immigration control on the other,102 to give a few examples. Moreover, investigative duties could be recast so as to require states to examine the wider systems and structures in which a violation occurred and thus to pursue systemic and structural measures towards securing non-recurrence. This would be vital, for example, in ensuring that states acknowledge and address issues such as institutional racism or cultures of brutality within law enforcement, thereby targeting the rotten orchard as well as the bad apples. For the preventive and protective re-orientation to take hold, the assumptions built into the ECtHR’s coercive human rights doctrine must be challenged. There is an assumption built into the Court’s coercive human rights doctrine that criminal law (enforcement) is a ‘practical and effective’ tool of protection,103 and to this should be added the assumption that other tools – those which are implicitly or explicitly excluded from the Court’s elaboration of positive obligations – are less ‘practical and effective’. It is not clear that this is the case. The case for systematically examining the (opportunity) costs of this diversion from alternative means of protection, and exploring the types of concrete alternatives that the Court has adopted and can further adopt, is strong and indeed urgent, as examples of coercive and carceral tools failing persons in vulnerable situations abound.104 VI. CONCLUSION

This chapter has unpacked three key dangers that the ECtHR’s coercive human rights doctrine gives rise to: coercive overreach, dilution and diversion. It has highlighted the potential, within this doctrine, of both coercive overreach and dilution of the human rights standards at play. The danger of coercive overreach relates to the potentially excessive reach of the Court’s demands of 100 On this, see Report of the United Nations Special Rapporteur on Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, ‘Relevance of the prohibition of torture and other cruel, inhuman or degrading treatment or punishment to the context of domestic violence’, UN Doc A/74/148, para 73. 101 ibid paras 23 and 79. 102 On the complex interplay between coercion and protection in the immigration context and the application of art 4 ECHR, see Stoyanova, ch 10 in this volume. 103 For a critical take on this, see Peršak, ch 7 in this volume. 104 See, for example, M Oppenheim, ‘Police Accused of “Systemic Failure” to Protect Victims of Domestic Abuse and Sexual Violence’, The Independent, 20 March 2019; M Bulman, ‘Government under Fire for “Outrageous” Treatment of Modern Slavery Victims Facing Deportation from UK’, The Independent, 11 June 2018. See also the comments made by Pinto, ch 8 in this volume.

202  Natasa Mavronicola criminalisation and punishment, as well as to the institutional overreach of its assessments of criminal wrongfulness and liability. The danger of dilution emerges in the (potential) narrowing effect of conceptualising human rights violations through a criminal lens, as exemplified in the application of an ‘honest belief’ filter in the assessment of absolute necessity under Article 2. A danger of dilution also pertains to the prospect that the emphasis on criminal liability and punishment for violations of rights such as those in Articles 2 and 3 ECHR may shift the Court’s current evidentiary standards in a criminal law direction that would offer states an undue benefit of the doubt. This chapter has also underlined the danger of diversion. This danger lies in the way in which the Court diagnoses and seeks to cure, and prospectively to curb, violations of fundamental human rights and their causes. A criminal law focus may offer an incomplete account of what has gone wrong and how to fix it, diverting the Court from potentially more effective tools of protection. There is therefore a need to revisit the assumptions built into coercive human rights doctrine and to examine the opportunity costs of the doctrine’s coercive slant more closely. At the heart of the warnings offered in this chapter is a preoccupation with protection, rather than coercion, and with upholding the Convention’s integrity, so that it speaks with ‘one voice’.105 While the analysis in this chapter challenges the coercive ‘sting’ in the ECtHR’s positive obligations doctrine, it by no means seeks to oppose an expansive approach to positive obligations. On the contrary, I hope that this chapter might serve as an invitation to re-imagine positive obligations through a protective, rather than coercive, re-orientation.



105 Dworkin

(n 25) 213.

10 Separating Protection from the Exigencies of the Criminal Law Achievements and Challenges under Article 4 ECHR VLADISLAVA STOYANOVA

I. INTRODUCTION

I

t is uncontroversial that states are under the positive obligation to ­criminalise abuses falling within the definitional scope of Article 4 of the European Convention on Human Rights (ECHR) (the right not to be subjected to slavery, servitude, forced labour or human trafficking).1 Criminalisation has been considered necessary to ensure that individuals are protected from harm inflicted by other individuals.2 Although criminal law only becomes relevant post factum (ie, once harm has been sustained), criminalisation is generally

1 Siliadin v France App No 73316/01 (ECtHR, 26 July 2005), para 112; CN and V v France, App No 67724/09 (ECtHR, 11 October 2012), para 105; Rantsev v Cyprus and Russia App No 25965/04 (ECtHR, 7 January 2010), para 285. See V Stoyanova, ‘Article 4 of the ECHR and the Obligation of Criminalizing Slavery, Servitude, Forced Labour and Human Trafficking’ (2014) 3(2) Cambridge Journal of International and Comparative​e Law 407; V Stoyanova, Human Trafficking and Slavery Reconsidered: Conceptual Limits and States’ Positive Obligations in European Law (Cambridge, Cambridge University Press, 2017) 331. 2 Article 1 ECHR imposes an obligation upon the state to ensure the rights, which implies not only refraining from harmful conduct, but also taking proactive protective measures. This has provided a justification for the development of positive obligations under the ECHR. See J Akandij-Kombe, Positive Obligations under the European Convention on Human Rights (Strasbourg, CoE Publishing, 2007) 8; Stoyanova, Human Trafficking and Slavery Reconsidered (n 1) 320.

204  Vladislava Stoyanova viewed as a deterrence tool.3 It is similarly uncontroversial that Article 4 has a procedural aspect: it can thus trigger the positive obligation of conducting effective criminal investigation.4 This implies that the national authorities have to conduct effective criminal investigation when faced with an arguable claim that an individual has been subjected to abuses that might qualify as slavery, servitude, forced labour or human trafficking. These two positive obligations (to criminalise and to conduct a criminal investigation) are intimately linked to the national criminal law and they seem to be underpinned by the understanding that criminal law and its effective application are contributory to the protection of human rights.5 This chapter aims to explore another type of positive obligation under Article 4 and how its link with criminal law might be detrimental rather than beneficial. In particular, it focuses on the obligation upon states to identify migrants as victims of harm that falls within the scope of Article 4 ECHR and to extend protection and assistance to them as victims independently of any criminal proceedings and any actual or potential participation by them in any criminal proceedings.6 The concrete questions under review are: is the ECtHR case law supportive of the separation between identification and protection, on the one hand, and any criminal proceedings, on the other? If identification and the ensuing protection are separated from the exigencies of the criminal law, what could be the positive and negative repercussions? Such repercussions can be examined both from the perspective of the state and from the perspective of the individual. Three preliminary clarifications are due in terms of the scope of the enquiry. These concern the specific focus on migrants and the question as to why migrant victims of slavery, servitude, forced labour and human trafficking should be identified and assisted independently of any criminal proceedings. Certainly, any victim of crime, irrespective of his or her migration and

3 J Rogers, ‘Applying the Doctrine of Positive Obligations in the European Convention on Human Rights to Domestic Substantive Criminal Law in Domestic Proceedings’ [2003] ­Criminal Law Review 690, 695. See also A Ashworth, Positive Obligations in Criminal Law (Oxford, Hart Publishing, 2013) 196–211. Applying criminal law as a tool for protecting human rights is not unproblematic. For a useful account of the problems, see F Tulkens, ‘The Paradoxical Relationship between ­Criminal Law and Human Rights’ (2011) 9 Journal of International Criminal Justice 577. 4 See Stoyanova, Human Trafficking and Slavery Reconsidered (n 1) 351. Admittedly, the procedural aspect of the ECHR substantive provisions might not be limited to criminal proceedings and it might include civil and administrative proceedings that aim at the identification and punishment of those responsible. See, for example, Janowiec and Others v Russia [GC] App No 55508/07 and 29520/09 (ECtHR, 21 October 2013), para 143. 5 However, this understanding is controversial. See L Lazarus, ‘Positive Obligations and ­Criminal Justice: Duties to Protect or Coerce’ in L Zadner and J Roberts (eds), Principles and Values in ­Criminal Law and Criminal Justice: Essays in Honour of Andrew Ashworth (Oxford, Oxford University Press, 2012) 136. 6 Assistance and protection are used here interchangeably and are generally understood to mean social assistance in terms of shelter, medical and psychological help, legal help, referral to asylum proceedings and referral to other migration proceedings that might regulate victims’ migration status.

Separating Protection from the Exigencies of the Criminal Law  205 residence status,7 can (and should) benefit from protection and assistance.8 However, there are some peculiarities when the victim happens to be a migrant. First, migrants are more vulnerable to some specific crimes; an example of this is human trafficking and severe forms of labour exploitation.9 This vulnerability is linked to their migration status.10 Second, because of this status and the related unwillingness to approach the national authorities,11 it is necessary to actually identify them as victims. For example, the Council of Europe (CoE) Convention on Action against Trafficking in Human Beings (hereinafter the CoE Anti-trafficking Convention) contains specific provisions aimed at the identification of migrants as victims of human trafficking. Victims of no other crime have been designated as being in need of such a specific identification procedure. This need has arisen because of their migration status.12 More specifically, Article 10 of the CoE Anti-trafficking Convention demands that States Parties shall adopt ‘such legislative or other measures as may be necessary to identify victims as appropriate in collaboration with other Parties and relevant support organizations’, shall train their competent authorities in identifying victims and shall suspend deportation upon reasonable grounds to believe that a person has been a victim until the completion of the identification process.13 In addition, Article 13 provides for a recovery and reflection period for migrants who might be victims of human trafficking. The extension of this period that implies suspension of the deportation proceedings against the migrant victim is contingent on the initiation of the formal identification of the person as a victim. In relation to this identification, the Explanatory Report to the CoE Anti-trafficking Convention clarifies that: The identification process provided for in Article 10 is independent of any ­criminal proceedings against those responsible for the trafficking. A criminal conviction is

7 I use the term ‘victim of crime’ in a general fashion, irrespective of whether a perpetrator has been convicted after a criminal trial. 8 For this reason at both the EU and CoE levels, specific legal instruments have been adopted addressing assistance and protection for victims of crime. See, for example, Directive 2012/29/EU of the European Parliament and of the Council of 25 October 2012 establishing minimum standards on the rights, support and protection of victims of crime. See also CoE Convention on the Compensation of Victims of Violent Crimes ETS No 116. 9 Nationals can be also victims of these crimes; however, empirical studies do point to the specific vulnerabilities of migrants: Severe Forms of Labour Exploitation (EU Fundamental Rights Agency, 2015). All the judgments delivered under art 4 ECHR and discussed in this chapter involve migrants. 10 Severe Forms of Labour Exploitation (n 9). 11 G Noll, ‘Why Human Rights Fail Undocumented Migrants’ (2010) 12 European Journal of Migration and Law 143. 12 An argument can also be made that the need arises because of the ambiguities as to the meaning of human trafficking or because, related to this, the victim himself or herself cannot understand that he or she is in fact a victim of any crime. 13 For a detailed examination of these provisions, see Stoyanova, Human Trafficking and Slavery Reconsidered (n 1) 101.

206  Vladislava Stoyanova therefore unnecessary for either starting or completing the identification process. (Emphasis added)14

In accordance with the approach outlined in the Explanatory Report, the Group of Experts on Action against Trafficking in Human Beings (GRETA), the body mandated to monitor the implementation of the CoE Anti-trafficking Convention, has taken the position in its reports that the national authorities have to ‘ensure that the identification of victims of THB [trafficking in human beings] is not made conditional on their co-operation in the investigation and criminal proceedings or the initiation of criminal proceedings’ (emphasis added)’.15 Yet, the actual text of the treaty is not that unequivocal as to the watertight separation between identification and criminal proceedings.16 The existing state practice also adds to this uncertainty.17 Crucially, the CoE Anti-trafficking Convention provides a lot of leeway to states as to how to organise their identification procedures and whether there have to be any procedural safeguards incorporated therein that can ensure the independence of procedures from the perspective of the individual.18 The Explanatory Report to the CoE Anti-Trafficking Convention acknowledges this scope of discretion left for states by clarifying that by competent authority, it means ‘the public authorities which may have contact with trafficking victims, such as the police, the labour inspectorate, customs, the immigration authorities and embassies and consulates’.19 As a result, different States Parties to the CoE Anti-trafficking Convention have mandated different bodies, including prosecutors and national investigating authorities,20 with the task of formally identifying victims. The primary objective of these authorities (ie, prosecuting and investigating authorities, the police etc) is the effective and successful application of the criminal law, in which context the victim might be simply instrumentalised. As a consequence, individuals who are not instrumental for achieving the criminal 14 The Explanatory Report is available at: https://rm.coe.int/16800d3812. 15 Report Sweden, 27 May 2014, GRETA(2014)11, para 142. In its report on Sweden, GRETA expressed its concern that ‘the criminal law-based approach to victim identification leaves victims of THB without formal identification and outside of the scope of the protection measures provided for under the Convention’. See also Report Ireland, 26 September 2013, GRETA(2013)15, para 165, where GRETA urged the national authorities to guarantee that ‘in practice identification is dissociated from the suspected victim’s co-operation in the investigation’. 16 For a detailed analysis, see Stoyanova, Human Trafficking and Slavery Reconsidered (n 1) 102. 17 The state practice can be identified in the country reports published by GRETA. See Report Sweden, 27 May 2014, GRETA(2014)11, para 142; Report Ireland, 26 September 2013, GRETA(2013)15, para 165’. 18 The incorporation of any concrete procedural guarantees (eg, the possibility to challenge a negative decision) into the procedure for victim identification was rejected when the CoE Anti-­ trafficking Convention was negotiated. Contributions by the delegation of Norway and by the observer of Mexico, Restricted CAHTEH(2003) 8 Rev.2, Addendum II, 1 December 2003 Strasbourg; Final Activity Report CAHTEH(2005) RAP 8, 16 March 2005, 13–14 (documents on file with the author). 19 Explanatory Report (n 14) para 129. 20 See Report Belgium, GRETA (2013)14, 25 September 2013, 35; Report Bulgaria, GRETA (2011)19, 14 December 2011, 35.

Separating Protection from the Exigencies of the Criminal Law  207 law objectives (by, for example, acting as witnesses or at least filing official criminal complaints) might not be formally identified and/or assisted. This is arguably contrary to the rationale of the CoE Anti-trafficking Convention, where victim identification is primarily perceived as a tool for ensuring that victims, including presumed victims,21 access assistance and protection. If the national identification procedure does not achieve this, due to its exclusive or dominant focus on prosecution and conviction of traffickers, then it might be doubtful as to whether it lives up to the requirements of the Convention.22 As will be suggested below, the independence of victim identification procedures from any criminal proceedings might also be demanded so that states comply with their positive obligations under Article 4 ECHR. The development of positive obligations under the ECHR more generally has been justified by the argument that the ECHR ‘is intended to guarantee not rights that are theoretical or illusory but rights that are practical and effective’.23 The principle of effectiveness can thus be successfully utilised for substantiating an argument that if victim identification procedures do not actually serve the objective of victim protection and assistance, due to their predominant utilisation for serving the criminal law objectives (identification and prosecution of alleged criminals), they might not be effective and thus might fail to comply with the positive obligations imposed by Article 4 ECHR. A third peculiarity related to the victims’ migration status is that often migrant victims of crime do not have secure status in the country; as a consequence, they are in need of a specific form of protection – suspension of deportation proceedings or granting of a residence permit. This is illustrated not only in the text of the CoE Anti-trafficking Convention,24 but also in the more recent CoE Convention on Preventing and Combating Violence against Women and Domestic Violence (hereinafter the Istanbul Convention).25 Specific provisions have been inserted therein precisely because of the awareness that migrants who are victims of crimes, including domestic violence, might require

21 See arts 10(2) and 13 of the CoE Anti-trafficking Convention that refer to ‘reasonable grounds to believe that the person concerned is a victim’. 22 Here we should take note of the three objectives of the CoE Anti-trafficking Convention as stated in art 1(1): to prevent and combat trafficking in human beings; to design a comprehensive framework for the protection and assistance of victims; and to promote international cooperation. Despite the strong focus on the effective application of criminal law, the preamble to the Convention suggests that protection of victims is the paramount objective. 23 Airey v Ireland App No 6289/73, 9 October 1979, para 24. See A Mowbray, The Development of Positive Obligations under the European Convention on Human Rights by the European Court of Human Rights (Oxford, Hart Publishing, 2004) 221; L Lavrysen, Human Rights in a Positive State (Cambridge, Intersentia, 2016) 146. 24 See arts 10(2) and 13 of the CoE Anti-trafficking Convention. 25 V Stoyanova, ‘A Stark Choice: Domestic Violence or Deportation. The Immigration Status of Victims of Domestic Violence under the Istanbul Convention’ (2018) 20(1) European Journal of Migration and Law 53; V Stoyanova, ‘On the Bride’s Side? Victims of Domestic Violence and their Residence Rights under EU and Council of Europe Law’ (2019) 37(4) Netherlands Quarterly of Human Rights 311.

208  Vladislava Stoyanova specific protection in the form of non-deportation or residence permits.26 It follows that in circumstances where migrants are victims of crime, an interesting interaction arises between criminal law, national immigration law and human rights law. Therefore, the answer to the question posed at the beginning of the chapter as to the protection of victims that is independent from the exigencies of the criminal law will be influenced by the operation and the rationales of the national immigration legislation. II.  OVERVIEW OF THE CASE LAW UNDER ARTICLE 4

Compared to other provisions of the ECHR, Article 4 has produced limited judicial output.27 The first case in which the European Court of Human Rights (ECtHR) had the chance to apply this provision in the context of harm inflicted by private actors was Siliadin v France decided in 2005.28 This judgment marked the first time that the ECtHR had the opportunity to clarify that Article 4 imposes a positive obligation upon states to criminalise servitude and forced labour.29 Almost five years later, the Court delivered the widely applauded judgment of Rantsev v Cyprus and Russia,30 which signified two important developments: first, the material scope of Article 4 was extended to cover human trafficking;31 and, second, the positive obligations under Article 4 were determined to be much more far-reaching than simple criminalisation, since states are also required to take additional protective measures. These will be examined in detail below. In 2012, the ECtHR rendered two important judgments under Article 4. In particular, CN and V v France clarified the distinction between forced labour and servitude, and reiterated the finding in Siliadin v France that states are under the positive obligation to specifically criminalise the abuses covered by Article 4 ECHR.32 In CN v UK, the Court further elaborated on the meaning of 26 See art 59 of the CoE Convention on Preventing and Combating Violence against Women and Domestic Violence, CETS No. 210. 27 There is a separate line of case law under art 4 addressing forced labour demanded by the state. See, for example, Chitos v Greece App No 51637/12 (ECtHR, 4 June 2015). 28 Siliadin v France (n 1). 29 H Cullen, ‘Siliadin v France: Positive Obligations under Article 4of the European Convention on Human Rights’ (2006) 6(3) Human Rights Law Review 585. For the changes introduced in the French substantive criminal law after Siliadin v France, see B Bourgeois, ‘Statutory Progress and Obstacles to Achieving an Effective Criminal Legislation against the Modern-Day Forms of Slavery: The Case of France’ (2017) 38(3) Michigan Journal of International Law 455. 30 Rantsev v Cyprus and Russia App No 25965/04 (ECtHR, 7 January 2010), para 282. 31 For a critical analysis of this addition, see J Allain, ‘Rantsev v Cyprus and Russia: The European Court of Human Rights and Trafficking as Slavery’ (2010) 10(3) Human Rights Law Review 546; V Stoyanova, ‘Dancing on the Borders of Article 4: Human Trafficking and the European Court of Human Rights in the Rantsev Case’ (2012) 30(2) Netherlands Quarterly of Human Rights 163. See also the Dissenting Opinion of Judge Koskelo in SM v Croatia App No 60561/14 (ECtHR, 19 July 2018) (referred to the Grand Chamber). 32 CN and V v France (n 1) para 91: ‘In the light of these criteria the Court observes that servitude corresponds to a special type of forced or compulsory labour or, in other words, “aggravated”

Separating Protection from the Exigencies of the Criminal Law  209 s­ ervitude and emphasised that this is a specific offence, which ‘involves a complex set of dynamics, involving both overt and more subtle forms of coercion, to force compliance’.33 Most importantly, it became clear that in order for states to comply with the procedural limb of Article 4 (the obligation to investigate), an understanding of the subtle ways in which an individual can fall under the control of another is required.34 M and Others v Italy and Bulgaria also has to be mentioned since, although the complaint by the applicants under Article 4 was found to be ­inadmissible, the Court re-affirmed the principles emerging from Rantsev v Cyprus and Russia.35 In 2016, the judgment in LE v Greece was delivered. The determination made at the national level that the applicant was a victim of human trafficking was not challenged by the Court. The core of the case instead concerns states’ positive obligations under Article 4,36 including the obligation to identify victims. LE v Greece will be examined in detail below. In 2017, the Court delivered two judgments under Article 4. Although J and Others v Austria37 did not raise any challenging definitional questions, as I will demonstrate below, it is of crucial importance for discussing the interrelationship between victim identification and criminal proceedings. Chowdury and Others v Greece38 was the first case in which the Court found that exploitation of irregular migrant labour amounts to forced labour. The previous cases (Siliadin v France and CN and V v France) where the Court determined that the factual circumstances amounted to forced labour involved children who provided domestic services. Chowdury is also the first case where the Court found that the victims were subjected to forced labour, but not to servitude. In terms of the obligation to identify victims, it is also an important source, although, as I will show below, there are also reasons to be critical of it. III.  THE POSITIVE OBLIGATION TO IDENTIFY VICTIMS IN THE CASE LAW

A.  Laying the Foundations: Rantsev v Cyprus and Russia As already suggested above, Rantsev was the first case that offered the Court the opportunity to extend the scope of states’ positive obligations under Article 4 forced or compulsory labour. As a matter of fact, the fundamental distinguishing feature between servitude and forced or compulsory labour within the meaning of Article 4 of the Convention lies in the victim’s feeling that their condition is permanent and that the situation is unlikely to change. It is sufficient that this feeling be based on the above-mentioned objective criteria or brought about or kept alive by those responsible for the situation.’ 33 CN v UK App No 4239/08 (ECtHR, 13 November 2012), para 80. 34 ibid para 80. 35 M and Others v Italy App No 40020/03 (ECtHR, 31 July 2012), paras 146–70. 36 LE v Greece App No 71545/12 (ECtHR, 21 January 2016). 37 J and Others v Austria App No 58216/12 (ECtHR, 17 January 2017). 38 Chowdury and Others v Greece App No 21884/15 (ECtHR, 30 March 2017).

210  Vladislava Stoyanova beyond the simple criminalisation of abusive conduct. The case was about a young woman from Russia who entered Cyprus on an artiste visa regime to work as a dancer in a club. There were reports that migrant women who entered the country on this type of visa would also work as prostitutes and were abused in this context. Soon after commencing her employment, Miss Rantsev left her employer, who subsequently found her and brought her to the police station to have her immigration status declared illegal since she had arguably breached her visa conditions. The police officers returned her back to her employer without making any further enquiries. The next morning, she was found dead under the balcony of the apartment where her employer had brought her. Both Cyprus and Russia were found to be in violation of Article 4 ECHR. In Ranstev, the Court formulated two positive obligations that are of interest here. First, it held that: In assessing whether there has been a violation of Article 4, the relevant legal and regulatory framework in place must be taken into account [references omitted]. The Court considers that the spectrum of safeguards set out in national legislation must be adequate to ensure the practical and effective protection of the rights of victims or potential victims of trafficking.39

The Court also added that: As with Article 2 and 3 of the Convention, Article 4 may, in certain circumstances, require a State to take operational measures to protect victims, or potential victims, of trafficking [references omitted]. In order for a positive obligation to take operational measures to arise in the circumstances of a particular case, it must be demonstrated that the State authorities were aware, or ought to have been aware, of circumstances giving rise to a credible suspicion that an identified individual had been, or was at real and immediate risk of being, trafficked or exploited.40

These two positive obligations (ie, the obligation to adopt relevant legal and regularly framework, and the obligation to take operational measures) have been developed and applied before in the case law.41 Notably, issues have arisen as to their distinction, to which we will turn below. At this stage, it is important

39 Rantsev (n 30) para 284. 40 ibid para 286. 41 L Lavrysen, ‘Protection by the Law: The Positive Obligation to Develop a Legal Framework to Adequately Protect the ECHR Rights’ in E Brems and Y Haeck (eds), Human Rights and Civil Rights in the 21st Century (Dordrecht, Springer, 2017) 69; V Stoyanova, ‘Causation between State Omission and Harm within the Framework of Positive Obligations under the European Convention on Human Rights’ (2018) 18(2) Human Rights Law Review 309. As to the obligation of taking operational measures to protect victims, see F Ebert and R Sijniensky, ‘Preventing Violations of the Right to Life in the European and Inter-American Human Rights Systems: From the Osman Test to a Coherent Doctrine on Risk Prevention?’ (2015) 15 Human Rights Law Review 343. See also V Stoyanova, ‘Due Diligence versus Positive Obligations: Critical Reflections on the Council of Europe Convention on Violence against Women’ in J Niemi, L Peroni and V Stoyanova (eds), International Law and Violence against Women: Europe and the Istanbul Convention (Abingdon, Routledge, 2020) 95.

Separating Protection from the Exigencies of the Criminal Law  211 to highlight that the innovation in Rantsev instead lies in their application ­specifically to Article 4. The ECtHR concluded that Cyprus had failed to adopt a legal and ­regulatory framework that could offer effective protection to Miss Rantseva. More specifically, there were deficiencies in the national immigration legislation regulating the visa regime under which she entered the country.42 In relation to the positive obligation of taking protective operational measures, the ECtHR concluded that: [T]here were sufficient indicators available to the police authorities, against the general background of trafficking issues in Cyprus, for them to have been aware of circumstances giving rise to a credible suspicion that Ms Rantseva was, or was at real and immediate risk of being, a victim of trafficking or exploitation.43

Despite these indicators, the national authorities failed to take measures to specifically protect her.44 B.  Missed Opportunity: LE v Greece The next case where the ECtHR had the chance to consider and apply these two positive obligations in the context of Article 4 was LE v Greece.45 The applicant, a Nigerian woman born in 1982, entered Greece in 2004 accompanied by KA, who had allegedly promised her that in Greece she would work in bars and nightclubs. As a result of these arrangements and once in Greece, she was told that she owed him €40,000. Upon arrival, KA confiscated her passport and forced her into prostitution for about two years. Throughout this period, she was in contact with the non-governmental organisation (NGO) Nea Zoi, which provides practical and psychological support to women who have been forced into prostitution. In the period from August 2005 to November 2006, LE was arrested three times for breaching the laws on prostitution and the laws on entry and residence of aliens in Greece. These proceedings resulted in acquittals. During the above-mentioned period, she was also issued with a deportation order; however, the deportation was suspended on the ground that it was deemed not feasible. On 29 November 2006, while still in detention pending

42 Rantsev (n 30) paras 291–93. 43 ibid para 296. 44 ibid para 298. 45 See V Stoyanova, ‘L.E. v Greece: Human Trafficking and the Scope of States’ Positive Obligations under the ECHR (2016) 3 European Human Rights Law Review 290. The case ­ of GJ v Spain (App No 59172/12, Inadmissible Decision 21 June 2016) could be considered as another missed opportunity. The application was found to be inadmissible on technical grounds. For an analysis, see V Stoyanova, ‘Protection of Victims of Human Trafficking and the ECtHR’s Inadmissibility Decision in G.J. v Spain’, 2016, http://www.asylumlawdatabase.eu/en/journal/ protection-victims-human-trafficking-and-ecthr%E2%80%99s-inadmissibility-decision-gj-v-spain.

212  Vladislava Stoyanova deportation, LE, with the support of Nea Zoi, filed a criminal complaint against KA and his partner DJ, accusing them of forcing her into prostitution. At this point, LE also claimed that she was a victim of human trafficking. The prosecutor at the Athens Criminal Court rejected the claim. The applicant requested a re-examination of her complaint, arguing that the investigation into her case was insufficient given that important evidence was disregarded. In particular, the testimony provided by Nea Zoi was not included in the record. Her request was successful, in that the prosecutor at the Athens Criminal Court was ordered to bring criminal proceedings against KA and DJ for human trafficking. Equally importantly, on 21 August 2007, the prosecutor officially recognised the status of the applicant as a victim of human trafficking and a couple of days later, the suspension of the deportation proceedings initially opened against her was approved. In the period 2008–11, criminal proceedings were conducted at the national level; however, only DJ could be arrested and prosecuted. The national court found DJ not guilty, since it was determined that DJ was not KA’s accomplice, but rather another of his victims who had been sexually exploited. The applicant complained before the ECtHR that the delay by the police to investigate her case and the deficiencies in the conduct of the investigation had led to the impossibility of prosecuting KA, who was still at large, and to the incorrect acquittal of DJ. The applicant argued that the national court did not assess the facts correctly since, as she claimed, she was forced into prostitution not only by KA, but also by DJ. She added that the initial rejection of her complaint by the prosecutor had serious consequences because she was not formally recognised as a victim of human trafficking and was not granted a special residence permit in Greece. In LE v Greece, the ECtHR reviewed how the respondent government fulfilled its obligation to adopt an appropriate legal and administrative framework and to take protective operational measures. As to the first point, the Court found no problems with the national substantive criminal law. Sadly, the applicant herself did not challenge the effectiveness of the national legislative framework in a broader sense, ie, in areas beyond criminal law.46 Such a challenge was warranted, since the story of LE exposes serious deficiencies in the national identification procedure for victims of human trafficking. More specifically, the relevant national legislation incorporated an identification procedure according to which the prosecutor can grant the formal status of a victim of human trafficking. The identification of the specific applicant was conducted exclusively in relation to the criminal proceedings against the alleged perpetrators. As occurred in the particular circumstances of LE, it was on the very day that the prosecutor at the Athens Criminal Court instituted criminal 46 However, this should not have prevented the Court from examining the effectiveness of the national regulatory framework, since the Court has reiterated that ‘it is the master of the characterization of the given in law to the facts of a case’. Söderman v Sweden [GC] App No 5786/08 (ECtHR, 12 November 2013), para 57.

Separating Protection from the Exigencies of the Criminal Law  213 proceedings against the alleged offenders for the crime of human trafficking that the applicant was formally recognised as a victim of human trafficking. Three days after this official recognition, the public prosecutor at the Athens Court of Appeal approved the suspension of her deportation. The identification was thus intimately linked to the criminal investigation. As a consequence, her initial complaint that she was forced into prostitution did not lead to formal identification; criminal proceedings had to be initiated for this purpose. This deficiency in the identification procedure was not scrutinised by the ECtHR through the lens of the positive obligation of adopting an effective regulatory framework. Thus, the Court in this case arguably succumbed to the assumption that victim identification is a facet of the criminal proceedings. One can object to my critique of the judgment by observing that the Court did find a failure in the national identification procedure. In particular, Greece was found to be in violation of its positive obligations under Article 4, since there was a nine-month gap between the point in time when the applicant filed a criminal complaint against the alleged traffickers and her formal recognition as a victim of human trafficking by the prosecutor.47 However, this failure was scrutinised in the context of the positive obligation of taking protective operational measures. At this point, it is relevant to clarify the distinction between the two positive obligations, namely to take protective operational measures, on the one hand, and to adopt an effective regulatory framework, on the other. The former concerns the requirement of personal protection of one or more identifiable individuals, who are at immediate risk of ill-treatment. The latter concerns the requirement of affording general protection and the particular applicant before the European Court might be a representative victim of the respondent state’s failure to afford such protection.48 Following this distinction, in Rantsev v Cyprus and Russia, the ECtHR found that Cyprus had failed to fulfil its obligation to put in place an appropriate legislative and administrative framework since, generally, the artiste visa regime was not only inadequate, but in fact structured the relation between employers and migrant women in a way that made the latter very vulnerable.49 Cyprus also failed to fulfil its positive obligation to take protective operational measures, since the police authorities should have been aware of ‘circumstances giving rise to a credible suspicion that Ms Rantseva was, or was at real and immediate risk

47 LE v Greece (n 36) paras 77 and 78. 48 See, for example, Mastromatteo v Italy [GC] App No 37703 (ECtHR, 24 October 2002), para 69; Eremia v The Republic of Moldova App No 3564/11 (ECtHR, 28 May 2013), para 56: ‘As recalled earlier (see paragraphs 48–52 above), the States’ positive obligations under Article 3 include, one the one hand, setting up a legislative framework aimed at preventing and punishing ill-treatment by private individuals and, on the other hand, when aware of an imminent risk of ill-treatment of an identified individual or when ill-treatment has already occurred, to apply the relevant laws in practice, thus affording protection to the victims and punishing those responsible for ill-treatment.’ 49 Rantsev (n 30) paras 290–93. See Stoyanova (n 31).

214  Vladislava Stoyanova of being, a victim of trafficking, or exploitation’.50 At the relevant time, the national authorities had concrete indicators that Miss Rantseva specifically might have been trafficked or exploited and yet they failed to take preventive measures. In LE v Greece, the nine-month delay in recognising the applicant as a victim of human trafficking is presented by the Court as a failure to protect the specific individual and not as a general structural failure of the victim identification procedure in Greece. The Court’s determination that the relevant legislation in force at the material time in Greece offered the applicant practical and effective protection51 is hard to accept – in fact, it was precisely the relevant national legislation which envisioned the particular identification procedure to which LE was subjected, and this identification procedure was contingent of the exigencies of the criminal law. As a result of this very contingency, there was a delay. In conclusion, in LE v Greece, the Court missed an important opportunity to review more generally the victim of human trafficking identification procedure in a way that could potentially distance it from the criminal law rationale so that it could better serve the purpose of protecting and assisting victims. C.  Not the Right Type of Case: J and Others v Austria In J and Others v Austria, the Court specifically engaged with the issue of the interaction between criminal proceedings and the identification of victims. The application was lodged by three women nationals of the Philippines. As they described their story, the first and the third applicant were recruited by an employment agency to work as au pairs in Dubai, the United Arab Emirates. The second applicant did not use the services of an agency; instead, she travelled to Dubai at the suggestion of the first applicant. Their passports were taken away. They were subjected to ill-treatment and exploitation by their employers. On 2 July 2010, the applicants’ employer took them along on a short holiday trip in Austria. They stayed in the same hotel as their employer, took care of the children and performed other domestic duties. Their passports remained with their employer. One or two days after their arrival in Austria, they were subjected to extreme forms of verbal abuse and threats, since one of the children went missing. The night following the incident, the applicants left the hotel with the help of a hotel employee with whom they had become acquainted. It was only in July 2011 when the applicants decided to turn to the police and filed criminal complaints against their employer. They also claimed that they were victims of human trafficking and were willing to cooperate with the police.



50 Rantsev 51 LE

(n 30) para 296. v Greece (n 36) para 72.

Separating Protection from the Exigencies of the Criminal Law  215 Criminal proceedings were initiated but subsequently discontinued, since the alleged offences had been committed abroad and not by Austrian citizens. The applicants complained to the ECtHR that Austria had failed to undertake an effective investigation into their allegations that they had been victims of human trafficking. One of the core issues concerning the procedural limb of Article 4 was whether Austria was under the obligation to investigate the crimes allegedly committed abroad. The Court determined in J and Others that Article 4 ECHR does not impose such a requirement: ‘under the Convention, there was no obligation incumbent on Austria to investigate the applicants’ recruitment in the Philippines or the alleged exploitation in the United Arab Emirates’.52 Nor were any failures found in respect of the Austrian authorities’ investigation of the harbouring and receipt of the applicants in Austria.53 The efforts by Austria in terms of investigation were found to be sufficient. Most importantly, according to the Court, in light of the fact that the police was alerted only approximately one year after the alleged events, when the employers had long left Austria, no unreasonable expectations could be raised against Austria in terms of its investigation. A valuable aspect of the Court’s reasoning in J and Others v Austria concerns the clear formulation of the positive obligation of identifying and supporting individuals as potential victims of human trafficking.54 It was easily found by the Court that Austria had not failed to fulfil this obligation for the following reasons: From the point when the applicants turned to the police, they were immediately treated as (potential) victims of human trafficking. They were interviewed by specially trained police officers, were granted residence and work permits in order to regularize their stay in Austria, and a personal data disclosure ban was imposed on the Central Register so their whereabouts were untraceable by the general public. During the domestic proceedings, the applicants were supported by the NGO LEFÖ, which is funded by the Government especially to provide assistance to victims of human trafficking.55

Crucially, the case did provide the Court with an opportunity to frame in very lucid and firm terms that Article 4 ECHR generates a positive obligation upon states to identify and support (potential) victims of trafficking and that, for this purpose, states have to build a legal and administrative framework.56 Just as significantly, the Court made it clear that the identification and the assistance of victims is independent of any criminal proceedings. While the latter are intended

52 J and Others v Austria (n 37) para 114. 53 ibid para 116. 54 The first concrete question that the Court identified in the case was ‘whether the Austrian authorities complied with their positive obligation to identify and support the applicant as ­(potential) victims of human trafficking’. See J and Others v Austria (n 37) para 109. 55 ibid paras 110 and 111. 56 ibid paras 109–11.

216  Vladislava Stoyanova to identify and potentially prosecute alleged traffickers, the former have a very different purpose (ie, identification and assistance of victims). More specifically, the Court stated the following: The applicants argued that the Austrian authorities had accepted that they were victims of the crime of human trafficking by treating them as such … However, the Court does not consider that the elements of the offence of human trafficking had been fulfilled merely because the Austrian authorities treated the applicants as (potential) victims of human trafficking … Such special treatment did not presuppose official confirmation that the offence had been established, and was independent of the authorities’ duty to investigate. Indeed, (potential) victims need support even before the offence of human trafficking is formally established, otherwise this would run counter to the whole purpose of victim protection in trafficking cases. The question of whether the elements of the crime had been fulfilled would have to have been answered in subsequent criminal proceedings. (Emphasis added)57

The applicants attempted to make an argument that since they were identified as victims of human trafficking for the purposes of assistance, this should necessarily trigger the criminal law framework; in other words, their identification and ensuring assistance should necessarily trigger the obligation upon the state to investigate events that occurred abroad.58 It is very positive that the Court rejected this type of argumentation and explicitly clarified in its reasoning that the two procedures (identification for the purposes of assistance and investigation for the purpose of identification of alleged criminals) are distinct. Therefore, in light of J and Others v Austria, Article 4 imposes an autonomous positive obligation of identifying potential victims. This identification should serve the purpose of victim protection, not the purpose of identification and prosecution of alleged perpetrators. Undoubtedly, the two procedures with their specific purposes might in certain circumstances work in harmony with each other and mutually reinforce each other.59 However, such a harmony and reinforcement should not be simply assumed. In certain circumstances, the rationales that drive the criminal proceedings might work against the victims’ interests of protection and assistance. Such a conflict arises when victim identification and assistance is dependent on the initiation and continuation of criminal proceedings. It should be also borne in mind that there might be no criminal proceedings at all since no perpetrators can be identified or, even if there are, proceedings might be terminated due to an absence of relevant evidence or due to an absence

57 ibid para 115. 58 ibid para 89. 59 For example, successful criminal proceedings might be instrumental for the victim to obtain civil remedies. Criminal proceedings might be also instrumental for a migrant victim to obtain a residence permit in light of states’ obligations under art 14 of the CoE Anti-trafficking Convention and under Council Directive 2004/81/EC of 29 April 2004 on the residence permit issued to thirdcountry nationals who are victims of trafficking in human beings or who have been the subject of an action to facilitate illegal immigration, and who cooperate with the competent authorities.

Separating Protection from the Exigencies of the Criminal Law  217 of evidence capable of proving the offence to the required standard of proof (which is demanding in the realm of criminal law). Even under these circumstances where criminal proceedings might be prevented or excluded, victims still might need to be assisted and protected, for which formal identification as a victim is crucial. In sum, the technical requirements that normally underpin the criminal law regime should not obstruct the identification and assistance of victims. Judge Pinto de Albuquerque (joined by Judge Tsotsoria) attached a Concurring Opinion to J and Others v Austria that is also worth examining, since it seems to be supportive of the applicants’ argument in favour of merging the victim identification and the criminal proceedings. The core of the Concurring Opinion can be found in the following paragraph: Allegedly, the applicants were forced to work in Austria and abroad and were trafficked for that purpose on Austrian soil. The domestic authorities disputed this fact, but nonetheless provided social support to them as if they had been victims of trafficking. This contradictory position is exemplary of the strengths and the weaknesses of the Austrian system: effective in victim protection, ineffective in punishing the perpetrators. Ultimately, the domestic authorities failed in the present case to investigate fully the denounced facts and, eventually, to bring those responsible to justice. However, the applicants also bear responsibility for this failure in view of the tardiness of their contact with the domestic authorities. Little more could be done at that time and in the specific circumstances of the case than activating the domestic and international warning notice systems. This is why I was nonetheless able to vote for the finding of no violation. (Emphasis added)60

Contrary to what Judge Pinto de Albuquerque suggests, there is nothing contradictory in the actions undertaken by the Austrian authorities. Victim identification for the purposes of providing social assistance and suspension of deportation proceedings should not be converged with the criminal proceedings; the latter have a different purpose. The dangers of convergence are clear: victims are likely to be assisted only if they are useful for the criminal proceedings. Re-assuringly, Austria has avoided this danger. As is also evident from GRETA reports, the problem in the state practice is rather the convergence of victim identification and criminal proceedings, as a consequence of which victims are not officially identified and offered assistance unless they are somehow instrumental for the criminal proceedings. Arguably, this state practice is to a certain extent excusable since, as mentioned in section I above, the text of the CoE Anti-trafficking Convention is not sufficiently robust to ensure a clear distinction between the two procedures.61 Importantly, the ECtHR reasoning in J and Others could ‘remedy’ the weaknesses of the CoE Anti-trafficking Convention and strengthen the position in favour of divergence. 60 Concurring Opinion of Judge Pinto de Albuquerque, joined by Judge Tsotsoria in J and Others v Austria (n 37) para 61. 61 Stoyanova, Human Trafficking and Slavery Reconsidered (n 1) 91.

218  Vladislava Stoyanova D.  Positive, But Not Entirely Helpful: Chowdury and Others v Greece62 Chowdury and Others v Greece is the most recent case where the ECtHR had the chance to apply the positive obligation of adopting effective legal and regulatory framework and of taking protective operational measures. The case concerned 42 Bangladeshi nationals in Greece with undocumented status. They were recruited to work on a strawberry farm in Manolada, Greece, and were promised wages of €22 for seven hours’ labour and €3 for each overtime hour. They worked in plastic greenhouses picking strawberries every day from 7 am until 7 pm in scorching heat under the supervision of armed guards. They lived in makeshift tents of cardboard boxes and nylon without running water and toilets. The workers were never paid their wages, for which they went on strike a number of times. Despite not being paid, they continued to work since they were afraid that if they were to leave, they would never be paid. After the recruitment of other migrants by the same employer, the Bangladeshi nationals again went on strike to demand their wages. At this point, one of the armed guards opened fire and seriously injured many of them. After this incident, the employer and the guards were convicted for grievous bodily harm and unlawful use of firearms (sentences that were subsequently commuted to a minimum financial penalty), but were acquitted of the charge of trafficking in human beings. The 42 Bangladeshi migrants argued before the ECtHR that they were subjected to forced labour and human trafficking, and that Greece had failed to fulfil its positive obligation under Article 4 to protect them against these abuses, to conduct effective investigation and to punish the perpetrators. The Court agreed that the factual circumstances amounted to forced labour and human trafficking.63 The Court first reviewed the positive obligation of putting in place an ­appropriate legal and regulatory framework. Very similarly to the approach in LE v Greece, the examination of this obligation did not go beyond the national criminal law framework. The tendency thus seems to be that criminal coercion is the first lens through which positive obligations under Article 4 are scrutinised. As a consequence, no issue was raised as to the relevance and the role of other national regulatory frameworks beyond criminal law, including the national regulatory framework for identifying and protecting victims of human trafficking, which, as already suggested in the context of LE v Greece, can be challenged due to its dependence on the exigencies of the criminal law. The second type of positive obligation under review in the judgment was the obligation to adopt protective operational measures. The Court emphasised that the authorities were well aware of the situation of the migrant workers in the Manolada region and of the abuses to which they were exposed, including 62 I intervened in the case by submitting a third-party-intervention on behalf of my law faculty (co-authored with Eleni Karageorgiou). 63 For a critical assessment of this qualification, see V Stoyanova, ‘Sweet Taste with Bitter Roots: Forced Labour and Chowdury v Greece’ (2018) 1 European Human Rights Law Review 67.

Separating Protection from the Exigencies of the Criminal Law  219 the refusals by the employers to pay their wages.64 Despite this awareness, the authorities’ response was limited and thus the Court concluded that ‘the operational measures taken by the authorities were not sufficient to prevent human trafficking or to protect the applicants from the treatment to which they were subjected’.65 Two comments are due in relation to this conclusion: the first one relates to the distinction between the positive obligation of taking protective operational measures, on the one hand, and the positive obligation of adopting an effective regulatory framework, on the other; and the second one concerns the standard of ‘sufficient’ measures. As to the first comment and contrary to the Court’s approach, the factual circumstances of Chowdury and Others v Greece did not call for protective operational measures. There was no immediate risk to the applicants that could trigger the obligation of taking measures targeted at specific individuals.66 Rather, the issue was one of systemic failures by the national authorities to identity and assist the migrants, not to mention that the applicants’ submissions suggested that the state was tacitly supporting the usage of migrant labour at the strawberry farms in the region and the related exploitative practices.67 Reshaping the case as one invoking the positive obligation of taking protective operational measures is not only misleading but also unhelpful, because the systemic deficiencies could not be exposed. As already suggested above, the second comment regarding the Court’s conclusion concerns the standard of ‘sufficient’ measures. The Court never indicated against what standard the sufficiency of the measures already ­ undertaken and the sufficiency of the measures that should have been undertaken was to be measured. Admittedly, this is a general problem in the Court’s case law involving states’ positive obligations68 and not a problem specific to Chowdury and Others v Greece.69 What is perhaps distinctive about the cases under Article 4 ECHR is that there is a relatively developed regulatory framework that can provide guidance as to the standards against which the sufficiency of the measures can be assessed. This framework is incorporated into the CoE ­Anti-trafficking Convention and, in particular, its provisions that requires states 64 Chowdury (n 38) paras 110–15. 65 ibid para 115. 66 For the role of the immediacy standard in the context of positive obligations, see V Stoyanova, ‘Causation between State Omission and Harm’ (n 41); see also V Stoyanova, ‘Fault, Knowledge and Risk within the Framework of Positive Obligations under the European Convention on Human Rights’ (2020) Leiden Journal of International Law (forthcoming); V Stoyanova, ‘Common Law Tort of Negligence as a Tool for Deconstructing Positive Obligations under the European Convention on Human Rights’ (2020) 24(5) International Journal of Human Rights 632. 67 Document on file with the author. 68 V Stoyanova, ‘The Disjunctive Structure of Positive Rights under the ECHR’ (2018) 87 Nordic Journal of International Law 344. 69 See, for example, the Dissenting Opinion of Judge Kalaydjieva in Söderman v Sweden (n 46), where she observed that ‘in the absence of criteria defining the required “acceptable level of protection” in specific terms, a comparison between the failed and the undefined remedies will inevitably lead to dissatisfaction “notwithstanding the respondent State’s margin of appreciation” in this area’.

220  Vladislava Stoyanova to formally identify victims of human trafficking and to provide them with specific forms of assistance.70 Very positively, in Chowdury and Others v Greece, the Court explicitly stated: [T]he member States’ positive obligations under Article 4 of the Convention must be construed in the light of the Council of Europe’s Anti-Trafficking Convention and be seen as requiring, in addition to prevention, victim protection and investigation, together with the criminalization of a criminal offence and effective prosecution of any act aimed at maintaining a person in such a situation. The Court is guided by that Convention and the manner in which it has been interpreted by GRETA.71

In its previous judgments, the Court drew from the CoE Anti-trafficking Convention to strengthen states’ positive obligations under Article 4.72 Still, Chowdury manifests one innovative feature. More specifically, the Court referred to the work of the CoE Anti-trafficking Convention’s monitoring body, GRETA, which will increase the importance of the interpretations offered by GRETA in its reports.73 Yet, the establishment of this interpretative approach at a general level is insufficient. When the Court specifically addressed the positive obligations of Greece, it did not use the CoE Anti-trafficking Convention and the provisions therein that impose a requirement upon the state to ‘adopt such legislative or other measures as may be necessary’ to identify and assist victims. To recap, the reasoning in Chowdury and Others v Greece has laudable aspects; however, there are also reasons to be critical of it. IV.  CHALLENGES AHEAD

The above review of the judgments leaves us with the following conclusions. Criminal law remains the primary framework invoked, which implies that the positive obligation to criminalise abuses falling within the scope of Article 4 and the related positive obligation to conduct effective criminal investigation will continue to shape the case law under Article 4. This is to a certain extent understandable in light of the grave form of abuses that Article 4 intends to address; this gravity can justify the recourse to criminal law.

70 Using the CoE Anti-trafficking Convention as an instrument that can guide the interpretation of the positive obligations under art 4 ECHR will be consistent with one of the interpretative principles generally endorsed by the ECtHR, namely that the provisions of the ECHR are never considered as ‘the sole framework of reference for the interpretation of the rights and freedoms enshrined therein’. Demir and Baykara v Turkey [GC] App No 34503/97 (ECtHR, 12 November 2008), para 67. 71 Chowdury (n 38) para 104. 72 Rantsev (n 30) paras 285, 287 and 296; LE v Greece (n 36) para 71. 73 For further elaboration on this point, see Stoyanova (n 63).

Separating Protection from the Exigencies of the Criminal Law  221 The full potential of the positive obligation of adopting effective regulatory frameworks that extend beyond the realm of national substantive and procedural criminal law remains to be explored and further developed. Two challenges are likely to transpire within this development. First, any future developments will have to include a better framing of the distinction between the positive obligation of adopting an effective regulatory framework and the positive obligation of taking protective operational measures.74 Contrary to the current approach in the case law, the identification of and assistance given to victims who are not exposed to an immediate risk of harm should fall within the scope of the first challenge. Identification and assistance are intended to provide general structures of protection and not only individually targeted protection to victims who are at immediate risk. Here I am not suggesting that victims who are at immediate risk should not receive specific protection tailored to the particular type and degree of risk to which they are exposed. The second challenge concerns states’ migration control interests that include detection and deportation of migrants who do not have the right to stay under the applicable national legislation. Further developments in the case law on positive obligations under Article 4 will also have to tackle any resistance that might be exercised by states’ immigration control interests. This is an area where the Court has not yet had the opportunity to venture.75 The core question here is whether the content of the positive obligation to identify and assist victims under Article 4 can include non-deportation and under what conditions. If it can, this can clearly undermine states’ migration control interests. Given the sensitivity of the issue and the leniency of the Court in this area,76 a delicate balancing of different considerations might have to be undertaken. At this juncture, it is pertinent to go back to the text of the CoE Anti-trafficking Convention, where an interesting interaction between criminal law and immigration law can be observed.77 In particular, states have agreed to restrain their migration control interests by allowing irregular migrants in relation to whom

74 It needs to be acknowledged that the distinction between these two positive obligations is generally not clear in the case law and continues to cause controversies. This has happened, for example, in the context of cases concerning domestic violence. See Separate Opinion of Judge Pinto de Albuqerque in Valiuliene v Lithuania App No 33234/07 (ECtHR, 26 March 2013). See also Talpis v Italy App No 41237/14 (ECtHR, 2 March 2017), paras 116–18. See Stoyanova, ‘Due Diligence versus Positive Obligations’ (n 41) 95. 75 However, there have been applications under art 4, in which the applicants have argued for different reasons that art 4 could be used for preventing deportations. See, for example, OGO v UK App No 13950/12, Decision 18 February 2014 (struck out of the list). The applicant in OGO argued, inter alia, that ‘her expulsion would breach Article 4 of the Convention both because it would expose her to a real risk of re-trafficking in Nigeria and because it would make it impossible for the British police to conduct an effective criminal investigation into her trafficking claim as required by the Trafficking Convention’. See also LR v UK App No 49113/09, Decision 14 June 2011 (struck out of the list); and FA v UK App No 20658/11, Decision 10 September 2013 (inadmissible). 76 V Stoyanova, ‘Populism, Exceptionality and the Rights of Migrants to Family Life under the European Convention on Human Rights’ (2018) 10(2) European Journal of Legal Studies 83. 77 Stoyanova, Human Trafficking and Slavery Reconsidered (n 1) 185.

222  Vladislava Stoyanova there are reasonable grounds to believe they are victims of human trafficking to stay (suspension of deportation) until the conclusive determination as to their victim status.78 However, the CoE Anti-trafficking Convention has left some scope for controversy as to whether the victim has to ‘pay’ for this suspension by assisting the criminal proceedings. For various reasons, states have powerful incentives to ensure effective criminal investigations into the crime of human trafficking and the related crime of human smuggling. It is not likely that effectiveness can be guaranteed without the participation of the victims and, in this sense, states need the victims. If non-deportation is entirely detached from the exigencies of the criminal law, states might be much more reluctant to restrain their migration control interests. Turning to the other side of the equation, ie, the interests of the victims, arguably, criminal proceedings are also in their interests. This seems to be assumption underlying the positive obligation to investigate alleged crimes (ie, the procedural limb of Article 4). However, it also has to be considered that the guarantee of non-deportation might be utilised to undermine the prospects of successful criminal proceedings. More specifically, an argument can be made that victims participate and testify so that they are not deported. This argument was made at the domestic level in Chowdury and Others v Greece. One of the reasons for the Greek public prosecutor rejecting the application of the migrants for the charge of human trafficking was that ‘they had made statements to the police after learning that they would receive residence permits as victims of human trafficking’.79 This problem was not addressed in the Court’s reasoning. To conclude, in any future development of the positive obligation of identifying victims that includes assistance and non-deportation, a careful balancing act might have to be achieved between the different rationales and interests underpinning immigration law, criminal law and the individuals’ protection needs.80 Such future development will have to balance the state interests in ensuring the removal of migrants who do not have the right to remain and in effective prosecution of crimes, on the one hand, and the individuals’ interests in being formally identified as victims so that they can be protected, assisted and allowed to remain on the territory, on the other.

78 Articles 10(2) and 13(1) of the CoE Anti-trafficking Convention. 79 Chowdury (n 38) para 14. 80 On the issue of balancing in the context of positive obligations under absolute rights, such as art 4(1) ECHR, see Stoyanova, Human Trafficking and Slavery Reconsidered (n 1) 279 and 324.

11 The Limitations of a Criminal Law Approach in a Transitional Justice Context BRICE DICKSON

I. INTRODUCTION

T

o say that something is illegal is, of course, to speak ambiguously. It could mean that the conduct in question is criminal, that it is a civil wrong, that it is an administrative or regulatory wrong, or that it is a breach of human rights. Lawyers see these distinctions as self-evident, but most non-lawyers do not. Students of law probably spend less than 10 per cent of their time on criminal law before acquiring a law degree, yet when nonlawyers speak of the need for justice to be done and to be seen to be done, they are usually implying that those who have done wrong should be brought to book in a criminal setting, where victims of their actions can have ‘their day in court’.1 In a transitional justice situation – where a society has been bedevilled by systematic violent conflict, whatever the basis for it – the temptation to resort to the criminal justice system when trying to cement a solution to the conflict is even greater. ‘Identify and punish the wrongdoers’ is often the mantra. But it quickly becomes apparent that resort to the criminal justice system might not be enough or might even be counter-productive.2 For a start, the criminal justice system might itself be one of the reasons why conflict broke out in the

1 On the various meanings of ‘lawful’ and ‘unlawful’, see P Harris, An Introduction to Law, 8th edn (Cambridge, Cambridge University Press, 2016) chs 2 and 9–12. 2 See generally C Turner, Violence, Law and the Impossibility of Transitional Justice (Abingdon, Routledge, 2016); C Murphy, The Conceptual Foundations of Transitional Justice (Cambridge, Cambridge University Press, 2017), esp ch 2 (‘The Problem of Transitional Justice’), where the author argues that the problems addressed by transitional justice are not the same as those addressed by criminal or retributive justice.

224  Brice Dickson first place or why it continued longer than it needed to. In addition, convicting people of crimes might be too difficult, perhaps because there were so many crimes committed, the alleged perpetrators are dead or it is very unlikely that sufficient evidence can be gathered to convince a court that a particular person committed a particular crime. Besides, criminal law is an especially individualistic species of law: it tends to hold people rather than groups or organisations to account for their involvement in crimes and usually requires that involvement to have been intentional, or at least reckless or very careless, and to have been a direct cause of the harm suffered. In terms of holding to account those who may be indirectly responsible for the harm suffered – for example, leaders of illegal paramilitary organisations, elected representatives of political parties or clerics who are supportive of an extreme religious ideology – the criminal law is a notoriously blunt instrument. Criminologists continue to argue over what causes crime and how those who commit crimes can best be deterred from repeating their actions. A few radicals believe that people should not be sent to prison for their crimes unless they are a very clear danger to a specific person or to the public at large.3 Some hold that it is not even appropriate to talk of ‘punishing’ a criminal because what he or she most needs is education, a job, a relationship and a purpose in life. In the context of transitional justice situations, these arguments are magnified. Greater creativity is called for in such scenarios because the harm that has been suffered is usually much more significant and has usually been caused not for personal gain or as a result of a loss of temper, but in the name of some larger goal, such as a political manifesto, a religious creed or an errant hate-based philosophy. At the same time, the likelihood of recidivism is likely to be less if the grievances in whose name the unlawful actions were taken have since been addressed. ‘Freedom fighting’, many would argue, should not be considered a crime at all and the law should be adapted to recognise that point.4 This chapter reflects on when it may be inappropriate to deploy the criminal law in a context where a society is moving out of serious conflict towards peaceful co-existence. It looks first at international principles concerning transitional justice before considering what aspects of criminal law may need to be tampered with in order to maintain the rule of law in a troubled society. It will draw on the experience of Northern Ireland in these respects and will use the jurisprudence of the European Court of Human Rights as a guide to when it is consistent with international human rights standards to not pursue a criminal justice agenda when seeking to hold people to account for killings. 3 See, eg, CRIO Publications Collective, Abolition Now! Ten Years of Strategy and Struggle against the Prison Industrial Complex (Chico, CA, AK Press, 2008). 4 See generally T Halliday, L Karpik and M Feeley, Fighting for Political Freedom: Comparative Studies of the Liberal Complex and Political Liberalism (Oxford, Hart Publishing, 2007). This examines 16 case studies from around the world. The first national declaration of human rights – proclaimed in France in 1789 – included ‘resistance against oppression’ as one of the four natural and imprescriptible rights of man, the others being liberty, property and safety (art 2).

The Limitations of a Criminal Law Approach  225 II.  INTERNATIONAL PRINCIPLES ON COUNTER-TERRORISM AND TRANSITIONAL JUSTICE

According to the United Nations (UN), ‘the international community has elaborated 19 international legal instruments to prevent terrorist acts’.5 An important turning point was the passing of Security Council Resolution 1373 on 28 September 2001.6 Amongst other things this called on UN Member States to ensure that terrorist acts are dealt with as serious criminal offences in domestic law. It was supplemented by three further Resolutions.7 Importantly, the UN set up a Counter-Terrorism Committee to monitor the compliance of states with these Resolutions. The Committee has produced a number of global surveys of the implementation of Resolution 1373, the most recent dating from 2015.8 Terrorism is a global phenomenon and it is therefore right and proper that the UN takes a keen interest in persuading states to deal with it effectively. However, serious internal conflicts are also a global phenomenon and almost always overlap with terrorism. The UN has not yet produced any declarations or conventions on how to manage internal conflict-resolution processes, even in situations where the violence involved has greatly diminished or has been temporarily suspended because of political talks or a mutually agreed ceasefire. The main document it has come up with is a Guidance Note issued in 2010 by then Secretary-General Ban Ki-Moon.9 Entitled United Nations Approach to Transitional Justice, the Note sets out 10 Guiding Principles on how the UN intends to approach transitional justice processes and mechanisms. While the document says that the centrality of victims in the design and implementation of transitional justice processes and mechanisms is essential,10 it adds that transitional justice measures, such as prosecution initiatives and

5 See the wording preceding the list of instruments at https://www.un.org/counterterrorism/ international-legal-instruments. 6 See https://undocs.org/S/RES/1373(2001). 7 Resolution 1456 (2003) required states to ensure that their counter-terrorism measures complied with their obligations under international law, in particular, international human rights law, refugee law and humanitarian law: see https://undocs.org/S/RES/1456(2003). Resolution 1566 (2004) defined for the first time what the UN means by terrorism, namely ‘criminal acts, including against civilians, committed with the intent to cause death or serious bodily injury, or taking of hostages, with the purpose to provoke a state of terror in the general public or in a group of persons or particular persons, intimidate a population or compel a government or an international organisation to do or to abstain from doing any act’: see https://undocs.org/S/RES/1566(2004), para 3. Resolution 1624 (2005) required states to adopt measures prohibiting incitement to terrorist action and denying safe haven to persons credibly suspected of being involved in such conduct: see https://undocs.org/S/ RES/1624(2005). 8 Global Survey of the Implementation of Security Council Resolution 1373 (2001) by Member States, https://www.un.org/sc/ctc/wp-content/uploads/2016/10/Global-Implementation-Survey-1373_EN.pdf. 9 See https://www.un.org/ruleoflaw/files/TJ_Guidance_Note_March_2010FINAL.pdf. This builds upon the Report by Secretary-General Kofi Annan in 2004 entitled The Rule of Law and Transitional Justice in Conflict and Post-conflict Societies, S/2004/616. 10 ibid Guiding Principle 6.

226  Brice Dickson institutional reform, should be interdependent with broader efforts aimed at strengthening the administration of justice and security within national systems while paying due regard to any indigenous and informal traditions concerning administering justice or settling disputes which exist there.11 The Guidance Note does not otherwise suggest particular ways in which national criminal justice systems should be altered in order to provide transitional justice. Nor does it specify exactly how the centrality of victims in the design of transitional justice mechanisms is to be realised in practice. It leaves open the possibility that victims’ wishes, after having been expressed and considered, can be overridden because there are even more important considerations to take into account. III.  TAMPERING WITH THE CRIMINAL LAW IN A CONFLICTED SOCIETY

Tampering with the criminal law, broadly defined, is almost always essential if conflict resolution processes, of which transitional justice is usually a part, are to succeed. If we take Northern Ireland as an example of a jurisdiction still going through a conflict resolution process, we can see many instances of how the ‘normal’ criminal justice system has been tampered with in order to support and develop the conflict resolution agenda. Alterations have occurred alongside more general reforms that would in any event have been deemed essential to keep pace with modern thinking on matters such as forensic science, human rights and penal policy. The general reforms have been wide-ranging and are intended to be long term in nature. The parallel tampering has been narrowly focused and intended to be temporary. This is a common pattern discernible in other transitional justice programmes playing out in conflict resolution processes around the world, South Africa and Colombia being just two examples.12 Even at the very start of the so-called ‘troubles’ in Northern Ireland in the late 1960s, steps were taken to try to end the violence by creating ‘exceptions’ to the normal criminal law. In May 1969, the government of Northern Ireland announced that there would be an amnesty concerning all offences connected with demonstrations occurring since 5 October 1968.13 A scheme was put in place to allow for unlawfully held weapons to be handed in without fear of prosecutions being initiated, but ‘the results were very discouraging’.14 For a 11 ibid Guiding Principle 7. 12 Space constraints do not permit the development of this point. For recent contributions to the vast literature on the South African and Colombian conflict resolution processes, see F Venter, ‘The Limits of Transformation in South Africa’s Constitutional Democracy’ (2018) 34 South African Journal on Human Rights 143; G Anders and O Zenker (eds), Transition and Justice: Negotiating the Terms of New Beginnings in Africa (Oxford, Wiley Blackwell, 2015) chs 4, 5 and 6 (on police violence, ‘toilet wars’ and land restitution); AL Sánchez-Mejía, Victims’ Rights in Flux: Criminal Justice Reform in Colombia (Dordrecht, Springer, 2017); R Suttner, Recovering Democracy in South Africa (Johannesburg, Jacana Media (Pty) Ltd, 2015). 13 Irish Times, 7 May 1969, 1. 14 ‘Freeland warns of arms possession’, Irish Times, 23 October 1969, 4.

The Limitations of a Criminal Law Approach  227 short period, the police in Northern Ireland, who unlike officers in Great Britain had always carried guns, were disarmed,15 but this policy was quickly reversed because the threat against police officers was still very real. From 1971 to 1975, a system of internment was deployed, whereby people could be arrested and detained without trial merely because they were considered to be security risks. In the early 1970s, a huge exception was made to the rule that all serious crimes had to be tried by a judge and jury: instead, judge-only courts were established, nicknamed ‘Diplock courts’ after the senior British judge who had recommended their creation, to try people accused of ‘scheduled offences’, ie, those connected with the troubles. As a quid pro quo, defendants in such cases were, unlike their counterparts in jury trials, given the right to a reasoned judgment from the trial judge as well as an appeal to a higher court without first needing to seek permission for the appeal. For some offences, especially possession offences, the burden and standard of proof were altered so that, for example, defendants had to show on the balance of probabilities that their possession of weapons was innocent, rather than the prosecution having to show beyond reasonable doubt that it was for a criminal purpose. The right of defendants to rely on their silence in the face of police questioning was also qualified. On the other hand, statements made by defendants were admissible as evidence only if the police could show beyond reasonable doubt that they were not the product of ill-treatment. Persons suspected of terrorist offences were, for a while, given more rights while in custody than persons suspected of other offences: in 1987, they acquired the right to have someone informed of their detention and the right of access to legal advice,16 rights that were not conferred on non-terrorist suspects until 1990.17 The literature on whether these various changes to the criminal justice system were justified and effective is vast.18 The point being stressed here is simply that adjustments were made because of the need to reduce the serious violence that was occurring all over Northern Ireland. To this day, opinions vary as to which of the adjustments worked and which did not. Regular reviews of the need for the adjustments were conducted by independent reviewers and parliamentarians, but some reviews were not as thorough and sincere as they should have been and very few adjustments were removed. When a peace settlement was eventually agreed in 1998 in the form of the Belfast (Good Friday) Agreement, there were further compromises with an

15 This was in line with recommendation 16 in the Report of the Advisory Committee on Police in Northern Ireland, 1969, Cmd 535, chaired by Lord Hunt. 16 Northern Ireland (Emergency Provisions) Act 1987, ss 14 and 15. 17 Police and Criminal Evidence (NI) Order 1989, arts 57 and 59. 18 See, eg, A Jennings (ed), Justice under Fire: The Abuse of Civil Liberties in Northern Ireland (London, Pluto Press, 1988); JD Jackson and S Doran, Judge without Jury: Diplock Courts in the Adversary System (Oxford, Clarendon Press, 1995); B Dickson, ‘Miscarriages of Justice in Northern Ireland’ in C Walker and K Starmer (eds), Miscarriages of Justice: A Review of Justice in Error (London, Blackstone, 1999) 287–303.

228  Brice Dickson orthodox approach to the criminal law. Most significant of all was the agreement between the British and Irish governments to allow for the accelerated release of prisoners convicted of scheduled offences, provided the prisoners were affiliated to organisations that were maintaining ‘a complete and unequivocal ceasefire’.19 Under this scheme, any qualifying prisoners who remained in custody two years after its commencement were released at that point, and anyone subsequently convicted of a troubles-related offence committed prior to the Good Friday Agreement was to serve no more than two years in prison.20 While technically not an ‘amnesty’, as it applied to people already convicted of crimes,21 this early release scheme was difficult for many voters to swallow when they were asked to endorse the 1998 Agreement in a referendum. Some people baulked at the fact that just because the motivation of these prisoners had been political rather than, say, sexual, financial or personal, they were being released back into the community years before their due release date, despite having committed horrific crimes. Moreover, these prisoners were promised support prior to and after release, including assistance with employment opportunities, re-skilling and further education. Such support was not always as readily available to so-called ‘ordinary decent criminals’ upon their release. Despite reservations about the prisoner release scheme, 71.1 per cent of voters in Northern Ireland endorsed the package in a referendum and in a simultaneous referendum in the Republic of Ireland, 94.4 per cent endorsed it.22 The Belfast (Good Friday) Agreement also led to changes to policing ­structures, the prosecution system, the Prison Service and services for victims.23 A Police Ombudsman – to conduct completely independent investigations of complaints against the police – was appointed in 2000 as a result of a report compiled in 1997.24 A particularly controversial ‘exception’ was made to the criminal law in 2005 when, in place of enacting legislation granting immunity 19 Belfast (Good Friday) Agreement, section on ‘Prisoners’, para 2. 20 Northern Ireland (Sentences) Act 1998, ss 3–16. See also ‘Maze Emptied as Terrorist Prisoners Walk Free’, The Guardian, 28 July 2000, https://www.theguardian.com/uk/2000/jul/28/northernireland, which reveals that within 22 months of the Good Friday Agreement, 428 terrorists, including 143 serving life sentences, had been released. 21 For a conspectus on what amnesty laws reveal about the struggle against impunity for crimes, see L Mallinder, ‘Atrocity, Accountability, and Amnesty in a “Post-human Rights World’, Political Settlements Research Programme Working Paper Series, 09/2017, https://ssrn.com/abstract=3051142. 22 These referendums occurred on 22 May 1998. 23 See A New Beginning: Policing in Northern Ireland (September 1999), the report of the Independent Commission on Policing for Northern Ireland, chaired by Chris Patten; the press release accompanying the publication of the Criminal Justice Review in March 2000: https://cain.ulster. ac.uk/issues/law/cjr/press30300.htm; Michelle Butler, ‘The Northern Ireland Prison Reform Programme: Progress Made and Challenges Remaining’ (2017), https://pure.qub.ac.uk/portal/ files/122986874/KESS_Policy_Briefing_NI_Penal_Reform_Programme.pdf; and the Victims and Survivors Services (www.victimsservice.org/about-us/what-we-do) and the Commission for Victims and Survivors (https://www.cvsni.org). 24 M Hayes, A Police Ombudsman for Northern Ireland? A Review of the Police Complaints System in Northern Ireland. See also Police (NI) Act 1998, ss 50–65; and Police (NI) Act 2000, ss 62–66.

The Limitations of a Criminal Law Approach  229 from prosecution to members of unlawful paramilitary organisations who were ‘on the run’ for crimes allegedly committed during the conflict, the UK government issued ‘letters of comfort’ to suspects assuring them that they were no longer wanted by the police anywhere in the UK. Although the scheme was not secret, it did not hit the headlines until 2014, when the trial of an alleged ex-member of the IRA, John Downey, was stopped because the judge deemed it an abuse of process given that Mr Downey had previously received one of the letters of comfort.25 A subsequent inquiry concluded that while the scheme had been legal, it had been allowed to evolve without any reference to a proper structure or policy, leading to considerable scope for error.26 The British government later announced that recipients of the letters should no longer rely upon them.27 Mr Downey himself has since been extradited from the Republic of Ireland to Northern Ireland to face charges relating to the murder of two British soldiers in 1972.28 Notwithstanding the reforms already adopted,29 there remain features of Northern Ireland’s criminal justice system which apply specifically to the stillsimmering conflict.30 The police retain a small number of special powers based on fortnightly authorisations issued by a senior officer.31 There are still juryless courts to try people for offences relating to the conflict.32 In recent months, suggestions have been made that British soldiers who may have killed people during the conflict should be granted immunity from prosecution unless there are exceptional circumstances.33

25 R v Downey (21 February 2014), https://www.judiciary.uk/wp-content/uploads/JCO/ Documents/Judgments/r-v-downey-abuse-judgment.pdf. 26 The (Hallett) Independent Review into the On the Runs Administrative Scheme, HC 380 (2014). In some other cases, the prerogative of mercy was used to ‘exonerate’ convicted terrorists. 27 See https://www.gov.uk/government/speeches/government-implementation-of-the-hallett-report. 28 See https://www.bbc.co.uk/news/uk-northern-ireland-50022543. 29 These are comprehensively analysed in A-M McAlinden and C Dwyer (eds), Criminal Justice in Transition: The Northern Ireland Context (Oxford, Hart Publishing, 2015). 30 Since the announcement of the Belfast (Good Friday) Agreement on 10 April 1998, there have been 160 deaths in Northern Ireland connected to the conflict that was supposedly ended. See P Nolan, ‘Lyra McKee Became 160th Person Murdered since Good Friday Agreement … They All Deserve to Be Remembered’, Belfast Telegraph, 2 May 2019, https://www.belfasttelegraph.co.uk/ opinion/news-analysis/paul-nolan-lyra-mckee-became-160th-person-murdered-since-good-fridayagreement-they-all-deserve-to-be-remembered-38070512.html. 31 Justice and Security (NI) Act 2007, ss 21–42. 32 ibid ss 1–9. Even today, British soldiers accused of troubles-related crimes are tried in such juryless courts: In the Matter of an Application by Dennis Hutchings for Judicial Review [2019] UKSC 26. 33 See the debate amongst UK MPs held on 20 May 2019, available at: https://hansard. parliament.uk/commons/2019-05-20/debates/06752875-A995-4952-A1C8-89E342B72FA6/ ImmunityForSoldiers. A video of the debate is also available at: https://www.youtube.com/ watch?v=jNV5MWbkMoQ. See also ‘Investigations into Fatalities in Northern Ireland Involving British Military Personnel’, 7th Report of the House of Commons Select Committee on Defence, 2016–17, HC 1064, and the government’s response to that report: 3rd Special Report of the Committee, 2017–19, HC 549.

230  Brice Dickson IV.  RESPECT FOR HUMAN RIGHTS AND FOR THE PRINCIPLE OF LEGALITY

When exceptions are made to the ordinary criminal law in an attempt to deal with a particularly challenging problem such as terrorism or internal conflicts, efforts must be made to ensure that in doing so, human rights are not abused. The Security Council’s Resolution 1456 (2003), already mentioned, re-affirmed that point. Throughout the conflict in Northern Ireland, the standards against which alleged assessments of human rights could be measured were not just those applicable within domestic law, but also those applicable under the European Convention on Human Rights (ECHR). Although the UK was prominent in the drafting of that treaty and was the first country to ratify it in 1952, it did not accept the right of individual petition to the European Court of Human Rights until 1966 and it did not allow Convention rights to be pleaded in domestic courts until the Human Rights Act 1998 came into force on 2 October 2000. Elsewhere I have attempted to evaluate the role played by the ECHR during the conflict in Northern Ireland, concluding that it had little impact either on the levels of violence manifested by illegal paramilitary organisations or on the nature of the responses to that violence by state bodies.34 By the time the European Court began to issue judgments condemning some of the state’s actions as violations of Convention rights, the violence was largely over and the peace process was under way. Some ‘exceptions’ to the criminal justice system in Northern Ireland received very little attention from the European Court of Human Rights – the juryless ‘Diplock’ courts is a good example – and most of those that were examined were not condemned by the Court as violations of human rights. These included the use of internment,35 the use of special police and army powers to stop, question, arrest and detain,36 and the use of force against alleged terrorist suspects.37 Many changes to the criminal justice system were protected against challenge because they were introduced in line with the ECHR’s provision permitting derogation from the Convention.38 Remarkably, in no case did the Court (or the European Commission of Human Rights, which functioned until 1998) find that the UK had violated the substantive right to life by allowing more force to be used than was necessary. Substantive violations were found only as regards the planning and control of operations.39 Violations of the procedural rights 34 See B Dickson, The European Convention on Human Rights and the Conflict in Northern Ireland (Oxford, Oxford University Press, 2010) passim. 35 Ireland v UK (1979–80) 2 EHRR 25. 36 See, eg, O’Hara v UK (2002) 34 EHRR 812 (on arrest powers); Brannigan and McBride v UK (1994) 17 EHRR 539 (on detention powers), 37 Stewart v UK (1985) 7 EHRR 453; Kelly (John) v UK (1993) 16 EHRR CD 20 (a decision by the European Commission of Human Rights that an application was inadmissible). 38 Article 15. 39 Most notably in McCann v UK (1996) 21 EHRR 97, where in 1995 the Court found by 10:9 that the UK’s security services had violated art 2 by not properly planning an arrest operation targeting IRA operatives in Gibraltar in 1988.

The Limitations of a Criminal Law Approach  231 which the Court has said are guaranteed by those articles were also established, but only late in the day.40 Today problems persist regarding how best to deal with what occurred during the conflict in Northern Ireland. In early 2019 there were still 1,425 deaths for which no one had yet been held to account.41 The central issue is to what extent the normal operation of the criminal justice system should be put to one side for the sake of a greater good, namely the reconciliation of the divided communities. It is the same issue which faced the Constitutional Court of South Africa in 1996, when relatives of Steve Biko, founder of the Black Consciousness movement, together with other victims of security force brutality, challenged the legislative provision42 which permitted a Committee to grant an amnesty to perpetrators of unlawful acts (in Biko’s case, his murder).43 They argued that an amnesty would be a violation of their constitutional right to have disputes settled by a court of law or another independent forum. The Constitutional Court accepted that the amnesty breached that right, but then ruled that the breach was justifiable because it was permitted by the country’s new Interim Constitution, which stressed the need for national unity and reconciliation. Amnesties, the Court said, were a crucial component of the negotiated peace settlement in South Africa, without which the Interim Constitution itself would not have come into being.44 No exactly comparable case has come before the courts of Northern Ireland or the UK concerning provisions in the Belfast (Good Friday) Agreement. The closest is perhaps Re Williamson’s Application, where a woman whose parents had been killed by a bomb planted by the IRA in 1993 applied for judicial review of the decision taken in 2000 by Mo Mowlam, the then Secretary of State for Northern Ireland, to release one of the bombers under the statutory early release scheme. The applicant argued that the IRA was not maintaining ‘a complete and unequivocal ceasefire’ and therefore its prisoners should no longer qualify for release under the scheme.45 She pointed to the fact that in 1999, the IRA had allegedly been involved in a murder and the importation of

40 The seminal case is McKerr v UK (2002) 34 EHRR 20, where in 2001 the Court ruled by 7:0 that a death caused by police officers had not been properly investigated for the purposes of art 2. The Court came to the same conclusion in several concurrent and subsequent cases, most recently in Hemsworth v UK App No 58559/09, Judgment of 16 July 2013. See also McDonnell v UK App No 19563/11, Judgment of 9 December 2014, a case about the death of a prisoner in Northern Ireland. 41 Report of the PSNI Chief Constable to the NI Policing Board, 7 February 2019, https://www.psni. police.uk/news/Latest-News/07022019-chief-constables-report-northern-ireland-policing-board. 42 Promotion of National Unity and Reconciliation Act 34 of 1995, s 20(7). 43 AZAPO v President of the Republic of South Africa 1996 (4) SALR 671 (CC). See the impressive student thesis on this case by KL Martin, ‘Tackling the Question of Legitimacy in Transitional Justice: Steve Biko and the Post-apartheid Reconciliation Process in South Africa’ (2015), available at: https://repository.upenn.edu/cgi/viewcontent.cgi?article=1225&context=curej. 44 AZAPO (n 42) 698. 45 See Northern Ireland (Sentences) Act 1998, s 3(4), (8) and (9).

232  Brice Dickson weapons from the US. Both the first instance judge and the Court of Appeal held that their role was not to second-guess the correctness of the Secretary of State’s decision, but merely to rule on whether she had gone through the correct process in considering what decision to make. They acknowledged the political dimension to the decision in question, with Carswell LCJ saying: It is part of the democratic process that such decisions should be taken by a minister responsible to Parliament, and so long as the manner in which they are taken is in accordance with the proper principles the courts should not and will not step outside their proper function of review.46

It was therefore lawful for the Secretary of State ‘to assess whether there had been a “systemic”, or root and branch, breakdown in the ceasefire and to take into account whether the events [in 1999] were evidence of a co-ordinated return to violence or represented something less than that’.47 This approach can be seen as an example of ‘the principle of legality’ at work: an action, or inaction, is lawful provided it is supportable by existing legal rules and standards. This means that, once it is within the parameters of the lawful, a judgement as to what to prioritise can appropriately assess political expediency as being more important than victims’ interests. The long and the short of it is that if societies which have suffered a long bout of conflict are to move forward in a harmonious and progressive fashion, they need to have mechanisms in place for putting the past behind them. This does not mean that they must strive to forget the past; rather, it entails trying to learn from the past so that the violence which occurred is unlikely to break out again. In the application of human rights standards, some compromises may have to be made to allow the society to move on. Thankfully, this is not as difficult to square with international human rights law as might be imagined, partly because human rights treaties tend to permit rights to be qualified but also because, while most treaties are clear on what rights need to be protected, they are not prescriptive about the remedies that should be available if that protection fails.48 It is worth exploring this last point in more detail. Do relatives of persons unlawfully killed have a right to see people prosecuted for those killings? V.  HUMAN RIGHTS AND THE CRIMINAL LAW

The purpose of criminal law is to allow the state to punish someone for breaking its rules. The process is not victim-driven, since that is the purpose of civil law. Standard textbooks on criminal law have little if anything to say about

46 Re Williamson’s Application [2007] NICA 7, 16. 47 ibid 14. 48 Article 13 ECHR simply provides that everyone whose rights are violated must have ‘an effective remedy before a national authority’.

The Limitations of a Criminal Law Approach  233 human rights standards. It is only in textbooks on the processing of alleged criminals – their arrest, detention, trial and punishment – that we find mention of human rights standards. The principal reason for this is that in all countries, the main categories of criminal offences were agreed long before human rights standards were developed. Moreover, criminal law serves purposes beyond just protecting human rights. Criminal law is society’s compact with itself whereby punishment is meted out when society deems it appropriate. A lot is done to ensure that the punishment processes are based on society’s interests. In the UK, for example, one of the two tests for whether a person should be prosecuted for a crime is whether doing so would be in the public interest, the other being whether there is ‘a realistic prospect of conviction’, although the standard required for eventual conviction is proof ‘beyond reasonable doubt’.49 The Sentencing Guidelines in England and Wales usually require a judge to take into account the accused’s degree of culpability (high, medium or lesser), the level of harm he or she has caused (greater or lesser), aggravating factors (such as trying to blame others for the crime) and factors reflecting personal mitigation (such as steps already taken to address the offending behaviour).50 Where the offence has led to loss or damage, the judge must also consider whether to require the convicted person to pay compensation.51 When a court is determining not how someone should be punished for a crime, but how a violation of human rights should be remedied, the questions it asks itself are very different. It pays much less attention to retribution and much more to compensation and the prevention of recurrences. The primary purpose of human rights law is to vindicate acknowledged rights through effective remedies, not to ensure that those who violated those rights are dealt with through the criminal justice system.52 The crucial question in the context of transitional justice is then: does the victim ever have the right to have someone held to account under the criminal law for the violation in question? To date, as we shall see, the European Court of Human Rights has been reluctant to recognise such a right, even if the right violated was the right to life. This is despite the fact that the Court has greatly expanded the category of situations in which states can be held to have violated the right to life. We will briefly consider that expansion before examining situations in which the Court has not insisted on criminal prosecutions for such violations.

49 See the Code for Crown Prosecutors, available at: https://www.cps.gov.uk/publication/ code-crown-prosecutors. 50 See, eg, the 185 sets of guidelines relating to specific crimes tried in the Crown Court, available at: https://www.sentencingcouncil.org.uk/crown-court. 51 Powers of Criminal Courts (Sentencing) Act 2000, s 130, as amended. 52 ‘The Court does not act as a court of appeal in relation to national courts; it does not rehear cases, it cannot quash, vary or revise their decisions’: European Court of Human Rights, ‘Questions and Answers’, https://www.echr.coe.int/Documents/Questions_Answers_ENG.pdf, 12.

234  Brice Dickson VI.  EXPANDING ACCOUNTABILITY FOR VIOLATIONS OF THE RIGHT TO LIFE

It is now more than 20 years since the European Court of Human Rights read into Article 2 three state duties which are not immediately apparent on the face of the provision: (1) the duty to take reasonable and proportionate measures to avert loss of life occurring through criminal conduct;53 (2) the duty to properly plan and control operations where state forces may need to use lethal force;54 and (3) the duty to thoroughly investigate a killing after it has occurred.55 There are two instances of the European Court recently embellishing the first and third of these duties. The first instance is Tagayeva v Russia,56 where the Court held for the first time that a state had violated Article 2 because it had not properly averted a foreseeable terrorist attack, which resulted in many fatalities. The case involved large-scale hostage-taking by Chechen terrorists in a school in Beslan, North Ossetia, over a three-day period in 2004. In order to rescue the hostages, Russian forces stormed the school, leading to hundreds of children being killed or injured. Citing what it had previously said in Osman v UK57 and Mastromatteo v Italy,58 but applying those statements to the particular context of counterterrorist activities, the Court found that the intelligence available to the Russian authorities prior to the attack was very specific and that adequate measures therefore ought to have been taken to minimise the known risk. Instead, a large group of illegally armed rebels had been able to access and occupy the school without confronting any preventative security arrangements.59 The Court also found that the investigation into the events had failed to adequately examine the use of lethal force by the state agents involved and that in Russia there appeared to be widespread immunity in respect of harm caused during counter-terrorist operations, thereby leaving a dangerous gap in the regulatory framework for such situations.60 53 Osman v UK (2000) 29 EHRR 245, Judgment of 28 October 1998. The case involved the police’s alleged failure to prevent a stalker from murdering two people. 54 McCann v UK (n 38) paras 150 and 201–14. See also B Dickson, ‘The Planning and Control of Operations Involving the Use of Lethal Force’ in L Early et al (eds), The Right to Life under Article 2 of the European Convention on Human Rights (Oisterwijk, Wolf Legal Publishers, 2016) 47–59. 55 McCann v UK (n 38) para 161; see also McKerr v UK (n 39), which is discussed below at nn 81, 85 and 91. For the European Court’s own summary of the current state of the relevant law (as of 31 August 2019), see its Guide on Article 2 of the European Convention on Human Rights, available at: https://www.echr.coe.int/Documents/Guide_Art_2_ENG.pdf; the procedural obligations are dealt with at 29–42. 56 Tagayeva v Russia App Nos 26562/07 and 6 others, Judgment of 13 April 2017. The report of this case runs to almost 90,000 words. 57 Osman v UK (n 52) para 116. 58 Mastromatteo v Italy App No 37703/97, Judgment of 24 October 2002 (GC), para 69. The case involved a man killed by prisoners who robbed a bank while they were out of prison on leave. 59 Tagayeva v Russia (n 55) paras 481–93. 60 ibid paras 592–99.

The Limitations of a Criminal Law Approach  235 The second example is Šilih v Slovenia,61 where the Court’s Grand Chamber confirmed that the procedural obligation to carry out an effective investigation under Article 2 ‘has evolved into a separate and autonomous duty’ and can now be considered as ‘a detachable obligation’ even if the death took place before the date when the state in question became bound by the ECHR (the so-called ‘critical date’).62 One of two conditions needs to be satisfied before such a detachable duty arises: either (a) the death must have occurred no longer than 10 years before the critical date and much of the investigation must have taken place, or ought to have taken place, after the critical date; or (b) the need to ensure the real and effective protection of the guarantees and underlying values of the ECHR must be sufficient to give rise to a connection between the death and the critical date.63 The second of these alternative conditions is very vague and, despite the Grand Chamber’s amplification of it in Janowiec v Russia,64 the vagueness continues to this day. In R (Keyu) v Secretary of State for Foreign Affairs, the UK Supreme Court held that the ‘underlying values’ test could not apply if the deaths in question occurred before the ECHR itself was adopted on 4 November 1950.65 One of the judges, Lady Hale, held that the investigative duty did not apply at all in the case before her because the claimants were seeking historical truth rather than civil or criminal liability.66 This point is important because it reminds us that ‘truth’ processes are different from ‘justice’ processes: many relatives would prefer to be told exactly what happened to their loved one rather than seek legal remedies through long drawn-out processes that tend to be confrontational as well as aleatory.67 The expansion of the scope of Article 2 has required many states to develop their law – but not necessarily their criminal law – to ensure that lethal incidents are prevented and investigated in ways that should re-assure the public that the chances of such incidents occurring again are small. The expansion does not mean that states are under a duty to prosecute all those who may be criminally liable under domestic law for the incidents in question.

61 Šilih v Slovenia (2009) 49 EHRR 37, Judgment of 9 April 2009. 62 ibid para 159. 63 ibid paras 161–63. 64 Janowiec v Russia (2014) 58 EHRR 30, Judgment of 21 October 2013. The case involved the massacre of more than 4,000 Polish soldiers in the Katyn forest in 1943. By 13:4, the Court found no violation by Russia of art 2 and by 12:5 no violation of art 3. 65 R (Keyu) v Secretary of State for Foreign Affairs [2015] UKSC 69, [2016] AC 1355, [88] per Lord Neuberger, [256]–[258] per Lord Kerr and [301] per Lady Hale. The case involved multiple deaths caused by British soldiers in Malaya in 1948. Four of the judges took the ‘critical date’ in this case to be 1966 (when the UK first allowed the right of individual petition to Strasbourg), while Lady Hale took it to be 1953, when the UK’s ratification of the ECHR meant that other Council of Europe States could lodge complaints against it to the European Commission of Human Rights. 66 ibid. Lady Hale, alone amongst the judges, did however hold that the government’s decision to refuse a new inquiry into the deaths was irrational or disproportionate. 67 See also Murphy (n 2).

236  Brice Dickson VII.  INVESTIGATIONS NOT FOLLOWED BY PROSECUTIONS

When the European Court first asserted in McCann v UK that investigations of alleged violations of Article 2 had to take place, it was in the context of undercover British soldiers having shot dead three members of the IRA in Gibraltar in 1988.68 The Grand Chamber put it thus: [A] general legal prohibition of arbitrary killing by the agents of the State would be ineffective, in practice, if there existed no procedure for reviewing the lawfulness of the use of lethal force by State authorities. The obligation to protect the right to life under [Article 2] … requires by implication that there should be some form of effective official investigation when individuals have been killed as a result of the use of force by, inter alios, agents of the State.69

Earlier, the Court had cited Article 9 of the UN’s Principles on the Effective Prevention and Investigation of Extra-legal, Arbitrary and Summary Executions (1989),70 which provide that: There shall be a thorough, prompt and impartial investigation of all suspected cases of extra-legal, arbitrary and summary executions, including cases where complaints by relatives or other reliable reports suggest unnatural death in the above circumstances.

As is its normal practice regarding international law, whether hard or soft, the Court did not directly apply the UN Principles, but it clearly borrowed from them and effectively read Article 9 into the ECHR. The UN Principles provide a lot more detail about investigating extra-judicial executions. They do not specify that a state is under a duty to prosecute persons suspected of such killings, but they do say that: Governments shall ensure that persons identified by the investigation as having participated in extra-legal, arbitrary or summary executions in any territory under their jurisdiction are brought to justice. (Emphasis added)71

The UN’s Basic Principles on the Use of Force and Firearms by Law Enforcement Officials, issued a year later (1990), are more specific: Governments shall ensure that arbitrary or abusive use of force and firearms by law enforcement officials is punished as a criminal offence under their law. (Emphasis added)72

This was cited by the Grand Chamber of the European Court in Nachova v Bulgaria,73 where there was a ruling that Article 2 had been violated because of 68 McCann v UK (n 38). 69 ibid para 161. On the facts of this case, the Court found that there had been an adequate investigation into the killings. 70 Adopted on 24 May 1989 by Economic and Social Council Resolution 1989/65. 71 Article 18. 72 UN, Basic Principles on the Use of Force and Firearms by Law Enforcement Officials (1990), para 7. 73 Nachova v Bulgaria (2006) 42 EHRR 43, Judgment of 6 July 2005.

The Limitations of a Criminal Law Approach  237 the lack of an effective investigation into the fatal shooting of two Roma men by military police in 1996. The European Court has so far stopped short of explicitly stating that if an investigation discloses evidence that a particular person may have killed ­someone in violation of Article 2 of the ECHR, that person must in all cases be prosecuted. However, it has come extremely close to doing so. In Öneryildiz v Turkey,74 a case where 39 people were killed in an explosion and landslide at a former rubbish tip in Istanbul, the Grand Chamber said: [W]hen lives have been lost as a result of events occurring under the responsibility of the public authorities, which are often the only entities to have sufficient relevant knowledge to identify and establish the complex phenomena that might have caused such incidents [and] … [w]here it is established that the negligence attributable to State officials or bodies on that account goes beyond an error of judgment or carelessness, in that the authorities in question, fully realising the likely consequences and disregarding the powers vested in them, failed to take measures that were necessary and sufficient to avert the risks inherent in a dangerous activity … the fact that those responsible for endangering life have not been charged with a criminal offence or prosecuted may amount to a violation of Article 2, irrespective of any other types of remedy which individuals may exercise on their own initiative. (Emphasis added)75

However, the Grand Chamber added that: It should in no way be inferred from the foregoing that Article 2 may entail the right for an applicant to have third parties prosecuted or sentenced for a criminal offence … or an absolute obligation for all prosecutions to result in conviction, or indeed in a particular sentence … On the other hand, the national courts should not under any circumstances be prepared to allow life-endangering offences to go unpunished. This is essential for maintaining public confidence and ensuring adherence to the rule of law and for preventing any appearance of tolerance of or collusion in unlawful acts. (Emphasis added)76

Here the Grand Chamber appears to want to have its cake and eat it. On the one hand, it requires all life-endangering offences to be punished, while on the other hand, it affirms that applicants have no right to have anyone prosecuted or sentenced for those offences. It does not say how punishment is to occur without such prosecutions and sentences. In the later case of Aydan v Turkey, the Court ruled that if a member of the security forces carelessly discharges his or her weapon, it is a violation of Article 2 if that person is prosecuted and found to have behaved in that fashion, but is then not punished in any way.77 It has also ruled in Jelić v Croatia, a case concerning the killing of a Serbian man by armed and masked Croatians in 1991, 74 Öneryildiz v Turkey App No 48939/99, Judgment of 30 November 2004. 75 ibid para 93 (references omitted). 76 ibid para 96 (references omitted). By ‘third parties’, the Grand Chamber presumably means persons other than representatives of the state. 77 Aydan v Turkey App No 16281/10, Judgment of 12 March 2013 (available only in French).

238  Brice Dickson that there is a violation of the investigative duty in Article 2 if the authorities fail to follow up credible leads relating to the identification of the alleged killers and as a result do not initiate any prosecution.78 Moreover, the Court said it was no excuse for Croatia to argue that at the time in question, its authorities were faced with the duty to investigate multiple killings arising from the war in their country.79 Nor, importantly, was it an excuse that a senior police officer had already been convicted of war crimes against civilians in Croatia. On that point, referring to provisions in the Statutes creating various international criminal courts,80 the European Court said: The punishment of superiors for the failure to take necessary and reasonable measures to prevent or punish war crimes committed by their subordinates cannot exonerate the latter from their own criminal responsibility.81

The Court repeated that the essential purpose of an Article 2-compliant investigation is ‘to secure the effective implementation of the domestic laws which protect the right to life and, in those cases involving State agents or bodies, to ensure their accountability for deaths occurring under their responsibility’.82 In two other cases concerning deaths occurring during major conflicts, the European Court has found that the investigations undertaken following the death were adequate to comply with Article 2’s stringent standards, even though no one was brought to justice as a result. In Palić v Bosnia and Herzegovina, the Court was faced with an application from a widow who claimed that there had not been a proper investigation into the disappearance and death of her husband at the hands of Serbian forces during the war in Bosnia and Herzegovina in 1995. On the facts, the Court held that there had been no violation of Article 2.83 It noted that the investigation had finally led, after 14 years, to the identification of the remains of Mr Palić and to the issuing of international arrest warrants

78 Jelić v Croatia App No 57856/11, Judgment of 12 June 2014. 79 According to the Annual Report of the European Court of Human Rights for 2014 (at 96), Croatia had by then ‘opened investigations into 3,436 alleged perpetrators of war crimes against a background of 13,749 reported victims of war’. 80 Statute of the International Criminal Court, art 25; Statute of the International Criminal ­Tribunal for Rwanda, art 6; Statute of the International Criminal Tribunal for the Former ­Yugoslavia, art 7. 81 Jelić v Croatia (n 77) para 88, cross-referring to, amongst others, art 25 of the Statute of the International Criminal Court. 82 ibid para 73, citing, amongst other cases, Jordan v UK (2003) 37 EHRR 2, paras 105–09 (a case decided alongside McKerr v UK (n 39). It is unclear why the Court in this instance limited accountability for deaths to cases involving state agents or bodies. In relation to the investigative duty under art 3 ECHR, it is abundantly clear that the duty to investigate effectively applies to all allegations of torture or inhuman or degrading treatment, regardless of who is the alleged perpetrator. This is the interpretation placed on ECHR law by the UK Supreme Court in Commissioner of Police of the Metropolis v DSD [2018] UKSC 11, [2019] AC 196, eg, at [16]–[43] and [59]–[62] per Lord Kerr (with whom Lord Neuberger and Lady Hale agreed). 83 Palić v Bosnia and Herzegovina App No 4704/04, Judgment of 15 February 2011.

The Limitations of a Criminal Law Approach  239 against two suspects in Serbia, whom Serbia could not extradite because they were Serbian nationals. The Court again asserted that: In these circumstances … the domestic criminal investigation was effective in the sense that it was capable of leading to the identification and punishment of those responsible for the disappearance and death of Mr Palić, notwithstanding the fact that there have not yet been any convictions in this connection. The procedural obligation under Article 2 is indeed not an obligation of result, but of means.84

Similarly, in Gürtekin v Cyprus, the European Court declared inadmissible applications from the relatives of three men who had disappeared during intercommunal conflicts in Cyprus in 1963–64 and whose bodies were unearthed between 2006 and 2009 during exhumation activities conducted by the UN’s Committee on Missing Persons.85 Again, the relatives complained of a lack of effective investigation, but the Court concluded as follows: The Court can understand that it must be frustrating for the applicants that potential suspects have been named and, in two instances, located and questioned but that no further steps apparently were going to be taken. However, Article 2 cannot be interpreted so as to impose a requirement on the authorities to launch a prosecution irrespective of the evidence which is available … Insofar as the applicants argued that, at the very least, the decision that the evidence was insufficient to justify a prosecution should have been submitted for decision by a court, the Court does not consider that the procedural obligation in Article 2 necessarily requires that there should be judicial review of investigative decisions as such … it is not for the Court to micro-manage the functioning of, and procedures applied in, criminal investigative and justice systems in Contracting States which may well vary in their approach and policies. No one model can be imposed.86

In this context, one could also cite Armani Da Silva v UK, where the Grand Chamber was faced with an applicant who claimed that there had been a violation of Article 2 in that no police officer had been prosecuted in relation to the fatal shooting of her cousin on the London Underground in 2005.87 The killing had been a case of mistaken identity and the only ‘penalty’ suffered by the police was a fine for breaching health and safety laws. The Court nevertheless held that Article 2 had not been violated. In doing so, it explained that its previous position, which was that an investigation should be capable of leading to the identification and punishment of those responsible, needed to be attenuated: the obligation to punish applies only if this is ‘appropriate’.88 It pointed out that to

84 ibid para 65, citing in support of the last statement Jordan v UK (2003) 37 EHRR 2, Judgment of 4 May 2001, para 107. 85 Gürtekin v Cyprus App Nos 60441/13, 68206/13 and 68667/13, decision of 11 March 2014. 86 ibid paras 27 and 28, citing, in support of the last statement, McKerr v UK (n 39) para 143. 87 Armani Da Silva v UK (2016) 63 EHRR 12, Judgment of 30 March 2016. 88 Here the Grand Chamber cited its own previous judgment in Giuliani and Gaggio v Italy App No 23458/02.

240  Brice Dickson date, the Court had never found a violation of Article 2 in a situation where there had been a decision not to prosecute in the wake of an investigation that was compliant with Article 2. Only ‘institutional deficiencies’ in criminal justice or prosecution systems had led to a finding of violation. It went on to rule that one of the prosecution tests used in England and Wales (that the evidence showed there was a ‘realistic prospect of conviction’) fell within the state’s margin of appreciation and so there was no ‘institutional deficiency’ amounting to a violation of Article 2. Institutional changes recommended by the Independent Police Complaints Commission had been implemented, the fine levied for the breach of health and safety laws was not manifestly too low, the victim’s family had been adequately involved in the investigation and ex gratia sums had been paid promptly to them. It is worth noting that the second prosecution test used in England and Wales (that prosecution must be in the public interest) was not scrutinised by the European Court in Da Silva because it comes into play only if the ‘realistic prospect of conviction’ test is satisfied. One imagines that it too, prima facie, would fall within the margin of appreciation doctrine, though this does not mean that in particular cases, the European Court could not conclude that an assessment that a prosecution would not be in the public interest was a violation of Article 2. VIII.  MAINTAINING EFFECTIVENESS, INDEPENDENCE, CO-OPERATION AND RESPECT

Notwithstanding the three examples just given of Palić, Gürtekin and Da Silva, where the lack of prosecution was held to be justifiable, the European Court continues to stress that effective independent investigations must always still occur. Thus, even though under Šilih v Slovenia89 the lapse of a long period of time between the occurrence of a death and the state in which it occurred becoming bound by the ECHR might exonerate a state from having to comply with its investigative duty, it is still the case that the duty can be ‘revived’ if new information comes to light suggesting that an earlier investigation was not effective. This much is clear from the European Court’s judgment in another case from Northern Ireland, Brecknell v UK.90 The same point was stressed in Harrison v UK, where the European Court was considering an application from relatives of the 96 people who died in a crush at Hillsborough football stadium in England in 1989.91 The Court ruled that the UK government was responding satisfactorily to the new information which had emerged and that the relatives’ 89 Šilih v Slovenia (n 60). 90 Brecknell v UK (2008) 46 EHRR 42, Judgment of 27 November 2007, para 66. The investigative duty did revive in that case, but the Court held that it had been fulfilled, except in relation to an early part of the investigation conducted by the Royal Ulster Constabulary (see also the text at n 95 below). 91 Harrison v UK App Nos 44301/13, 44379/13 and 44384/13, decision of 25 March 2014.

The Limitations of a Criminal Law Approach  241 application alleging a breach of Article 2 was premature because the state’s new investigation was still incomplete. Whether an investigation into a death is sufficiently independent in a postconflict situation can also be problematic, given that some of the investigators may have links to organisations allegedly involved in the death in question. Unless the criminal justice system in question has been radically reformed since the end of the conflict, there is the possibility that defects in that system which became apparent during the conflict will continue to undermine the objectiveness, thoroughness and reliability of post-conflict investigations. Moreover, it has to be remembered that the European Court in McKerr v UK said that ‘independence’ means ‘not only a lack of hierarchical or institutional connection but also a practical independence’.92 The Grand Chamber of the European Court has confirmed in Tunç v Turkey93 that the sufficiency of an investigation in this regard is highly fact-sensitive: The adequacy of the degree of independence is assessed in the light of all the circumstances, which are necessarily specific to each case … Where an issue arises concerning the independence and impartiality of an investigation, the correct approach consists in examining whether and to what extent the disputed circumstance has compromised the investigation’s effectiveness and its ability to shed light on the circumstances of the death and to punish those responsible … In this regard, the Court considers it appropriate to specify that compliance with the procedural requirement of Article 2 is assessed on the basis of several essential parameters: the adequacy of the investigative measures, the promptness of the investigation, the involvement of the deceased person’s family and the independence of the investigation. These elements are interrelated and each of them, taken separately, does not amount to an end in itself … They are criteria which, taken jointly, enable the degree of effectiveness of the investigation to be assessed.94

The Court went on to cite a series of cases which fell on either side of the line95 and then held that in the case before it, the investigation was sufficiently independent. The question of independence has been the subject of considerable attention in cases arising within Northern Ireland, mostly in relation to whether the Police Service of Northern Ireland (PSNI), which was formed in 2001, is sufficiently independent of its predecessor body, the Royal Ulster Constabulary (RUC). Everyone who was in the RUC on the day that the PSNI was formed automatically became a member of the PSNI, provided they had sworn a new oath of allegiance. In Brecknell v UK, the European Court therefore suggested that the PSNI was independent of the RUC,96 but in later cases in the domestic



92 McKerr

v UK (n 39) para 112. v Turkey App No 24014/05, Judgment of 14 April 2015. 94 ibid paras 223–25. 95 ibid paras 227–31, citing five separate cases against Romania and two against Turkey. 96 Brecknell v UK (n 89). 93 Tunç

242  Brice Dickson courts, the generality of that suggestion has been successfully challenged.97 In its first decision on the matter, the Court of Appeal declared that the PSNI’s Legacy Investigations Branch had not yet demonstrated the practical independence required for it to be able to carry out an Article 2-compliant investigation into the death in question.98 In its second decision, the Court of Appeal held that the brother of the deceased had a procedural legitimate expectation under the common law that an overarching report would be compiled by an independent police team into the activities of the gang which committed the killing, but that there was no enforceable duty in that regard under Article 2 because of the passage of time since the death in 1976. Tunç v Turkey99 was not cited in either of those cases, but the same fact-sensitive approach was adopted. The Court of Appeal in McQuillan did cite Hackett v UK, an application to the European Court which was declared inadmissible as manifestly ill-founded, given that there was no appearance of a violation of Article 2 in the appointment of a PSNI officer alongside an officer from outside Northern Ireland to produce an independent report on whether new allegations of police collusion justified the re-opening of an earlier investigation into a death.100 All three dimensions of the required independence were satisfied in that case. There are other decisions by the European Court which suggest that it still retains a keenness to facilitate prosecutions in order to allow the loved ones of a deceased person to have the people responsible for the killing brought to justice. A startling example is the recent decision in Romeo Castaño v Belgium, where a Chamber ruled unanimously that Belgium had violated Article 2 by not extraditing to Spain a suspected Euskadi Ta Askatsuna (ETA) terrorist who was wanted for the murder of a senior army officer there 38 years earlier.101 The violation flowed from the fact that in the Court’s view, Belgium had not adequately cooperated with Spain in its attempt to prosecute someone for the murder. Earlier in 2019, the existence of such a duty had been confirmed by the Grand Chamber of the European Court in Güzelyurtlu v Cyprus and Turkey,102 where Turkey was found to have violated Article 2 by not adequately cooperating with Cyprus in relation to the investigation of three murders committed in Cyprus by suspects now living in the ‘Turkish Republic of Northern Cyprus’. In Romeo Castaño, the Court extended the 97 See, eg, McQuillan’s (Margaret) Application [2019] NICA 13 (the Court of Appeal’s seven-point summary of its conclusions as to the legal principles governing the requirement of ‘independence’ is at para 176); Barnard’s (Edward) Application [2019] NICA 38. 98 McQuillan’s (Margaret) Application (n 96) para 213. The PSNI is seeking to appeal this decision to the UK Supreme Court. 99 Tunç v Turkey (n 92). 100 Hackett v UK App No 34698/04, decision of 10 May 2005. 101 Romeo Castaño v Belgium App No 8351/17, Judgment of 9 July 2019 (available in French but not yet in English). See the comment published on 10 September 2019 by M Pinto at https:// strasbourgobservers.com/2019/09/10/romeo-castano-meticulously-elaborated-interpretations-forthe-sake-of-prosecution. One of Pinto’s conclusions is that, by finding Belgium in breach of the art 2 rights of the family of the deceased, the European Court may have lessened the art 3 rights of the accused person (because of the way she might be treated when sent back to Spain). 102 Güzelyurtlu v Cyprus and Turkey App No 36925/07, Judgment of 29 January 2019, para 178.

The Limitations of a Criminal Law Approach  243 duty to cooperate to a situation where what was alleged was not a failure by the respondent state to investigate a crime committed in another state, but a failure to extradite someone to that other state to stand trial there for the crime.103 The Court was at pains to point out that the finding of a violation of Article 2 on Belgium’s part did not necessarily imply that Belgium had an obligation to extradite the suspect to Spain: the violation was based on the inadequate factual basis relied upon by Belgium when refusing to extradite the suspect, but that did not replace Belgium’s obligation to ensure that, if extradited, the suspect would not run the risk of being treated in a way that violates Article 3 ECHR.104 In a similar vein, the Grand Chamber of the European Court has been reluctant to accept that a state can grant an amnesty to persons who have committed ‘grave breaches of human rights’. In Marguš v Croatia,105 another case arising out of the Balkan wars in the early 1990s, a former commander of the Croatian army had been convicted in 2007 of war crimes committed against civilians in 1991. He argued that he had been tried twice for the same crimes because in 1997, proceedings against him had been withdrawn in compliance with a General Amnesty Act 1996. But the European Court held that the ne bis in idem principle did not apply here and that, more generally: A growing tendency in international law is to see such amnesties as unacceptable because they are incompatible with the unanimously recognised obligation of States to prosecute and punish grave breaches of fundamental human rights. Even if it were to be accepted that amnesties are possible where there are some particular circumstances, such as a reconciliation process and/or a form of compensation to the victims, the amnesty granted to the applicant in the instant case would still not be acceptable since there is nothing to indicate that there were any such circumstances.106

The Grand Chamber issued this opinion having cited a wide range of international law sources on amnesties.107 Its statement strongly implies that if, for the purpose of promoting reconciliation and the ongoing peace process in Northern Ireland, a law were to be passed in the UK granting an amnesty for all killings carried out in connection with the troubles in Northern Ireland before, say, 1998, the European Court might hold that law to be a sufficient justification for the state’s refusal to prosecute and/or punish someone for a killing, even though there had not been a thorough and effective investigation of that killing. This would be even likelier given that a law has now been passed awarding a special ‘victim’s payment’ to victims of the conflict in Northern Ireland.108 However, the Court might not so hold in a situation where the killing was part of ‘a grave breach of fundamental rights’, a term which the European Court has chosen not

103 Romeo

Castaño v Belgium (n 100) para 80. para 92. 105 Marguš v Croatia (2016) 62 EHRR 17, Judgment of 27 May 2014. 106 ibid para 130. 107 The citations fill 26 pages of the pdf version of the Court’s judgment. 108 See the Victims’ Payments Regulations 2020, SI 103. 104 ibid

244  Brice Dickson to define. The four Geneva Conventions of 1949, dealing with wartime situations, all include ‘wilful killing’ within the definition of grave breaches of those Conventions,109 but that may not be the stance preferred in international human rights law. As regards the obligation to punish a killer once he or she has been found guilty of homicide, the European Court of Human Rights is usually content to leave the degree of punishment to the discretion of the state in question, but it will intervene if the punishment is disproportionately high or low, given the crime committed. Thus, in Enukidze and Girgvliani v Georgia,110 the Court found a violation of Article 2 when four senior officials from the Ministry of the Interior were released from prison after serving just three-and-a-half years for the brutal murder of a young man. This ‘unreasonable leniency deprived the criminal prosecution of the four officers of any remedial effect under Article 2 of the Convention’.111 In May 2020 judgment was issued in the comparable case of Makuchyan v Azerbaijan and Hungary, where a convicted murderer serving a life sentence in Hungary was transferred to his home state of Azerbaijan to continue his sentence, but was immediately released and lauded there after receiving a presidential pardon.112 The Court found no violation by Hungary but a violation by Azerbaijan of the procedural obligation in Article 2. IX. CONCLUSION

In summary, it is clear that in a transitional justice context, the European Court of Human Rights does not require prosecutions to ensue in all cases where there has been an Article 2-compliant investigation into a death. Even a very thorough investigation may not produce enough evidence to justify the launching of criminal proceedings against one or more alleged perpetrators, and in cases where such evidence does arguably exist, there may be other public interest reasons for not going down the prosecution channel. A limitation period may apply, even in relation to charges of murder.113 In a transitional justice context, those public interest considerations could include factors which do not arise in an ‘ordinary’ criminal justice context.114 Societal reconciliation and healing may

109 Geneva Convention No 1, art 50; Geneva Convention No 2, art 51; Geneva Convention No 3, art 130; Geneva Convention No 4, art 147. 110 Enukidze and Girgvliani v Georgia App No 25091/07, Judgment of 26 April 2001. 111 ibid para 275. 112 App No 17247/13, Judgment of 26 May 2020. 113 eg, in Poland, art 101§1(1) of the Penal Code prevents the crime of murder from being punishable after 30 years. I am grateful to Dr Marek Martyniszyn for this reference. 114 Within ‘ordinary’ criminal justice contexts, there may be particularly strong reasons for state prosecutions to be initiated even if the victim of a crime – such as domestic violence – does not support that course of action; see, eg, Volodina v Russia, App No 41261/17, Judgment of 9 July 2019, paras 92–111.

The Limitations of a Criminal Law Approach  245 be greater priorities than pursuing prosecutions. I would suggest that if this is true for the conflict in Northern Ireland, it may also be true for other extant conflicts within the Council of Europe area, such as those in Turkey, Georgia, Ukraine and Nagorno-Karabakh. Whether there has been an adequate attempt to bring people to justice will largely depend on the particular circumstances of each case. It is possible, as hinted at in Marguš v Croatia,115 that the European Court will still require some minimum ‘satisfaction’ to be accorded to victims in such situations. This may require one or more schemes to be put in place at the national level for such purposes as facilitating truth-telling outside of the criminal justice process, allowing access to special compensation funds established for categories of victims associated with the conflict, or promoting reconciliation amongst different communities whose interests were supposedly represented by protagonists actively involved in the conflict. In this manner, the priority accorded to ‘extraordinary’ public interest considerations could be tempered by an affirmation that the conflict nevertheless created victims of human rights abuses who at the very least deserve official recognition and acknowledgement of their suffering. Apart from deaths occurring as a result of ‘grave breaches of fundamental rights’, the state may be able to dispense with a prosecution so long as it takes all reasonable steps to identify the perpetrators and does not in some way collude in covering up what happened. As time passes, the likelihood of a successful prosecution will in any event diminish, given the decreasing accuracy of witnesses’ memories and the degrading of forensic evidence. In addition, if the country as a whole might be held back from implementing an agreed peace process if it were to spend a large amount of time and money on planning for prosecutions that are rarely if ever going to succeed, the European Court might well hold that it is within a state’s margin of appreciation to limit by law the initiation of prosecutions. It appears to be within that margin for states to reduce the punishments awarded to persons found guilty after such prosecutions, so it is a small step from there to allowing states not to prosecute in the first place. As academic commentators have noted, even in situations where there have been ‘gross human rights violations’, victims may be more appreciative of an award of reparations for their suffering than of a ruling that someone must be p ­ rosecuted for what was done to so badly harm them.116

115 Marguš v Croatia (n 104). 116 See, eg, L Moffett, ‘Transitional Justice and Reparations: Remedying the Past?’ in C Lawther, L Moffett and D Jabobs (eds), Research Handbook on Transitional Justice (Cheltenham, Edward Elgar, 2017) ch 19. It is also possible for criminal trials themselves to have beneficial reparative effects on victims; see, eg, RF Leyus, The Reparative Effects of Human Rights Trials: Lessons from Argentina (Abingdon, Routledge, 2017). See also Alina Balta, ch 4 in this volume.

246

Part IV

Uncharted Waters for the ECtHR’s Coercive Duties Doctrine

248

12 Preventive Obligations, Risk and Coercive Overreach LIORA LAZARUS

I

n 2012, I argued that the term ‘positive obligations’ in a criminal justice context masked a ‘coercive sting’. Consequently, in the context where protective duties required the state to coerce private individuals in order to protect the rights of those at risk of harm, I suggested that a more appropriate descriptor would be the term ‘coercive duties’. I called for more attention to the prospect of ‘coercive overreach’ in the framing of these duties and of how they might be received in an operational context within a broader politics of security, a factor that I believed was receiving insufficient attention from human rights scholars.1 Tulkens was similarly concerned at the time to scrutinise the ‘paradoxical relationship between criminal law and human rights’, arguing that ‘human rights have both a defensive and offensive role, a role both neutralizing and triggering the criminal law’. She warned that ‘it appears the reserved and essentially negative obligation to use the criminal law only as a last resort is being replaced, at least in part, by a positive obligation to favour it and to implement it in practical terms’.2 Over the last eight years, scrutiny of the rise of positive obligations has intensified and a rich critical literature has emerged on the relationship between positive obligations and criminal law.3 As reflected in this volume, there is now 1 L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce’ in L Zedner and J Roberts (eds), Principles and Values in Criminal Law and Criminal Justice: Essays in Honour of Andrew Ashworth (Oxford, Oxford University Press, 2012) 136. A few scholarly voices were an exception to this rule at the time: P van Kempen, Repressie door mensenrechten (Nijmegen, Wolf Legal Publishers, 2008); Françoise Tulkens, ‘The Paradoxical Relationship between Criminal Law and Human Rights’ (2011) Journal of International Criminal Justice 9, 577. 2 Tulkens (n 1) 594. 3 Many of the leading scholars in this terrain are contributors to this volume: L Lavrysen, Human Rights in a Positive State (Cambridge, Intersentia 2017), N Mavronicola, ‘Taking Life and Liberty Seriously: Reconsidering Criminal Liability under Article 2 of the ECHR’ (2017) 80(6) Modern Law Review 1026; V Stoyanova, Human Trafficking and Slavery Reconsidered: Conceptual Limits and States’ Positive Obligations in European Law (Cambridge, Cambridge University Press, 2017); M Pinto, ‘Awakening the Leviathan through Human Rights Law: How Human Rights Bodies Trigger the Application of Criminal Law’ (2018) 34 Utrecht Journal of International and European

250  Liora Lazarus considerable awareness of a potential turn to ‘criminal fetishism’ flowing from the recognition of ‘coercive duties’,4 while at the same time a nuanced account of where courts may well be guilty of ‘coercive underreach’.5 However, this chapter will not look at the turn to criminal law in the sense of criminalisation, but will turn its focus to the scope of discretion afforded under preventive operational obligations, which raise similar challenges. While analysing exemplary judgments to demonstrate problems with the scope of these obligations, the chapter will also outline two case studies. The first will draw on empirical research into how preventive operational obligations developed under Article 2 of the European Convention on Human Rights (ECHR) have shaped policing practice in the UK, and will explore the use of ‘Osman warnings’. The second will look at how preventive obligations developed under Article 4 ECHR were used to legitimate expansive risk-averse measures in the drafting of the Modern Slavery Act in the UK. The focus here is relevant to coercive overreach as it reflects on the way in which human rights might shape the culture of law enforcement within a criminal context, or the responses of the state in extremely violent conditions. To this extent, the chapter has a bearing on the debate on criminal law enforcement, because of its implications for the practice of policing, law enforcement and the shaping of operational discretion, rather than the shaping of the criminal law itself. Moreover, it has implications beyond the realm of criminal law enforcement as it also applies to conditions of non-international armed conflict and the specialised area of terrorism prevention. I.  PREVENTIVE OPERATIONAL OBLIGATIONS

As is well reported in the literature,6 Osman v UK is the starting point for a discussion on the scope of operational preventive obligations.7 Osman established both Law 161. See also A Ashworth, Positive Obligations in Criminal Law (Oxford, Hart Publishing, 2013); K Engle, ‘Anti-impunity and the Turn to Criminal Law in Human Rights’ (2015) 100(5) Cornell Law Review 1069; K Kamber, Prosecuting Human Rights Offences (Leiden, Brill, 2017); A Huneeus, “International Criminal Law by Other Means: The Quasi-criminal Jurisdiction of the Human Rights Courts’ (2013) 107 American Journal of International Law 1; FC Ebert and RI Sijniensky, ‘Preventing Violations of the Right to Life in the European and Inter-American Human Rights Systems: From the Osman Test to a Coherent Doctrine on Risk Prevention?’ (2015) 15 Human Rights Law Review 343. It should also be noted that earlier works remain important in this field, in particular C Droege, Positive Verpflichtungen der Staaten in der Europäischen Menschenrechtskonvention (Dordrecht, Springer, 2003); H Cullen, ‘Siliadin v France: Positive Obligations under Article 4 of the European Convention on Human Rights’ (2006) 6(3) Human Rights Law Review 585; P van Kempen, Repressie door mensenrechten (Nijmegen, Wolf Legal Publishers, 2008); J Allain, ‘Rantsev v Cyprus and Russia: The European Court of Human Rights and Trafficking as Slavery’ (2010) 10(3) Human Rights Law Review 546; A Seibert-Fohr, Prosecuting Serious Human Rights Violations (Oxford, Oxford University Press, 2009). 4 See Mattia Pinto, ch 8 in this volume. 5 Mavronicola (n 3). 6 Ebert and Sijniensky (n 3). 7 Osman v UK (2000) 29 EHRR 245.

Preventive Obligations, Risk and Coercive Overreach  251 general and specific obligations relating to the right to life. The general obligations related to the systematic obligations to establish an ‘effective’ criminal law and criminal ‘law enforcement machinery’ capable of preventing, suppressing and sanctioning breaches of the right to life.8 In addition, Osman established specific obligations ‘to take preventive operational measures to protect an individual whose life is at risk from the criminal acts of another’.9 In order for these specific obligations to arise in a way that does not impose a ‘disproportionate burden on the authorities’, it ‘must be established … that the authorities knew or ought to have known at the time of the existence of a real and immediate risk to the life of an identified individual or individuals from the criminal acts of a third party and that they failed to take measures within the scope of their powers which, judged reasonably, might have been expected to avoid that risk’.10 Much has been written on the operational obligation flowing from Osman. Ebert and Sijniensky argue that the adoption of the test within ECtHR as well as the Inter-American Court of Human Rights jurisprudence has given rise to ‘conceptual confusion’, whereby a set of ‘different sub-tests’ are lumped together under one label. They also argue that the individualised focus of the test negates broader sets of ‘structural risks’ that also need to be addressed by states.11 Stoyanova argues that insufficient attention has been paid to questions of causation, and the specific causal nexus between ‘harm and state omissions within the framework of positive obligations’.12 She is critical of the Court’s tendency to merge ‘issues of knowledge, reasonableness and causation in its assessment of state responsibility’ in respect of operational obligations.13 Lavrysen is more optimistic about the flexibility of the causation doctrine under the ECtHR jurisprudence, arguing that the ‘open ended fair balance test’ is ‘sufficiently flexible to accommodate any policy considerations in favour of restricting State liability’.14 While these emerging debates on omissions and causation and their bearing on the scope of the state’s specific operational obligations are highly instructive, they do not address the question of how the operational preventive obligation might be constrained by the rights of those who are potential objects of coercive 8 ibid para 115. 9 ibid. 10 ibid para 116. 11 Ebert and Sijniensky (n 3) 365–67. It is welcome that a recent dissenting judgment of Judge Pinto de Albuquerque (in Volodina v Russia App No 41261/17, 9 July 2019) acknowledged the need to tailor the Osman test of ‘real and immediate risk’ to ‘awareness’ of the broader structural risks to women in domestic violence cases. This broadening of foreseeability of risk will also be discussed further below. See also the Concurring Opinion of Judge Pinto de Albuquerque in V ­ aliuliene v ­Lithuania App No 33234/07, 26 March 2013; C Heri, ‘Volodina, Art 3, and Russia’s Systemic ­Problem Regarding Domestic Violence’, Strasbourg Observers, 30 July 2019. 12 V Stoyanova, ‘Causation between State Omission and Harm within the Framework of Positive Obligations under the European Convention on Human Rights’ (2018) 18 Human Rights Law Review 309. 13 ibid 345. 14 L Lavrysen, ‘Causation and Positive Obligations under the European Convention on Human Rights: A Reply to Vladislava Stoyanova’ (2018) 18 Human Rights Law Review 705, 717.

252  Liora Lazarus obligations, such as those who threaten harm. This slightly different question is also part of the Osman test on the scope of positive obligations, as the Court noted ‘the need to ensure that the police exercise their powers to control and prevent crime in a manner that fully respects the due process and other guarantees which legitimately place restraints on the scope of their action to investigate crime and bring offenders to justice, including the guarantees under Articles 5 and 8 of the Convention’.15 What is clear from this dictum is that the balance between a negative right and a positive operational obligation is part of determining the scope of the latter, also implying that the scope of the negative rights of the potential object of the obligation must be defined in order for the scope of the positive obligation to be resolved. I argued some time ago that such a balancing process was underdeveloped in Convention and UK case law, in contrast to the balancing processes in jurisprudence under domestic jurisdictions such as South Africa and Germany.16 One of the reasons we should be concerned about this lack of coherent balancing is that the operational context of policing, law enforcement or state action in violent conditions is sometimes highly charged. Indeed, as I will show later in this chapter, state agents tasked with the maintenance of security are already institutionally oriented towards a practice of risk aversion and operate in an ambient security politics of risk intolerance. They are also often operating on a thin and partial understanding of human rights law. Indeed, as will be discussed further below, in-depth qualitative work by Richard Martin on the way in which human rights influence policing operations shows clear evidence that low-ranking officers are most knowledgeable in relation to human rights where the Article 2 obligations require them to act in protection of life.17 This is a crucial element of the application of human rights obligations in practice, which remains underexamined by courts and scholars. The in-built tendency within criminal justice and military institutions to satisfy a ‘voracious security ideal’ is why I have advocated for consistent vigilance in the construction of coercive obligations (as well as the shaping of the right to security).18 In this environment, the way in which courts shape, weigh and balance countervailing rights considerations, as well as the rhetorical framing of these obligations, will play a part in the state’s understanding of its general operational response to coercive obligations claims. 15 Osman v UK (n 7) 116. 16 L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce’ in L Zadner and J Roberts (eds), Principles and Values in Criminal Law and Criminal Justice: Essays in Honour of Andrew Ashworth (Oxford, Oxford University Press, 2012) 136. 17 R Martin, Policing Human Rights (Oxford, Oxford University Press, forthcoming) ch 4 (on file with the author). This is a concern to which I will return in section III below. 18 L Lazarus, ‘Insecurity and Human Rights’ in D Akande, J Kuosmanen and D Roser (eds), Human Rights and 21st Century Challenges: Poverty, Conflict and the Environment (Oxford, Oxford University Press, 2019); L Lazarus, ‘The Right to Security – Securing Rights or Securitizing Rights’ in Dickinson et al (eds), Examining Critical Perspectives on Human Rights (Cambridge, Cambridge University Press, 2012).

Preventive Obligations, Risk and Coercive Overreach  253 At a theoretical level, the relationship between the scope of a positive obligation and the limit of a negative right is not straightforward. Moreover, as Mavronicola crisply demonstrates, the interplay between rights and obligations is even more complex when the positive obligation arises as a consequence of an absolute right.19 But the complexity of this task cannot be a reason to bypass its requirements. Just as the proportionality exercise requires the balancing of interests, so the shaping of a positive obligation ought to be defined by similar questions of weight and reconciliation. Yet, as I argued previously, courts in a variety of jurisdictions are sometimes inclined to bypass this balancing requirement or to assume at the outset that the positive obligation takes priority over any countervailing negative rights.20 It is my view that a failure to undertake a rigorous evaluation of these countervailing interests gives rise to significant potential for coercive overreach by states in the determination by courts of the scope of the preventive obligations. As Pinto also argues in this volume on the ECtHR, courts ‘send an ambiguous message to domestic criminal justice systems, which are required to enforce coercive human rights obligations while guaranteeing due process rights … It is likely that this mixed message leads to undesirable outcomes’.21 Consequently, the interplay between opaquely reasoned positive obligations and how these obligations are understood and applied in security practice requires serious examination. II.  RISK AVERSION AND PREVENTIVE OBLIGATIONS

Given the ambient operational and political environment in which declarations of coercive obligations are received, the failure by courts to delineate and re-assert the existence of countervailing rights and principles when defining positive obligations raises concerns. There are few signs that the ambient securitised political environment in which these judicial claims are made is showing signs of abating.22 Moreover, there are growing indications that the strict temporality requirements of the Osman foreseeability test are being elasticated. Indeed, Judge Pinto de Albuquerque has suggested that the ‘real and immediate risk’ test in Osman is unsuited to the context of broader structural risk such as domestic violence. His dissenting judgment in Volodina v Russia is an excellent example

19 N Mavronicola, ‘What is an “Absolute Right”? Deciphering Absoluteness in the Context of Article 3 of the European Convention on Human Rights’ (2012) 12(4) Human Rights Law Review 723. 20 Lazarus (n 1). 21 See Pinto, ch 8 in this volume. 22 L Lazarus and BJ Goold, ‘Security and Human Rights: Finding a Language of Resilience and Inclusion’ in B Goold and L Lazarus (eds), Security and Human Rights (Oxford, Hart Publishing 2019); Robert Skidelsky, ‘The Terrorism Paradox’, Project Syndicate 20 January 2020, https://www. project-syndicate.org/commentary/western-terrorism-fear-inversely-proportional-to-risk-by-robertskidelsky-2020-01?barrier=accesspaylog.

254  Liora Lazarus of a gender-sensitive reading of the Convention and of ECtHR jurisprudence. Consequently, he suggests moving away from the tighter ‘real and immediate risk’ test to one based in general awareness of gender violence. The crucial part of his judgment is a call for a move away from strict foreseeability of risk to one grounded in a ‘wider window of time’: [T]he Osman test fails to achieve its purpose if taken word for word. A ‘real and immediate risk’ in the context of domestic violence infers that the risk, namely the batterer, is already in the direct vicinity of the victim and about to strike the first blow. Were the test to be applied in such a manner, two concerns arise. Firstly, any protective action offered by the State would be too late and secondly, the State would have a legitimate excuse for not acting in a timely manner, since it is implausible to assume that the victim will be constantly accompanied by a State agent who may jump in to help. Hence, the ‘immediacy’ of the Osman test does not serve well in the context of domestic violence. I would rather propose … that the standard by which the State is held accountable is whether a State knows or ought to know that a segment of its population, such as women, is subject to repeated violence and fails to prevent harm from befalling members of that group of people when they face a present (but not yet imminent) risk. As a consequence, the due-diligence standard against which the State’s action or inaction is assessed spans across a wider window of time, starting from the moment when a risk of domestic violence is present, but not yet imminent. It is baffling to witness that the majority have accepted CEDAW’s approach that genderbased violence need not pose an ‘immediate’ threat in paragraph 77 of the present judgment, yet goes on to apply the original Osman test in paragraphs 56 and 98. This failure to adopt a coherent approach creates the risk that victims of domestic violence will fail to be protected since the positive obligation on the authorities to act would be triggered too late. (Emphasis added)23

While this reading of the temporal frameworks of the foreseeability test here is to be applauded, it is not accompanied at any point in Pinto de Albuquerque’s judgment by any consideration or even allusion to the existence of countervailing rights of suspected attackers. Indeed, he goes on to call for a Russian legislative scheme of ‘preventive detention of the perpetrator where this is deemed necessary’ and ‘adequate training for judges, prosecutors and lawenforcement officers, so that they will implement the legislative measures on domestic violence adequately’.24 In respect of this training, he calls for a ‘change in mentality’ and ‘sensitisation training, which presents domestic violence as a human-rights violation triggering a responsibility to protect, investigate and prosecute perpetrators’.25 Nowhere in his discussion is there an assertion of the need to balance any such duties with the negative rights of potential assailants or to incorporate this balance in the training programme of the law enforcement officials involved. 23 Volodina v Russia App No 41261/17, 9 July 2019, Separate Opinion of Judge Pinto de Albuquerque, para 12. 24 ibid, Separate Opinion of Judge Pinto de Albuquerque, paras 17, 18 and 20. 25 ibid, Separate Opinion of Judge Pinto de Albuquerque, para 18.

Preventive Obligations, Risk and Coercive Overreach  255 So I repeat my concern that human rights institutions are acting with insufficient attention to the danger that their specific judgments or declarations inadvertently promote a culture of risk aversion that overwhelms a culture of rights. Human rights institutions need to send clear signals that there are limits to the prevention of risk that are embedded in the architecture of a human rights order. This might involve vivid rhetorical devices which promote the toleration of risk, much like Aharon Barak’s famous phrase when speaking about the prohibition on torture in the Israeli counterterrorism context: ‘sometimes, a democracy must fight with one hand tied behind its back’.26 Perhaps the sharpest example of this potential for coercive overreach in the context of preventive obligations comes in the extreme case of Tagayeva v Russia27 decided by the ECtHR in 2017. This case dealt with a terrorist attack, siege and storming of a school in Beslan, North Ossetia for three days in September 2004. Many hostages were taken and hundreds were killed. The deaths resulted both from the actions of the terrorist attackers and from the actions taken by the state’s attempt to storm the school and rescue the hostages. The result was two claims arising from Article 2 of the Convention. One related to the state’s positive obligations to prevent threats to life, while the other arose from the state’s use of lethal force. The operational conflict in this context could not have been starker and, indeed, more reflective of a common operational dilemma. On the question of the state’s use of lethal force to rescue the hostages, the Court found that prima facie the use of force had been justified under Article 2, but found a violation because of the use of indiscriminate weapons such as explosives, which it regarded as not being ‘absolutely necessary’. It also found a violation of Article 2 because of failures in the operation and command structures of the Russian authorities to ‘take such feasible precautions’ to ‘coordinate and communicate the important details relevant to the rescue operation to the key structures involved and to plan in advance for the necessary equipment and logistics’.28 The judgment has been praised, most notably in the partly dissenting opinion by Pinto de Alberquerque, for resisting the paradigm of war in the interpretation of the right to life in this context and for not following the approach of the ECtHR in Hassan v UK, where it applied the international humanitarian law paradigm to imply further restrictions to the right to liberty.29 However, for our purposes here, the key point of the case is that the ECtHR found that Russia had breached its positive obligations under Article 2 for failing to take adequate measures to prevent a clearly knowable risk of terrorist attack. In the expert evidence provided, it had been argued that Beslan was clearly under 26 HC 5100/94, Pub Committee Against Torture in Israel v Government of Israel, 53(4) PD 817, 845; A Barak, ‘A Judge on Judging: The Role of the Supreme Court in a Democracy’ (2002–03) 116 Harvard Law Review 148. 27 Tagayeva v Russia App No 26562/07, 13 April 2017 (Final 18 September 2017). 28 ibid para 574. 29 Hassan v UK App No 29750/09, 16 September 2014. See further L Hill-Cawthorne, ‘The Grand Chamber Judgment in Hassan v UK’, EJIL:Talk!, 16 September 2014.

256  Liora Lazarus a ‘real and immediate threat of an attack on a school on 1 September’30 because of known terrorist activity in the area. Consequently, the applicants argued that the Russian authorities had failed to take reasonable preventive measures despite having ‘knowledge of a real and immediate threat to life’.31 The ECtHR agreed with the applicants and found that the Russian authorities had breached the operational preventive obligation. While the general security threats were evidently high in the area, diffuse evidence was presented of the specific attack taking place. Nevertheless, the Court accepted that there was ‘sufficiently specific information’ of a knowable and concrete risk of such an attack that the Russian authorities knew or ought to have known about.32 Having accepted this concrete risk existed, the Court then went on to leave a remarkably wide margin of discretion as to what preventive actions would have been suitable or reasonable in the circumstances. The crucial paragraph in the case is as follows: The Court reiterates that in the preparation of responses to unlawful and dangerous acts in highly volatile circumstances, competent law-enforcement services such as the police must be afforded a degree of discretion in taking operational decisions. Such decisions are almost always complicated, and the police, who have access to information and intelligence not available to the general public, will usually be in the best position to make them (see P.F. and E.F. v. the United Kingdom (dec.), no. 28326/09, § 41, 23 November 2010). This is especially so in respect of counter-terrorist activity, where the authorities often face organised and highly secretive networks, whose members are prepared to inflict maximum damage to civilians, even at the cost of their own lives. In the face of an urgent need to avert serious adverse consequences, whether the authorities choose to use a passive approach of ensuring security of the potential targets or more active intervention to disrupt the menace, is a question of tactical choice. However, such measures should be able, when judged reasonably, to prevent or minimise the known risk. With regard to the above arguments, the Court finds that in the case at issue the Russian authorities failed to take such measures. (Emphasis added)33

Commentators have praised the Tagayeva decision for placing victims at the centre of its concerns34 and as the ‘high water mark’ for the protection of hostages’.35 Indeed, Galani celebrates the fact that ‘the Court successfully translated an international legal obligation into the human rights framework leaving no doubt that states, bound by the Convention, will be in breach of their human rights obligations if they fail to prevent terrorist attacks against individuals 30 Tagayeva v Russia (n 27) para 437. 31 ibid para 478. 32 See ibid para 491. In this respect, the Court lowered the standard applied in Finogenov v Russia App No 18299/03 and 27311/03, ECHR 2011. 33 Tagyaeva v Russia (n 27) para 492. 34 J Gavron and J Clifford, ‘Victims Placed at the Centre in Beslan School Siege Judgment (Tagayeva v Russia)’ Strasbourg Observers, 24 May 2017. 35 S Galani, ‘Terrorist Hostage Taking and Human Rights: Protecting Victims of Terrorism under the European Convention on Human Rights’ (2019) 19 Human Rights Law Review 149.

Preventive Obligations, Risk and Coercive Overreach  257 or society as a whole’.36 But the case raises a number of concerns. First, the judgment was replete with references to the complexity of counter-terrorist operations, suggesting a general deference to operational concerns. Second and crucially, the permissive nature of the powers of prevention in the Court’s framing has alarming implications. From the context and phrasing of the wording, it is hard to escape the conclusion that the Court is suggesting that states could use lethal pre-emptive force in order to avoid falling foul of the Convention. Indeed, when read together with other parts of the judgment, this conclusion becomes even more difficult to avoid. In paragraph 486, the Court argues that it could ‘reasonably be expected that some preventive and protective measures’ would ‘include a range of other security steps, in order to detect, deter and neutralise the terrorists as soon as possible and with minimal risk to life’. It is unclear from this passage whether the Court is concerned with the risk to life of the terrorist attackers or their potential victims in the event of the attack occurring. Certainly, the term ‘neutralise’ in respect of the terrorist attackers appears to suggest that the Court is here concerned with victim life. Moreover, there is little reference in this part of the judgment to the countervailing right to life of the terrorists and how this might be balanced against the operational duty to prevent or neutralise terrorist attacks. It is important at this juncture to let this prospect sink in. This is a human rights court that in the same decision explicitly rejected the idea that the laws of armed conflict may apply to the use of lethal force, which is exacting in respect of the use of lethal force in light of the negative right to life of the hostages, but nevertheless constructs an obligation which has the potential to legitimise pre-emptive lethal force towards terrorist attackers. Let us contemplate for a moment how this judgment might be received by the respondent state in this instance. As is evident from its ‘counter-terrorist’ operations in the Northern Caucasus, Russia is hardly a state known for its restraint on use of force or for its embedded commitment to human rights. The prospect of Russia using this decision to legitimise pre-emptive use of lethal force is hardly remote. Moreover, the prospect that this decision might be used for a general legitimation of preemptive lethal force across international law by those intent on seeking it is also plausible. Not only did the Court lower the evidential standards for the existence of a ‘concrete risk’ of an attack, it was also silent on the temporal framework in which a pre-emptive measure would be viewed as legitimate.37 Indeed, there was a noticeable lack of critical engagement with the whole question of ‘pre-emptive’ action, or a recognition of the implications of such a decision under international laws of self-defence.38 36 ibid 158. 37 Indeed, if the Court were to apply the ‘wider window of time’ test proposed by Judge Pinto de Alberquerque in Volodina v Russia (n 23) to these factual conditions, the implications of this reasoning would be even more concerning. 38 See in general on the international law on self-defence T Ruys, ‘Armed Attack’ and Article 51 of the UN Charter: Evolutions in Customary Law and Practice (Cambridge, Cambridge U ­ niversity

258  Liora Lazarus Tagayeva demonstrates the ongoing risk that human rights to protection through coercive or forcible means easily slide into broader legitimation of coercive force. The question is how to delimit these duties in scope in order to resist this risk-averse human rights logic. A stark contrast can be made here with the Aviation Security Act decision of the German Federal Constitutional Court, which directly confronts, and transparently balances, the potential conflict between negative rights protections and the protective duties of the state. The German Federal Constitutional Court (FCC) struck down powers under the Aviation Security Act to shoot down aircraft which had been hijacked with the intention of using the aircraft ‘as weapons in crimes against human lives’. The Court took the view that the essence of the right to life consisted in the human dignity of the innocent passengers on the plane. The Aviation Security Act was viewed as a violation of the (explicitly Kantian) conception of human dignity because it rendered these innocent passengers mere objects of state action. Importantly, having established the shape of the right to life of the passengers on the hijacked plane, the FCC rejected the government’s argument that the relevant powers under the Aviation Security Act could be justified as a consequence of the state’s protective duties. Thus, it argued as follows: Finally, para 14.3 of the Aviation Security Act also cannot be justified by invoking the state’s duty to protect those against whose lives the aircraft … is intended to be used. In complying with such duties of protection, the state and its bodies have a broad margin of assessment, valuation and organisation … Unlike fundamental rights in their function as subjective rights of defence [against the state], the state’s duties to protect which result from the objective contents of the fundamental rights are, in principle, not defined … How the state bodies comply with such duties of protection is to be decided, as a matter of principle, by themselves on their own responsibility … This also applies to their duty to protect human life. It is true that especially as regards this protected interest, in cases with a particular combination of circumstances, if effective protection of life cannot be achieved otherwise, the possibilities of choosing the means of complying with the duty of protection can be restricted to the choice of one particular means … The choice, however, can only be between means the use of which is in harmony with the constitution.39

The FCC’s approach here is clear. Preventive positive obligations to protect life give rise to broad discretion in which a range of policies or actions might be adopted to protect life of those at risk of lethal attack. The FCC accepts that these choices might be ‘restricted to the choice of one particular means’ Press, 2010); D Bethlehem, ‘Principles Relevant to the Scope of a State’s Right of Self-Defense against and Imminent or Actual Armed Attack by Nonstate Actors’ (2012) 106 American Journal of International Law 770; O Corten, ‘A Plea against the Abusive Invocation of Self-Defence as a Response to Terrorism’, EJIL:Talk!, 14 July 2016; Rt Hon J Wright, ‘The Modern Law of Self-Defence’, Speech at International Institute for Strategic Studies, EJIL:Talk!, 11 January 2017; M Milanovic, ‘What is an Imminent Armed Attack?’, EJIL:Talk!, 12 January 2017; J Green, ‘Initial Thoughts on the UK Attorney General’s Self-Defence Speech’, EJIL:Talk!, 13 January 2017. 39 BVerfGE (2006) 1 BvR 357/05, 15 February 2006, at [136]; see also [120]–[121].

Preventive Obligations, Risk and Coercive Overreach  259 where ‘effective protection of life cannot be achieved otherwise’. Nevertheless, the choice can only be made between ‘means’ that are consistent with the Constitution’. In the FCC’s logic, a choice to kill a passenger on a hijacked plane was inconsistent with human dignity and consequently with the Constitution. As a result, a pre-emptive strike ‘cannot be justified by invoking the state’s duty to protect those against whose lives the aircraft … is intended to be used’. There are a number who disagree with the outcome and reasoning of the FCC decision, not least in respect of the conception of human dignity or its distinction between the lives of the hijackers and the lives of the passengers.40 But the point of this example is not to defend the way in which the Court defined the right to life itself; rather, it is to show how the balancing process might be structured where positive obligations compete with negative rights. The first point is at a technical level of rights reasoning. Courts in general need to take account of the competing obligations, principles or rights explicitly and early on when determining the contours of the coercive obligation. The second point is that at a rhetorical level, the court or human rights institution needs to signal in broad terms that the prospect of balance is unavoidable; in short, they need to promote a stereoscopic view of protective obligations. When dealing with prospective risk, we also need a third rhetorical tier in this exercise. This must emphasise that risk is not only tolerated within a human rights order, but that acceptance of risk may also be required at times by the existence of a countervailing negative right. Tolerance of risk is certainly inherent to the Osman test, where the Court noted that ‘not every claimed risk to life can entail for the authorities a Convention requirement to take operational measure to prevent that risk from materialising’.41 As a result, language of ‘reasonable burdens’ is explicit and repeated across the case law, as is the limiting device of a ‘real and immediate risk’ to establish the existence of a positive obligation.42 But these limiting devices normally refer to the plausibility or reasonableness of the duty in practice. What is less often asserted is that the extent of the duty to prevent risks to the rights of potential victims is also constrained by competing negative rights. Consequently, it is not always explicit from the case law that a reasonable burden of risk avoidance must also be reconciled with a competing negative right. Nevertheless, the constraining effect of negative rights on positive obligations is an essential principle that requires consideration when specifying the scope of positive obligations. As Mavronicola argues in respect of Article 3

40 O Lepsius, ‘Human Dignity and the Downing of Aircraft: The German Federal Constitutional Court Strikes Down a Prominent Anti-terrorism Provision in the New Air-Transport Security Act’ (2006) 7 German Law Journal 761; K Möller, ‘The Right to Life between Absolute and Proportional Protection’, LSE Legal Studies Working Paper No 13/2010. 41 Osman v UK (n 7) para 116. 42 See in depth Stoyanova (n 12) 338–45. See also Mastromatteo v Italy App No 37703/97, 24 ­October 2002.

260  Liora Lazarus ECHR, ‘the Court should not impose a duty to take action which it is unlawful to take under the Convention’.43 Consequently: [W]hile the imperative to protect individuals from torture, inhumanity and degradation may be very strong indeed, it cannot displace non-displaceable negative obligations. As to displaceable negative obligations, with respect to rights which may be lawfully infringed, these may be displaced insofar as necessary and p ­ roportionate – a principle applicable generally in relation to interference with these rights.44

Adopting or promoting a view of protective obligations that fails to face up to the potential conflict between the obligation to avoid risk and countervailing negative rights is, as a consequence, an incomplete and monoscopic account of positive obligations in this context.45 III.  CASE STUDIES ON PREVENTIVE OBLIGATIONS

This monoscopic approach is not limited to cases dealing with extreme conflict such as Tagayeva. Examples of coercive overreach – where courts either ignore opposing negative rights or understate their weight relative to positive obligations – are also evident in more prosaic circumstances. I have been challenged a number of times since first writing in this field to provide concrete evidence of the state’s potential for coercive overreach in response to a judicial assertion of a positive coercive obligation. One piece of evidence presented in my earlier contribution was the domestic UK case of R v G.46 Here Baroness Hale raised the positive obligation to protect against rape as a reason for the Court to defeat a potential rights claim of a teenage applicant who argued that he could have been prosecuted for a lesser offence because the teenage victim had convincingly lied about her age. He argued that his prosecution and conviction for the graver sexual offence was a violation of his Article 8 right not to be stigmatised. Hale suggested that the court would itself be in breach of the state’s positive obligation had it considered the applicant’s claim. This was, as I argued, an overblown rhetorical assertion of the positive duties on states without legal foundation.47 Similarly, the Separate Opinion of Judge Pinto de Albuquerque in Volodina v Russia (discussed earlier) is another example of a monoscopic view of protective

43 N Mavronicola, Torture, Inhumanity and Degradation under Article 3 of the ECHR: Absolute Rights and Absolute Wrongs (Oxford, Hart Publishing, forthcoming) ch 3, sub-s 3.5 (on file with the author). 44 ibid. 45 See also on this conflict S Smet, ‘The “Absolute” Prohibition of Torture and Inhuman or Degrading Treatment in Article 3 ECHR: Truly a Question of Scope Only?’ in E Brems and J Gerards (eds), Shaping Rights in the ECHR: The Role of the European Court of Human Rights in Determining the Scope of Human Rights (Cambridge, Cambridge University Press, 2013); or S Smet, ‘Conflicts between Absolute Rights: A Reply to Steven Greer’ (2013) 13 Human Rights Law Review 469. 46 R v G [2008] UKHL 37, [2009] 1 AC 92. 47 Lazarus (n 1).

Preventive Obligations, Risk and Coercive Overreach  261 obligations. But there are other instructive examples in the practice of security outside of the courts. A.  Policing and Osman Warnings The first example comes from the policing operational context and concerns the use (and political reception) of so-called ‘Osman warnings’. After Osman was decided, the UK police developed procedures known as ‘threat to life notices’, which became termed in policing practice and popular language as ‘Osman warnings’. The practice has become standardised within policing since this time and is elaborated in the Home Office Threat to Life Policy, which in turn is based on the National Police Chiefs Council guidelines.48 Furthermore, each Police Authority publishes its own ‘Threat to Life’ guidelines.49 The documents are partly available to the public and are partly categorised as ‘officially sensitive’ or ‘internal use only’, but there is sufficient available public information to get a sense of the operational guidance issued. What is evident on a reading of these documents is that the right to life, the Osman decision and subsequent domestic law implementing Osman under the Human Rights Act50 all feature heavily at the outset of the document. What is also clear is that the ‘Osman’ positive obligation is described as relative or as ‘not absolute’ in the sense of not imposing an unreasonable or impractical obligation to forestall any risk. So the message about reasonable operational burdens, and the limits on potential liability for breach of the Osman obligation, is clearly relayed in these documents. A similar message is set out in the National College of Policing Professional Guidance Risk register. Under Principle 2, it is stated that ‘maintaining or achieving the safety, security and wellbeing of individuals and communities is a primary consideration in risk decision making’. In this provision, it is also asserted that ‘even though making a decision involves the risk of criticism if harm occurs, this risk … is secondary to the primary responsibility to protect life’. The duty is also tempered in the reasonableness language of the Osman duty, where it is argued that ‘police officers are entitled to take other considerations into account’ and that the ‘state is not expected to undertake an unduly burdensome obligation: it is not obliged to satisfy an absolute standard requiring the risk to be averted, regardless of all other considerations’.51

48 HM Home Office, Threat to Life Policy, Version 6.0, 12 January 2017, https://assets.publishing. service.gov.uk/government/uploads/system/uploads/attachment_data/file/583105/Threat-to-life_ policy-v6.pdf. 49 eg, Dorset Police, Threat to Life Policy, 24 February 2004, https://www.dorset.police.uk/ media/54485/threats-to-life-policy-p07-2004-v23-external.pdf. 50 eg, Van Colle v Chief Constable of the Hertfordshire Police; Smith v Chief Constable of Sussex Police [2008] UKHL 50 (HL). 51 National College of Policing Professional Guidance on Risk, Principle 2, https://www.app. college.police.uk/app-content/risk-2/risk/#principle-2.

262  Liora Lazarus However, what none of these documents broadcast is that the operational duty is also constrained by the rights of the potential object of the protective obligation. It is not explicit from the publicly available guidance how the rights of the suspect may condition or shape the measures adopted to avert the risk in question. This is particularly problematic in the light of evidence about police practice and knowledge of human rights. Richard Martin’s outstanding empirical work showcasing interviews with community officers demonstrates that although human rights knowledge is generally patchy, most officers are aware of their positive obligation requirements under Article 2. As he makes clear: [T]he rare instance where explicit reference to human rights was made occurred when incidents involved a risk to a person’s life. This type of incident was labelled an ‘Article 2 issue’, alluding to the positive duty imposed on public authorities by Article 2 ECHR, requiring them to take positive steps to protect threats to life, including where the threat or harm emanates from private actors.52

Martin goes on to quote Neighbourhood Constable Fiona on the ‘mention of human rights’: The odd time you’d maybe hear it mentioned in custody but it’s never really mentioned out on the ground too much, it’s never really mentioned in the station and I doubt that we’d be able to name all the human rights. The only one is article 2 right to life, you’d hear it a lot on the radio: ‘Is there an article 2 issue here?’ Am I putting on blue lights, basically … If there’s somebody in a house with a knife or whatever, is there an article 2 issue here, and then you get supervision coming on giving the whole spiel about ‘Oh, article 2’. But you’re so used to hearing that that you zone out.53

Martin also demonstrates that the Article 2 obligation raises serious anxieties for police officers acting in a highly charged operational context where quick decisions must be made which cut both ways with respect to Article 2. He uses the testimony of another officer to bear this out: Say you’re told not to go into a house unless there’s article 2 issues, you hear that all the time, ‘Contain unless there article 2 issues’. And we’re sort of going okay, theoretically if you were to disobey that and go in, justifying in your head there’s article 2 issues, and it goes horribly wrong, they’ll say you shouldn’t have went in, you misread the situation. By the same token, if you stand outside and don’t go into the situation and it goes horribly wrong and somebody does get seriously injured or killed they’ll use that same piece to say you should have went in, you had the personal responsibility that there was an article 2 issue there.54

The operational context here provides a vivid insight into the Article 2 dilemma where the officer is having to balance the negative aspect of the right to life against the positive duty under the same right. The testimony provided by Martin here illustrates exactly why concrete guidance and training in this

52 Martin 53 ibid

8. 54 ibid 9.

(n 17) ch 4 at 8–9 of draft file.

Preventive Obligations, Risk and Coercive Overreach  263 context is so important, and it is also illustrative of what a risk-averse culture is likely to produce if positive obligations continue to spur this. But without a strong message in police guidance about how such a dilemma is to be resolved, it is plausible that the police officer in this position would take the action most likely to protect the life of the innocent (potential) victim. What is also clear is that Osman warnings have entered into the law-and-order debate and are viewed by the national British press as a measure of public insecurity. It is notable that Freedom of Information requests are frequently directed at police authorities by the press as regards the number of ‘Osman warnings’ issued annually.55 The numbers are then reported in the national press as an indication of alarm about public safety. Hence, the Scottish Daily Record included the headline: ‘Police Scotland Issue More Than 1,000 “Threat to Life” Warnings in Five Years’ and the Daily Telegraph raised alarm with the headline: ‘The Number of People Told Their Lives are under Threat Grows, New Figures Show’ and goes on to raise alarm about public safety: Forces across England and Wales issued more than 776 so-called ‘Osman warnings’, or ‘threat to life’ notices, last year. That figure represents a 14 per cent increase compared to 2016 when at least 679 warnings were made, a Freedom of Information request has shown. And, because only 31 of the 43 forces in England and Wales responded to the request the real number of threat to life warnings police made are likely to be far higher. Of those, only 20 had the information available. Osman warnings are issued if police have intelligence of a real and immediate threat to the life of an individual. In such cases officers may not have enough evidence to make an arrest and so warn the potential victim through a notice or letter that their life is endanger [sic]. The figures offer an insight into how violent crime in the country is escalating at an alarming rate. The extent of that increase is illustrated starkly by figures from the capital’s Metropolitan Police Service, Britain’s biggest force.56

What these anecdotal references point to, and invite further empirical research into, is a number of contradictory patterns. On the one hand, while the existence of Osman warnings may be viewed, institutionally, as a measured response to an obligation to protect life, it is also clear that they have become indicators in a wider political context of public insecurity. There is also a possibility, which needs to be empirically tested, that the rise of ‘Osman warnings’ is linked to a

55 See, eg, Nottinghamshire Police Freedom of Information Unit, ‘Osman Warnings (Threats to Life Warnings) – FOI 005723/12’, https://www.nottinghamshire.police.uk/document/osmanwarnings-threats-life-warnings-foi-00572312; Metropolitan Police, ‘Freedom of Information Request – Number of Osman Warnings’, https://www.met.police.uk/SysSiteAssets/foi-media/ metropolitan-police/disclosure_2018/march_2018/information-rights-unit---osman-warningsor-other-warnings-on-threats-to-life-issued-by-the-mps; Police Scotland, ‘Disclosure Log – Article Reference 10-0744 – Statistics – Threat to Life Notices (Osman Warnings)’, https://www.scotland. police.uk/access-to-information/freedom-of-information/disclosure-log/0/2019/april/19-0744. 56 S Bird, ‘The Number of People Told Their Lives are under Threat Grows, New Figures Show’, Daily Telegraph, 2 June 2018, https://www.telegraph.co.uk/news/2018/06/02/number-people-toldlives-threat-grows-new-figures-show.

264  Liora Lazarus risk-averse operational culture within policing, which itself has been generated by the human rights obligation in the first place. The risk aversion here is not related only to risk to life, but is also importantly related to fear of liability for breach of a human rights obligation. Whether this risk aversion within police culture manifests in violation of the rights of suspects is a matter which needs to be tested empirically, but it is plausible that the culture generated may result in more extreme responses (certainly this is something evident in the de Menezes case).57 B.  Preventive Risk Orders and Modern Slavery A second example of the potential for positive obligations to give rise to preemptive risk aversion (potentially going too far) lies in the legislative process of the UK Modern Slavery Act. This legislation is already the subject of an excellent analysis by Mattia Pinto in this volume, who is rightly concerned with the way the ‘British government explicitly invoked … positive duties’ under the ECHR to validate coercive prevention and risk orders under the Act.58 This legislation went through a careful consultation process steered under the Joint Parliamentary Committee on the Draft Modern Slavery Bill. I was asked to give advice to this Committee on the inclusion of Slavery and Trafficking Risk Prevention Orders. The particular characteristics of these orders raised concern about scope and legal certainty, and the undifferentiated categories of individuals who could be caught under them. I presented this concern, alongside Lucia Zedner, in oral and written evidence to the Committee.59 We both argued that the preventive orders were too vague and over-inclusive, and risked falling foul of the legality and quality-of-law tests under ECtHR jurisprudence. My argument rested on the claim that positive obligations to prevent modern slavery under Article 4 ECHR could not be used as a basis to undermine well-established principles of legality: [P]ositive obligations under the ECHR have a range of dimensions in relation to human trafficking, and give rise to specific preventive obligations where a real and immediate risk exists to an individual that their rights will be breached under Article 4. The obligations are not meant to place an impossible or disproportionate burden on State authorities which is why a breach of the specific obligations arises only where the State knew or ought to have known of a real and immediate risk to a specified individual. These obligations are also meant to be fulfilled in a manner which fully respects the due process and other guarantees of the Convention. To be clear, the UK’s positive obligation arising out of Article 4 must be executed in such a way that also takes into regard the rights of the defendants. The balance has to be 57 Da Silva v UK App No 5878/08, 30 March 2016. 58 Quoted from Pinto, ch 8 in this volume. 59 L Lazarus, Written submission to the Joint Parliamentary Committee on the Draft Modern Slavery Bill, 28 February 2014, www.parliament.uk/business/committees/committees-a-z/joint-select/ draft-modern-slavery-bill/written-evidence.

Preventive Obligations, Risk and Coercive Overreach  265 struck between these two imperatives. It is thus a question of reconciling the preventive obligations that the Strasbourg case law clearly gives rise to, and the constraints on such preventive action.60

My proposals for tighter specification found favour with the Committee and were supported in a report of the Committee to the Home Secretary, Theresa May.61 However, the Home Office rejected the call for safeguards, relying on existing Convention duties under the Human Rights Act as evidence that such considerations were already met within the Act: The Government has reviewed the Committee’s arguments about legal certainty and safeguards with care. The Government will be writing to the Joint Committee on Human Rights to set out in detail why we believe our approach is fully compliant with our human rights obligations. We believe we have set substantial and appropriate safeguards to ensure that the Orders will only be used in appropriate circumstances.62

The reasoning presented by the government was not accepted by the Committee, which supported their advisor Maya Sikand’s view that ‘the Home Office cannot simply rely upon the Court’s duties under section 6 of the Human Rights Act’ to those subject to prevention or risk orders ‘to rectify a lack of legal certainty on the face of the Bill’.63 The stand-off between the Committee and the Home Office is a telling example of how legislative provisions relating to prevention have been framed to override competing human rights concerns, even where the government is made aware of these competing claims. Clearly the government was not moved by the argument of legality arising from competing rights and there was little scope to shift its position. On the contrary, the need for broadly framed risk prevention orders was re-asserted and the view of law enforcement was used as a reason to do so: When giving evidence to the Committee, law enforcement bodies indicated that STROs could be particularly effective in scenarios where law enforcement would seek

60 ibid paras 3 and 4. 61 ‘As Professor Lazarus said, the effect of this principle is that any rules interfering with these Convention rights must be sufficiently certain and accessible to allow people to understand what is expected of them and when an interference with their daily lives will be justified. Given the potential for restrictions of everyday behaviours or rights to result from the imposition of the Prevention and Risk Orders, there is a high threshold requirement of legal certainty: first over the threshold requirements for the imposition of an Order; and second, clarity as to the contents of the Order and the effects of such an Order being imposed upon an individual. But this is also a practical issue for magistrates, who will be required to assess whether an Order is necessary but with no guidance on the risk factors they should consider to make the imposition of an Order proportionate, or on the possible restrictions they could impose which would be proportionate to the actual risk presented by the defendant in front of them. This should be rectified.’ Joint Committee on the Draft Modern Slavery Bill, Report on the Draft Modern Slavery Bill, HL Paper 166, para 52 (emphasis added). 62 Government Response to the Report from the Joint Committee on the Draft Modern Slavery Bill Session 2013–14 HL Paper 166/HC 1019, 9. 63 Joint Committee on the Draft Modern Slavery Bill (n 61) para 53.

266  Liora Lazarus to restrict the behaviour of individuals who are at the periphery of modern slavery investigations, where there may be insufficient evidence to convict an individual, but the risk is still clear. They also suggested that they could be useful to control the behaviour of others who posed a risk but where it was difficult to prosecute, such as brothel keepers who advertise internationally for women, and move victims backwards and forwards across borders.64

What we can take away from this second example is that legislation may itself take a predominantly risk-averse approach where positive rights provide a legitimation of a particular coercive approach.65 Whether positive human rights obligations prompt such coercive action and thereby intensify the net of social control or whether they are used as ex post facto justifications of proposed coercive policy is unclear. What is clear is that the legislators here gave more weight to the obligation arising out of Article 4 ECHR than they did to the general legality requirements which apply across the Convention as a whole. They explicitly ignored the view of the Committee and its advisors on this question. If courts also fail to send clear messages about balance, they may also exacerbate this problem. IV. CONCLUSION

The ECtHR as well as many domestic constitutional courts have a number of rhetorical devices that they use as signals to states. The rhetorical signifiers of ‘necessary in a democratic society’, ‘European consensus’, ‘living tree’, or ‘rule of law’ are all devices of this kind. These devices work most importantly at the level of general judicial legitimation rather than as fine-grained legal devices which can provide specific guidance in concrete cases. The coercive human rights jurisprudence is clearly developing similarly, with phrases such as ‘effective criminal law’ and ‘measures when judged reasonably that prevent or minimise a known risk’ doing similar work. This chapter is calling upon courts, human rights institutions and scholars to develop a persuasive and consistent rhetorical language which prioritises a rights culture over a culture of risk. In order to do this, courts must consistently re-assert the limits that negative rights present to the scope of preventive obligations aimed at risk avoidance. Most importantly, this language needs to develop in the context of operational prevention, just as much as it might play a role in the context of criminalisation. Without it, coercive rights risk crowding out other rights in a risk-averse culture of insecurity.

64 Government Response to the Report from the Joint Committee on the Draft Modern Slavery Bill (n 62) 7. 65 See also Mattia Pinto and Vladislava Stoyanova, chs 8 and 10 in this volume respectively.

13 Coercive Human Rights and Unlawfully Obtained Evidence in Domestic Criminal Proceedings KELLY M PITCHER*

I. INTRODUCTION

A

n often-made argument in domestic debates concerning the question of how to address procedural violations committed in the pre-trial phase of criminal proceedings is that it is a reason for the trial judge in criminal proceedings to respond to such a violation (for example, by excluding the evidence obtained thereby) that it violated the individual rights of the accused.1 While the contours of the argument may vary in terms of what is to count as an individual right in this regard such as to trigger a particular judicial response, it is fair to say that the argument is at its most compelling when the individual right to have been violated constitutes a right of an elevated status, for example, a human right of the accused.2 However, the rights of the accused are not the only set of human rights standards capable of informing the determination of whether to attach consequences to a pre-trial procedural violation. In the Netherlands, the trial judge in criminal proceedings is required in some cases (and permitted in others), in the determination of whether to exclude evidence * This chapter develops certain ideas presented in an earlier piece of writing (KM Pitcher, Judicial Responses to Pre-trial Procedural Violations in International Criminal Proceedings (The Hague, TMC Asser Press, 2018)). I am grateful to Laurens Lavrysen and Natasa Mavronicola for inviting me to contribute to this volume and for their editorial assistance, and to Ellen Gijselaar for her insightful comments on an earlier draft. This chapter incorporates developments in the case law up to 1 July 2019. 1 The most well-known articulation of this argument, otherwise known as the ‘protective principle’, is that offered by A Ashworth in his oft-cited article ‘Excluding Evidence as Protecting Rights’ [1977] Crim LR 723. 2 See generally P Roberts, ‘Excluding Evidence as Protecting Constitutional or Human Rights?’ in L Zedner and JV Roberts (eds), Principles and Values in Criminal Law and Criminal Justice: Essays in Honour of Andrew Ashworth (Oxford, Oxford University Press, 2012).

268  Kelly M Pitcher obtained unlawfully, to take into account human rights standards that protect other persons, that is, the victims of crime.3 Specifically, the judge must (or may) take into account the positive obligations arising from the European Convention on Human Rights (hereinafter ECHR or ‘the Convention’). The idea is that the exclusion of unlawfully obtained evidence may be inconsistent with the positive obligations flowing from the ECHR because it prevents the effective punishment of ‘human rights offences’ – criminal acts that violate the substantive norm underlying a Convention right.4 Indeed, the exclusion of evidence may lead to an acquittal if any remaining evidence is insufficient to found a conviction.5 Proponents of an approach which treats the positive obligations flowing from the ECHR as a reason not to exclude unlawfully obtained evidence thus read a duty to punish into such obligations, whereby the concern is that the lack of punishment that results from the court attaching consequences to a pre-trial procedural violation would amount to a failure to discharge this duty. This chapter will consider when not punishing a human rights offence will mean a failure to discharge the positive obligations arising from the Convention right in question and what this means for the judge in domestic criminal proceedings confronted with pre-trial impropriety. It does so against the backdrop of a broader discussion that has gained momentum in recent years, in which the ‘coercive’ dimension of human rights, which sees rights as demanding state action rather than as a limitation thereon,6 is being appraised from a variety of perspectives, including international human rights law itself,7 as well as fundamental principles of criminal law and criminal procedure.8 A concern to have arisen in this regard is what Liora Lazarus has coined the ‘risk of coercive overreach’.9 It is widely accepted that states’ human rights obligations are not limited to abstaining from conduct that violates the human rights of persons within their jurisdiction; such obligations also require states to take measures to prevent human rights violations and to adequately address such violations

3 However, this is not to say that such human rights standards exist in order to provide redress to the individual victim or his or her next of kin. See the discussion of the rationales underlying the positive obligation resting on the state in the criminal law-enforcement sphere in section III.A. 4 To borrow from K Kamber, Prosecuting Human Rights Offences: Rethinking the Sword Function of Human Rights Law (Leiden, Brill, 2017) 19. ‘Human rights offences’ is a more inclusive term than human rights violations, in that it also covers criminal acts that violate the substantive norm underlying a Convention right committed by private individuals. 5 This line of reasoning applies to other judicial responses too: a stay of proceedings prevents the case from reaching a verdict on the merits (and therefore the establishment of criminal responsibility) and a significant reduction of sentence precludes the imposition of punishment commensurate with the criminal responsibility established. 6 L Lazarus, ‘Positive Obligations and Criminal Justice: Duties to Protect or Coerce?’ in Zedner and Roberts (n 2) 136. 7 See, eg, N Mavronicola, ‘Taking Life and Liberty Seriously: Reconsidering Criminal Liability under Article 2 of the ECHR’ (2017) 80 MLR 1026; and A Seibert-Fohr, Prosecuting Serious Human Rights Violations (Oxford, Oxford University Press, 2009). 8 See, eg, Kamber (n 4); and Nina Peršak, ch 7 in this volume. 9 See generally Lazarus (n 6).

Coercive Human Rights and Unlawfully Obtained Evidence  269 when they occur.10 Moreover, this may require taking measures in the criminal law and criminal law-enforcement sphere, ie, the criminalisation of certain types of conduct (which may be characterised as a ‘substantive’ obligation), as well as criminal investigation and prosecution (which may be characterised as a ‘procedural’ obligation). However, the concern is that in the absence of identifiable and delimited obligations in this regard, and of appropriate language that properly captures the coercive nature of these measures, there is a risk that the case law setting forth such positive obligations will enable states to make excessive demands for ‘forms of [state] coercion directed at perpetrators of harm (or potential perpetrators of harm) which restrict rights and liberties to a far greater extent than the normal general limitations we might experience by living in a fair, socially democratic, and redistributive society’,11 under the guise of human rights protection. Thus, even if the European Court of Human Rights (hereinafter ECtHR or ‘the Court’) itself could not be accused of coercive overreach by demanding punishment at all costs, its case law may have unduly coercive implications, in that it may be construed by states as providing a basis on which to legitimately pursue an overly coercive agenda. The same may be said of the case law of higher domestic courts informing lower courts of their human rights obligations. All of this calls for a closer examination of an approach to the question of how to address pre-trial impropriety that treats the procedural positive obligations flowing from the ECHR as a reason not to attach consequences to a procedural violation, on the basis that doing so would be inconsistent with a duty to punish. Section II of this chapter sets out in more detail the Dutch judicial response to pre-trial procedural violations and the role of positive obligations therein. It is in the context of the determination of whether to exclude unlawfully obtained evidence that the trial judge is expressly required or permitted to take the positive obligations flowing from the ECHR into account; accordingly, the focus in this section is on this particular response to pre-trial procedural impropriety. While the Dutch approach is but an example of how positive obligations might be incorporated into the determination of whether to attach consequences to a pre-trial procedural violation, it provides a useful basis for discussing the issues raised in the previous paragraph (a discussion which is the focus of a later section). Section III sets out the relevant case law and doctrine on the procedural positive obligations flowing from the ECHR (it is the procedural positive obligations flowing from the Convention that are implicated by the issue that lies at the heart of this chapter), thereby focusing on Articles 2 and 3 ECHR (it is in the context of these provisions that the state’s duty to take measures in the criminal law enforcement sphere – hereinafter referred to as the ‘procedural 10 Such ‘positive obligations’ or ‘affirmative duties’ are derived from substantive rights, read in conjunction with art 1 ECHR (see, eg, J-F Akandji-Kombe, Positive Obligations under the European Convention on Human Rights (Strasbourg, Council of Europe, 2007) 8). 11 Lazarus (n 6) 146.

270  Kelly M Pitcher coercive obligation’ – has matured).12 It considers when not punishing a human rights offence falling under Article 2 or 3 ECHR will mean a failure to discharge the procedural coercive obligation, given its scope and meaning.13 In light of the reason for enquiring into the procedural coercive duty on the state in this chapter, in this section consideration will also be given to the case law of the ECtHR addressing the relationship between this duty and the rights of the accused.14 Section IV considers the implications of the case law and doctrine set out in section III for the determination of whether to attach consequences to a pre-trial procedural violation, ie, what this means for the trial judge in domestic criminal proceedings confronted with such a violation. Finally, building on the findings in the preceding section, section V discusses the Dutch law and practice as an example of how the procedural coercive obligation might enter the analysis of whether to attach consequences to pre-trial procedural violations, in light of the risk of coercive overreach. II.  COERCIVE HUMAN RIGHTS IN DOMESTIC CRIMINAL PROCEEDINGS: THE DUTCH EXAMPLE

In the Netherlands, the judicial response to pre-trial procedural violations is governed by Article 359a of the Dutch Code of Criminal Procedure (CCP), as interpreted in the case law of the Dutch Supreme Court. Article 359a CCP provides for three responses within the criminal trial to pre-trial procedural violations: sentence reduction, the non-use (ie, exclusion) of the evidence obtained thereby, and a declaration that the prosecution is inadmissible (a procedural step akin to a stay of proceedings). In addition, the Supreme Court has held that a simple declaration that a procedural violation has occurred may suffice.15 While it is in the context of the determination of whether to exclude evidence that the trial judge is expressly required or permitted to take the positive obligations flowing from the ECHR into account, it is important to say something about the wider framework for determining the consequences (if any) to be attached to a pre-trial procedural violation. An essential aspect of this framework is the requirement that the court take into account certain factors (those set forth in the second paragraph of Article 359a CCP: the interest that the violated provision purports to protect, the seriousness of the violation and the prejudice caused by it) in the determination of the consequence (if any) to be attached to pre-trial impropriety, thereby forcing it to look beyond the fact of the 12 See Laurens Lavrysen, ch 2 in this volume. See also Kamber (n 4) 243. 13 See section III.A below. In light of the issue that lies at the heart of this chapter, attention will also be paid to the case law of the ECtHR addressing states’ procedural obligations in respect of human rights offences committed by private individuals. Domestic criminal proceedings, it is reasonable to assume, largely concern criminal acts committed by private individuals. 14 See section III.B below. 15 HR 30 maart 2004, ECLI:NL:HR:2004:AM2533, r.o. 3.6.1, NJ 2004/376 m.nt. Y Buruma.

Coercive Human Rights and Unlawfully Obtained Evidence  271 procedural violation to the individual circumstances of the case and preventing a ‘rigid, knee-jerk response to breaches of procedural rules’.16 In practice, this means that in determining a defence application for a particular consequence to be attached to a pre-trial procedural violation, the court, in assessing whether the ‘test’ for that consequence has been met, is required to take into account the aforementioned statutory factors and all of the particular circumstances of the case.17 Each of these tests, it should be noted, is stringent. Moreover, the process of taking such factors into account may lead the court to conclude that a simple declaration that an irreparable violation has occurred would suffice (on account of there being minimal or no actual prejudice, for example). It is to the test for exclusion that we now turn. According to the Supreme Court, evidence which has been obtained unlawfully is subject to exclusion when an important provision or principle (of criminal procedure) has been seriously breached and the evidence was obtained by the breach.18 At the same time, it has held that the exclusion of evidence thereby obtained is not the inevitable consequence of impropriety in the course of a criminal investigation, given that Article 359a CCP confers on the court a power to attach consequences to such violations, not a duty.19 Regarding the first aspect of the test, ie, the requirement that an important provision or principle be seriously breached,20 it is clear that a violation of the right to privacy within the meaning of Article 8 ECHR will not necessarily satisfy this criterion.21 For this position, the Supreme Court relies on the case law of the ECtHR, from which it follows that the use of evidence obtained thereby need not violate the right to a fair trial within the meaning of Article 6 ECHR.22 As to what may, or should, satisfy this criterion, according to the Supreme Court, there are three categories of procedural violation that require (or may require) the exclusion of evidence, depending on the category. First, exclusion may be necessary to secure the accused’s right to a fair trial within the meaning of Article 6 ECHR, as also interpreted by the ECtHR.23 By way of example, the Supreme Court points to its case law on the right to

16 R Kuiper, Vormfouten. Juridische consequenties van vormverzuimen in strafzaken (Deventer, Kluwer, 2014) 605. See also 53 f. 17 HR 30 maart 2004, ECLI:NL:HR:2004:AM2533, r.o. 3.6.2, NJ 2004/376 m.nt. Y Buruma; HR 19 februari 2013, ECLI:NL:HR:2013:BY5322, r.o. 2.4.1, NJ 2013/308 m.nt. BF Keulen. Accordingly, other facts and circumstances than those explicitly provided for in art 359a(2) CCP may be taken into account. 18 HR 30 maart 2004, ECLI:NL:HR:2004:AM2533, r.o. 3.6.4, NJ 2004/376 m.nt. Y Buruma; HR 19 februari 2013, ECLI:NL:HR:2013:BY5322, r.o. 2.4.2, NJ 2013/308 m.nt. BF Keulen. 19 ibid. 20 The second aspect – the requirement of causal connection – will not be considered any further in this chapter, aside from observing that the Supreme Court has interpreted this requirement strictly. 21 HR 19 februari 2013, ECLI:NL:HR:2013:BY5322, r.o. 2.4.2, NJ 2013/308 m.nt. BF Keulen. 22 Oft-cited authorities in this regard are Khan v UK 2000-V; (2001) 31 EHRR 45 and PG and JH v UK 2001-IX; (2008) 46 EHRR 51. 23 HR 19 februari 2013, ECLI:NL:HR:2013:BY5322, r.o. 2.4.4, NJ 2013/308 m.nt. BF Keulen.

272  Kelly M Pitcher legal assistance prior to questioning purporting to implement the ECtHR’s Salduz case law,24 and its case law on the use of statements made by a suspect to an undercover agent pretending to be a fellow detainee,25 which draws on the ECtHR’s judgment in Allan v UK.26 Once a violation of this nature has been established, there is very limited room for the court to nevertheless decide not to exclude the evidence pursuant to the factors set out in the second paragraph of Article 359a of the CCP. It is worth noting here that the national case law cited by the Supreme Court is based on case law of the ECtHR that adopts a principled, ‘interventionist’ approach to the admissibility (or use) of evidence, ie, in which it has held that the use of certain types of unlawfully obtained evidence automatically, or nearly so, renders the proceedings unfair.27 More is said about the ECtHR’s case law in this regard below. Second, where the accused’s right to a fair trial is not directly at stake, but another important provision or principle (of criminal procedure) has been seriously breached, exclusion may be necessary as a means of upholding the rule of law and preventing the public authorities charged with the investigation and prosecution of crime from acting unlawfully.28 The Supreme Court thus relies on the deterrence rationale for this category of exclusion.29 It refers to two types of situations that are likely to fall into this category: a ‘very intrusive’ breach of a fundamental right of the accused and cases in which the use of the evidence obtained ‘fundamentally devalues’ the interest served by the protection of legal and other types of professional privilege. Whether a very intrusive breach of a fundamental right of the accused, once established, requires the exclusion of the evidence obtained thereby depends on the factors set out in the second paragraph of Article 359a CCP and the individual circumstances of the case.30 Moreover, in this context, the court ‘may’ take into consideration whether the (deterrent) benefits of exclusion outweigh the negative effects that it may be expected to have on such ‘heavily weighted interests’ as the need to establish the truth and the need to punish perpetrators of (‘possibly very serious’) crime, and, where applicable, on the rights of victims or their next of kin (‘in light also of the p ­ ositive obligations flowing from the ECHR of effective punishment’).31 24 HR 30 juni 2009, ECLI:NL:HR:2009:BH3079, NJ 2009/349 m.nt. TM Schalken. See also Salduz v Turkey [GC] ECHR 2008, (2009) 49 EHRR 19. 25 HR 28 maart 2006, ECLI:NL:HR:2006:AU5471, NJ 2007/38 m.nt. TM Schalken. 26 Allan v UK 2002-IX; (2003) 36 EHRR 12. 27 Regarding this ‘interventionist’ approach, see, eg, J Jackson, ‘Human Rights, Constitutional Law and Exclusionary Safeguards in Ireland’ in P Roberts and J Hunter (eds), Criminal Evidence and Human Rights: Reimagining Common Law Procedural Traditions (Oxford, Hart Publishing, 2012) 136. 28 HR 19 februari 2013, ECLI:NL:HR:2013:BY5322, r.o. 2.4.5, NJ 2013/308 m.nt. BF Keulen. 29 See also BF Keulen in his annotation to HR 19 februari 2013, ECLI:NL:HR:2013:BY5322, NJ 2013/308 and T Schalken in his annotation to HR 9 september 2014, ECLI:NL:HR:2014:2650, NJ 2014/420. 30 However, in respect of the second type of violation, there is very limited room for the Court to nevertheless decide not to exclude the evidence. See, eg, Kuiper (n 16) 526. 31 HR 19 februari 2013, ECLI:NL:HR:2013:BY5322, r.o. 2.4.5, NJ 2013/308 m.nt. BF Keulen.

Coercive Human Rights and Unlawfully Obtained Evidence  273 In other words, what is envisaged in respect of evidence obtained by such a breach is a cost–benefit analysis (or balancing exercise), in which crime control or instrumentalist objectives concerned with the need to ensure that those who have committed a criminal offence are convicted function as a reason not to exclude the evidence. Given that, according to the Supreme Court, taking into account such objectives is also justified in light of the procedural coercive obligation,32 in this second category of procedural violation that may require exclusion, the procedural coercive obligation may be said to function as a reason not to exclude unlawfully obtained evidence.33 In the third category of pre-trial impropriety that may require the exclusion of evidence, which concerns procedural violations that form part of a wider pattern (ie, are structural in nature), the court is required34 to undertake the same cost–benefit analysis that it may undertake in the context of the second category of exclusion. In two of the three categories of exclusion, then, the positive obligations flowing from the ECHR constitute a reason not to exclude unlawfully obtained evidence (where applicable), that is, a potential cost in the cost–benefit analysis of whether to exclude such evidence. This particular feature of the exclusionary mechanism is in keeping with the overall framework for determining the consequences to be attached to a pre-trial procedural violation under Article 359a CCP, underlying which is the notion that such violations should not invariably lead to a particular response and that in determining how to respond, the court should also be mindful of facts and circumstances that might militate against a (certain) response. Also worth noting here is that, in addition to the factors set out in the second paragraph of Article 359a CCP, which the court is required to take into account, the court may take into account the seriousness of the offence with which the accused is charged as a reason not to attach a (particular) consequence to pre-trial impropriety.35 In the literature, the procedural coercive obligation has been linked to this factor.36 More is said about the relationship between this factor and the inclusion of the procedural coercive obligation in the test for excluding evidence in section V below. Suffice to say for now that both

32 The wording adopted by the Supreme Court is somewhat ambiguous as to whether positive obligations may be utilised to justify taking into account all of the objectives mentioned, or only the rights of victims or their next of kin. In a leading textbook on Dutch criminal procedure, the authors argue that it is the latter. See GJM Corstens/MJ Borgers and T Kooijmans, Het Nederlands strafprocesrecht, 9th edn (Deventer, Wolters Kluwer, 2018) 875. However, either way, it is fair to say that the Supreme Court sees the procedural coercive obligation as a matter of importance for victims. 33 See also Kuiper (n 16) 349 ff. However, it should also be noted that the wording adopted by the Supreme Court suggests that it is not exhaustive of the considerations that may militate against exclusion and, moreover, recognises that the procedural coercive obligation will not always be applicable. 34 HR 19 februari 2013, ECLI:NL:HR:2013:BY5322, r.o. 2.4.6, NJ 2013/308 m.nt. BF Keulen. 35 See, eg, HR 25 juni 2002, ECLI:NL:HR:2002:AD9204, r.o. 3.7–3.9, NJ 2002/625 m.nt. TM Schalken. 36 See Kuiper (n 16) 346 and 351; and M Samadi, ‘Het toezicht op de strafvorderlijke overheid: een modern artikel 359a?’ (2018) Platform Modernisering Strafvordering, fn 83 and accompanying text.

274  Kelly M Pitcher the seriousness of the offence and the procedural coercive obligation form part of a framework that heavily restricts the ability of the trial judge to attach legal consequences to pre-trial impropriety. III.  THE PROCEDURAL COERCIVE OBLIGATION UNDER THE ECHR

A.  Scope and Meaning As stated in section I above, under the ECHR, in the context of the positive obligations flowing therefrom, states are increasingly being required to take measures in the criminal law and criminal law-enforcement sphere. The ECtHR is thus increasingly imposing coercive obligations of both a substantive and a procedural nature on states. It is the latter type of obligation with which we are concerned here and, as already noted, it is in the context of Articles 2 and 3 ECHR that the procedural coercive obligation has matured. Accordingly, in the overview below, the focus is on these provisions37 and, because the case law under each provision may be said to be equally representative of the scope and meaning of the procedural coercive obligation,38 they will be dealt with together. Before looking to the content of the procedural coercive obligation, it is important to note that whether the obligation is applicable ultimately depends on the gravity of the human rights violation in question. For example, a violation of the right to life need not give rise to the obligation; it depends on the facts and circumstances of the particular case. On account of its intentional nature, murder, it seems, does give rise to a duty to conduct a criminal investigation and, if need be, prosecute.39 But death as a result of negligence or error of judgement need not do so (save in case of gross negligence),40 although this is not to say that in such cases, there would be no obligation to investigate at all. In such cases – and in cases involving other Convention rights – other measures, such as a civil action, may suffice.41 Although it all depends on the circumstances of the particular case,42 it is fair to say that only in respect of the most serious human rights offences will the institution of criminal proceedings be required.43 Under the case law of the ECtHR, then, human rights protection can be secured 37 However, it is important to note that it is not only in respect of these rights that states may be required to take measures in the criminal law-enforcement sphere. See Kamber (n 4) 326 ff. 38 Indeed, the Court has held that the scope and meaning of the procedural coercive obligation is the same under each. See, eg, Assenov and Others v Bulgaria 1998-VIII; (1999) 28 EHRR 652, para 102. 39 See, eg, Öneryıldız v Turkey [GC] 2004-XII; (2005) 41 EHRR 20, paras 91–96. 40 See, eg, ibid paras 92–93. 41 See, eg, DJ Harris, M O’Boyle, EP Bates and CM Buckley, Harris, O’Boyle & Warbrick: Law of the European Convention on Human Rights, 3rd edn (Oxford, Oxford University Press, 2014) 215. See also Akandji-Kombe (n 10) 35. 42 See, eg, Akandji-Kombe (n 10) 35. 43 See, eg, Seibert-Fohr (n 7) 224; and D Shelton, Remedies in International Human Rights Law, 2nd edn (Oxford, Oxford University Press, 2005) 395.

Coercive Human Rights and Unlawfully Obtained Evidence  275 in a variety of ways, ie, it does not require recourse to the criminal law in all cases.44 Regarding the institution of criminal proceedings, the ECtHR speaks of the need to ‘apply’ the criminal law provisions protecting Convention rights (as ‘backed up by law-enforcement machinery for the prevention, suppression and punishment of breaches of such provisions’)45 ‘in practice’, through effective investigation and prosecution. According to the ECtHR: ‘In order to be “effective” … an investigation … must be capable of leading to the identification and punishment of those responsible.’46 It is not solely when state authorities have committed the human rights offence that the state is under a duty to implement the criminal law and law-enforcement machinery (assuming that the human rights offence in question is of such a nature as to require such a response). Indeed, in cases involving human rights offences by private individuals also, there is a procedural coercive obligation on the state entailing an effective investigation capable of identifying and punishing those responsible.47 However, as Krešimir Kamber observes, while in cases in which the human rights offence has been committed by a private individual, the procedural requirements are the same as in cases in which it has been committed by state agents,48 the rationale underlying the procedural ­coercive obligation appears to be different. Whereas in cases in which the human rights offence was committed by state agents, an effective investigation ‘serves an important interest of protection from arbitrary and excessive use of force by state authorities’, Kamber argues that in cases involving a private individual, this is (logically) not a pertinent consideration.49 In cases involving a private individual, an effective investigation instead ‘serves … to ensure that acts of ill-treatment do not remain ignored by the relevant authorities and to provide effective protection against acts of ill-treatment’.50 While there would indeed appear to be a discrepancy in this regard, as will be seen below, regardless of who has committed the human rights offence, the question for the Court is ­ultimately whether the state can be said to have acted with impunity. An essential characteristic of the procedural coercive obligation is that it is an obligation of means, not result. In Kopylov v Russia, an Article 3 case involving

44 For a normative defence of this position, see, eg, Seibert-Fohr (n 7) 198 ff. 45 See, eg, Armani Da Silva v UK [GC] App No 5878/08 (ECtHR, 30 March 2016); (2016) 63 EHRR 12, para 95; Valiulienė v Lithuania App No 33234/07 (ECtHR, 26 March 2013), para 75. 46 See, eg, Ramsahai and Others v The Netherlands [GC] 2007-II; (2008) 46 EHRR 43, para 324; and Cestaro v Italy App No 6884/11 (ECtHR, 7 April 2015), para 204. 47 See Beganović v Croatia App No 46423/06 (ECtHR, 25 June 2009), paras 69–71; Nikolić v Croatia App No 5096/12 (ECtHR, 29 January 2015), para 39; and Mladenović v Serbia App No 1099/08 (ECtHR, 22 May 2012), para 51. See also the Court’s comments in Denis Vasilyev v Russia App No 32704/04 (ECtHR, 17 December 2009), para 100. 48 Kamber (n 4) 241. Accordingly, in setting out the procedural coercive obligation in this section, no separate mention will be made in the main text of the case law concerning human rights offences committed by private individuals. 49 ibid 241 f. 50 Beganović v Croatia (n 47) para 79. See also Valiulienė v Lithuania (n 45) para 77.

276  Kelly M Pitcher ill-treatment at the hands of the police, the Court explained the means-orientated character of the procedural coercive obligation as follows: An obligation to investigate ‘is not an obligation of result, but of means’: not every investigation should necessarily be successful or come to a conclusion which coincides with the claimant’s account of events; however, it should in principle be capable of leading to the establishment of the facts of the case and, if the allegations prove to be true, to the identification and punishment of those responsible.51

In Öneryıldız v Turkey, in which the procedural coercive duty was held to extend to the trial stage of criminal proceedings, the Court similarly held that Article 2 ECHR does not entail ‘an absolute obligation for all prosecutions to result in conviction, or indeed in a particular sentence’.52 However, in the same breath, it held that ‘national courts should not under any circumstances be prepared to allow life-endangering offences to go unpunished’.53 In what may be seen as an attempt to reconcile these two aspects of its case law on the procedural coercive obligation (albeit in the context of another Convention right, namely Article 8 ECHR), the Court in Söderman v Sweden held that ‘there is no absolute right to obtain the prosecution or conviction of any particular person where there were no culpable failures in seeking to hold perpetrators of criminal offences accountable’ (emphasis added).54 According to Kamber, the means-oriented character of the procedural coercive obligation should accordingly be taken to mean that ‘as long as the authorities have deployed their utmost efforts in the investigation and prosecution of those responsible, the inherent procedural difficulties or failures, such as those related to the limited epistemological capacities, cannot be held against them’.55 Regarding its position that national courts should not be prepared to allow serious human rights offences to go unpunished, it is worth noting that the ECtHR is increasingly looking to the sentence imposed in domestic criminal proceedings in determining whether a state has discharged its positive obligations. Thus, the Court has held that: [W]hile the Court should grant substantial deference to the national courts in the choice of appropriate sanctions for ill-treatment and homicide by State agents, it must exercise a certain power of review and intervene in cases of manifest disproportion between the gravity of the act and the punishment imposed. Were it to be otherwise, the States’ duty to carry out an effective investigation would lose much of its meaning, and the right enshrined by Article 2, despite its fundamental importance, would be ineffective in practice.56 51 Kopylov v Russia App No 3933/04 (ECtHR, 29 July 2010), para 132. See also Beganović v Croatia (n 47) para 71 for a case involving private violence. 52 Öneryıldız v Turkey (n 39) paras 95–96. 53 ibid para 96. 54 Söderman v Sweden [GC] ECHR 2013, para 83. 55 Kamber (n 4) 310. See also Seibert-Fohr (n 7) 134. 56 Nikolova and Velichkova v Bulgaria App No 7888/03 (ECtHR, 20 December 2007); (2009) 48 EHRR 40, paras 61–62. See also Cestaro v Italy (n 46) para 207.

Coercive Human Rights and Unlawfully Obtained Evidence  277 In her book Prosecuting Serious Human Rights Violations, Anja Seibert-Fohr argues that with the ECtHR increasingly seeking ‘to ensure that those responsible for serious human rights … [offences] serve an adequate sentence as a matter of general human rights protection’, the assertion that the procedural coercive obligation is one of means, not result ‘is losing ground’.57 While this is doubtless true and while it may even be possible to speak of a duty to punish – and an obligation of result – in concrete cases (ie, those cases in which the Court has established state liability on the basis of manifest disproportion between the gravity of the act and the punishment imposed), it is understandable that the ECtHR has continued to depict the procedural coercive obligation more generally as one of means rather than result.58 Among other things, this descriptor reflects the need, in prosecuting those alleged to be responsible for serious human rights offences, to respect the right of the accused to a fair trial and, importantly, conveys the message that what is therefore not envisaged under the procedural coercive obligation is punishment at all costs.59 More is said about the relationship between the procedural coercive obligation and the rights of the accused below. On the one hand, then, the procedural coercive obligation on the state is considered to be an obligation of means, not result. On the other hand, the ECtHR is increasingly looking to the punishment imposed in determining whether the state in question has discharged the procedural coercive obligation resting on it. Recalling here the reason for enquiring into the procedural coercive duty in this chapter in the first place, it should be noted that responses within the criminal trial (ie, judicial responses) to pre-trial procedural violations may prevent conviction and/or the imposition of commensurate punishment. Accordingly, the question arises as to when not convicting or not punishing commensurately will entail a violation of the procedural coercive obligation under the case law of the ECtHR. It is to this question that we now turn, in which regard it is instructive to look more closely at the rationale underlying the procedural coercive obligation. Against this backdrop, consideration will then be given to the ECtHR’s case law addressing failures on the part of the courts in the context of the procedural coercive obligation. However, before doing so, it is worth emphasising that the responsibility for a violation of the procedural coercive obligation need not always lie with the courts alone; in a given case, multiple state organs may be said to bear responsibility for such a violation.60 Moreover, it is possible for failings on the part of the investigative and prosecuting authorities to prejudice the court proceedings to such a degree that those

57 Seibert-Fohr (n 7) 138. 58 See, eg, Armani Da Silva v UK (n 45) paras 233 and 238. 59 Seibert-Fohr (n 7) 225 and also 115 f and 202 f. 60 See, eg, Nikolova and Velichkova v Bulgaria (n 56) paras 59–63. In this case, the military prosecution authorities and the courts both bore responsibility.

278  Kelly M Pitcher authorities, and not the courts to have subsequently acquitted the defendant, may be said to bear responsibility in this regard.61 Generally speaking, the procedural coercive obligation may be rationalised in two ways: as protecting human rights in general, whereby the preventive function of punishment is emphasised, or as protecting the individual victim, whereby the remedial function of punishment is emphasised.62 The first rationalisation, then, sees prosecution as a means of preventing the commission of human rights offences in the future, whereas the second sees it as a means of providing redress to the individual victim for harm suffered.63 Traditionally when the ECtHR has required the application in practice of the criminal law provisions protecting Convention rights, it has done so as a matter of general human rights protection, ie, as a matter of general deterrence or prevention. Thus, according to the Court, the fact that ‘national courts should not under any circumstances be prepared to allow … [serious human rights] offences to go unpunished’, which ‘is essential for maintaining public confidence and ensuring adherence to the rule of law and for preventing any appearance of tolerance of or collusion in unlawful acts’, means that its: [T]ask [is] therefore … [to review] whether and to what extent the courts, in reaching their conclusion, may be deemed to have submitted the case to the careful scrutiny required … so that the deterrent effect of the judicial system in place and the significance of the role it is required to play in preventing [serious human rights offences] are not undermined. (Emphasis added)64

At the same time, the Court has consistently rejected the claim that there exists an individual right to demand the prosecution or punishment of human rights offences.65 While it is true that the Court has sometimes cited retributive goals as underlying the procedural coercive obligation,66 this should not be taken to mean that what is envisaged in this regard is vengeance or revenge for the particular victim in question, or that there is an individual right to demand 61 See in this regard Kamber (n 4) 323 ff. For this reason, and because the responsibility for a violation of the procedural coercive obligation need not always lie with the courts alone, the depiction of the procedural coercive duty on the state as a duty to punish is unfortunate; it appears to imply that the responsibility for such a violation always ultimately lies with the courts. 62 Seibert-Fohr (n 7) 190. 63 Nevertheless, the distinction is not always easy to draw, as the notions are, as Seibert-Fohr observes, often mixed; ibid 282. See also Shelton (n 43) 396 and 465. Nevertheless, as Seibert-Fohr observes, even if prosecution and punishment is in the interest of a victim, ‘it does not follow that this is the result of [the] victim’s human right’; Seibert-Fohr (n 7) 208. 64 Öneryıldız v Turkey (n 39) para 96. See also Armani Da Silva v UK (n 45) para 239; and Cestaro v Italy (n 46) para 205. For a case in which the human rights offence in question was committed by a private individual, see Beganović v Croatia (n 47) para 77. 65 See, eg, Öneryıldız v Turkey (n 39) para 96. See also Armani Da Silva v UK (n 45) para 238; and Tadić v Croatia App No 10633/15 (ECtHR, 23 November 2017), para 66. See also Seibert-Fohr (n 7) 148 f. Nevertheless, there are cases in which criminal punishment has seemingly been regarded as an individual remedy for serious human rights offences. See Seibert-Fohr (n 7) 138 and 150 f, discussing Nikolova and Velichkova v Bulgaria (n 56). 66 See, eg, Jelić v Croatia App No 57856/11 (ECtHR, 12 June 2014); (2015) 61 EHRR 43, para 90.

Coercive Human Rights and Unlawfully Obtained Evidence  279 punishment. As Seibert-Fohr observes, even where at the national level retribution is considered to be a function of punishment, it does not follow that it is ‘exercised in the interest of the particular victim’, since ‘[r]etribution, as a matter of re-establishing the rule of law, is in the interest of society’.67 To the extent that the cases in question (and the language adopted therein) allow for a more ‘victim-centred’ reading of the notion of retribution in the context of the procedural coercive obligation, it is worth observing that such cases are open to criticism on the basis that it is not obvious from an international human rights law perspective why criminal punishment is necessary to protect the victim of a crime against, for example, his or her right to life.68 In any case, it is clear from the case law that the notion of general deterrence continues to constitute the primary justification for the procedural coercive obligation, while cases in which retributive goals feature in this regard remain relatively rare.69 General human rights protection, then, is concerned with the ability of the judicial system in place to prevent the commission of further human rights offences, and it is indeed when national courts have ‘allowed’ human rights offences to go unpunished – having applied the criminal law system in place ‘far from rigorously’70 or ‘fostered a “sense of impunity” instead of showing that such acts could in no way be tolerated’71 – that the Court has found a violation of the procedural coercive duty. More concretely, a national court may be said to have allowed a human rights offence to go unpunished when it has failed to engage, or genuinely engage, with the question of the defendant’s criminal responsibility.72 Indeed, while the Court does not see it its role as addressing ‘such issues of domestic law concerning individual criminal responsibility, that being a matter for assessment by the national courts, or to deliver guilty or not guilty verdicts in that regard’,73 it is prepared to address the efforts made by the national court, in terms of making the necessary enquiries and establishing the necessary facts in this regard, in order to determine such responsibility. However, it should be noted that even in cases in which the national courts have acquitted the individual or organ in question and the Court has found a violation of the procedural coercive obligation, it is shortcomings in the conduct of the court proceedings with which the Court appears to take issue, not the fact of the acquittal.74 A national court will also be considered to have allowed a human 67 Seibert-Fohr (n 7) 208. See also Kamber (n 4) 59. 68 See Seibert-Fohr (n 7) 151. See also 206 ff and Shelton (n 43) 12 and 396. However, a different view on this matter is possible, as is apparent from Kamber’s overview of the literature in this regard. See Kamber (n 4) 56 ff. 69 See also Lavrysen, ch 2 in this volume. 70 See, eg, Ali and Ayşe Duran v Turkey App No 42942/02 (ECtHR, 8 April 2008), para 72. 71 See, eg, Kopylov v Russia (n 51) para 141. 72 See, eg, Öneryıldız v Turkey (n 39) para 116. 73 ibid para 116. 74 See, eg, Galotskin v Greece App No 2945/07 (ECtHR, 14 January 2010), para 49. For a case in which the national courts’ conduct did not result in a violation of the procedural coercive obligation (despite an acquittal), see, eg, Otašević v Serbia App No 32198/07 (ECtHR, 5 February 2013), para 34.

280  Kelly M Pitcher rights offence to go unpunished in cases of ‘manifest disproportion between the gravity of the act and the punishment imposed’,75 especially when accompanied by the use of ‘sham reasons’ by the national court in order to justify the sentence in question,76 ie, the ‘de facto [use of] the legislative and punitive provisions of domestic law in order to prevent … effective conviction’.77 The ECtHR’s approach to failings on the part of national courts is in keeping with its approach to the procedural coercive duty on states more generally. This approach is captured well in the Grand Chamber’s observation in Söderman v Sweden (as set out above and affirmed in an extensive and persuasive study by Kamber) that ‘there is no absolute right to obtain the prosecution or conviction of any particular person where there were no culpable failures in seeking to hold perpetrators of criminal offences accountable’.78 ‘Culpable failures’ are also depicted in the case law as ‘significant flaws’,79 ‘any deficiency in the investigation which undermines its [effectiveness]’,80 ‘culpable disregard, discernible bad faith or lack of will on the part of the police or prosecuting authorities’81 and ‘sheer unwillingness to establish the truth and punish those responsible’.82 However, on their own, ‘objective difficulties that can be attributed to the passage of time or the loss of evidence’83 or ‘the inherent procedural difficulties or failures, such as those related to … limited epistemological capacities’84 will not in all likelihood undermine the effectiveness of the investigation. This is especially so when the authorities can otherwise be said to have acted diligently or promptly,85 submitted the case to careful scrutiny86 and provided sufficient reasoning for their decisions.87 As such, it is culpable failures in the investigation of serious human rights offences signifying impunity (that is, exemption from punishment) that the ECtHR takes issue with.88 And while it would appear possible for a singular incident to give rise to impunity such as to trigger state liability, in determining whether there was, in the circumstances, a procedural coercive duty on the state in question, a pertinent consideration would nevertheless appear

75 See n 56 and the accompanying text. 76 See Lavrysen, ch 2 in this volume. 77 Cestaro v Italy (n 46) para 224. 78 Söderman v Sweden (n 54) para 83. See also Kamber (n 4) 438. 79 See, eg, Söderman v Sweden (n 54) paras 90–91. 80 Armani Da Silva v UK (n 45) para 233. 81 Szula v UK App No 18727/06 (ECtHR, 4 January 2007). In Gäfgen v Germany [GC] ECHR 2010, (2011) 52 EHRR 1, para 124, the Court speaks of ‘attempting to conceal’ state violence. 82 Abakarova v Russia App No 16664/07 (ECtHR, 15 October 2015), para 98. 83 ibid. 84 See n 55 and the accompanying text. 85 See Kamber (n 4) 338, discussing Szula v UK (n 81) and Radionova v Russia App No 36082/02 (ECtHR, 26 March 2009). 86 Öneryıldız v Turkey (n 39) para 96. 87 See, eg, Radionova v Russia (n 85). 88 It is worth emphasising that this is no different in cases involving (violent) human rights offences committed by private individuals. See Beganović v Croatia (n 47) para 77, incorporating Ali and Ayşe Duran v Turkey (n 70) paras 61 and 62; and n 50 and the accompanying text.

Coercive Human Rights and Unlawfully Obtained Evidence  281 to be whether there existed a climate or culture of impunity. Thus, Seibert-Fohr argues that it is ‘[e]specially where there is a climate of impunity which gives rise to further serious human rights violations … [that] there is undoubtedly a duty to punish grave human rights abuses’ and that the ‘pronouncements of the international human rights bodies [including the ECtHR] should be understood in this sense since these usually concern systematic failures to prosecute serious human rights abuses’ (emphasis added).89 This, in turn, is consistent with the primary rationale of the procedural coercive obligation – general human rights protection – which is concerned with the ability of the judicial system in place to prevent the commission of further human rights offences. B.  Relationship to the Rights of the Accused Recalling again the reason for enquiring into the procedural coercive duty on the state in this chapter, it is instructive to consider the Court’s case law addressing the relationship between this duty and the rights of the accused, and the right of the accused to a fair trial in particular, in circumstances in which the duty has been invoked by the victim of a human rights offence (or his or her next of kin). The Court’s findings in Gäfgen v Germany are illustrative in this regard, given also that it addresses this relationship in the particular context of unlawfully obtained evidence. In that case, the procedural coercive obligation was considered not only to function as a reason not to exclude evidence (the function that may most readily come to mind in this context). Indeed, in affirming that serious human rights offences require two measures of redress – ‘a thorough and effective investigation capable of leading to the identification and punishment of those responsible’ and ‘an award of compensation’90 – the Court, sitting as a Grand Chamber, also entertained the notion that such obligations could constitute a reason to exclude evidence. In that case, it should be recalled that the serious human rights offence under consideration was not the child’s murder (for which Gäfgen had been convicted), but the ill-treatment suffered by the applicant at the hands of the police in the context of the investigation into the child’s disappearance (and which led to him disclosing the location of the child’s body). According to the Court, since ‘the question as to what measures of redress are appropriate and sufficient in order to remedy a breach of a Convention right depends on all the circumstances of the case’: It would not … exclude the possibility that in cases in which the deployment of a method of investigation prohibited by Article 3 led to disadvantages for an applicant in criminal proceedings against him, appropriate and sufficient redress for that 89 Seibert-Fohr (n 7) 202. Nevertheless, she recognises that it is not solely where there is a climate of impunity that the ECtHR will intervene, with ‘more recent cases … [being] concerned with singular incidents’. However, she is critical in this regard (at 140 f). 90 Gäfgen v Germany (n 81) paras 116–19.

282  Kelly M Pitcher breach may have to entail, in addition to the above-mentioned requirements, measures of restitution addressing the issue of the continuing impact of that prohibited method of investigation on the trial, in particular the exclusion of evidence obtained by breaching Article 3.91

However, the Grand Chamber did not consider it necessary to rule on whether, in the circumstances, this was required, given that the criminal and disciplinary sanctions imposed by the domestic authorities in respect of the impugned conduct could already be said to fall foul of the requirement of redress as established in the Court’s case law.92 Moreover, in Gäfgen, the applicant had complained that he had not had a fair trial under Article 6 ECHR, given that his conviction for the child’s murder had been based on evidence obtained in violation of Article 3 ECHR. Not excluding that evidence, he argued, violated his right to a fair trial. In response to the German government’s argument (in reliance on the procedural coercive obligation arising under Article 2 ECHR) that ‘they were obliged under the Convention to apply the criminal law against a murderer, and thus to protect the right to life’, the Grand Chamber held that while The Convention indeed requires that the right to life be safeguarded by the Contracting States … it does not oblige States to do so by conduct that violates the absolute prohibition of inhuman treatment under Article 3 or in a manner that breaches the right of every defendant to a fair trial under Article 6.93

Immediately thereafter, it acknowledged its well-established case law holding that the use of statements obtained (directly) by a violation of Article 3 ECHR will render the trial unfair.94 According to the Grand Chamber, the rationale for the exclusion of such evidence lies, among other things, in the need to deter the investigative authorities from adopting methods which violate Article 3 ECHR.95 It follows from the Grand Chamber’s findings in Gäfgen that in the case of the use of evidence in respect of which the ECtHR has adopted an ‘interventionist’ approach (notwithstanding its general approach to matters of evidence),96 there is no room (or, in the case of the use of the evidence which does not automatically – but will in most cases – render the proceedings unfair,97 very limited

91 ibid paras 128 and 116 (citations omitted). 92 ibid para 129. 93 ibid para 177 (citations omitted). 94 ibid para 178. 95 ibid. 96 See n 27 and the accompanying text. The ECtHR’s general approach, as first set out in Schenk v Switzerland Series A No 140 (1988); (1991) 13 EHRR 242, is that the admissibility of evidence is primarily a matter for regulation under national law and that its role is to determine whether the proceedings as a whole have been fair. 97 Indeed, depending on the type of unlawfully obtained evidence used, it may be more appropriate to speak of a ‘near-automatic’ rather than an automatic exclusionary rule, eg, where use of the evidence will ‘in principle’ irretrievably prejudice the rights of the defence. See Salduz v Turkey (n 24) para 55.

Coercive Human Rights and Unlawfully Obtained Evidence  283 room) to nevertheless conclude that the trial was fair on account of the procedural coercive duty on the state. However, in the case of evidence in respect of which the ECtHR does not adopt an interventionist approach, as is the case in relation to evidence obtained by violation of the right to privacy enshrined in Article 8 ECHR,98 the question of whether the use of the evidence rendered the trial unfair depends on a variety of factors, including (but not limited to) whether and how the evidence was used and whether the rights of the defence were respected (in particular, whether the defence was given an opportunity to challenge the authenticity of the evidence and to oppose its use).99 This ‘holistic’ approach adopted by the ECtHR in the context of Article 6 ECHR (so called because the ultimate question for the Court is whether the proceedings ‘as a whole’ were fair),100 entailing as it does an assessment of a wide variety of factors, would appear to give significant scope for accommodating the procedural coercive obligation on the state as a reason not to exclude evidence. Here it should be noted that in the context of this approach, the Court already attaches importance to the ‘quality’ or ‘strength’ of the evidence thereby obtained,101 thus opening the door to taking into account the public interest in the investigation and punishment of crime.102 At the same time, the scope of this obligation appears to be circumscribed by that which is ‘necessary and sufficient’ in order to be able to say that the trial was fair.103 This means observance of the Court’s (more) interventionist case law, but also of the safeguards that apply to the use of evidence in respect of which the Court does not take an interventionist approach, principally the opportunity of the defence to challenge the authenticity of the evidence and oppose its use. It can thus be said that there is a minimum core of fairness under Article 6 ECHR that is non-negotiable, even in the face of what at first glance may appear to be a compelling case for conviction or punishment. Of course, what comprises and, in particular, what should comprise this minimum core may be a matter of some debate, but a recognition of a core of fairness that is resistant to the coercive ‘push’ would seem essential for conveying the message that what is not envisaged under the procedural coercive obligation is punishment at all costs.104

98 See in this regard the cases cited in n 22 above. 99 See Gäfgen v Germany (n 81) paras 162–68. 100 ibid. 101 See, eg, Bykov v Russia [GC] App No 4378/02 (ECtHR, 10 March 2009), para 90. See also Khan v UK (n 22) para 37. 102 KM Pitcher, Judicial Responses to Pre-trial Procedural Violations in International Criminal Proceedings (The Hague, TMC Asser Press, 2018) 59 f and 63 f. 103 See Gäfgen v Germany (n 81) para 178. 104 See also nn 58–59 and the accompanying text. This in turn is essential for ensuring the credibility and legitimacy of prosecutions for human rights offences. See, eg, Shelton (n 43) 397; and Seibert-Fohr (n 7) 225. And see the Grand Chamber in Gäfgen v Germany (n 81) para 177.

284  Kelly M Pitcher IV.  IMPLICATIONS OF THE PROCEDURAL COERCIVE OBLIGATION FOR THE EXCLUSION OF EVIDENCE

The question is now what this all means for a judge in domestic criminal proceedings, seised with a case in which the crime charged involves a serious human rights offence (for example, murder), such that there is a procedural coercive duty on the state, and being called upon to provide relief in respect of pre-trial procedural violations, with such relief having the potential to prevent conviction and/or punishment. However, in seeking to distil from the Court’s case law implications for the judge in domestic proceedings in this respect, it is important to be mindful of the nature of the standards set by the Court. Such standards tend to be framed negatively, reflecting the margin of appreciation afforded to states, the ‘culpable failures’ standard being a case in point. That standard gives expression to what states should not do, and determining on this basis what states should do in order to avoid a violation of the procedural coercive obligation is not a straightforward matter. With the difficulties of transposing standards set by supranational bodies to the domestic context105 well in mind, the following observations may nevertheless be made. As stated in the previous section, the scope of the procedural coercive obligation is circumscribed by the right to a fair trial enshrined in Article 6 ECHR, ie, the minimum core thereof. Among other things, this means that the procedural coercive obligation cannot be invoked to justify the use of evidence which, in the Court’s case law, automatically renders the proceedings unfair, or to override the essential safeguards that apply to the use of evidence in respect of which the Court does not take an interventionist approach. However, when it is not the minimum guarantees under Article 6 ECHR that are (directly) at stake, but some other conception of fairness or principle affording protection to the individual to have been drawn into the criminal process from improper interference by the authorities (such as the notion inherent to the rule of law that the authorities are also bound by the law), or some other principle of criminal procedure or criminal justice (such as the need to preserve the integrity of the proceedings), the question is whether, and to what extent, the procedural coercive duty on the state may or should constitute a reason not to exclude evidence. While the ECtHR does not appear to have pronounced itself on this particular issue, its case law is nevertheless instructive in this regard. It is clear from the previous section that not punishing, or not punishing commensurately (provided the punishment is not manifestly disproportionate), a serious human rights offence will not always mean a failure to discharge the procedural coercive obligation. To begin with, the procedural coercive obligation is one of means, not result. Moreover, in the case law of the ECtHR, it is

105 See in this regard Lavrysen, ch 2 in this volume, where the author observes that ‘a downside of the Court’s deferential approach is that it risks providing little guidance to domestic authorities’.

Coercive Human Rights and Unlawfully Obtained Evidence  285 culpable failures on the part of the state authorities that have triggered state liability under this obligation. In other words, it is when national courts allow human rights offence to go unpunished – as apparent from, for example, a failure to genuinely engage with the question of the defendant’s criminal responsibility or a manifest disproportion between the gravity of the act underlying the human rights offence in question and the sentence imposed – that the ECtHR will intervene. Given the reason for enquiring into the procedural coercive obligation in this chapter, it is worth emphasising that not punishing will not necessarily be the result of a culpable failure, understood as a (discernible) flaw or deficiency, or (flagrant) disregard, bad faith or unwillingness on the part of the state authorities. Insofar as culpable failures may be depicted in terms of impunity, not punishing will not always amount to impunity. Indeed, it may simply be the collateral damage of a decision legitimately premised on some principle of criminal procedure or criminal justice. For example, the unlawful manner in which evidence was obtained may have such a bearing on its reliability that its use would be inconsistent with the notion that the criminal justice system should protect those who are innocent from wrongful conviction. The decision to exclude evidence may thus be premised on legitimate epistemic considerations, ie, considerations concerned with promoting accurate verdicts.106 In light of the ECtHR’s case law holding that ‘objective difficulties’ will not likely undermine the effectiveness of the investigation, it seems unlikely that the exclusion of evidence on this basis (where it results in an acquittal) would raise issues under the ECHR, assuming, of course, that the decision to exclude was made in good faith. However, the justification for the exclusion of unlawfully obtained evidence need not lie in epistemic considerations, and the question then is how a decision to exclude evidence premised on non-epistemic considerations – ie, considerations which are concerned with criminal justice objectives other than the promotion of accurate verdicts, such as the need to preserve the integrity of the criminal justice system – relates to the procedural coercive obligation, in circumstances in which the decision has resulted in an acquittal. As apparent from the previous section, for the ECtHR, the ultimate question is whether the judicial system in place is capable of preventing further human rights offences. In turn, the deterrent effect of the judicial system in place would appear to be tied up with its ability to maintain public confidence, ensure adherence to the rule of law and prevent the appearance of tolerance of unlawful acts.107 Regarding the question just raised, two observations may be made. 106 M Redmayne, ‘Confronting Confrontation’ in Roberts and Hunter (n 27) 289. It is important to note here that the decision to admit (ie, not to exclude) evidence may also be premised on epistemic considerations, that is, the (corresponding) aim of ensuring that those who have committed a criminal offence are convicted. It might thus be argued that the fact that the (unlawfully obtained) evidence is reliable means that it should be used. In the literature, this is known as the ‘reliability’ argument, pursuant to which ‘determining the truth of the criminal charges is the sole purpose of the criminal trial, and evidence should be admitted or excluded solely on that basis’. Ashworth (n 1) 723. 107 See the Grand Chamber’s findings in Öneryıldız v Turkey (n 39).

286  Kelly M Pitcher First,  decisions by national courts premised on non-epistemic considerations may be aimed at reinforcing goals closely related to those seemingly lying at the heart of a judicial system capable of deterring human rights offences (and in any case underlying the procedural coercive obligation).108 Indeed, in theoretical discussions at the national level on unlawfully obtained evidence, the (non-epistemic) rationales for exclusion – for example, the deterrence rationale,109 the protective principle110 and the integrity rationale111 – are often expressly linked to such goals. Thus, the exclusion of unlawfully obtained evidence may be seen as a means of re-establishing the rule of law or by which the judiciary can distance itself from executive unlawfulness, or of promoting public confidence in (or the legitimacy of) the criminal justice system. Of course, decisions based on epistemic considerations may also promote (most of) the aforementioned goals (and this includes decisions to use evidence which has been unlawfully obtained), but the point is that the achievement of these goals and, by extension, of a criminal justice system that may be said to be legitimate is dependent on more than ensuring that those who commit criminal offences are convicted and punished. Indeed, adherence to the rule of law requires more than enforcing (or reinforcing) the legal rules in place criminalising wrongdoing and conferring on the authorities powers to investigate and prosecute crime; it also requires observance of the legal norms regulating the exercise of such powers.112 The same may be said of maintaining public confidence. Because the ECtHR is committed to maintaining public confidence, ensuring adherence to the rule of law and preventing the appearance of tolerance of unlawful acts in the context of the procedural coercive obligation, there would appear to be room thereunder to accommodate decisions to exclude unlawfully obtained evidence. Moreover, given that it is tied up with these goals, it may be argued that the deterrent value of the judicial system in place is dependent on more than the conviction and punishment of those who have committed human rights offences alone. The point here is not so much that human rights protection can be secured in a

108 That is, whether as autonomous goals or as elements of the notion of deterrence. While in Öneryıldız v Turkey the Court seemed to link the deterrence value of the judicial system in place to its ability to maintain public confidence, ensure adherence to the rule of law and prevent the appearance of tolerance of unlawful acts, in other judgments, this link is not so explicit. See, eg, Nikolova and Velichkova v Bulgaria (n 56) paras 57 and 60. 109 Pursuant to this rationale, courts should attach legal consequences to such violations to deter such violations in the future. 110 See n 1 and the accompanying text. 111 For an overview of this argument, see P Mirfield, Silence, Confessions and Improperly Obtained Evidence (Oxford, Oxford University Press, 1997) 23 ff. 112 This argument is captured well in the oft-cited quote from Judge Brandeis’ dissenting opinion in Olmstead v US 277 US 438 (1928): ‘Decency, security, and liberty alike demand that government officials shall be subjected to the same rules of conduct that are commands to the citizen. In a government of laws, existence of the government will be imperiled if it fails to observe the law scrupulously. Our government is the potent, the omnipresent teacher. For good or for ill, it teaches the whole people by its example. Crime is contagious. If the government becomes a lawbreaker, it breeds contempt for law; it invites every man to become a law unto himself; it invites anarchy.’

Coercive Human Rights and Unlawfully Obtained Evidence  287 variety of ways and need not entail recourse to the criminal law,113 but that other goals than the need to convict and punish those who have committed a human rights offence – ie, those which in a criminal justice context might be depicted as non-epistemic in nature – may well (also) bolster the deterrent effect of the judicial system in place. Indeed, the exclusion of evidence obtained unlawfully by state authorities, for example, by violation of Article 3 ECHR, may deter those same authorities from doing the same in the future.114 However, it may bolster the judicial system’s ability to prevent human rights offences committed by private individuals also, albeit more indirectly, on the basis of the idea that private citizens are more likely to abide by the law (including that which criminalises human rights offences) when the government does so too (and is made to face the consequences when it does not).115 The second observation worth making here is that an acquittal will not automatically undermine the deterrent effect of the judicial system already in place. Thus, the fact that a prosecution for a serious human rights offence (for example, murder) has not resulted in punishment will not necessarily mean that the criminal law system in question is not being implemented effectively and, by extension, not providing for effective deterrence. A relevant consideration for the Court in this regard would seem to be whether, at the time of adjudication, there existed a culture of impunity in respect of the type of violence in question, whether committed by agents of the state or private individuals.116 Against the backdrop of a climate or culture of impunity, it may be more difficult for courts to avoid the appearance of tolerance of or collusion in unlawful acts where not punishing is not the result of a decision aimed at protecting the individual on trial from wrongful conviction, but of one premised on non-epistemic considerations, such as the need to preserve the integrity of the proceedings. Returning to the question posed at the beginning of this section, it is trite to observe that courts should not allow human rights offences to go unpunished in circumstances in which there is a convincing case for establishing criminal responsibility and imposing punishment. However, under the Court’s case law, it would seem possible for a court to take a case seriously and make decisions that ultimately prevent it from doing either or both of the aforementioned. Because what matters is whether the judicial system in place is capable of deterring the commission of human rights offences, courts need not shy away from making decisions that ultimately prevent them from establishing criminal responsibility and/or punishing in an individual case (and which may be unpopular among certain stakeholders), but which are aimed at preserving the legitimacy of the criminal justice system in the long run. Nevertheless, courts should remain mindful of the context in which they are operating. In particular, where there is a

113 See

in this regard nn 43–44 and the accompanying text. n 95 and the accompanying text. 115 See n 112 above. 116 See in this regard n 89 and the accompanying text. 114 See

288  Kelly M Pitcher culture or climate of impunity in respect of the type of violence in question, they should ask themselves what would do more to bolster the deterrent value of the system: establishing criminal responsibility (assuming, of course, there is sufficient evidence to do so) and imposing commensurate punishment, or making decisions aimed at promoting (more long-term) non-epistemic goals (but which prevent the court from establishing criminal responsibility and/or punishing in the case at hand). Moreover, in order to avoid the appearance of tolerance of or collusion in unlawful acts, courts would do well to give reasons for (ie, explain carefully) their decisions to exclude evidence, whether premised on epistemic or non-epistemic considerations. V.  THE PROCEDURAL COERCIVE OBLIGATION AS A REASON NOT TO EXCLUDE EVIDENCE AND THE RISK OF COERCIVE OVERREACH: RETURNING TO THE DUTCH EXAMPLE

Having set out the procedural coercive obligation under the ECHR and discussed the implications thereof for the exclusion of evidence in domestic criminal proceedings, the question is now how the Dutch law and practice, as an example of how the procedural coercive obligation might be treated as a reason not to exclude evidence which has been obtained unlawfully, is to be evaluated in light of the risk of coercive overreach. In keeping with the focus of this chapter – how state authorities, and courts in particular, treat their coercive human rights obligations – the question is how the Dutch Supreme Court’s inclusion of the procedural coercive obligation in the test for excluding unlawfully obtained evidence is to be assessed in light of the risk of that obligation being used by lower courts to justify an overly coercive approach to pre-trial impropriety. It should be recalled that there is potential for coercive overreach when coercive human rights obligations are not defined properly. As Lazarus observes, what is needed in this regard is careful framing of the scope of such obligations, reflecting how much coercion is required in order to fulfil them.117 It was seen above that in two of the three categories of exclusion prescribed by the Dutch Supreme Court, the positive obligations flowing from the ECHR constitute a reason not to exclude evidence, while in one category – ie, that premised on the ECtHR’s case law regarding pre-trial procedural violations which more or less automatically violate the right to a fair trial – they do not play a role at all. The Dutch test for exclusion therefore recognises that the procedural coercive obligation is circumscribed by the right to a fair trial (ie, by the minimum core thereof) and therefore that what is not envisaged under the procedural coercive obligation is conviction or punishment at all costs. Moreover, regarding the other two categories of exclusion, in which the procedural coercive obligation



117 Lazarus

(n 6) 147 f.

Coercive Human Rights and Unlawfully Obtained Evidence  289 enters the determination through a cost–benefit analysis, the Supreme Court’s case law seems to recognise that the procedural coercive obligation will not always apply. Nevertheless, the manner in which the procedural coercive obligation is otherwise depicted in this context is a cause for concern. It should be recalled that in the case law, the Supreme Court links the procedural coercive obligation, which it depicts as an obligation of ‘punishment’, to the rights of victims (as well as other instrumentalist considerations). Moreover, it makes no mention of the rationale of general human rights protection, which, it should be recalled, sees the procedural coercive obligation as a mechanism by which to prevent human rights offences in the future. All of this allows for an interpretation of the procedural coercive obligation that sees prosecution and punishment as a form of redress for the individual victim or his or her next of kin. Such an interpretation would not only be at odds with the ECtHR’s case law as regards the primary rationale of the procedural coercive obligation (in which, moreover, the Court has consistently rejected the claim that there exists an individual right to demand the prosecution or punishment of a human rights offence). Indeed, an understanding of the procedural coercive obligation as a means of providing individual redress might also divert attention away from alternative, less coercive measures of human rights protection, ie, measures which lie outside the criminal law-enforcement sphere.118 This is because, against the backdrop of an understanding of the procedural coercive obligation as a matter of redress for the individual victim and to the extent that this understanding sees punishment as a necessary means of repairing the harm caused to the victim in terms of his or her ‘value or dignity’ (ie, the ‘moral injury’ caused),119 there would always seem to be cause for punishment (in circumstances in which there is sufficient evidence to do so, at least). From the perspective that accountability for human rights offences can be secured in a variety of ways and that prosecution of serious human rights offences ‘is not an end in itself’, an approach which treats recourse to the criminal law as the primary means of providing human rights protection may be said to be overly coercive.120 However, in and of itself, a ‘more’ coercive approach than is strictly necessary need not be contrary to human rights law. This is because the Court’s case law with respect to the procedural coercive obligation may also be understood as providing for minimum requirements,121 and states are therefore not precluded from adopting a more coercive approach to human rights offences than that required under the case law. At the same time, and further to the observations made in the previous paragraph, an approach which is overly coercive sits

118 See, eg, Seibert-Fohr (n 7) 198. 119 As Hampton appears to see it: J Hampton, ‘Correction Harms versus Righting Wrongs: The Goal of Retribution (1992) 39 UCLA Law Review 1659, 1666. 120 Seibert-Fohr (n 7) 214 and 286. See also n 44 and the accompanying text. 121 The Court thus speaks of ‘the minimum threshold of the investigation’s effectiveness’. See, eg, Armani Da Silva v UK (n 45) para 234.

290  Kelly M Pitcher somewhat uneasily with the idea that the deterrent effect of the judicial system in place is dependent on more than the conviction and punishment of those who have committed human right offences alone. Thus, even if an approach which may be said to be overly coercive need not be contrary to the Court’s case law with respect to the procedural coercive obligation (on the understanding that such case law provides for minimum requirements), it may nevertheless be said to be problematic from the perspective of those national goals of criminal justice wholly unrelated to the epistemic goal of ensuring that those who have committed a criminal offence are convicted. It may thus be lamented for giving primacy to crime control or instrumentalist objectives over, for example, the (non-epistemic) notion inherent to the rule of law that the authorities are also bound by the law, or of preserving the integrity of the proceedings. The question of which goals of criminal justice should take precedence lies at the heart of the debate surrounding the question of how pre-trial impropriety should be addressed. A question often raised in this regard is whether or not the seriousness of the offence – ie, the fact that the crime with which the accused is charged is (very) serious – is a factor to which the judge determining whether to exclude unlawfully obtained evidence should attach significant importance. In the Dutch literature, it has been argued that the procedural coercive obligation reinforces the notion that this is a factor to which the court should have due regard as a reason not to exclude evidence.122 Further support for this notion has been sought in the fact that the ECtHR itself has, on occasion, entertained the notion that the public interest in the investigation and prosecution of crime is a factor relevant to the determination of whether the use of unlawfully obtained evidence violated the right to a fair trial.123 However, this case law has been criticised for putting ‘pragmatism before principle’ (and, moreover, for signifying a departure from earlier case law in which the Court seemed to distance itself from the inclusion of such considerations in the determination of Article 6 ECHR).124 At the domestic level too, including in the Netherlands, the notion and practice of taking into account the seriousness of the offence as a reason not to exclude evidence is not uncontroversial, due, among other things, to its capacity to lead to consequentialist reasoning. The same may be said of taking into account the strength of the evidence – ie, of attaching importance to the probative value of the evidence – or the fact that it is reliable.125 This need not be objectionable, of course, if what is envisaged is an approach to pre-trial impropriety that gives primacy to crime control or instrumentalist objectives. But if such objectives are 122 See, eg, Kuiper (n 16) 343 ff. 123 ibid 344. 124 See A Ashworth, ‘The Exclusion of Evidence Obtained by Violating a Fundamental Right: Pragmatism before Principle in Strasbourg Jurisprudence’ in Roberts and Hunter (n 27). 125 Indeed, both factors may be said to be relevant to the epistemic aim of ensuring that those who have committed a criminal offence are convicted. It is where the evidence is sufficiently probative or, at least, has not been rendered unreliable by the unlawful manner in which it was obtained, that the seriousness of the offence takes on meaning.

Coercive Human Rights and Unlawfully Obtained Evidence  291 the justification for a restrictive approach to pre-trial procedural violations, this needs to be acknowledged, however uncomfortable a truth this may be. With respect to the cost–benefit analysis incorporated into the Dutch test for exclusion, it is worth observing here that what is troubling is not so much the idea that the state’s human rights obligations could constitute a reason not to exclude evidence, as the idea that they were incorporated into the analysis in an attempt to justify from a rights perspective an approach that was already putting instrumentalist objectives first.126 However, whatever the justification for a coercive course of action, articulation thereof is essential for a focused and meaningful discussion on what is lost and gained when such a course is adopted.

126 It is worth noting in this regard that the cost–benefit analysis through which the procedural coercive obligation may be factored into the equation was introduced at a time when the response was already (highly) restrictive and therefore could fairly be rationalised primarily, if not solely, in terms of crime control or instrumentalist objectives, as indeed it was in some quarters. See, eg, MJ Borgers, ‘De toekomst van artikel 359a Sv’ (2012) Delikt en Delinkwent 257, 265 f. However, it is important to note that it has also been rationalised in terms of rights protection. Yet, it is questionable whether this account of the Dutch judicial response to pre-trial procedural violations captures its true nature. See Pitcher (n 102) 119 ff.

292

Postscript Coercive Human Rights in Times of Coronavirus NATASA MAVRONICOLA AND LAURENS LAVRYSEN

W

hen we were putting the finishing touches to the manuscript of this volume in early March 2020, we hardly expected that we would be writing such a postscript with COVID-19 having so rapidly altered our lives in such a radical way. At the time of writing this postscript in May 2020, states across the world have mobilised substantial institutional and financial resources towards addressing the profound and complex challenges posed by the COVID-19 pandemic. Many of the measures taken in this context have a strong coercive dimension, in particular the extensive implementation of ‘lockdown’ or ‘quarantine’ measures, often backed up by novel criminal law provisions and penalties, and sometimes implemented by heavy-handed policing.1 The taking of such measures can be, and has been, framed as involving the discharge of positive obligations under key human rights, notably the right to life and physical integrity.2 In this regard, it is tenable that courts may consider that states bear a positive obligation to take certain ‘lockdown’ measures in response to the pandemic or that governments may more broadly invoke the rhetorical power of positive obligations to protect the right to life and physical integrity to justify such coercive measures. In this context, the readiness with which duties to protect are translated into duties to coerce – that is, to mobilise the sharp end of the state and its law-enforcement mechanisms – warrants particular critical reflection. In the spirit of this volume to subject the coercive approach to positive obligations to ‘continued critical scrutiny’, at least three sets of concerns should be raised.

1 Bonavero Institute of Human Rights, ‘Assessment of Legislative and Regulatory Responses to the COVID-19 Pandemic across 11 Jurisdictions’, Bonavero Report No 3/2020, 6 May 2020, 11 and 13. 2 See, for example, Joint Committee of Human Rights, ‘The Government’s Response to COVID-19: Human Rights Implications’, https://committees.parliament.uk/work/218/the-governments-responseto-covid19-human-rights-implications.

294  Natasa Mavronicola and Laurens Lavrysen First, framing ‘lockdown’ measures as being taken in fulfilment of positive obligations under human rights law should not obscure or divert from the imperative of properly scrutinising the far-reaching and often unequally experienced restrictions on human rights that result from such measures and the way in which they are implemented, let alone delegitimise such scrutiny. The protection of public health may justify certain restrictions of human rights, perhaps even ‘exceptional’ measures against the backdrop of a public health emergency, but it does not give licence to states to act in ways that are unlawful under the Convention (such as inhuman or degrading treatment, or restrictions on qualified rights that are not necessary and proportionate, or, in the case of lawful derogation under Article 15, that go further than what is strictly required by the exigencies of the emergency situation). Nor can the scope of positive obligations to protect the right to life and physical integrity encompass such measures.3 Reflecting on the danger of coercive overreach, it is particularly important to remain vigilant in respect of the possibility that certain governments may use the public health emergency as an opportunity or pretext to (further) entrench illiberalism in the name of the protection of the right to life and physical integrity. In addition, even where governments employ far-reaching coercive powers in a good faith manner to respond to the threat to the right to life and physical integrity, such powers may become excessive or even abusive if they remain in place for longer than the public health emergency requires.4 Second, in the context of an emphasis on the coercive dimension of the protection of the right to life and physical integrity, the pandemic also ­underscores the importance of ascertaining and addressing the less immediate but potentially grave indirect harms – including harms to bodily and mental integrity – arising out of certain coercive measures. This is particularly relevant, in the current circumstances, where the risk-mitigation pursued by restricting people’s movement gives rise to or exacerbates other risks to life and physical and mental integrity (mental health and domestic abuse being two prime examples). Third, the pandemic highlights how important it is not to reduce positive obligations to protect the right to life and physical integrity to the taking of coercive measures. Positive obligations are relevant in shaping access to healthcare and to protective equipment, requiring the protection of vulnerable patients or persons put in harm’s way,5 and indeed demanding measures to address risks created or exacerbated by ‘lockdown’ or quarantine conditions. Moreover, positive obligations may also be relevant in the context of the investigation of deaths resulting from alleged negligence or systemic and structural failings in

3 See also N Mavronicola, ‘Positive Obligations in Crisis’, Strasbourg Observers, 7 April 2020. 4 A Donald and P Leach, ‘Human Rights: The Essential Frame of Reference in the Global Response to COVID-19’, Verfassungsblog, 12 May 2020. 5 E Stubbins Bates, ‘Article 2 ECHR’s Positive Obligations: How Can Human Rights Law Inform the Protection of Health Care Personnel and Vulnerable Patients in the COVID-19 Pandemic?’, Opinio Juris, 1 April 2020.

Postscript  295 responding to the COVID-19 outbreak.6 Contemplating the operation of the investigative obligation in this context makes it necessary to acknowledge the limitations of the criminal law. Reducing such investigation to the mere determination of individual criminal responsibility may not allow for the proper identification of the structural deficiencies (the ‘rotten orchard’) at play that must be addressed in order to remedy the ongoing crisis or to be better prepared for future crises. More broadly, as the current crisis dramatically confronts us with the interdependence of human rights, it is vital that the human rights approach to the pandemic takes socio-economic rights, and notably the right to health,7 more seriously. We write this postscript in a sombre mood, but with an element of hope. We believe that the current circumstances deepen the importance of critically scrutinising coercive human rights, but also that they are capable of hastening the adoption of a wider outlook on what a protective approach to human rights can and ought to entail.

6 P Bowen, ‘Learning Lessons the Hard Way: Article 2 Duties to Investigate the Government’s response to the Covid-19 Pandemic’, UK Constitutional Law Blog, 29 April 2020, https:// ukconstitutionallaw.org/2020/04/29/paul-bowen-qc-learning-lessons-the-hard-way-article-2-dutiesto-investigate-the-governments-response-to-the-covid-19-pandemic. 7 See, eg, European Committee of Social Rights, ‘Statement of Interpretation on the Right to Protection of Health in Times of Pandemic’ (adopted by the Committee on 21 April 2020); UN Committee on Economic, Social and Cultural Rights, ‘General Comment No 14: The Right to the Highest Attainable Standard of Health (Art 12)’, 11 August 2000, E/C/12/2000/4.

296

Index A Amnesties  47, 65, 200, 231, 243 Anti-impunity agenda international trend  8–12 promise and pitfalls of the anti-impunity agenda  8–12 suggested areas for future research  26 Arbitrary deprivation of liberty (Art 5)  1 C ‘Carceral feminism’  13, 103–4 Case law Art 3 and vulnerability  95–7 deterrence as key factor general context  56–8 limits on national policy set by Convention  66–8 obligation to criminalise  58–61 obligation to investigate  61–5 overview  55–6 emergence of ‘coercive human rights law’  161–2 empirical issue of effectiveness  150–2 gross human rights violations common denominators  84–7 concluding remarks  92 implications for victims  75–80 implications of retributive responses  87 individual and general measures adopted by Court  88–9 obligation to investigate  91–2 overview  71–2 retributive responses as reparations  72–5 selected cases  83–4 significance of finding state responsibility  90 hate speech freedom of expression (Art 10)  121–3 obligation to prosecute  124–8 ‘horizontal effect’ of human rights jurisprudence  147–8

impact on coercive human rights in domestic criminal law encouragement of ‘culture of conviction’  169–71 example of Modern Slavery Act 2015  172–81 pitfalls of the jurisprudence  181–2 state obligations in criminal matters  166–9 increasing influence on national criminal law systems  141 no distinction between substance and procedure  31 obligation to criminalise other Convention provisions  41–3 privacy (Art 8)  39–41 prohibition of slavery, servitude and forced labour (Art 4)  37–9 prohibition of torture or ill-treatment (Art 3)  35–7 right to life (Art 2)  33–5 obligation to punish limitations  44–5 main developments  46–50 obligation upon states to identify migrants as victims of harm identification of victims – Chowdury and Others v Greece  218–20 identification of victims – J and Others v Austria  214–17 identification of victims – LE v Greece  211–14 identification of victims – Rantsev v Cyprus and Russia  209–11 overview  208–9 Children coercive power of vulnerability ‘ambiguity’ of role of state  105–6 non-sexual abuse  103–5 overview  98 sexual abuse  99–103 integration into family  148 positive obligations doctrine  40 prohibition of slavery, servitude and forced labour (Art 4)  209, 214

298  Index right to life (Art 2)  234 risk of coercive overreach  111–16 Coercive human rights balancing of ‘shield’ and ‘sword’ functions  2 broader international trend anti-impunity agenda  8–12 promise and pitfalls of the anti-impunity agenda  12–14 dangers of aligning human rights protection with criminal law need for protection rather than coercion  201–2 overview  183–4 positive duties applicable in different contexts  184–6 problem of diversion  198–201 risk of coercive overreach  187–92 risk of diluting human rights standards  192–8 underlying premise  184 development by ECtHR  1–2 historical context in ECtHR changes from late 1990s  7–8 development of positive obligations doctrine  6–7 X and Y v The Netherlands  4–6 impact of ECtHR case law on domestic criminal law encouragement of ‘culture of conviction’  169–71 example of Modern Slavery Act 2015  172–81 pitfalls of the jurisprudence  181–2 state obligations in criminal matters  166–9 implications see Implications of coercive human rights justifications see Justifications limitations of criminal law approach see Limitations of criminal law approach relevance during COVID-19 pandemic framing ‘lockdown’ measures  294 harms to bodily and mental integrity  294 overview  293 protection of life  294–5 subject of growing debate and academic interest  2–4

suggested areas for future research evolution and impact of doctrine in systemic terms  24–5 overview  22 reception, perception and effects of doctrine  22–3 what is excluded or missed  26 three central themes implications  18–21 justifications  16–17 limitations of a criminal law approach  21 Coercive overreach central theme of coercive human rights  18–21 dangers of aligning human rights protection with criminal law overview  187–8 prohibition of torture or ill-treatment (Art 3)  188–92 right to life (Art 2)  188–92 framing ‘lockdown’ measures for COVID-19  294 preventive operational obligations  255 overview  260–1 policing and Osman warnings  262–4 preventative risk orders and modern slavery  264–6 procedural coercive obligation under ECHR  288–91 risk from asserting coercive duties  55–6 use of positive obligations doctrine  53 vulnerability  111–16 Conviction see ‘Culture of conviction’ Council of Europe assistance and protection for victims of crime  205 complaints against the European Commission of Human Rights  235 consensus towards criminalisation  61 document on the effectiveness of the Convention  153 emphasis on rehabilitation and reintegration  67 impact of enlargement  82 monitoring mechanism on racism and intolerance  118 systematic approach to hate speech  121 Counter-terrorism coercive overreach of Art 2  190 ‘culture of conviction’  173, 182

Index  299 culture of risk aversion  255 international principles on counter-terrorism  225–6 preventative obligations  250–7 principle of legality  230 punitive trends  164–5 COVID-19 pandemic framing ‘lockdown’ measures  294 harms to bodily and mental integrity  294 overview  293 protection of life  294–5 Criminalisation as last resort content of principle  144 legitimately expressed concerns  142 meaning and effect  143–4 negation by positive obligations doctrine  157–9 origins and development  145–7 overview  142–3 ‘principle of legislative ethics’  144–5 role on European law systems  143 theoretical perspectives  145 ‘Culture of control’  164 ‘Culture of conviction’ encouragement by ECtHR  169–71 impact of ECtHR case law on domestic criminal law example of Modern Slavery Act 2015  172–81 state obligations in criminal matters  166–9 overview  161–3 pitfalls of the jurisprudence  181–2 rise of penal state in Europe  163–6 D Deterrence anti-impunity view  12 ‘effective deterrence’ central theme  16–17 coercive power of vulnerability  97–103, 109 meaning of effectiveness  153–4 obligation to criminalise  58–9 positive obligations doctrine  65, 150–2 privacy (Art 8)  39 unlawfully obtained evidence  287 victim protection  20 as key factor general context  56–8 limits on national policy set by Convention  66–8

obligation to criminalise  58–61 obligation to investigate  61–5 overview  55–6 obligation to punish  48–50 substance and procedure distinguished  31 Domestic criminal justice systems binding effect of Convention  5 coercive power of vulnerability  105–6, 109 Convention as limitation of policy  66–8 ‘culture of conviction’ overview  161–3 rise of penal state in Europe  163–6 failure to hold police officers accountable  36 failure to protect members od LGBTIQA+  110 impact of ECtHR case law on coercive human rights encouragement of ‘culture of conviction’  169–71 example of Modern Slavery Act 2015  172–81 pitfalls of the jurisprudence  181–2 state obligations in criminal matters  166–9 inadequacies of criminal legislation  23 obligation to punish  44–7, 50–3 procedural coercive obligation under ECHR  276–81 tailoring of criminal law protection  41 unlawfully obtained evidence example of Dutch judicial response  270–4 overview  267–70 Domestic violence coercive power of vulnerability  106–9 historical context of coercive human rights  8 limitations of a criminal law approach  21 obligation to punish  49 preventive operational obligations  253–5 Drittwirkung doctrine  5–6 E ‘Effective deterrence’  150–2 central theme  16–17 coercive power of vulnerability  97–103, 109 meaning of effectiveness  153–4 obligation to criminalise  58–9 positive obligations doctrine  65 privacy (Art 8)  39

300  Index unlawfully obtained evidence  287 victim protection  20 Enforcement broader international trend  8–12 dangers of aligning human rights protection with criminal law need for protection rather than coercion  201–2 overview  183–4 positive duties applicable in different contexts  184–6 problem of diversion  198–201 risk of coercive overreach  187–92 risk of diluting human rights standards  192–8 underlying premise  184 obligation to investigate  61–5 suggested areas for future research  23 European Commission against Racism and Intolerance (ECRI) establishment  118 General Policy Recommendation on hate speech  118 importance  119 key pillar of policy  130 list of hate-speech offences  136–7 non-criminal law responses to hate speech  134 specialised mechanism for the fight against racism  17, 118 state-monitoring reports  120 systematic approach to hate speech  121 European Convention on Human Rights (ECHR) dangers of aligning human rights protection with criminal law need for protection rather than coercion  201–2 positive duties applicable in different contexts  184–6 problem of diversion  198–201 risk of coercive overreach  187–92 risk of diluting human rights standards  192–8 underlying premise  184 limits on national policy set by Convention  66–8 positive obligations doctrine  148 protection of the rights of suspects  1 underlying rationale  147 European Court of Human Rights (ECtHR) case law see Case law development of coercive duties  1–2

historical context of coercive human rights changes from late 1990s  7–8 development of positive obligations doctrine  6–7 X and Y v The Netherlands  4–6 procedural coercive obligation exclusion of evidence and coercive overreach  288–91 implications for exclusion of evidence  284–8 relationship to rights of the accused  281–3 scope and meaning  274–81 reparations just satisfaction requirements  82–3 legal framework  80–2 suggested areas for future research  25 use of positive obligations doctrine coercive overreach  53 obligation to criminalise  32–4 obligation to punish  44–53 overview  29–30 substance and procedure distinguished  30–2 F Fair trial (Art 6) positive obligations doctrine  147 procedural coercive obligation under ECHR implications for exclusion of evidence  284 relationship to rights of the accused  281–3, 282–3 protection of the rights of suspects  1 unlawfully obtained evidence  271–2 Feminist perspectives anti-impunity view  12–13 critique of the public/private divide  8 punitive fervour  165 scepticism about seeking emancipation through the criminal law  13 Freedom of association and assembly (Art 11)  41 Freedom of expression (Art 10) emphasis of case law  118 hate speech cases before ECtHR  121–3 scope of positive obligations  133–5 obligation to criminalise  41 Freedom of religion (Art 9) hate speech obligation to prosecute  125–6 positive obligations doctrine  131 obligation to criminalise  41

Index  301 G ‘Governance feminism’  13 Gross human rights violations see also prohibition of torture or ill-treatment (Art 3) ; Right to life (Art 2) common denominators  84–7 concluding remarks  92 historical context of coercive human rights  7–8 implications for victims  75–80 implications of retributive responses  87 individual and general measures adopted by Court  88–9 justification for positive obligations doctrine  150 legal framework at ECtHR  80–2 obligation to investigate  91–2 overview  71–2 procedural coercive obligation under ECHR  274 reparations just satisfaction requirements  82–3 retributive responses as reparations  72–5 selected cases  83–4 significance of finding state responsibility  90 H Hate speech burning contemporary issue  138 case law freedom of expression (Art 10)  121–3 obligation to prosecute  124–8 ECRI list of hate-speech offences  136–7 growing phenomenon  117 obligation to prosecute case law  124–8 discharging the obligation to prosecute  135–7 relevant cases  132–5 overview  118 positive obligations doctrine  128–32 role of CoE  119–21 sexual orientation or gender identity  110 victim protection  117–18 Human trafficking and slavery see Prohibition of slavery, servitude and forced labour (Art 4) I Implications of coercive human rights central theme  18–21 coercive overreach  187, 191

‘culture of conviction’  162, 182 danger of dilution  193 enforcement  250 hate speech  118 last resort principle  156 obligation to punish  32 preoccupation with extra-legal implications  3 preventative obligations  257 problem of diversion  199–200 procedural coercive obligation under ECHR  284–8 response to COVID-19  293 retribution through reparations  71–3, 75, 78–80 systematic and institutional implications  24–5 unlawfully obtained evidence  269 International developments anti-impunity agenda broader international trend  8–12 promise and pitfalls of the anti-impunity agenda  12–14 principles on counter-terrorism  225–6 J Justifications central theme of coercive human rights  16–17 coercive overreach  187–8 coercive power of vulnerability  97, 111, 113–14 ‘culture of conviction’  164 ‘effective deterrence’  150–2 exclusion of evidence  273, 285, 288 hate speech  119, 121, 130 last resort principle  145–6, 150–6, 159 measures to combat COVID-19  293–4 Modern Slavery Act 2015  172, 174 need for further research  22–3 obligation to criminalise  56–7 other Convention provisions  41–3 privacy (Art 8)  39–41 prohibition of slavery, servitude and forced labour (Art 4)  37–9 prohibition of torture or ill-treatment (Art 3)  35–7 right to life (Art 2)  33–5 obligation to punish  45, 48–9 positive obligations doctrine  150 procedural coercive obligations  279 requirement of criminalisation  56

302  Index restrictive approach to pre-trial procedural violations  291 retributive responses as reparations  73 state’s refusal to prosecute and/or punish  243–4 transitional justice  236–40 use of force by agents of the state  194–5, 198, 255, 258–9 victim protection  180 L Last resort principle see Criminalisation as last resort Legality principle key principle of criminal law  143 preventive orders  264–5 prohibition of slavery, servitude and forced labour (Art 4)  266 transitional justice  230–2 LGBTQIA+ community  94, 110–11 Limitations of criminal law approach central theme  21 hate speech  118–19, 133 on national policy set by Convention  66–8 need for further research  22–3 obligation to punish  44–5 response to COVID-19  295 transitional justice effective independent investigations  240–4 international principles on counter-terrorism  225–6 investigations without prosecution  236–40 legality principle  230–2 necessity of tampering with criminal law  226–9 overview  223–4 relationship between human rights standards and criminal law  232–3 right to life (Art 2)  234–5 O Obligation to criminalise dangers of aligning human rights protection with criminal law need for protection rather than coercion  201–2 overview  183–4 positive duties applicable in different contexts  184–6 problem of diversion  198–201

risk of coercive overreach  187–92 risk of diluting human rights standards  192–8 deterrence as key factor  58–61 gross human rights violations  72–5 hate speech discharging the obligation to prosecute  135–7 obligation to prosecute  124–8 relevant cases  132–5 impact of ECtHR case law on domestic criminal law encouragement of ‘culture of conviction’  169–71 interrelationship between substance and procedure  30–2 prohibition of slavery, servitude and forced labour (Art 4)  204 rise of penal state in Europe  163–6 transitional justice investigations without prosecution  236–40 summary and conclusions  244–5 underlying rationale  57–8 use of positive obligations doctrine by ECtHR focus on violence against the person  43–4 other Convention provisions  41–3 overview  32 privacy (Art 8)  39–41 prohibition of slavery, servitude and forced labour (Art 4)  37–9 prohibition of torture or ill-treatment (Art 3)  35–7 right to life (Art 2)  33–5 Obligation to investigate deterrence as key factor  61–5 gross human rights violations  91–2 principle of effectiveness  46–50 procedural coercive obligation under ECHR  275 prohibition of slavery, servitude and forced labour (Art 4)  204 transitional justice effective independent investigations  240–4 investigations without prosecution  236–40 Obligation to punish effect of unlawfully obtained evidence example of Dutch judicial response  270–4 overview  267–70

Index  303 gross human rights violations  72–5 interrelationship between substance and procedure  30–2 limitations  44–5 main developments  44–5 procedural coercive obligation under ECHR implications for exclusion of evidence  287 relationship to rights of the accused  283 scope and meaning  277–8 rise of penal state in Europe  163–6 scope of obligation  50–3 transitional justice  244 Osman warnings  20, 250, 262–4 Overreach see Coercive overreach P Pardons  47, 65, 244 Positive obligations doctrine deterrence as key factor general context  56–8 limits on national policy set by Convention  66–8 obligation to criminalise  58–61 obligation to investigate  61–5 overview  55–6 ‘effective deterrence’  150–2 historical context of coercive human rights  6–7 ‘horizontal effect’ of human rights jurisprudence  147–8 increasing influence on national criminal law systems  141 inherent in the text of the Convention  148–50 justifications  150 meaning and effect  141 negation of ultima ratio principle  157–9 obligation upon states to identify migrants as victims of harm challenges ahead  220–2 identification of victims – Chowdury and Others v Greece  218–20 identification of victims – J and Others v Austria  214–17 identification of victims – LE v Greece  211–14 identification of victims – Rantsev v Cyprus and Russia  209–11 overview of case law  208–9 preliminary clarifications  204–8

procedural coercive obligation under ECHR exclusion of evidence and coercive overreach  288–91 implications for exclusion of evidence  284–8 relationship to rights of the accused  281–3 scope and meaning  274–81 relevance during COVID-19 pandemic framing ‘lockdown’ measures  294 harms to bodily and mental integrity  294 overview  293 protection of life  294–5 theoretical perspectives circularity of reasoning  155–6 expressivity of invoking criminal law  154–5 meaning of effectiveness  153–4 sociologically informed practical reasons  156 use by ECtHR coercive overreach  53 obligation to criminalise  32–4 obligation to punish  44–53 overview  29–30 substance and procedure distinguished  30–2 Preventative risk orders  178–9, 264–6 Preventive operational obligations case studies overview  260–1 policing and Osman warnings  262–4 preventative risk orders and modern slavery  264–6 coercive overreach right to life (Art 2)  255 domestic violence  253–5 need for persuasive and consistent rhetorical language  266 overview  249–50 prohibition of torture or ill-treatment (Art 3)  259–60 right to life (Art 2) application of IHL paradigm  255 broad discretion in adopting policies or actions  258–9 existence of knowable and concrete risk  255–7 impact of Tagayeva judgment  257–8 importance of human dignity  258–9 need for acceptance of risk  259

304  Index Osman v UK as starting point  250–3 policing and Osman warnings  262–4 Privacy (Art 8) coercive power of vulnerability case law  97 risk of coercive overreach  112, 114 sexual abuse of children  99–103 deterrence as key factor  58 duty to criminalise  167 hate speech discharging the obligation to prosecute  135–7 obligation to prosecute  124–8 positive obligations doctrine  129–31 scope of positive obligations  133–5 interrelationship between substance and procedure  31–2 obligation to criminalise  39–41 obligation to investigate  61 positive obligations doctrine  148–9 procedural coercive obligation under ECHR  276 protection of the rights of suspects  1 unlawfully obtained evidence  271 X and Y v The Netherlands  4–6 Procedural coercive obligation example of Dutch judicial response  273 exclusion of evidence and coercive overreach  288–91 implications for exclusion of evidence  284–8 relationship to rights of the accused  281–3 scope and meaning  274–81 Prohibition of discrimination (Art 14) hate speech obligation to prosecute  126–7 positive obligations doctrine  131 obligation to criminalise  41–2 positive obligations doctrine  148 Prohibition of slavery, servitude and forced labour (Art 4) anti-impunity view  14 children  214 deterrence as key factor  60 duty to criminalise  167 historical context of coercive human rights  8 limitations of a criminal law approach  21 Modern Slavery Act 2015 as example of domestic law ‘culture of conviction’  179–81 overview  172–4 preventive orders  178–9

prosecution of victims  176–8 sentencing  174–6 obligation to criminalise  37–9 obligation to investigate  61 obligation upon states to identify migrants as victims of harm challenges ahead  220–2 identification of victims – Chowdury and Others v Greece  218–20 identification of victims – J and Others v Austria  214–17 identification of victims – LE v Greece  211–14 identification of victims – Rantsev v Cyprus and Russia  209–11 overview of case law  208–9 preliminary clarifications  204–8 preventative risk orders and modern slavery  264–6 Prohibition of torture or ill-treatment (Art 3) see also Gross human rights violations amnesties or pardons  65 anti-impunity view  12 coercive power of vulnerability case law  95–7 effect of coercive obligations  116 non-sexual abuse of children  103–5 risk of coercive overreach  111–16 sexual abuse of children  98–103 sexual and domestic violence  106–9 culture of risk aversion  255 dangers of aligning human rights protection with criminal law positive duties applicable in different contexts  184–6 problem of diversion  199 risk of coercive overreach  187–92 risk of diluting human rights standards  192–8 underlying premise  184 deterrence as key factor  58–9 duty to criminalise  167 hate speech  128–9 discharging the obligation to prosecute  135–7 obligation to prosecute  131–2 interrelationship between substance and procedure  31–2 obligation to criminalise  35–7 obligation to investigate  61–2

Index  305 obligation to punish main developments  46–50 scope of obligation  52 positive obligations doctrine  149 preventive operational obligations  259–60 procedural coercive obligation under ECHR implications for exclusion of evidence  287 relationship to rights of the accused  281–2 scope and meaning  275–6 protection of the rights of suspects  1 transitional justice  243 unlawfully obtained evidence overview  270 R Rape see Sexual violence Reparations gross human rights violations common denominators  84–7 concluding remarks  92 implications for victims  75–80 implications of retributive responses  87 individual and general measures adopted by Court  88–9 just satisfaction requirements  82–3 legal framework at ECtHR  80–2 obligation to investigate  91–2 overview  71–2 retributive responses as reparations  72–5 selected cases  83–4 significance of finding state responsibility  90 justifications  17 limitations on criminal approach  245 Retribution gross human rights violations common denominators  84–7 concluding remarks  92 implications for victims  75–80 implications of retributive responses  87 individual and general measures adopted by Court  88–9 just satisfaction requirements  82–3 legal framework at ECtHR  80–2 obligation to investigate  91–2 overview  71–2 retributive responses as reparations  72–5 selected cases  83–4 significance of finding state responsibility  90 obligation to punish  49–50

Right to life (Art 2) see also Gross human rights violations amnesties or pardons  65 coercive power of vulnerability  113 dangers of aligning human rights protection with criminal law positive duties applicable in different contexts  184–6 problem of diversion  199 risk of coercive overreach  187–92 risk of diluting human rights standards  192–8 underlying premise  184 deterrence as key factor  58–9 duty to criminalise  167 hate speech discharging the obligation to prosecute  135–7 obligation to prosecute  126–7 obligation to criminalise  33–5 obligation to investigate  61–2 obligation to punish main developments  46–50 scope of obligation  51 positive obligations doctrine  147, 149 preventive operational obligations application of IHL paradigm  255 broad discretion in adopting policies or actions  258–9 existence of knowable and concrete risk  255–7 impact of Tagayeva judgment  257–8 importance of human dignity  258–9 need for acceptance of risk  259 Osman v UK as starting point  250–3 policing and Osman warnings  262–4 procedural coercive obligation under ECHR  274 relationship to rights of the accused  282 relevance during COVID-19 pandemic  294–5 transitional justice effective independent investigations  241–4 expanding accountability  234–5 investigations without prosecution  236–40 unlawfully obtained evidence overview  270 Right to property (Art 1, Protocol 1)  42–3 Risk see Coercive overreach

306  Index Rule of law anti-impunity agenda  12 ‘effectiveness’ as justification for criminalisation  155 exclusion of evidence  284–6 fair trial (Art 6)  272 HR response to grave violations  168 obligation to punish  49–50 origins of criminalisation as last resort  146–7 procedural coercive obligation  278, 290 response to life-endangering offences  237 retributive responses  79, 86, 279 rhetorical device for ECtHR  266 S Sexual violence coercive overreach  111–16 coercive power of vulnerability adults  106–9 children  99–103 historical context of coercive human rights  8 interrelationship between substance and procedure  31–2 positive obligations doctrine  40 Slavery see Prohibition of slavery, servitude and forced labour (Art 4) T Terrorism see Counter-terrorism Theoretical perspectives criminalisation as last resort  145 content of principle  144 legitimately expressed concerns  142 meaning and effect  143–4 negation by positive obligations doctrine  157–9 origins and development  145–7 overview  142–3 ‘principle of legislative ethics’  144–5 role on European law systems  143 theoretical perspectives  145 positive obligations doctrine circularity of reasoning  155–6 expressivity of invoking criminal law  154–5 meaning of effectiveness  153–4 sociologically informed practical reasons  156 retributive responses as reparations  72–5 retributivist theory of criminal punishment  17

Torture or ill-treatment see Prohibition of torture or ill-treatment (Art 3) Transitional justice anti-impunity view  14 human rights-compatibility  14 international principles on counter-terrorism  225–6 limitations  21 limitations of a criminal law approach  21 necessity of tampering with criminal law  226–9 obligation to investigate effective independent investigations  240–4 investigations without prosecution  236–40 overview  223–4 relationship between human rights standards and criminal law  232–3 respect for human rights and legality principle  230–2 right to life (Art 2) expanding accountability  234–5 investigations without prosecution  236–40 summary and conclusions  244–5 U Ultima ratio see Criminalisation as last resort Unlawfully obtained evidence example of Dutch judicial response  270–4 overview  267–70 procedural coercive obligation under ECHR exclusion of evidence and coercive overreach  288–91 implications for exclusion of evidence  284–8 V Victim protection dangers of aligning human rights protection with criminal law  201–2 ‘effective deterrence’  20 gross human rights violations common denominators  84–7 concluding remarks  92 implications for victims  75–80 implications of retributive responses  87 individual and general measures adopted by Court  88–9 just satisfaction requirements  82–3 legal framework at ECtHR  80–2

Index  307 obligation to investigate  91–2 overview  71–2 retributive responses as reparations  72–5 selected cases  83–4 significance of finding state responsibility  90 hate speech  117–18 limitations of a criminal law approach  21 Modern Slavery Act 2015 as example of domestic law  176–8 obligation to punish  49–50 obligation upon states to identify migrants as victims of harm challenges ahead  220–2 identification of victims – Chowdury and Others v Greece  218–20 identification of victims – J and Others v Austria  214–17 identification of victims – LE v Greece  211–14

identification of victims – Rantsev v Cyprus and Russia  209–11 overview of case law  208–9 preliminary clarifications  204–8 substance and procedure distinguished  31 Vulnerability Art 3 case law  95–7 CoviD-19 patients  294 effect of coercive obligations  116 examples of coercive power adults  106–9 children  98–106 LGBTQIA+ community  110–11 overview  95–7 immigration status  176, 205 link with hate speech  120, 130 resort to criminalisation and penal severity  165 risk of coercive overreach  111–16 shaping of coercive measures  93–5 situational vulnerability  134 trafficking  173

308