Anti-Terrorism Criminal Law: Where Emergency Regime Meets the Investigative Agenda

Abstract

This article aims to show how reform of the law on terrorism not only has the power to create new criminal procedures, it can also create a distinct, parallel field operating alongside general criminal law. This parallel configuration presents certain unique features and processes which merit their own typology – namely, anti-terrorism criminal law (ATCL). First, the article discusses how states have responded to terrorism through reform of four key arenas: military law, immigration law, administrative law and criminal law. Comparison is then drawn between the United States and Israel in their respective approaches, showing that Israel has executed far more sweeping and significant reforms over the last four decades, mainly in criminal procedure. Examples are given to illustrate how Israel's evolving anti-terrorism legislation – and specifically, the new Counter-Terrorism Law of 2016 – changed the criminal procedural landscape to such a degree that it constituted the new field of ATCL. I contend that this move was anti-liberal in its definition and targeting of terror suspects, and in its pursuit of emergency aims and intelligence gathering rather than liberal criminal law objectives. Further, I show that liberal theory struggles to explain the integrated change model that Israel has implemented in its counter-terrorism reforms, and that the theoretical framings of Carl Schmitt and Michel Foucault may explain it more effectively.

1. Introduction

Democratic states have long faced serious threats from international and domestic terrorism.¹ They have responded to such threats by using four primary law-enforcement methods, which may be combined: (i) a military intervention, implemented by reforming the military law and the military tribunal powers;² (ii) immigration restrictions, implemented by reforming immigration law, expanding the powers of the immigration services to detain terror suspects, restricting the granting of permission to foreigners to enter and stay, and so on; (iii) administrative constraints, implemented by administrative writs, additional executive branch powers, and similar; and (iv) criminal processes and criminal law ³ – introducing reform in the areas of offences,⁴ evidence law, procedure and punishment.⁵

To illustrate this point and highlight the different approaches that countries can take to design and use criminal law to combat terror, the article analyses the respective responses of the United States (US) and Israel to the 9/11 terror attack. The US reformed its immigration law,⁶ first by regulations and then by enacting the US Patriot Act, thereby expanding the power of the Immigration and Naturalization Services to detain terror suspects.⁷ At the same time, the US revised its military law, defining terror suspects as enemy combatants,⁸ and enabling their trial by military tribunals.⁹ It also adopted the administrative method, authorising the President to ‘use all necessary and appropriate force against those nations, organizations, or persons he determines planned, authorized, committed, or aided the terrorist attacks that occurred on September 11, 2001, or harbored such organizations or persons’.¹⁰ Yet, in the fourth method – reforming criminal processes – there was just one minor reform: the pre-emptive testimony of ‘material witnesses’.¹¹ The material witness reform authorises courts to issue special warrants that enable a person to be arrested and detained if their testimony is deemed important in a criminal proceeding, and if it is shown that it may become impracticable to secure the presence of the person by subpoena.

Like the US, Israel has also responded to the terror threat by developing and reforming a variety of existing methods of law enforcement. For example, Israel reformed military law so that terror suspects became unlawful enemy combatants.¹² It also reformed immigration law,¹³ by restricting marriage between residents and foreign residents from Judea, Samaria and Gaza (family unification permits) on the basis that, in the past, these permits had been used by terrorists to plan and execute terrorist attacks. The change in the immigration law prevented the issuing of such permits except in a few exceptional cases (for example, involving minors). Israel also issued a new administrative order, which enables the detention of suspects without any evidence when the only purpose of the detention is to prevent a future threat.¹⁴

Yet, if we examine the Israeli counter-terrorism toolkit after 9/11, unlike the US, we can see comprehensive usage of criminal law. The question arises, then: Why? To explain why Israel has made such widespread and significant use of criminal law, our analysis needs to stretch back over the last four decades – that is, to before 9/11. Here, I examine one aspect of Israel's reform in the area I term anti-terrorism criminal law (ATCL): the change in criminal procedures for the adjudication of terror suspects.¹⁵ I describe and explain the particular characteristics of ATCL and seek to understand them by applying different theoretical lenses, liberal theory, and otherwise.

I offer the comparison with the United States not in order to claim that ATCL analysis can be applied in any country, but rather to show two test cases: one featuring intensive use of criminal law to combat terror; the other featuring minor use of criminal law for the same purpose. Because terror is recognised as an international issue, this renders ATCL one of the tools that a state may consider.

The events of 9/11 and the terror attacks in the US and Europe between 2000 and 2010 compelled countries to regulate ways of combating terror. Now it is possible to make a comparison and identify the measures that different countries implemented. Specifically, Israel had experience in combating terror using criminal law even 20 years prior to the 9/11 attack, and other states may similarly have had measures in place before 2001. Although Israel continued to take the same measures – criminal laws, before and after 9/11–9/11 is still important for the Israeli case, as a time point to check and measure it, this time after other countries had experience of coping with terror.

I make this brief comparison in the context of 9/11, a moment in time when the US needed to respond to a terror event on an unprecedented scale. This context enables us to see how, following this event, both the US and Israel decided to reform the law –Israel with decades of experience in combating terror, the US with less experience in this field. The discussion about Israel post-9/11 must also address the reforms it made progressively during the decades preceding that event in its fight against terror.

Although it includes a comparison with the US process relating to the 9/11 events, the main analysis is focused on the Israeli case while other countries chose different paths in different contexts. With regard to the question of applicability of ATCL, other countries may apply the ATCL model, but this is not central to the purpose of this article. The scope of this article is to focus on the Israeli case, which is unique, as the comparison with the US shows.

This article analyses the three primary approaches that have been applied by different states to reform general criminal law and, specifically, ATCL. I term these the evidential and substantive model, the penal model, and the procedural model (henceforth, the ‘substantive’, ‘punishment’ and ‘procedural’ models).¹⁶ Targeting any of these models at one group of people – in the Israeli case, terror suspects – challenges the basic principles of law, particularly its generality, equality and certainty.¹⁷ The cumulative effect of applying criminal law reforms to one targeted group – reforms such as preventing attorney–suspect meetings; disregarding the duty to record interrogations; extending detention periods; obstructing the suspect's presence at court hearings; and even concealing court detention decisions from the suspect – has created a new, unique and separate legal field of ATCL within general criminal law.

Criminal law is growing increasingly specialised in its various branches. The question is: Why is the difference between ATCL and general criminal law more significant than differences between other ‘independent’ branches of criminal law, such as white-collar criminal law or the ‘war on drugs’? While some branches – such as those dealing with drugs offences, sex offences or white-collar offences – are self-evident aspects of criminal law, the place of ATCL is not so clear-cut. Should it fall under criminal law (law of the enemy) or military law? Why should this branch be the only one with a unique procedure? The professional expertise associated with ATCL is the result of a unique procedure with which the lawyer must be well acquainted and which involves a security classification process that the lawyer must go through in order to represent in these cases, and not only as a substantive branch within criminal law.¹⁸

Against this backdrop, my research questions are:

• • What are the features of using the procedural model against one specific group of terror suspects?

• • What are the cumulative effects of procedures that allow the state to hold the suspect in the investigation facility in isolation for almost three weeks?

• • Are the principles of the liberal framework – particularly, individual protection and rule of law¹⁹ – applied in ATCL?

• • Can the liberal framework explain ATCL, or is there a break away from the liberal framework that can be explained by another framework?

• • By what theory, liberal or otherwise, can the characteristics of ATCL be understood?²⁰

The article proceeds as follows. Section 2 describes the anti-terrorism legislation that has reformed Israeli criminal procedure over the last four decades and the impact of this reform on terror suspects, in particular. Section 3 analyses the three aforementioned approaches used by legislators to effect reform in criminal law – the substantive, punishment, and procedural models – in the context of Israeli legislative history. Section 4 explores Israel's new Counter-Terrorism Law and demonstrates the integrated use of the three aforementioned models. Section 5 explains the use of procedural and integrated models against one specific group – namely, terror suspects. It examines these models using basic principles of law, noting incongruences with the principles of equality, generality and certainty. Questions surrounding this reform are also examined. Who is the subject of the reform? How often is it used, and for what purpose – for example, for emergency procedures alone? What problems are associated with its use?

Section 5 also describes two main characteristics of ATCL: its emergency aspect (such as the severity of the offences)²¹ and its investigation-and-intelligence aspect (the aim of ATCL being to protect the investigation process and intelligence gathering). Drawing on primary research from interviews, as discussed in this section, I contend that liberal theory has difficulty in explaining these characteristics and that Carl Schmitt's emergency theory and Michel Foucault's power theory regarding the ‘investigation mechanism’ in ATCL do a better job. The results of the interviews were published previously in 2019 in Hebrew.²² Section 6 summarises and concludes.

1.1. Liberalism and ATCL

According to the literature, the new sphere of law that I term ‘ATCL’ includes both the basic liberal principles and the relationship between the defendant and the community, as perceived in ordinary criminal law. Hence, liberal political theory is vital for the present discussion, but here the term ‘liberalism’ refers to a particular regime of rights enshrined in a constitutional order that protects the individual. The basis of the justification for liberal law is the protection of these rights, and the individual is the basic unit of state order. Another key aspect of this regime is the rule of law, according to which the state is subject to the law even in emergencies. It is a conception of the substantive rule of law, which – besides this concept in its formal sense – gives priority to certain rights stemming from it over vital social interests, protects the rights of the minority, and prevents the tyranny of the majority.²³

The liberal debate places both the conflict between national security and the rule of law and the protection of human rights at the centre. As a result, it emphasises the confrontation that is played out between the prosecution and the defence. From the perspective of the defence, human rights must be protected in the criminal process, regardless of whether this is within the specific area of the criminal-security process. According to the defence, the standard criminal investigation – intended to prove facts with evidence – must take place for all kinds of offence. From the perspective of the prosecution, the criminal-security procedure must, first and foremost, assist in maintaining state security. As a solution to this conflict, liberalism proposes that these opposing values should be balanced, weighing one against the other in accordance with the spirit and purpose of the law, relevant in the specific state and social reality in question, as well as the overall values of that society.²⁴

In this article the liberal model is used as an ‘ideal type’ methodology²⁵ in which the concept of liberalism is refined and rendered solely for the purpose of analytical inquiry. The article investigates the phenomenon of ATCL by comparing it with this ideal model, to understand how and where its legal arrangements are pertinent to the model or deviate from it. This analysis is not intended as a normative critique of the liberal regime or its anomalies but, rather, aims to characterise these anomalies as a descriptive model. That is, by comparison with the ideal type – the ‘pure’ liberal model – this study seeks to understand the actions and forces of the liberal regime and the rule of law by examining how the basic principles of that regime are preserved in criminal proceedings. By using this model, the study seeks to offer a conceptual and descriptive understanding of the field of study.²⁶

ATCL challenges some liberal principles and also certain principles from the tradition of common law that exist in ordinary criminal procedure:

1. (a) individualism or the principle of personal responsibility, which expresses the individual perspective and does not permit group conception of a suspect;

2. (b) the principle of adjudicating the criminal act as opposed to the criminal perpetrator, which means punishment for specific acts and not for characteristics of the identity of the accused;²⁷

3. (c) the principle of adversariality, according to which the parties confront each other and the judge decides. To resolve the dispute fairly, the judge does not participate in the dispute, and the contention between the parties must be balanced and considered;²⁸

4. (d) the principle of due process, which requires the rights of the accused in the prosecution process to be protected, to shield the individual from the power of the government; this procedure is the basis of litigation rights;²⁹

5. (e) the principle of generality, which is derived from equality before the law and means that, in criminal proceedings, the rights of every suspect and defendant must be upheld.

Scholars have identified a general trend in which criminal law is moving away from the basic liberal principles that tend to be seen as part of the criminal-security process. Combining criminal law with other areas of law such as immigration can also create a withdrawal from basic liberal principles in criminal law.³⁰ In recent decades, liberalism has undergone a crisis when it comes to shaping criminal law according to its principles. As part of this crisis, one can identify a shift in the rationale of criminal law towards one that is risk-based and focused on thwarting future offences rather than on past acts, at the cost of reducing procedural guarantees – a phenomenon that this article analyses and seeks to understand in the context of ATCL.³¹

Liberal democratic systems require special and stringent protection of the state's exceptional power to prosecute and convict suspects. Liberalism places the individual at its centre – as an autonomous being who makes decisions, as rights bearing, and in the light of the individual's perceptions of the common good. Alongside the role of the state in enabling individual autonomy, in the case of criminal law it interferes with the moral autonomy of its citizens. ‘Classic’ criminal law contains egalitarian norms that judge impartially between the state's claim and that of the defendant. It is the criminal procedure that allows a state to attribute guilt to a citizen, by maintaining neutrality on its side. Hence, the procedures must be special and precise to ensure fairness and equality. In contrast, in terrorism law there is a danger of deviating from neutrality, fairness and equality.³²

Therefore, to understand and characterise ATCL in the context of liberal democracies, we must examine the liberal discourse on emergencies.³³ This debate features two basic themes: (i) the idea that the individual and their protection take precedence, and (ii) the existence of a broad legal framework that applies also to emergencies. Subordination of all state activity to the law is intended to prevent the government from exercising its power arbitrarily and to protect individuals and their rights. Liberalism's treatment of emergencies does not exempt them from the rule of law but recognises only the need for the existence of ‘exceptional’ authority in such situations. The revoking of ordinary law is by virtue of a clause in the constitution or other general basic law, and therefore emergency situations are also protected within the framework of the law. The possibility of complete arbitrariness without law does not exist in liberalism, as the rule of law refers to the supremacy of the judiciary over the governing authorities. The arbitrariness and legal framework are maintained by adherence to the principle of separation of powers. The rule of law and the separation of powers ensure a clear political order, which limits the powers of the executive branch.

There are restrictions on liberal constitutionality, both in judicial review of the actions of the executive and through the supreme power of the law. Even in emergencies, liberal states are careful to maintain certain norms set by law. When the executive branch acts contrarily to these norms, the main brake is the judicial system and judicial review.³⁴

This article examines whether ATCL can be understood through the liberal paradigm and whether, as the study of forces in this procedure is expanded, the understanding of the procedure should be extended beyond the liberal conception. The use here of the term ‘liberalism’ is as a descriptive theory that seeks to understand the liberal state of emergency and the rule of law.

Liberalism approaches ATCL as a set of norms adapted to a state of emergency which repeals some of the usual laws and rights, and which exists by virtue of law. Apart from the rights expressly waived under criminal-security emergency laws, are the aforementioned basic liberal principles in criminal law – individualism, the adjudication of acts, adversariality, due process and generality – maintained within its framework? Do these basic principles exist in the liberal approach to the state of emergency defined in ATCL legislation, and are they part of the rule of law? Is there a broad legal framework that is maintained throughout the procedure (including these principles) and designed to limit the powers of the executive and protect the individual? Is it possible to propose a liberal model that can explain and characterise this procedure? The theoretical discussion at the end of the article examines these questions in the light of the findings presented below.

Liberal conceptualisation creates a dichotomy between legal and legitimate versus illegal and illegitimate. The powers that the liberal state can exercise are limited to the legitimate use of force, and it has a monopoly on the exercise of such power. This dichotomy explains the limitations of the liberal state in emergencies.³⁵ The use of political power by the state is legitimate if it acts legally, meaning that a state acts violently when it fails to act according to a legal standard.³⁶ Indeed, much of the liberal scholarship has dealt with the legitimacy of legal norms and broad considerations for assessing their legitimacy, such as the structure of political institutions and the justification of state actions as neutral and fair,³⁷ as well as political discourse and how law turns moral principles into questions of fact.³⁸

However, emergencies, in general, and ATCL, in particular, are characterised by the suspension of norms and the (temporary) creation of a state that does not follow regular legal procedures. The assessment of legitimacy is therefore concerned with a potential lack of clear norms in the liberal discourse, with an emphasis on the relationship between the individual and the state in criminal law, and it focuses on the exercise of political power as opposed to the set of norms (and their legitimacy conditions) within which it is exercised. A narrow definition of liberalism is applied in this article, centring on the rule of law and rights; other aspects are deliberately not included, as the analytical purpose of the article is to identify any deviation from the narrow liberal model. Thus, this narrow definition is used as a derivative of post-liberal theory, such as Carl Schmitt's theory, according to which the rule of law cannot always exist. The narrower the definition of liberalism, the stronger the argument. The question is whether liberalism successfully explains the complexity of the issue of legality and the use of force in the ATCL process, and to what extent liberalism struggles to characterise and identify forms of force that are not part of the dichotomy in this process.

It is to be noted that the critique of liberalism through Schmitt's approach, which is presented later in the study, is one of several critical approaches to the liberal idea. Among the approaches that criticise liberalism, Marxist and communitarian criticism should be noted in particular. According to the Marxist critique, liberalism in the modern age has strengthened civil-political rights over socio-economic rights as a means of maintaining the class, social and ethnic status quo. Karl Marx expressed doubts about the ability of rights to promote equality and social justice. He argued that certain rights reduce the state's ability to promote a truly free agenda. Marxist critics have argued that the rule of law and the rights discourse are merely tools for establishing classes in the name of the law; but, as they do not offer a discussion within the framework of the law, they are irrelevant here.³⁹

The communitarian critique of liberalism has focused on the individual's desire to be surrounded by a community and to strengthen the perception that a person is a ‘social creature’ whose existence depends on the public sphere and not just the private sphere. This approach criticises liberal individualism, and emphasises the importance and vitality of the political public sphere for individuals as a place where they might achieve self-realisation and stand up for their identity and abilities. While these criticisms are important for the discussion of liberalism, again, they are not relevant for our purposes, as they do not focus on questions of the rule of law and the discourse on rights and of emergency and decision making, which constitute the focus of discussion here.⁴⁰

In the context of communitarian critique, these concepts of the rule of law and the discourse of rights have also been analysed as tools that serve the individualistic, atomistic conception, and criticised for not giving sufficient space to the community and its importance to the individual. This critique also discusses these concepts outside the framework of the law and focuses on harm to the communal and social needs of the individual, which the liberal discourse creates through the law. As such, it is therefore also irrelevant to this article, which therefore focuses on liberal criticism through Schmitt's approach.

1.2. Schmitt's ‘emergency theory’

In the context of a state of emergency, questions have also been discussed in legal literature and political thought concerning the relationship between sovereignty and the rule of law. This relationship between emergency and sovereignty was at the centre of Carl Schmitt's writing.⁴¹ In accordance with his approach, ‘a sovereign is he who decides both when there is a state of emergency and how to respond to it’ – a situation in which the law is suspended and the legal process continues to exist only ostensibly because the judiciary collapses into the role of sovereign.⁴² A basic distinction in politics, in Schmitt's view, is between an ‘enemy’, whose sovereign defines an existential threat to be eradicated, and a ‘friend’.⁴³ The enemy can usually be most clearly identified in wartime, when an emergency is declared. Declaring a state of emergency allows the sovereign to suspend the existing law and use unlimited force to eradicate the ‘enemy’.

The sovereign, according to Schmitt, therefore, has the authority to decide the means required to eradicate the emergency situation. He contends that when state law is activated, it is a situation of routine and normalcy, in contrast to states of emergency when the law is suspended and a state of abnormality is created, similar to a state of war. According to this approach, the state of emergency is an exception to the rule. That a state of emergency exists is not the interesting feature; rather, what it offers is an insight into the political forces at play. Although there is a section in the law that allows the law to be revoked, it is formalistic because this exception indicates that there are routine events against which those of an exceptional nature can be distinguished. The problem with liberalism, according to Schmitt, is that it proposes a comprehensive and idealised governmental description of the regime, without any reference to its failures, difficulties and shortcomings. The literature has also addressed the sovereign's control and the ‘exposed life’ of the citizens under that control, who are regulated and contained by way of exclusion. It is the sovereign, then, who does the excluding. According to this approach, on the one hand, sovereignty affirms the law and, on the other, declares its own primacy in relation to the law.⁴⁴

The matter of decision making as an act of sovereignty is central to this approach. The sovereign is decisive and declares the state of emergency out of concern for the people and for the purpose of promoting its good. The decision itself constitutes the political moment and realisation of the political power. It is precisely in emergencies, according to this approach, that basic elements are under threat, calling for a sovereign decision. This moment is not subject to the rule of law. Indeed, in a moment of crisis when the liberal system is no longer able to function, the state opts to suspend the legal order and declare a state of emergency in which ordinary law is not valid. The state regards this option as preferable to its own elimination.⁴⁵

Schmitt's critique of liberalism centres on its focus on the individual and its disregard for the power of sovereign decision making in its own right. Schmitt's concern is for the existence of the political space, which allows individuals to develop and fulfil themselves. A regime of separation of powers that does not allow for effective decision making does not guarantee the continued existence of the state as a collective.⁴⁶

While this theory makes it possible to overcome the deficiencies of liberalism in the matter of sovereign decisions, it goes no further. Schmitt's approach refers primarily to the role of the sovereign, but it does not explain the nature of the exception made and the differences created between the ordinary criminal procedure and what I term the ATCL procedure. This article reveals rights that have been denied to suspects in security offences and shows how the basic principles of criminal law have been violated. Schmitt's approach does not seem to have the tools to explain the purposes and practices of ATCL. Therefore, in addition to the explanation of the sovereign decision, an explanation is also needed regarding the operation of this unique procedure, its goals, and the forces operating within it.

As an alternative to the suspension of the law in an emergency, there is another approach according to which, even in an emergency, liberty should not be denied without a fair procedure. In February 2009, for example, the European Court of Human Rights ruled in A and Others v United Kingdom that, even in an emergency which allows for flexibility around certain laws, a person's liberty should not be denied without giving them a fair opportunity to defend themselves in court.⁴⁷ The judicial procedure cannot be deemed fair, for example, if the detainee has not disclosed enough information to allow them to give ‘effective instructions’ to their lawyer. A similar determination can be found in the matter of Secretary of State for the Home Department v AF and Others, in which the United Kingdom House of Lords clarified that even when liberty is violated to a lesser extent, no confidential material should be relied upon in a manner that denies the right to a fair trial in any situation.⁴⁸ For example, in this case it was determined that, to ensure a fair trial, the authorities should disclose to the suspect the information he needed to compile for his defence. Therefore, cases that are based on confidential information cannot withstand this test, even in times of emergency.

1.3. Foucault's ‘power theory’

Given that intelligence investigation constitutes a central feature of ATCL procedures, the socio-political significance of this feature must be understood. The scholarship refers to such investigation in terms of social control, where the investigator exerts force to achieve social control over the investigation and the suspect. Rather than control using physical means, interrogators use psychological practices in the interrogations, based on the element of persuasion, which contributes to strengthening the legitimacy of the interrogation and enhancing the power exercised on behalf of the state. This and other investigative improvements were made possible as a result of the development of socio-psychological knowledge on the subject of interrogation, and are therefore related to the power of scientific discourse. This development creates the impression of legal neutrality, which justifies such activity from the state side. Therefore, the persuasion aspect of the investigation, based on knowledge from the fields of psychology and medicine, is closely linked to the realisation of state power.⁴⁹

This ‘political power’ perception of investigations conducted by state authorities – and, indeed, of security intelligence investigations – can be explained through Foucault's theory.⁵⁰ This article examines whether the ATCL process can also be understood through concepts of power and political goals, according to Foucault's approach, and especially as a mechanism for the use of force in modern society, looking beyond sovereignty and the limitations of the law (as per Schmitt's perspective). The advantage of Foucault's theory lies in removing the sovereign from the focus of the discussion; instead, it offers a conception of power relations. Foucault's historical analysis of centralised sovereignty shows how power techniques that deviated from state mechanisms were created after the French Revolution. The law and its limitations no longer constitute the only theory of power, and the new technology of power combines techniques of control and knowledge.

The power mechanisms operating in modern society, of which the sovereign is only one of its operators, appear in various forms (for example, in the institutions of modern society). Foucault defines power as a relationship in which each party has the capacity to act and a space for reaction, and each party enjoys a degree of freedom. Power relations exist between unequal parties when one party restricts the other. Power is exercised according to economic principles, and, according to Foucault, the power exerted last is the one that is most effective. These power mechanisms are activated elusively and intelligently to render them almost imperceptible – and therefore they do not provoke opposition.⁵¹ Political power is a system of relationships between individuals or groups, in which each side strives to change the other party's behaviour. Control, efficiency and obedience are not achieved by physical means but through ‘normalising’ techniques that adapt humans to the rules that are intended to govern society.

In examining the law and its restrictions (identified with the sovereign), Foucault focuses on mechanisms, practices and techniques of force to understand the components of the activation of that force, the forms of subordination, and the relationships between them. According to Foucault, to understand power relations, one should attempt to locate and analyse the mechanisms of power – those that solicit, supervise, control and organise the subordinate forces. However, in the mechanism proposed by Foucault, power is acquired not by ‘overpowering’ or eliminating subordinates but by ostensibly building them up – enabling them to self-manage, enlarge their sphere of operation, and organise their own hierarchical structures – but within a tightly controlled environment. In this mechanism the power of close supervision and imposition of general arrangements by regimes on their populations, for example, is exercised extremely effectively, as it operates in the expansive and inconspicuous context of everyday life (as opposed to the overt sphere of war and death). The power mechanism is operated through a series of interventions and regulated supervisory actions, which are, in fact, disciplinary towards the population. It is through these actions that power over day-to-day life is exercised – as Foucault termed it, the ‘bio-politics of the population’.⁵²

The power mechanism therefore comprises the various techniques applied to achieve control and supervision of the population or a particular group by monitoring and analysing it, penetrating it and influencing it, with the aim of creating an integrated and beneficial society. Effective, efficient and ‘normalised’ control is achieved by activating these techniques, which allow for continuous monitoring of individuals in the system as a central device. Knowledge constitutes a professional and institutionalised system that is competent to act for the purpose of re-establishing the pattern of compliance with norms. In everyday life, ‘grey’ mechanisms are applied to the individual that are self-evident yet imperceptible. It is the mechanism, its expert operators, their method of collecting information, and their schedules and patterns of action that together monitor human behaviour. In this sense it has become routine practice for governments to continually change the techniques they use to track, scan and monitor citizens, social movements and social processes in the name of security, through intelligence services. The development of interrogation methods through scientific knowledge has also contributed to the expansion of the mechanism: as noted earlier, in interrogations there has been a shift from the use of physical means to psychological persuasion tools, including the use of extraordinary interrogation methods with a medical dimension, even accompanied by physicians.

Is it possible to explain through these concepts, which developed under Foucault's theory, the differences between ordinary criminal procedure and the criminal-security procedure, and the differences between Israel and the United States in matters relating to ATCL? Does this theory explain the goals and practices of security regimes revealed here? These questions are addressed in the theoretical discussion at the end of the article.

2. The impact of anti-terrorism legislation on criminal procedure in Israel

Within the last four decades, anti-terrorism legislation has entirely reformed criminal procedure in Israel. Among other consequences, its impact has manifested in the following five phenomena:

1. (a) suspects being prevented from meeting with their attorney (1981);

2. (b) suspending the duty to record interrogations (2002);

3. (c) extending detention periods (2006);

4. (d) obstructing the suspect's attendance at court hearings (2006);

5. (e) concealing court detention decisions from the suspect (2006).

The discussion that follows is devoted to detailed descriptions of changes in legislation and in its actual workings in order to show the slow birth of ATCL, which is spread over four decades. The discussion shows how ATCL was created, examines how these reforms are to be understood and conceptualised, and explores the question of whether ATCL should be considered an ‘emergency model’ or an ‘investigation-and-intelligence model’.

To understand the characteristics of ATCL, let us survey each of the five legislative reforms in turn.

2.1. Prevention of meetings between suspect and attorney

The Criminal Procedure Order (1981) was the first legislative instrument to be enacted that protected the right of suspects to meet with their attorney. However, security offences were excluded from this protection:⁵³ attorney–client meetings could be suspended for a period of 15 days. In 1996⁵⁴ a number of significant changes were made to this rule, expressly excluding security offences, which received extensive attention in the legislation: a separate article was devoted to the rule for security suspects; the grounds (rationales or causes) for preventing suspects from meeting with their attorney were specified and detailed by law; a definition of ‘security offence suspect’ was added; the suspect was given the right to appeal against the administrative decision in this matter; the option to extend the period of detention to 21 days was added to the powers of the President of the District Court on the basis of an application filed with the approval of the Attorney General; and the Minister of Justice was authorised to produce new regulations to govern operations in this sphere.

The prevention of attorney–suspect meetings can be viewed as one of the first steps in the creation of a special procedure for security suspects, distinct from the ‘regular’ procedure pertaining to general criminal suspects. Furthermore, the Detention Act (1996) introduced new mechanisms of judicial supervision over the detention periods established (15 days, extendable to 21).⁵⁵

The importance of this legislation⁵⁶ cannot be overstated: it categorised ‘the individual suspect’ and ‘the group of suspects’ in the field of security offences into separate classes. Article 35(b) of the Detention Act defines the ‘security offence suspect’ in the singular as opposed to ‘security suspects’ in the plural. In essence, this definition – a central notion used widely in various laws and regulations – refers to a group of suspects and creates a separate detention procedure. That is, the definition of ‘security suspect’, which encompasses a list of offences, ultimately creates a special group. Hence, the Detention Act effectively created a group of security-offence suspects under this definition that could be used henceforth to make major procedural changes.

The objective of this Act was to give expression to Israel's Basic Law: Human Dignity and Liberty and to minimise the prevention of attorney–suspect meetings, but it had the opposite result in the case of ATCL, as the right to counsel was restricted.⁵⁷ By defining the security-offence suspect in Article 35(b),⁵⁸ the Act increased procedural divergence between criminal and security suspects in a declarative manner and constituted the justification for procedural changes in relation to this group, as discussed in Section 5.⁵⁹

2.2. Exemption from the duty to record interrogations

Another major legislative development in this field was the granting of an exemption from the duty to record security-suspect interrogations. The Criminal Procedure Act (Suspects Investigation) (2002) (Documentation Act) establishes the obligation to document all stages of such interrogations. According to this obligation, the police must document the suspect's investigation by any means – visual, audio or written – as long as it faithfully captures the course of the investigation from start to finish. Exemption from this obligation appears in Article 17 of the Documentation Act (as a Temporary Order)⁶⁰ in the specific case of investigations into security-offence suspects. In 2017, as a response to the Marmara case,⁶¹ the Turkel Committee Report (No 2, 2013),⁶² and the Chahanover Committee Report (2015),⁶³ the legislator added an option to the law for the prosecution to watch the investigation via CCTV from a control room – at any time, round the clock – instead of documenting it.

2.3. Extending detention periods

A further important development in this area was the Criminal Procedure Act (Enforcement Powers – Detention) (Detainee Suspected of Security Offence) (Temporary Order) 2006.⁶⁴ This ‘temporary’ order allows the prosecution of security suspects from Gaza, as a result of the disengagement of Israel from Gaza and the authorities’ increasing concerns about the consequent potential security threat and feared increase in the use of terrorism by this collective. Crucially, it is relatively easy in Gaza to find oneself classified as a security suspect – for example, if one is a teacher in a Hamas-run kindergarten or if one lives near the Israeli border and is approached by Hamas, asking whether any Israeli Defense Forces (IDF) personnel can be seen over there.⁶⁵

The legislature's goal was to expand the investigators’ power because of the special characteristics of the security investigation: lack of access to evidential information; lack of access to the infrastructure (for instance, if the act of terrorism is perpetrated in Gaza, to which Israel has no access); the nationalist–ideological character of security offences; lack of cooperation with other witnesses (because they usually belong to the group of the suspect); and the need for an investigation that is ‘continuous, without interruption’ in order to thwart and prevent terrorist attacks.⁶⁶

The legislation restricts the rights of security suspects beyond the periods stipulated in the general criminal procedure,⁶⁷ which already has a special reference to such suspects.⁶⁸ In effect, the Temporary Order permits violations of basic constitutional rights in criminal proceedings; for example, extended detention periods violate the liberty of security suspects. Other violations are discussed below.

2.4. Absence of the suspect in court hearings

Another striking development is the order relating to the exclusion of security suspects from court hearings. Article 5 of the Temporary Order allows courts to hold a hearing on extending the detention period without the security-offence suspect being present. It also permits courts to extend the maximum detention period to 20 days without the suspect being present for that decision (provided that in a previous hearing the suspect was present and the detention was extended by less than 20 days).⁶⁹

This procedure is conditional upon whether the court believes that a break in the investigation – pausing in order to bring the detainee to court – may impede the prevention of security offences or the prevention of loss of life. The Temporary Order also allows courts to hold a hearing on appeal for requests or reconsideration requests regarding the arrest decision in the absence of the detainee, and it is conditional upon the court's conviction that the break in the process may cause ‘real harm’ to the investigation.⁷⁰

In February 2010, in a request submitted by the Public Defender's Office, the Supreme Court decided that Article 5 of the Temporary Order (establishing the freedom to extend the detention period without the detainee's presence) was void. The Court stated that the article disproportionately affected the detainee's constitutional right to be present in arrest proceedings and thus contradicted the Basic Law: Human Dignity and Liberty. The Court emphasised that the scope of the right to due process is derived from the constitutional right to freedom and dignity, and highlighted the importance of the detainee's presence in court and effective judicial review of the detention process; and the importance of effective investigation balanced against the proportionality of the injury.⁷¹ Following this ruling, the Knesset (Israeli Parliament) amended the Law once again to re-enable hearings to be conducted without the presence of the detainee in exceptional cases.⁷² Also, the new Counter-Terrorism Law (2016) allows courts to hold hearings (again, in exceptional cases) while the detainee is absent.⁷³ This law is intended to unify the legislation in this field, reshape the basic definitions, and provide a variety of tools – criminal, administrative and public (which are discussed later in this article). The Temporary Law has become a permanent rule under this anti-terrorism law, with few changes.

2.5. The suspect not being made party to court decisions on detention

This procedure, which was regulated in Article 5(4) of the Temporary Order,⁷⁴ allows the law enforcement authorities and the courts not to inform the security detainee of decisions relating to the extension of the detention period. According to this legislation, in cases where the detainee did not appear at the hearing, as mentioned above, and the state submits a request for an extension, the court can decide not to inform the suspect of its decision. Again, according to this article, the court can use this option only if it is convinced that informing the detainee about the decision would impede efforts to prevent a security offence or prevent loss of life.

2.6. The cumulative effect: A new legal field

The cumulative effect of these relatively new procedures is the creation of a de facto unique and separate legal field within the general Israeli criminal law – ATCL – which is located at the intersection between the security field and the criminal law. Alongside its main feature – a 20-day investigation period during which the suspect's normal rights can be suspended – there are other specific features, which include the use of special lawyers (ATCL experts and security classification holders), no recording or videotaping of the confession,⁷⁵ and two phases of investigation (by the police and the General Security Service or Shin Bet). These methods, which were mentioned in the interviews as ‘law in practice’, are part of a discrete sphere standing alongside criminal law.

In this new sphere liberal principles and the rule of law do not apply.⁷⁶ Indeed, ATCL challenges the basic liberal principles of law: generality, equality, certainty, the rule of law, individualism, human rights, and the adjudication of acts – not the actors.⁷⁷

3. Existing models for effecting change in criminal law

Why and how did Israel apply such comprehensive and significant changes to its criminal law, and are the reforms really restricted to emergency cases only? To answer these questions, one must understand the models of reform available to states in the area of criminal legislation. As noted earlier, I contend that there are three such models that the legislator can use to effect reform: (i) the substantive model; (ii) the punishment model; and (iii) the procedural model.

In this section I analyse each model vis-à-vis Israeli legislative history, concentrating mainly on reforms that answer a need or a problem in society which has arisen as a result of past events. In this sense each of the models will be examined as a possible choice of the legislature. The choice will be discussed according to the formula prescribed by law and not by the application of the legislation after its enactment. Applying any of these models to one single group of people challenges the basic principles of liberal law, mainly when using the procedural model exclusively against one group – security suspects – for the first and only time, and not in general.

The three models I present and analyse here (substantive, punishment and procedural), to better understand how states effect change in criminal law, offer a descriptive argument to examine the routes that were open to the legislator against the route that was ultimately chosen.

3.1. The substantive model

According to this model, when a country wishes to address a worrying criminal phenomenon or a national crisis, the legislature can decide to change an essential element of criminal law related to the offence or its proof. The legislature thereby changes the definition of the offence, its component elements, the criminal defence or the manner of its proof.

To understand how this model operates, let us take an example from a different realm: the 2001 amendment of the offence of rape in Israel.⁷⁸ Before the amendment in question, section 345(a)(1) of the Penal Code defined ‘normal’ rape as ‘intercourse without the free consent of the woman, by the use of force, infliction of physical suffering, or the use of pressure or threat of one of these’. The increase in acts of rape and the rise in society's awareness of victims’ responses at the time of the rape (freezing, temporary inability to fend off the rapist, and so on) led to the revision (Amendment 61) of the Penal Code. In this amendment the legislator eliminated the requirement to use force (or the threat of doing so) from the definition of the crime of rape. Israel thus opted to frame the offence based on the central element of absence of consent from the woman. From that point on, a man who penetrates a woman without her consent, even if he is not using force or threats of any kind, is considered to have carried out rape. This framing would therefore extend to any intercourse without the consent of the woman, including rape within marriage, for instance.⁷⁹ In this example, the legislator used the substantive model, changing one of the basic elements in the definition of the offence so that there is now no need for the element of force to be present or proven in court in order to charge a man with rape.

Another means of using this model is to change the criminal protection. For example, in the case of farmers who tried to forcibly eject thieves or suspected thieves from their farms, in 2008 the legislator changed the protection article in the Penal Code by adding another defence to criminal liability. According to this defence, no person shall be criminally liable for an immediately necessary act for the purpose of ejecting people from their property who are intending to steal from it.⁸⁰ The legislator could have selected other options in attempting to deal with the growing phenomenon of thefts from farms and the aggressive defence of farmers, such as imposing harsher punishments for the offence of theft. However, the choice of the substantive model suggests that the legislator approached the problem as a defence issue in order to protect the farmers, placing them at the centre of their considerations. This model can also include changes in the law of evidence: if one of the required types of evidence is missing, there will be no charge. Changing the manner of proof can render the innocent guilty, or vice-versa. For example, moves such as lowering the standard of criminal proof or diminishing the possibility of presenting evidence in the presence of one party as a result of security (confidentiality) concerns may have a material effect on the outcome.

In sum, this model is concerned with the crucial elements of the offence, of the charge, and of the manner of proof. The legislator tends to apply this model in situations that it regards as critical social issues, when the aim is to change societal awareness. Thus, the model, despite being perceived as very basic, is comprehensive and extremely powerful because it paves the way for far-reaching social changes.

3.2. The punishment model

According to this model, the legislator can opt to change the punitive element of a law in response to social change. The punishment options are varied – from fines to imprisonment and even, in some countries, the death penalty. In most cases, when using this model, the reform will be targeted at the severity of the punishment rather than at its type.

For example, in 1996 and 2001 the Israeli legislator ordered the Penal Code to be amended in response to the increasingly light punishments imposed for offences of domestic violence.⁸¹ Sections 335(1A) and 382(c) were duly added. The declared purpose of these statutory provisions, then, was to toughen the punishment for such offences. Here, as part of the reform, the legislature directs the court with regard to the appropriate punishment considerations for domestic violence.

3.3. The procedural model

In this model the legislator uses legal procedure to reform the law. This can relate to pre-trial procedure, such as the investigation and the detention, or to the trial procedure itself. Using this model usually raises questions about the investigation, due process, representation and the inquisitorial or adversarial character of the system.

The scenario we have already examined in this article – the anti-terrorism legislation used to reform criminal procedure in Israel – is an example of this model in practice. By using this model, the legislator gives specific instructions relating to rules of procedure, which are directed towards both law enforcement agencies and the courts. The model answers the question of how to prosecute as a result of the reform. It does not address the question of the nature of the offence, the indictment, or the proof.

4. The three models combined: integrated implementation

A discussion of Israel's new Counter-Terrorism Law (2016) will serve to demonstrate the use of integrated implementation of all three models – substantive, punishment, and procedural, combined. This new law brought together the various laws dealing with terrorism to provide unified, comprehensive and coherent legislation that is up to date. It includes significant changes at various levels in the fight against terrorism: reforming the definitions (what is a terrorist organisation, what constitutes membership of a terrorist organisation, what classifies as an act of terrorism, and so on); reforming the substantive law by adding offences (such as incitement and threats); reforming the procedural rules; and reforming the corresponding punishments.

Unlike the process described above, whereby most of the reform regarding criminal anti-terrorism legislation in recent decades has been largely procedural, this law draws on all three models of criminal law reform. From this point of view, this law and its aims mark a turning point in Israel's approach to terrorism and criminal law. While, in the past, the reforms were focused largely on the procedural model, they now extend to include also the essence of the law, the definition of offences, and the punishments to be applied.

More specifically, the Israeli legislator used an integrated implementation approach to enact the Counter-Terrorism Law (2016). This move meant that most of the ATCL procedural rules from 2006 were absorbed into the new law, by applying the procedural model; and harsher and more extensive punishments for terror offences were adopted, using the punishment (penal) model. A heavy punishment tariff was applied to accessories in the commission of a terror crime, and life imprisonment was enacted for the principal actors in acts of mass terrorism that cause bodily injury to multiple victims. However, while ATCL emerged in the wake of significant procedural changes (that is, applying the procedural model of reform, which is unique to ATCL), from 2016 onwards the legislator drew on all three models in mapping out the Counter-Terrorism Law, taking an integrated approach.

The state justified the creation of the Counter-Terrorism Law with the need to address the threat of terrorism as a matter of emergency. It refers to the phenomenon of terrorism as a unique crime, explained both by the intensity of the injury it produces and by its scope and complexity.⁸² In addition, as a matter of principle, the state generally considers terrorism to be a grave phenomenon, by definition – that is, not an act characterised by various levels and degrees of complexity or severity. For instance, transporting a terrorist, even without knowing that they are a terrorist, is considered a serious terrorism offence. Therefore, giving a lift to a terrorist hitchhiker is classified as providing a service to a terrorist organisation also, on the presumption that the driver had their suspicions but refrained from taking action to test that suspicion, and thus is duly punishable.⁸³

Despite the definition of terror as a special ‘crime phenomenon’ and not only as mens rea or an ideology, the tools the state offers to handle it are of both a criminal and public nature. That is, new offences, severe procedures and harsher punishments are ‘criminal tools’, while ‘administrative’ detention (to which Israel resorts in cases where there is no evidence, holding detainees for up to six months) and special authority granted to the Attorney General and ministers are ‘public tools’. Alongside the selection of these tools, the government's explanation for the legislation – which distinguishes ‘anti-terror’ from other criminal issues – relies on the psychological effect of terrorism. This effect is based on the fear and the threat to public security that terrorism generates and its insidious undermining effect that damages daily life for so many people, beyond the direct physical and material damage it produces. Hence, the legislator positioned the phenomenon of terrorism as primarily exerting a public and psychological effect, which calls for public and psychological tools to manage it.⁸⁴

‘Terrorism’ is explained in the Counter-Terrorism Law as a ‘multifaceted enemy’,⁸⁵ which has various methods and goals.⁸⁶ According to this logic, then, the security-offence suspect is an enemy from the outset.⁸⁷ Here, criminal law becomes an instrument to be employed against the enemy and not just law in a democratic state – that is, it becomes ‘enemy criminal law’.⁸⁸

Whereas the focus in criminal law is on the act, here, the actor is at the centre. Using this field of law, the state automatically denounces every citizen acting against it as an enemy, from the very beginning of the investigation. For example, the law allows charities to be defined as terrorist organisations if there is any affiliation between them and terrorist activities, and also prescribes punishment for those who publicly express identification with a terrorist organisation.

Importantly, the broad definitions delineated under the Counter-Terrorism Law greatly expanded the boundaries of the terrorism phenomenon. Essentially, it offers one same format encompassing various forms of behaviour, corresponding with harsher punishment and suspension of the ordinary rules of law, without the necessary attention to the differential details or nuances of that behaviour. One example of this expansion is the inclusion of ‘surrounding organisations’ that support or are somehow linked to the organisation that carries out terrorism in the definition of ‘terrorist organisation’. Prior to this new legislation, such organisations were not included in the definition. This means that organisations that perform humanitarian activities that encourage and support the activities of a terrorist organisation, or belong to the same political movement, are now included in the definition of ‘terrorist organisation’, even though the surrounding organisations do not perform acts of terror themselves. Teachers working in a Hamas school or kindergarten, for example, would be caught by this definition.

The reform also extends the definition of a ‘body of persons’, adding a reference to ‘factions, groups and sub-organizations’ known by various names and engaging in different activities in relation to the terrorist organisation. As the state places significant emphasis, in the legislation, on defining the offender, it refers to military, political and humanitarian organisations in the same terms, to avoid acknowledging the activities (welfare, education, and so on) of humanitarian entities associated with political or terrorist organisations. Thus, criminal law applies equally, regardless of whether the activity is promoting terrorist activity per se, or engaging in humanitarian activities under a shared or related political movement. ‘Service provision’ for those humanitarian organisations is also included in the definitions of terror offences. Consequently, humanitarian organisations are declared terrorist organisations – even if their members do not seek to promote or facilitate the activities of a terrorist organisation and/or are unaware of the terrorist activities of the organisation in which they operate.

In its definition of ‘member of a terrorist organisation’ the legislator expands the scope considerably. Departing from the behavioural element of ‘taking an active part in an organisation’ from the previous law, the new legislation expands the definition to include anyone who merely expressed a notional agreement to join a terrorist organisation in the presence of someone they had reasonable grounds to believe was a member of that organisation, or who presented themselves as a terrorist organisation member or a representative sent on behalf of the organisation. So, in fact, the law now classifies as a ‘terrorist organisation member’ anyone who gave even passive agreement to join said organisation in the presence of a third party (who is not necessarily a member of that terrorist organisation); the same applies if they expressed even a vague statement to that effect to someone who is a member of a terrorist organisation.⁸⁹

Hence, the definition does not require actual participation in the organisation, and it even flips the burden of proof so that the suspect needs to demonstrate that they are not a member. Crucially, the definition does not relate to a specific date of any activity or alleged membership, and the membership period never ends. According to the explanation of the legislator, this is because it is difficult to prove terrorist organisation membership, as the process of becoming and remaining a member does not require the existence of a formal proceeding. As a solution to this difficulty, the new law blurred the distinctions between security offenders and humanitarian activists or innocent civilians, and imposed severe criminal sanctions and harsh criminal proceedings on both groups.

The definition of ‘act of terror’ was also expanded significantly. It includes both the act and the threat – hence, a threat designed to ‘provoke fear and panic’ is classified as an act of terror.⁹⁰ The definition expands the purpose of the terrorist act to include ‘prompting’ (rather than ‘forcing’) a government body or international organisation to take or refrain from taking a given action – thereby lending governmental authority to international organisations. In this definition, the risk element – in terms of potential damage to property, the economy or the environment – was also expanded. Thus, where weapons are used or there is a threat to use them, the law waives the purpose requirement. Another significant expansion can be found in the definition of the ‘act of terror carried out by a terrorist organization’. The current definition states that, if the act is performed by a terrorist organisation, there is a legal presumption that the motive and purpose already exist and, therefore, the prosecution does not need to prove those elements.⁹¹

The combined net result of these expansions is that many different scenarios, types of behaviour and acts – from the most innocuous and minor to the most severe and malign – are treated within the same framework. That is, the law's treatment is uniform, as it considers all such forms of behaviour as ‘severe’ actions that require the application of strict procedural processes and extreme sanctions.

Furthermore, it was not only the definitions of terror-related offences that changed under the new law, but also the associated maximum penalties: Article 37 of the 2016 Law doubles maximum penalties if the offence is considered an act of terrorism.⁹² Harsher prison sentences were also established for accessories who assist in the commission of a terror crime, including a life term for mass terrorism perpetrators who commit acts that cause bodily injury to multiple victims.

An examination of the effect of ATCL procedures on general criminal procedure is not central to this article, mainly because this understanding is hampered by the lack of disclosure surrounding the existence of criminal-security law and its characteristics. Despite this impediment, as part of its theoretical framework, the article does draw on the extant literature addressing this effect. This scholarship also examines the general implications of legislation in the criminal-security field and its interrelationship with general criminal procedure. Such studies consider the violation of specific rights, such as the legal presumption of risk in the detention of security suspects or the restriction of the right to silence in Northern Ireland.⁹³ Some of these studies also point to a substantial extension of special provisions from criminal-security law to ordinary general criminal law. American law scholars have also written about such expansion in the security context, which relies mainly on a similar discussion around leakage from the field of law dealing with illegal drugs into general procedure⁹⁴ and the extension of temporary arrangements in criminal-security legislation within general criminal law, and justifications for this. The concern has also been expressed that today's creation of separate rules for terror suspects could lead to separate and different criminal treatment of another group tomorrow.

The replacement of some temporary emergency powers with the permanent 2016 Law was in line with the general trend in emergency regimes. Israel now uses permanent law for the regulation of terror as an emergency, rather than temporary law or a ‘declaration of emergency’ model.⁹⁵

5. When the procedural model is used against one specific group: Terror suspects

As we have seen, Israel has made significant and comprehensive reforms in criminal law procedure over the last four decades for the adjudication of terror cases. I have shown that the reform in procedural criminal legislation has led to an entirely distinct field of criminal law: ATCL. The roots of procedural ATCL stem from the Mandate period and the establishment of the State of Israel, and this field of law remained relatively stable until the 1980s, undergoing almost no changes. Security offences in the first decades after Israel's establishment focused mainly on espionage and treason. Because of technological changes during the 1970s, the scope of the use of security legislation dwindled. It was in the 1980s that anti-terrorism procedural law was enacted: first, in the Prevention of Meeting Law (1981) and subsequently in the Investigation Documentation Law (2001). The 9/11 attacks and the increase in global terrorism, along with the 2005 disengagement from the Gaza Strip, prompted the enactment of the Temporary Order⁹⁶ discussed in Section 2.

Figure 1 shows three periods in the development of ATCL in Israel and the relevant model for each period. Here, we see the central role of the procedural model in this development.

Figure 1. How did the law arrive at ATCL?

This brief survey of Israel's employment of criminal law to regulate terror raises the question of whether it upholds the principles protected in general criminal law. The process by which the legislator distinguishes between general offences and security offences in the criminal process is vastly different from that of the special security offence provisions enacted in the past. As mentioned above, previously the legislator categorised security offences as criminal in nature, considering them severe and thus enabling criminal presumptions to be made, such as the presumption of dangerousness or even the need for harsher detention and sentencing. In contrast, the legislation in the field of security offences over the past four decades has been characterised by far-reaching procedural changes,⁹⁷ which have effectively created two procedural systems: the regular criminal procedural system for general criminal offenders, and the special criminal procedural system for suspected security offenders.

Over time, ATCL has developed unique features that distinguish it significantly from regular criminal law. For example, in ATCL there are special prosecution and defence attorneys in this field as a result of the different processes and security clearance issues involved. Because the Shin Bet is involved in the criminal processes in court, there are always two or more prosecution attorneys in court and only one for the defence. Another striking feature is that investigations are very lengthy and tend to end in a confession.⁹⁸ As mentioned earlier, there is a dual investigation: first by the Shin Bet and then by the police. The process of the former is largely undocumented (only summarised); the process of the latter is known as a ‘laundering investigation’, which gives the official, ‘cleaned-up’ account of events. This name points to the two phases of ATCL investigations, which are very different from those conducted under regular criminal law. Because of the exemption from the duty to record the investigation, there is no documentation of the confessions. Unlike the general criminal law, there are still ‘trials within a trial’ in this field, ostensibly to ensure free and voluntary confessions, even though ATCL investigations and trials almost always end in confessions. Because of the procedural character of ATCL, it turns into a game of ‘nothing to lose’, with both sides doing everything they can to achieve their desired outcomes. Because of the ex parte court hearing, judicial supervision is restricted and the materials related to the investigation are typically classified as confidential information – meaning that, to gain access to them, the defence attorney must approach the Supreme Court for permission.

Most of the offences involve passive behaviour and can be placed in the ‘second circle’ – that is, outside the sphere of hard-core acts such as carrying out a specific terrorist attack, or ‘ticking bombs’ such as conspiracy, aiding, weapons possession, contact with a foreign agent, speech offences or graffiti.⁹⁹ When such cases come to trial, there is little evidence, no exonerations, many plea bargains, and severe punishments. These tactics indicate that the procedural model is being used against one specific group – namely, terror suspects.¹⁰⁰

To analyse ATCL, scholars can turn to theoretical interpretive research together with qualitative research. The former presents criminal–security law and its analysis (the ‘law-in-books’ approach); the latter describes this process in reality (the ‘law-in-action’ approach), a description that is different from the legislative process. Applying both research methods in parallel provided a comparative avenue for the present study to describe the security criminal process and propose an explanation for the legislative developments. The qualitative method enabled subjective interpretation of the responses of the interviewees¹⁰¹ (prosecution lawyers and defence lawyers), all of which were anonymised, and an understanding of the socio-cultural reality of ATCL. It therefore enabled the phenomenon of ATCL and its characteristics to be investigated in the area where it is routinely applied. The method was also helpful in gaining a deeper understanding of the phenomenon, its interpretations and its layers, and enabled the key actors in the prosecution processes – defence and prosecution attorneys – to tell their stories and have their voices heard. These attorneys have first-hand knowledge of the ATCL phenomenon and its procedures, objectives and characteristics, such that focusing on these actors provides insights from a grassroots perspective. In particular, their testimony reveals scenarios in which the procedural model has been used to target terror suspects specifically.

The interviews quoted in this article were tailored to the purpose of the study: to understand the field in depth through the actions of the actors involved in the process of prosecuting suspects in security offences – from the prosecution and from the defence. These lawyers have a comprehensive understanding of the intersection between terrorism-related criminal proceedings and ordinary criminal proceedings. The attorneys who are engaged in defence representation are ‘returning players’ in the field; they are engaged in daily and ongoing practice, meaning that they regard it as a routine matter and a contemporary experience. As such, they are expert practitioners who understand both the language of general criminal procedure and the language of the criminal procedure of terrorism. Their deep acquaintance with the issue enables them to see the complexity at the micro-level, which is the perspective required to fulfil the objectives of the study.

One interviewee from the prosecution emphasised that the attitude towards the security suspect being investigated in the criminal investigation is, first and foremost, as a source of information. Therefore, the timing of the individual's disclosure (as a criminal being interrogated) and whether they will be allowed a meeting with a family member or lawyer are important questions for all involved. One of interviewees highlighted these issues when describing the process of ‘extracting information’ from the suspect, which may take a week or two, or even longer:

[A]lways think that the interrogee in front of you is a source of information. He is not a defendant, he is not a good man, he is not a bad man, it does not interest us what his punishment will be. [He is a] source of information because he comes from there and he knows a lot of things we do not yet know [so, the point at which he is allowed to be] exposed to sunlight … or to the wider world, … it could be relatives, it could be lawyers, it could be his other friends, when exactly he will be exposed, is a critical question for the ability to extract this information from him.

Talking about the special procedures in the Temporary Order, an interviewee from the defence made similar observations about the interrogee being a source of information, but from a different point of view:

If the purpose of an ordinary criminal investigation is to reveal the truth about a specific incident—whether or not a person committed the [acts] attributed to him—investigations into security offenses aim at gathering intelligence, gathering general facts. You see it because you always ask [the suspect]: who did it? Did you see anyone else at the crime scene? In the GSS investigations before the police investigation, … the detainees are broken so that they make a confession and [are then] brought to the police weak and powerless … [so that] all that needs to be done is to just sign a police statement. So the goals are different and the means of investigation are different.

According to this interviewee, both the questions asked in the interrogation and the means used to extract information are derived from the perception that the interrogee is a source of information necessary to thwart future offences, thereby creating a deep distinction between ordinary criminal procedure and security criminal procedure. This position of the defence attorney in many security cases shows that he discerns a major difference between prosecuting security suspects and ordinary suspects.

5.1. The liberal model

This study shows that the reasoning behind the legislation of ATCL was mainly that it was a response to an emergency situation. How does the liberal model explain the existence and characteristics of ATCL procedures in the context of the liberal debate? Two basic conceptions of liberalism are relevant here, from which its approach to emergencies are derived: one concerns the prioritising and protection of the individual (the foundational principle of liberalism being the individual and not the collective); the other is the existence of a broad legal framework which applies to all activities in the country: the ‘rule of law’. On the basis of these two conceptions, liberalism seeks to subject all state activities to a legal framework that prevents the government from exercising its authority arbitrarily and restricting individuals’ liberty and rights.

Liberalism's treatment of emergencies does not exempt them from the rule of law; even in such situations the legal framework must be maintained. While liberalism recognises the need for ‘exceptional’ authority in an emergency, this does not equate to the absence of law. In countries where the rule of law is a foundational idea there is a tendency to apply the law also to the exercise of arbitrary power. In the historical process of the development of liberalism, which culminated after the Second World War, liberal democracies preferred to approach the constitution as the legal source that would provide inherent and broad emergency authority to deal with states of emergency. The state of emergency in these countries is perceived as a disruption of the order, but not an illegal disruption.

The idea of the rule of law is achieved through the separation of powers, which preserves arbitrariness and ensures the freedom of the individual in the socio-political structure. It also expresses a clear political order, within which there are restrictions and restraints on the powers of the executive. According to liberalism, only an individual with sufficient power can restrain another's attempt to yield power over them, and therefore those exposed to the power of the sovereign cannot defend their liberty. The fear that liberalism seeks to tackle is that of harming the freedom of human beings, who are exposed to the arbitrary power of the sovereign. Liberalism especially emphasises the separation of powers between the authorities, using appropriate checks and balances, as a binding idea: they must balance and restrain each other in order to ensure that the individual remains free from the centralised state. Therefore, in the separation and balance of power between executive, legislature and judiciary, liberalism ensures the individual's ability to act freely within the framework of the law. What guarantees individuals’ freedom of action is the existence of a stable constitutional government operating within a legal framework that ensures this separation and balance between the three branches of government.

Liberalism is aware of the need for a decision in an emergency to protect the liberties of individuals. The exceptional situation is regulated by the norm through the possibility of declaring a state of emergency. Liberalism believes that it has created the appropriate framework for the individual's development and social activity by proposing the rule of law – not the rule of power. By law, in emergencies the state can repeal the regular law under a clause in the constitution, meaning that, even in emergencies, the action of the authority is within the law.

The United States represents an example of the functioning of the judiciary as a brake on the executive branch. The president enjoys broad emergency powers but, whenever the legal norm in emergencies has been violated, the Supreme Court has curtailed the range of action of the executive branch. For example, it ruled that the Court had the authority to decide on the constitutionality of detaining foreign nationals in Guantanamo Bay;¹⁰² that the special military committees that have tried illegal combatants violate the norm in military law;¹⁰³ and that law enforcement agencies are not allowed to deny detainees with US citizenship minimum legal protection¹⁰⁴ (such as the right to be represented by counsel). Similarly, the Supreme Court in Israel also restricted the powers of the authorities when it rejected the provision regarding the possibility of holding a hearing to extend detention without the detainee being present at the hearing, as described earlier.

In summary, according to liberalism, the law applies to everything, and it is impossible to exercise political power beyond its sphere. According to this approach, power is subject to restraint and institutional restrictions. Liberal constitutionality is twice limited: both through judicial review of the actions of the executive administration, and through the organisation of power in the law. In democratic/liberal states, where there is a rule of law, the tendency is to maintain certain norms, which are set by law, even in emergencies. When the executive branch acts contrary to these norms, the main barrier to restraint is the judicial system.

The two methods described above (law-in-books and law-in-action) revealed two primary characteristics of the ATCL research field: the emergency aspect and the investigation-and-intelligence aspect, respectively. The former, which emerges from the legislative analysis, is concerned with the fact that security offences constitute a unique arena in criminal law because of the severity of the offences. The regular criminal legislation and basic principles do not apply here. The latter, which emerges from the law-in-action analysis, is the goal of the procedural process. Unlike general criminal law, the security criminal procedure does not seek to adjudicate but to protect the investigation process and intelligence gathering.¹⁰⁵

Liberal theory has difficulty in explaining these ATCL features. Here, Schmitt's theory (relating to emergency powers), together with Foucault's theory (relating to power as manifest in the ‘investigation mechanism’) do a better job of explaining them in the context of Israel's security criminal law.

The analysis of ATCL in-books and in-action, and the conclusion regarding the creation of security criminal law as a separate discipline with its own characteristics (the emergency aspect, according to an examination of the legislation and case law, and the aspect of investigation and intelligence gathering, according to a study in the field) can replace the literature's previous analysis of ATCL, which stands at the intersection between security and human rights.

The liberal debate places the conflict between national security and the rule of law and protection of human rights at the centre.¹⁰⁶ As a result, it emphasises the confrontation that is played out between the prosecution and the defence.

From the perspective of the defence, human rights must be protected in the criminal process, regardless of whether it is within the specific arena of the criminal-security process. According to the defence, the usual criminal investigation, intended to prove evidence, must take place for all kinds of offence. From the perspective of the prosecution, the criminal-security procedure must, first and foremost, assist in maintaining state security. As a solution to this conflict, the liberal view proposes balancing these opposing values, weighing one against the other in accordance with the spirit and goal of the law, relevant in that specific state and that social reality, as well as the overall values that society holds.

The theoretical discussion in this article has examined the characteristics of Israeli ATCL, placing them in relation to the liberal discourse.¹⁰⁷ I argue that the liberal discourse does not offer a viable explanation for the distinct characteristics of this field – primarily, the lack of law and the lack of supervision in this process. The procedural legislation has created a separate area of criminal law to which general criminal law does not apply, and the liberal model cannot explain the existence of this sphere, as the basic approach of liberalism supports the primacy of the rule of law as a broad legal framework – one that applies to all state actions. Moreover, the liberal concept creates a dichotomy between legal and legitimate and illegal and illegitimate – that is, the state's use of political power is legitimate if it is used legally. Thus, state actions that are not based on a legal standard are considered violent.

Post-liberal theories, which criticise this discourse, enable us to overcome these difficulties. Thus, following the theoretical analysis of the liberal discussion, in the light of the difficulties therein I explain ATCL through the legal political philosophy of Carl Schmitt.¹⁰⁸ This move is mainly in the light of the characteristics of ATCL in-the-books, whereby the legislation of ATCL creates a sphere without law as a matter of emergency. Schmitt focuses on the sovereign's decision to declare a state of emergency and concludes that the sovereign decides on the exception. However, the explanation of ATCL offered by Schmitt's theory is also lacking, as it does not relate to the goal of criminal-security procedures and the distinction between these and general criminal procedures, particularly the activity of the everyday investigation system – not only as an emergency matter.

ATCL in practice (from the law-in-action perspective) shows that it focuses on the investigation-and-intelligence dimension and that the main aspect of the criminal-security process is intelligence gathering. Along with Schmitt's ‘state of emergency’ aspect, the other primary goal of the criminal-security process is to protect intelligence gathering, which explains the vast differences between general criminal law and ATCL.

In sum, both aspects of ATCL (state of emergency versus investigation and intelligence) sit in contrast to the relevant liberal facets of general criminal law. Notably, every time the Israeli legislator or the court expanded the rights of all (regular) suspects, security suspects were excluded from that expansion – in the Prevention of Meeting Law (1982), the Documentation of Investigations Law (2002), and the Temporary Order regarding Security-Offence Suspects (2006). Hence, when the legislator expanded the rights of ordinary suspects and determined by law their right to meet with their attorneys and have their investigations documented, these extended rights did not apply to security suspects. Also, the 2006 Temporary Order denied security suspects rights to which they had previously been entitled. Even in very recent times, after the Israeli Supreme Court declared a minimum cell size for every inmate (both suspects and prisoners), the government presented a new draft bill before parliament to exclude security suspects from this right.

The problem here lies in the simultaneous implementation of all these exceptions. These laws, acting together, have shaped the ATCL field and marked its exception from the general criminal process. The simultaneous operation of these laws has led to harsher treatment of criminal-security suspects in comparison with general criminal suspects and has denied them rights that are protected as a matter of principle in general criminal law procedures.

In this respect Schmitt notes the difficulties associated with liberalism (that a state of emergency is the exception that proves and demonstrates ‘everything’)¹⁰⁹ and contends that the problem lies in the fact that liberalism's otherwise holistic description of the rule of law makes no reference to its failures. Schmitt deems these failures systemic and necessary in the context of emergencies (that is, in ATCL): he claims that as the field (state of emergency) exists outside the realms of law and constitution, the law does not apply to it. Rather, the sovereign's decision is the reigning component, meaning that there is no general rule of law. Thus, if we apply Schmitt's theory, it helps to explain the sphere created in ATCL, to which general criminal law does not apply – an explanation superior to that of liberalism.

The other main aspect of ATCL emerges from an examination of the practical application of the legislation, taking the law-in-action approach. This is measured in the present study by qualitative methods, including interviews with prosecution and defence attorneys. The present findings on this approach reinforce, and add value to, those discovered by scrutinising law-on-the-books, showing that this field is also, in reality, excluded from the general criminal process.¹¹⁰ Specifically, general criminal lawyers do not represent suspects charged under ATCL, as it requires knowledge and special tools unique to this area. What is more, the defence rights of security suspects in criminal proceedings are also restricted in practice. Indeed, my findings show (in line with the extant scholarship) that the law fails to limit the use of criminal-government power¹¹¹ (state power in criminal law) and permits the law-enforcement authorities to exercise broad discretion without review.¹¹²

Moreover, these findings also suggest that criminal prosecution in ATCL not only serves criminal justice purposes but is also used in pursuit of gathering intelligence for investigation agencies. In the legalism process of Israel's ATCL procedure, both rhetoric and practice changed from ‘state of emergency and sovereign decision’ to ‘intelligence gathering’. The present findings indicate that most of the actors in this field believe that the intelligence goal behind ATCL (in preventing meetings, disregarding the duty to document investigations, and the Temporary Order) is fundamental. Investigation and intelligence are central organisational axes in criminal-security law. Contrary to the understanding of liberals, and even some of their opponents (including Schmitt) on this matter, this field is not only concerned with the sovereign's decision and state of emergency but also with the daily, routine mechanisms for collecting intelligence.

This means that the findings pertaining to ATCL in practice distinguish between two concepts: judgment in criminal law, and intelligence gathering of the Shin Bet. In the light of these findings and this distinction, Schmitt's perspective can also be considered insufficient to a certain extent. Instead of interpreting criminal law through the exception and the state of emergency, there is a legislative-administrative purpose that differs from that of general criminal law.

In sum, the analysis of ATCL in-the-books and in-action identifies the creation of security criminal law as a separate discipline with its own specific characteristics: primarily, the ‘emergency’ aspect (according to the legislation and case law examination) and the ‘investigation and intelligence gathering’ aspect (according to reality in the field). What is clear from the findings is that this is not a classic case of prosecution versus defence, unlike in general criminal justice. Rather, intelligence gathering is a distinct goal that changes the interaction of the players in the field. It is also clear that most of the cases prosecuted are for minor security offences peripheral to the more serious offences. The logic leading the field is not necessarily the principle of the state of grave emergency (hastily responding in order to prevent the use of bombs or missiles) but rather the logic of collecting information and intelligence as a quiet, routine practice. This is an extra-legal expansion, which the law does not directly address but enables. Although the general purpose of intelligence is to serve security interests, the routine nature of intelligence gathering and the fact that it is permitted outside the law are distinctive phenomena, in which the principle of intelligence is maintained as a central organising axis.

Meanwhile, Foucault's theory of power¹¹³ helps us to understand that, although the system presents itself as an emergency response, its purpose is not only to deal with emergencies but also to create an arena without law – a space which enables mechanisms for routine investigation that maintains the norms and order. In this arena we find continuous control and the use of force in the relationship between interrogator and subject.¹¹⁴ In this way, control is achieved both over the suspects and over the group to which they belong. These mechanisms track, analyse and influence the relevant population to achieve security, and ostensibly create an integrated and profitable society.

These mechanisms achieve effective control by employing continuous tracking of subjects as their main technique. Under ATCL, intelligence is routinely collected to build a domain of cumulative security knowledge. This intelligence gathering is based on the life stories of individual security suspects, but is also designed to accumulate information on the broader process of insurgency and terrorist activity. Knowledge is collected about these suspects as a personal matter but, once catalogued, it can be interwoven with other data, tested, and scanned into a broader picture of the security situation and the activities of groups and terrorist organisations. Governance techniques now centre on scanning and monitoring society, which means that detection of social movements and mobilisation processes is currently performed routinely, including via intelligence services.

Investigative control is achieved by measures imposed following the development of knowledge, as Foucault's theory suggests. In addition to the private information (of the suspects) and statistical information on populations, there are also mechanisms for accumulating intelligence and security knowledge as an independent purpose of ATCL.

6. Conclusion

The discussion here has shown that ATCL is not based on the general model of criminal prosecution but rather on the emergency and investigative logic of collecting information and intelligence. Liberal laws can regulate states of emergency through democratic means by imposing temporary limitations on human rights in the event of potential harm to national security. Thus, the state of emergency can be considered exceptional. Notwithstanding, I contend that the correct conception of ATCL is not only in relation to a state of emergency but also as standard practice within a system that maintains an ordinary way of life within the democratic context. The system presents itself as an emergency state, but its objective is not to deal with an emergency – rather, it is to produce an extra-legal arena that enables a ‘normalised’ routine intelligence interrogation mechanism to be operated. Finally, ongoing control and power are exerted in the system of relations between the interrogator and the security suspect.

Yet, precisely because the power of the information-gathering mechanism is centralised in the hands of the sovereign, the investigation-and-intelligence model does not explain the centrality of the sovereign. The role of the sovereign is significant and cannot be ignored. The concept of power in Foucault's theory is vague and almost metaphysical, making it difficult to understand the power used during investigation. Foucault's theory emphasises the existence of an investigation mechanism that uses power and control, but no more.

In summary, these two aspects can be identified in ATCL: the emergency dimension, as indicated from the findings regarding ‘ATCL-on-the-books’, and the investigative dimension, as indicated from the findings regarding ‘ATCL in-action’. Both of these aspects, together, explain ATCL, while emphasising the lack of law and the relationship of powers therein, as an alternative to the liberal claim, which focuses on the relationship between the individual and the state.

The comparison with the United States provides a number of understandings. First, unlike the US, Israel has executed comprehensive and significant reform in criminal law for decades in order to combat terror. Second, in Israel there is greater usage of the procedural model for one particular group of suspects – namely, terror suspects – and the new Counter-Terrorism Law (2016) integrated three distinct models to reform criminal law: what I term the substantive, punishment and procedural models. Third, the cumulative effect of the procedures created a unique and separate legal field in parallel with general criminal law – ATCL – with its unique characteristics that radically expand the powers of the authorities over terror suspects and detainees. Finally, as a result, Israel reformed criminal law in an anti-liberal manner which promotes emergency and intelligence-gathering rather than criminal law objectives.

This article illuminates the nexus of criminal law and terror, which can differ from country to country. This feature of ATCL can make for a unique adjudication process, which is distinct from that of the regular criminal process. As a result, the characteristics of criminal law are broader and include both citizens and enemies; law-and-order together with emergency interventions; investigation for the purpose of routine intelligence gathering; and liberal ideas alongside non-liberal practices. Ultimately, ATCL renders criminal law less liberal and, more importantly, presents serious flaws in liberal theory that need to be addressed.