The ECtHR’s suitability test in national security cases: Two models for balancing human rights and national security

Abstract

The European Court of Human Rights has often been criticized for lacking clarity and consistency in its reasoning of balancing human rights against conflicting public interests. To reconcile national security with human rights protection, the Court requires the interference with rights to be suitable for reaching the objective purported by the government. In this article I deal with how the Court conducts the suitability test in national security cases, in line with two models under which a few representative test considerations can be categorized: human rights priority model and national security priority model. To explain how each model works in a comparable sense, I follow the same analytic structure and examine the manner of the Court’s test and the intensity of its scrutiny. I argue that in compliance with the two models, the Court applies the suitability test in a consistent and predictable way in national security case law.

1. Introduction

Suitability is one of three tests that has evolved in judicial interpretations of the principle of proportionality,¹ which is often applied to examine whether or not the government’s interference with human rights can be concluded as justifiable. It requires government measures to be proven suitable to achieve the intended goal.² The principle is derived from the case law of the German Federal Constitutional Court,³ which has later been introduced into international human rights regime.⁴ The European Court of Human Rights (hereinafter the Court), though not strictly following the orthodox structure of analysis developed by German case law,⁵ has undoubtedly embraced the suitability test in its own decision-making process.⁶

Is national security the Achilles’ heel of the discourse on human rights protections? Facing practical needs and political pressure in recent decades, government authorities are reviewing norms, policies, and instruments as they attempt to strike a balance between national security and human rights. As the human rights watchdog, the Court has reviewed a number of disputes arising from the government’s national security actions that are allegedly in conflict with its human rights obligations. These cases have been filed under the ‘qualified rights’ category of the European Convention on Human Rights (ECHR),⁷ and occasionally, under its ‘limited rights’ category.⁸ Although the Court barely uses the term ‘suitability’ in its judgment, the effectiveness of government measures in reaching its declared goals has been a main element to be reviewed by the Court. For instance, in Big Brother Watch and Others v. the United Kingdom, the Court first acknowledged that ‘operating a bulk interception regime’ served the aim of ‘identifying hitherto unknown threats to national security’, especially in the context that advancements in technology have made it easier for terrorists and criminals to evade detection on the Internet, and electronic communications are transmitted via unpredictable routes.⁹

Notwithstanding that the suitability test is not an established research subject on its own, there is much insight from scholarly attempts to shed light on how Strasbourg organs have applied the principle of proportionality.¹⁰ One sort of attempts comprehensively looks into how proportionality principle is applied in case law under ‘qualified rights’ category.¹¹ The other sort is subject-specific and focused on the scope and conditions for government’s interference in certain rights.¹² The existing literature, however, does not address sufficiently the topic of national security, let alone any research finding about the review pattern depicted by the Court. Scholars often attribute the Court’s features of review to the doctrine of margin of appreciation, but their study has not deduced what does wide or narrow margin of appreciation mean specifically in case law. I will develop this argument in the next section. In practice, the Court applies general standards which have been established in its case law when it is confronted with the same issue again.¹³ There are arguably patterns when the Court reviews national security cases. In this light, I intend to find out the patterns by detailing how the margin of appreciation is unfolded in national security cases.

The main question explored in this article is how the Court applies suitability test in reaching its conclusion in national security cases. I unfold my analysis in three main parts. In the first part (Section 2), I indicate the limitations of using the doctrine of ‘margin of appreciation’ for clarifying the Court’s decision-making patterns. By taking national security case law for example, I argue that the doctrine needs to be specified in a normative sense in order to provide a more precise picture of the Court’s decision-making patterns. In the second part (Sections 3 and 4), I find that the Court checks the suitability of government’s interference with a clear preferential framing. In my reading, the Court applies different norms which demonstrate some identifiable tendency for prioritizing the protection of national security over that of human rights, or the other way around. According to where this tendency is heading, I propose two models from a broad categorical perspective and compare the norms adopted by the Court in each model. In the third part (Section 5), by taking a case-specific approach, I focus on how the Court mitigates the preferential norms in line with the model being applied, in so far as it is required by the concrete circumstances of a case. I also identify circumstances under which the government’s interference has been held not surviving the suitability test. For this contribution, I aim to provide a paradigm through which the Court’s review of the suitability in national security cases can be assessed in a relatively coherent manner. By locating the consistencies and portraying the patterns, it helps to increase the predictability of the Court’s decisions on national security issues, guiding state parties to comply with the Convention.

Given the considerable amount of case law of the Court, I limit the research scope to the merits of the judgments of Chamber and Grant Chamber,¹⁴ under the qualified rights of Articles 8, 10, and 11 and Article 2 of Protocol No. 4, as well as the limited rights of Articles 5 and 6. Data collected for this research can be categorized into two groups: the first group contains recent case judgments that are made after 1 December 2013.¹⁵ The data on cases concerning qualified rights was collected by searching the HUDOC database through setting its ‘keywords’ filter, sorted by date.¹⁶ The data on cases on limited rights was collected mostly by using the snowball method.¹⁷ The second group of collected data consists of cases having a significant impact on the subsequent cases. They were collected on the basis of the existing literature,¹⁸ and the Court’s case law citations, those which were taken note of when reading the first group of cases.

2. Margin of appreciation: A practical solution to identify decision-making patterns?

The Strasbourg organs have been accused of inconsistency in their judgments. The doctrine of margin of appreciation is often quoted by scholars to explain why the Court conducts its reviews of the proportionality principle in different manners, who aim to illustrate the underlying consistence of the Court’s decisions. The margin of appreciation refers to ‘the latitude a government enjoys in evaluating factual situations and in applying the provisions enumerated in international human rights treaties’.¹⁹ The Court applies the doctrine in a way that justifies its scrutiny to be either close or superficial, reflecting in the discretion enjoyed by the government.²⁰ In a general sense, a right embedded with essential elements for democratic society often induces the Court to proceed a close scrutiny.²¹ In this context, the Court regards democracy and democratic society as values and purpose to be sticked to by government authorities, which grant a greater importance to the rights closely related to it. It also corresponds to the requirement provided by the Convention that any restriction on rights needs to be proved necessary ‘in a democratic society’.²² When a case concerns national security matters, it undergoes a deferential review by the Court.²³ The Court is demonstrating deference to the states, as it contends that the respondent government is more equipped to make the assessment. But the deference does not become obsequious. A European consensus and common practice of states are among the arguments used by the Court to prevent the Convention from falling into cultural relativism.

Two major problems stem from attempts that solely adopt the margin of appreciation doctrine to make sense of the decision-making pattern of the Court in reviewing suitability in national security cases. The first problem results from a contradiction in the doctrine itself, i.e., the nature of the rights involved and the subject-matter of the case. The doctrine does not stipulate that either a deferential review or a close one should be applied when a national security case involves a government interference with a right accommodating democracy characteristic. Instead, the answer to this is to be found by turning to the Court’s adjudications.

The second problem, which is more critical than the first one, concerns the doctrine’s practicality. As a norm for clarifying the patterns, it provides merely a breadth of the margin of appreciation that is being measured roughly as either ‘wide’ or ‘narrow’, but this is not accurate enough to indicate how suitability test is to proceed in a given case. Does a wide margin of appreciation mean that the Court should accept the government’s allegations about the effectiveness of its measures, without question? Or instead, that the state shall provide evidence to substantiate its allegations? If a narrow margin means a strict review by the Court, how should a ‘strict review’ be carried out in a specific case? The doctrine is helpful to classify different types of scrutiny, but it fails to state what exactly differentiates a strict scrutiny from a less strict one.

Some scholars have tried to resolve the problems associated with the doctrine of margin of appreciation by proposing some concrete criteria. For instance, Janneke Gerards has identified four elements that vary according to the breadth of margin: the clarity of the aims alleged by the government, the temporal issues of assessing effectiveness, the level required to be deemed effective, and the burden of proof.²⁴ Jonas Christoffersen works on the law of evidence by proposing three foci: admission of evidence, burden of proof, and standard of proof.²⁵ Despite some difference in their foci, the above-mentioned efforts share one goal: to develop some pattern in line with which the Court’s reasoning and review decisions would be consistent and therefore, predictable. My following analyses aim to make a contribution to these meaningful attempts.

3. Two models for analysing suitability

In this section, I will explore how suitability test is conducted in case law by the Court, followed by shedding light on application scenarios in the next section. The Court normally examines whether or not the contested measure is necessary on the basis of the principle of proportionality.²⁶ The legal basis is that any restriction is required to be ‘necessary in a democratic society’, in accordance with the provisos of Articles 8, 10, and 11, and Article 2 of Protocol No. 4 (qualified rights). Although limited rights do not accommodate provisos in the same format, the Court does sometimes make an assessment on the proportionality of the contested measure.

3.1 The rationale for discerning the two models

From a broad categorical perspective, depending on the nature of the right being interfered with, the Court applies a wide or narrow margin of appreciation. To be specific, the Court aims at the right’s connections with some essential features of a democratic society. As I indicated in the last section, the wide or narrow margin of appreciation demonstrates a clear preferential framing, as it determines if the Court’s scrutiny will be intense or lenient. In this regard, a question may be asked why the Court invokes this element (i.e., essential features of a democratic society) to determine the intensity for reviewing a case.

First and foremost, it provides a legitimate ‘anchor’ for balancing the contested interests. In case law, two different sorts of interests are at stake: national security, and personal freedom and rights. As Stavros Tsakyrakis points out, the balance between the two could remain a rhetoric or a manifestation in the human rights adjudications.²⁷ Neither of them can be measured quantitatively; or as Tsakyrakis puts it, they are not ‘amenable to any meaningful form of quantification’.²⁸ For instance, when measuring the interest of protecting personal communication from interception and the interest of detecting a terrorist plot, one cannot be so sure in the mathematics sense which interest prevails.

Alternatively, a more convincing balance may be reached via another approach. Jeremy Waldron indicates that weighing or balancing ‘is not necessarily Benthamite quantification but any form of reasoning or argumentation about values in question’.²⁹ Balancing underlines the reasoning behind how the Court assigns priorities to one party over the other,³⁰ and in this regard, accommodating essential features of the democratic society remains as a rather strong consideration for the Court to prioritize human rights protection over national security protection. In national security case law, the Court identifies these essential features as ‘pluralism, tolerance, and broadmindedness’.

Secondly, democracy is a shared value in Europe which is broadly accepted as contributing to the maintenance of fundamental freedoms.³¹ The Court’s determination of whether nation security reconciles with human rights is achieved by interpreting the limitation clause of the said Articles (Articles 8, 10, 11, and Article 2 of Protocol No. 4). When interpreting an international treaty, purpose is one of the three factors to be considered according to the Vienna Convention on the Law of Treaties.³² In this regard, it is legally defendable for the Court to use ‘democratic society’ as a normative value when reviewing the case.³³ When it comes to the practice, the ‘democratic society’, as Alain Zysset observed, plays a normative role as the Court reviews cases that involve political rights.³⁴ Zysset reads the notion functioning to delimit the scope of right, and as a collective interest to be balanced in the proportionality stricto sensu, which is the last step of the proportionality analysis.³⁵ I agree with him on the former point, whereas regarding the latter, I argue that in national security case law the ‘democratic society’ sets the premise for the Court’s proportionality analysis.

In my observation, the Court has identified three rights that accommodate ‘pluralism, tolerance, and broadmindedness’ features of the democratic society in case law. They are the freedom of speech, the freedom of assembly and association, and the right to respect for private life. The Court prioritizes these rights over national security. The chief reason for this prioritization is that, once such essential features are set aside, democratic society would be more nominal than substantial. When it comes to the remaining rights, the Court assigns priority to national security. The underlying reasoning is that the state is the entity in which democracy is vested, and democracy will only survive when national authorities preserve the security of the state.³⁶

3.2 The decision-making pattern under each model

In accordance with the margin of appreciation applied by the Court, I argue that the Court adopts different norms with a clear preferential framing in its reasoning in national security cases. I make a hypothesis that under the narrow margin of appreciation, the norms reflect a general tendency for prioritizing the protection of human rights; I thereby name this scenario the ‘human rights priority model’. Correspondingly, the wide margin of appreciation implies a general tendency for prioritizing national security concerns, so I name this scenario the ‘national security priority model’. In terms of suitability test, I will analyse the Court’s decision-making in two aspects: the first is a normative one, concerning the desirability of the effectiveness of the government measure in question; the second is a factual one, relating to the Court’s evaluation of evidence and other materials that substantiate the government’s allegations.³⁷ In this sub-section, I will delve into the normative and factual aspects of each model, and from a broad categorical perspective make sense of the decision-making pattern of the Court’s review of the suitability.

3.2.1 Human rights priority model

To start with, the suitability test concerns the causal link between the means and the ends of the government action under scrutiny.³⁸ In my reading, under the human rights priority model, it requires the causality to be a relatively strong one from the normative perspective. To establish that, it requires the following: first, national security needs to be in a real danger. In the Stomakhin case, the Court indicated that as a general principle:

the adjective “necessary” implies the existence of a “pressing social need”, which must be convincingly established. Admittedly, it is first of all for the national authorities to assess whether there is such a need capable of justifying that interference … However, the margin of appreciation is coupled with supervision by the Court both of the law and the decisions applying the law …³⁹

How can a danger be recognized as real? There could be a ‘defence perimeter’, within the scope of which any danger may be real enough for the government to take actions. For instance, a political controversy, ranging from dissent opinions or criticisms against the government to separatist statements, often hits the nerves of the government which is prone to claim that national security is at stake. If such a case is brought to the Court, it would conclude in favour of the government only that the statements comprise incitement to violence.⁴⁰ In terms of incitement to violence, in the Mukhin case, the Court reiterated the norm set by its case law:

where the views expressed, for instance on political issues, do not comprise incitement to violence – in other words unless they advocate recourse to violent actions or bloody revenge, justify the commission of terrorist offences in pursuit of their supporters’ goals or can be interpreted as likely to encourage violence by expressing deep-seated and irrational hatred towards identified persons – contracting states must not restrict, on the basis of the aims set out in Article 10 § 2 concerning the protection of territorial integrity and national security …⁴¹

This is to say the state would be in fact asked to take a risk that those political oppositions might indirectly, partially or in a long run, contribute to a potential violence.

Secondly, a relatively strong causality requires the impugned interference to reach a high level of effectiveness for achieving the alleged purpose. For example, in the case law with regard to preserving state secret, the Court argued that ‘it was unnecessary to prevent the disclosure of certain information seeing that it had already been made public or had ceased to be confidential’.⁴² It is a matter of measuring the efficacy of the contested interference. In the Vereniging Weekblad Bluf! case, a journal containing a confidential report from Dutch security services had been circulated before the government took actions. In the opinion of the Court, although prohibiting the circulation might be able to reduce the damage by limiting its further spread, such a prohibition would not be sufficient enough due to its limited effectiveness.⁴³ The reason is that withdrawing the copies of journal could narrow down the scope or slow down the speed of the spread of confidential information, but it could no longer fulfil the aim of protecting the state secret now that the confidential information had been already disseminated.

It is worth noting that the government authority needs to submit its evidence to substantiate a sufficient link between interference and its purpose.⁴⁴ In terms of the factual aspect of the Model, the Court carries out a detailed review on the persuasiveness of the evidence available to it. In the Gîrleanu case, although the Court recognized that defining national interests forms ‘part of the inner core of State sovereignty’, it still examined whether the domestic courts took into consideration in their analysis some ‘specific element concerning the applicant’s conduct’.⁴⁵ In most cases, the evidence concerns whether or not there was a real danger to national security, rather than evaluating the degree of effectiveness in the government measures. This is probably because, as long as the danger to national security can be established in concrete terms, the question as to whether and to what extent the contested measure is effective to tackle that danger can be answered by restoring to common sense.⁴⁶ The Court does not necessarily need statistical evidence or experimental assessment reports to answer this question in some cases. For instance, in the Observer and Guardian case, the Court concluded that due to the fact that a book containing confidential material had already been published outside the UK, the interlocutory injunctions imposed on publishing details of this book in the UK could not prevent the classified information from being disclosed.⁴⁷ This assessment of effectiveness came straightforward as the Court did not have to find out how many books have been imported into the UK, or how many book stores were selling them.

To demonstrate the danger to be actual, the government sometimes puts forward expert conclusions,⁴⁸ or internal assessment reports.⁴⁹ Whilst the evidence submitted appears to be specific, well-organized, and impartial, the Court is not ready to agree with government authorities’ claims. In practice, the Court even scrutinized the method of the assessment. In the Smith and Grady case, the Court noted the survey methods used by government authorities on the issue concerning how accepting homosexuals in the army would affect the servicemen’s morale. The Court downplayed the independence and value of this survey, as it was conducted by ‘Ministry of Defence civil servants and service personnel’, and participated by ‘only a very small proportion of’ servicemen; and ‘interviews and the focus group discussions were not anonymous’.⁵⁰ In addition, the Court found that many questions in the attitude survey were asked in a biased way.⁵¹ In other cases, the Court re-evaluated the wording, form, and nature of articles published,⁵² or the activities or programmes of a political party,⁵³ as well as the wide and immediate context in which the government actions were adopted.⁵⁴

3.2.2 National security priority model

In contrast, the intensity of the Court’s scrutiny depicts a sharp decrease under the national security priority model. Firstly, a danger to national security does not need to be imminent. For instance, the Court concluded on the usefulness of the government’s secret surveillance in terrorist cases, without a prerequisite to be established that there should be a terrorist attack ‘just around the corner’.⁵⁵ It underlined the role played by intelligence measures in detecting threats, as the Court noted that ‘it (bulk interception) enabled the security services to adopt a proactive approach, looking for hitherto unknown dangers rather than investigating known ones’.⁵⁶ Secondly, the suitability requirement is satisfied as long as no major defects are identified in the efficacy of the means. In other words, the measure imposed by the government could fail the test only if it is manifestly ineffective. Such an assessment can be arguably quite straightforward by using common sense. This norm adopts a negative formulation, and it can be explained much more clearly from the other way around. I will elaborate this point later in Section 5.2.2. In addition, the means may be accepted as justifiable when they contribute to achieving their ends in the long run or in a step-by-step manner.⁵⁷

As to the factual pillar of the Model, the Court is generally ready to accept the government’s argument. In particular, the danger of a potential or cumulative nature has been often acknowledged by the Court when the case concerns one or more of the following issues: counter-terrorism, confidentiality of critical information,⁵⁸ operation of security services,⁵⁹ and furthermore, for the sake of due effectiveness of their operations, the secrecy of their working methods,⁶⁰ or the outcomes they have produced.⁶¹ For instance, in terms of mass surveillance, the Court acknowledged in general ‘the current threats facing many Contracting States (including the scourge of global terrorism and other serious crime …), advancements in technology which have made it easier for terrorists and criminals to evade detection on the Internet, and the unpredictability of the routes via which electronic communications are transmitted’, and concluded that ‘the decision to operate a bulk interception regime in order to identify hitherto unknown threats to national security is one which continues to fall within States’ margin of appreciation’.⁶²

With regard to the effectiveness of the means, it may not always be the case that the state has to demonstrate the actual effect of the government’s measure, or that the applicant needs to prove specifically the existence of a major defect in the government measure.⁶³ For example, in the Bartik case, Russian government did not specify the effect of the international travel ban on the applicant who used to have access to classified information, by providing factual, statistical, or empirical information. In the meantime, the applicant did not attempt to prove that travel ban would be only marginally effective, if not entirely ineffective.⁶⁴ Instead, based on the information in hand, the Court would more or less resort to presumptions and inferences to evaluate the effectiveness of the contested measure.⁶⁵

4. Case scenarios of two models

4.1 Scenarios of the human rights priority model

I indicated in Section 3.1 above that considerations in favour of ‘pluralism, tolerance, and broadmindedness’ normally outweigh national security interests.⁶⁶ The rights embedded with these values are the freedom of speech, freedom of assembly and association, and the right to respect for private life. Each of those rights has multiple aspects,⁶⁷ but not all these aspects carry those key democratic values. Also, circumstances in which those rights may be interfered with by the government for the interest of national security are not infinite.⁶⁸ There are in general four sets of circumstances in which states are under close scrutiny of the Court, namely when they:

1. 1. Ban or sanction political speech (in conflict with the freedom of speech, and freedom of assembly);

2. 2. Dissolve a political organization (in conflict with the freedom of association);

3. 3. Ban imparting classified information on public interests (in conflict with the freedom of speech);

4. 4. Discharge individuals from the army on the ground of homosexuality (in conflict with the right to respect for private life).

The preceding circumstances of government interference concern, as confirmed repeatedly by the Court in its reasonings, essential features of a democratic society – ‘pluralism, tolerance, and broadmindedness’.⁶⁹ To be specific, statements and interviews of a political nature, and the organization of political parties are regarded as providing the public with different political arguments, choices, approaches, and goals.⁷⁰ Tolerating and discussing different opinions, however disturbing or shocking they might appear to some people, are essential for sustaining democracy. Apart from political issues, the Court has also attached importance to exchanging opinions on matters of public interests,⁷¹ and thus recognized the essential role played by the press in a democratic society.⁷² A free flow of information concerning public interests can attract people’s attention on a given topic and contribute to public debate. As to the policy on discharging personnel from the army on the basis of sexual orientation, the ‘pluralism, tolerance, and broadmindedness’ supports one’s freedom to choose the way one lives their life, even though it might cause some discomfort to the moral schema of others.⁷³

4.2 Scenarios of the national security priority model

Under the national security priority model, a prominent feature is that the Court is often prepared to accept a state’s allegations about the suitability of its interference measures. The circumstances that would fall under this model vary, but some typical scenarios can be summarized as far as the Court’s case law is concerned:

1. 1. Secret surveillance, including intercepting the content of communication, intercepting the data of communication, and intelligence sharing (in conflict with the right to respect for private life);

2. 2. Personal information stored in the secret registers (in conflict with the right to respect for private life);

3. 3. International travel bans on retired personnel who used to have access to classified information (in conflict with the freedom of movement);

4. 4. In camera trial (in conflict with the right to a public hearing);

5. 5. Non-disclosure of sensitive material and information supporting the reasonableness of apprehension of the suspect (in conflict with the right against arbitrary arrest and detention);

6. 6. The use of secret evidence in a trial (in conflict with the equality of arms, the right to an adversarial hearing, the right to a fair trial, and the right to prepare the defence);

7. 7. Prolonged pre-trial detention of terrorist suspects (in conflict with the right to trial within a reasonable time or to be released pending trial); and

8. 8. Delayed access to the lawyer in terrorist cases (in conflict with the right of access to a lawyer).

By resorting to the common sense, some factors, such as the nature of a terrorist crime or the operational effectiveness of intelligence services, contribute substantially to the government’s interference surviving the suitability test. When any of such factors occurs in an instant case, the Court is ready to recognize that the security of the state is at stake.⁷⁴

4.3 A summary

From a broad categorical perspective, I argue that two models have evolved to scrutinize the suitability of a state’s interference for national security concerns in case law of the Court. Rather than discrediting the margin of appreciation doctrine, the two models are in fact put forward on the basis of the doctrine, with its application in national security cases being specified. One is the human rights priority model, under which more weight is assigned to certain rights and freedoms. The other is the national security priority model, under which the importance of protecting national security is placed higher than the protection of rights.

Under the two models, the stringency of normative and factual requirements faithfully reflects the Court’s weighing between security and liberty, which further impacts, to a great extent, its conclusion. The human rights priority model requires the establishment of a real danger to national security and the high level of effectiveness of the government’s measure adopted. By contrast, under the national security priority model, as long as no major defects in the government measure can be identified, the Court would decide in favour of the state.

5. Redressing the bias of each model

When reading through the case law, we may see that the Court does not always decide the contested measures to be suitable for protecting national security under the national security priority model; likewise, reviewing the suitability of the government’s interference under the human rights priority model does not promise that the Court’s decision would be in favour of the applicant. It should be empathized that, when the Court is reviewing a specific case, it takes concrete circumstances into account in examining the suitability of a challenged government interference. This is arguably one of the reasons why the Court’s reasoning appears to be inconsistent. Nevertheless, some of those considerations play the same role in the Court’s decision-making progress: to redress the identifiable bias in favour of protecting human rights or preserving national security in line with the model being adopted.

5.1 Redressing the bias towards human rights

5.1.1 Cases involving political speech

The government interference of banning or restricting political speech is not suitable unless the speech amounts to inciting violence. Overall, such a high threshold will lead to a larger number of expressions to be spared than to be curtailed due to national security concerns. When it comes to a specific case, the question for the Court to answer is whether or not a statement can be identified as constituting an incitement to violence.⁷⁵ It is a matter of predicting the danger of the contested expressions. In this regard, the Court adopts a proactive strategy to address national security threats.⁷⁶ In general, the state is not being expected to wait until negative consequences have materialized.

A simple scenario is the impugned statement, in a direct and explicit way, which calls for armed resistance against government, violent uprising against the authorities, or terrorist attacks alike.⁷⁷ In a number of complaints against Turkey, the Court decided that statements containing such a call were a prima facie incitement of violence.⁷⁸ Furthermore, in these Turkey cases, the Court confirmed its judgment that national security was in danger, by taking into account the fact that serious disturbances occurred in the south-east of the country.⁷⁹ A comparable scenario is that the content of a statement is not as extreme as that in the last scenario, yet it is politically opposing the state authorities. On this type of political speech, the Court has normally taken two factors into account: whether there is a tense climate in politics or society, and the identity of the person concerned.

A statement propagating an ‘us-vs-them’ mentality is an illustrative example. As a general principle, the speech promoting hatred or intolerance does not necessarily amount to incitement to violence.⁸⁰ Specifically, it involves remarks that attempt to antagonize and dichotomize social groups, such as non-Muslims versus Muslims, nationals versus immigrants, one ethnic group versus another, or minorities versus the central government.⁸¹ This sort of speech may be subject to legitimate restrictions, but not necessarily on the ground of safeguarding national security,⁸² as its content neither advocates recourse to violence nor justifies terrorist attacks.⁸³ However, in a context where violence and conflicts are ongoing or have recently ceased, the applicable legal ground for restrictions may be rather different as the above-mentioned statements now run the risk of rendering harmful consequences imminent and pressing.⁸⁴

For instance, in the Stomakhin case, by labelling Russia’s army and security forces as ‘maniacs’, ‘murderers’, and otherwise criminally minded personnel, the impugned publications incited the hatred of the Chechen people towards Russian government.⁸⁵ In light of recent disturbances and terrorist attacks following the Second Chechen War,⁸⁶ the negative emotional assessments made by these hatred statements remained no longer an issue of discrimination or encouraging sporadic hate crimes, but would justify and stir up violence against Russian army and security forces,⁸⁷ or in other words, a violent resistance in a separatism discourse.

As to the second factor, the occupation or capacity of the author or applicant, the Court takes it into account to measure the impact of the impugned statement. In both the Gürbüz and Bayar case and the Karatepe case, the Court looked into the personality of the authors and speakers. One was the leader of a terrorist organization,⁸⁸ and the other was a politician – a mayor of a big city.⁸⁹ According to the Court, their roles would make that their remarks have a profound influence on people. In other cases including Halis Doğan, Hocaoğulları, and Sürek (No 1), the applicants were the editors or the owner of a journal or newspaper that provided a platform for the contested statements to be widely accessible to the public.⁹⁰

5.1.2 Cases involving political organizations

The freedom of association grants an individual the right to found or join a political party, the aim of which includes competing with the current ruling party and putting forward political agenda different or even opposite to that of the ruling party. On the one hand, such a political party enjoys some benefit of the doubt from a broad categorical perspective: calling for a radical change in its programme does not necessarily impose a real danger to national security. In light of this, the Court renders that the dissolution of such an organization before it ever engages in any activities unsuitable. On the other hand, when the Court reviews a specific case, some concrete circumstances may indicate its political programme’s danger to national security is real even though the political party has not put it into practice.

In fact, the Court adopts a proactive strategy to evaluate the danger, meaning that the government’s interference does not have to be put on hold until a political party has seized power and implemented policies that go against democracy.⁹¹ The state’s preventative intervention can be legitimate. As a general principle, the danger to national security can be established when the change that a political party seeks to bring about goes against fundamental principles of the democratic society, or its means used for that purpose are illegal or otherwise undemocratic.⁹²

A party’s purpose and the approach are at the outset reflected in its statute and political programs announced. However, it is neither unthinkable nor unknown that an ambitious party might conceal its political agenda until it comes into power.⁹³ In this regard, ‘the acts and positions of the members and leaders’ of the party in question should be taken account of,⁹⁴ in order to examine whether or not the potential danger can be identified. For instance, in the case of Refah Partisi and Others v. Turkey, the Court concluded that the party’s real intentions were contrary to democratic principles, based on the reading of relevant speeches delivered by its leaders and members.⁹⁵ Those speeches suggested that the party would introduce Sharia into the regime,⁹⁶ establish a legal system on the basis of religious discrimination,⁹⁷ and resort to force to achieve these purposes.⁹⁸ In a more recent case, Ignatencu and Parti communiste roumain v. Romania, the Court rendered a similar decision. In this case, the party’s statute and political programs stated that the party respects the constitutional order and the principle of democracy, and opposes totalitarianism.⁹⁹ Notwithstanding that, based on an assessment of publications from the party’s chairman, the Court concluded that the party’s real intentions were against democratic principles.¹⁰⁰

5.1.3 Cases involving imparting state secrets concerning public interests

When the press publishes ‘a matter of public interests’ containing classified information, the Court normally holds that the freedom of the press outweighs preventing the leak of state secrets. Nevertheless, in terms of specific cases, it is a question about whether the classified information in question was published for the sake of public interests.¹⁰¹ In the Court’s case law, this means that the Court questions whether the published information contributes to an ongoing public debate,¹⁰² or whether it reveals misconduct of the government or the abuse of power.¹⁰³ In the absence of ‘public interests’, the case falls under the national security priority model, meaning that the Court resorts to the less intense scrutiny.¹⁰⁴

Another consideration is whether the contested disclosure would cause a ‘considerable damage’ to national security.¹⁰⁵ The Court takes into account the age of impugned information,¹⁰⁶ as well as its nature and content.¹⁰⁷ The assessment of these elements has so far led to the Court’s conclusion that the disclosure would not cause a considerable damage instead of the other way around.¹⁰⁸ Nevertheless, this consideration serves to prevent national security from suffering severe damage.

5.1.4 Cases involving the discharge of military personnel on the ground of homosexuality

In the opinion of a state, a person’s sexual orientation may collide with the security of the country. Several cases in the UK have been reviewed by the Court.¹⁰⁹ The Court believes that neither the presence of homosexual personnel in the army, nor the negative attitudes of heterosexual personnel towards them, would necessarily bring about serious damage to the army’s fighting power and operational effectiveness,¹¹⁰ although the Court did admit that it could lead to some relational difficulties amongst personnel.¹¹¹ With regard to the suitability test, the Court made only little effort to redress the model’s bias in favour of the right concerned. After all, such a scenario is quite exceptional in the sense that it involves an individual’s right to choose their way of life on the one hand, and national security on the other.

As a summary, the model’s bias towards protecting human rights due to the Court’s intense scrutiny on the suitability does not necessarily determine that the human rights prevail over national security in every single case. On the case-specific level, by taking account of the concrete circumstances, the Court redresses the bias in two ways: one is to adopt a proactive strategy to evaluate whether the alleged danger is real; and this is usually applied in the cases involving political speeches and parties. The other way is to limit the application scope of the model by specifying the circumstances appropriate for it. This method is commonly used in the cases concerning imparting information related to public interests, and government’s policy on homosexuality in army forces.

5.2 Redressing the bias towards national security

Under the national security priority model, the Court’s decision-making pattern favours protecting national security. On the case-specific level, in my observation, the Court often turns to another test under the principle of proportionality – the necessity test, to redress the bias towards national security.¹¹² Nevertheless, some case law did find the interference violating the Convention by examining its suitability. Therefore, in this part of the study I will investigate some situations in which the state’s interference measures failed the suitability test.

5.2.1 The decreasing danger

While the Court normally tends to respect the state’s judgment on the situations such as terrorist attacks or classified information leaks, the danger to national security that was established as real may decrease as time passes. Circumstantial changes can lead to a new estimate of the severity of the danger. For instance, the severity can simply decrease due to the later changes in situation, or evidence revealed at a later stage can prove that the state overestimated the danger. Under the national security priority model, the Court would, however, not take into account the gradually shifting severity of the danger until the shift becomes substantial enough. This is mainly because the Court accepts the danger to be potential and cumulative, and the government often describes the danger in a general and abstract way. Nevertheless, I find that the change of situation could reach a point where the alleged danger becomes evidently too slight to satisfy the presumptive suitability.

In the case law, such a point is sometimes reached due to decisive changes of conditions. Those changes occur either in the political context or through social transformation. In the case of Segerstedt-Wiberg and Others v. Sweden, two applicants had their information collected and stored in the Swedish Security Police register, in connection with their involvement in left-wing activities 30 years ago.¹¹³ While such information could be useful for the Security Police to fulfil its duty during the Cold War, being a left-wing sympathizer is no longer regarded as a potential danger to national security in the present day.¹¹⁴ Based on this consideration, the Court concluded that the continued storage of the applicants’ information was unsuitable. Similarly, the regime change along with a different political ideology can be seen as a decisive factor by the Court. In the Turek v. Slovakia case, the Court ruled that disclosing the documents that used to be classified as secrets by the security services under former regimes, which were communist regimes, did not entail an actual danger of divulging the functions or operation methods of the security agencies under current regimes, which are liberal democratic regimes.¹¹⁵

5.2.2 Lack of efficacy

If the government’s impugned measure is deemed ineffective to fulfil its alleged ends, the Court will conclude that the measure is not suitable. By dint of the experience, common sense, and precedent, the Court has a priori recognized the relationship between certain means and ends of the government’s interference with human rights in protecting national security, without an explicit explanation about how the relationship had been established. However, serious defects may sometimes be found in the specific case, that result in a failure of achieving the desired effects. In these circumstances, the Court does not consider it its task to figure out how such defects might have occurred or whether they can be justified. Instead, it would simply conclude that the contested interference is not able to realize its aim.

For example, in the case of Bartik v. Russia concerning Article 2 of Protocol No. 4, a major defect in the government measure was pointed out by the Court, as it ruled that there was a lack of efficacy of an international travel ban that intended to prevent the divulgence of state secrets.¹¹⁶ In the Court’s view, the retired personnel who had access to classified information and were subject to the travel ban could still reveal the classified information via other means of communication which were not monitored by the government.¹¹⁷ That is to say, the classified information could still be disclosed to foreigners at home or abroad if the applicant wished to do so. Due to a serious lack of efficacy, the Court came to the conclusion that the link was missing between the means and ends.¹¹⁸

5.2.3. Claims lacking in substance about the danger

How specific should the government be to substantiate its claims about the danger to national security? In the case law of the Court, the answer seems to be contingent on which type of rights the government interferes with, that is, qualified rights or limited rights. The Court tends to accept the state’s general argument in relation to its interference with qualified rights, whereas declining explicitly similar state submissions concerning an interference with some limited rights. The latter occurs in scenarios such as the prolonged pre-trial detention of terrorist suspects (as to Article 5(3)),¹¹⁹ a trial held in camera (as to Article 6(1)),¹²⁰ and the delayed access to lawyer in terrorist cases (as to Articles 6(1) and 6(3)).¹²¹

In a specific case, the state is expected to present grounds with a higher degree of specificity that support the government’s claim about an alleged danger to national security. The Court expects that the state explains the alleged danger in more details, through an individual and case-specific assessment. For instance, in the case of Belashev v. Russia, in defending its action of excluding the public from the contested trial, the state simply claimed that the classified information was contained in the case file,¹²² but it did not further illuminate what materials were confidential, and how they were related to the applicant’s offences.¹²³ By contrast, in the Welke and Białek v. Poland case, the state further indicated that the classified materials included secret recordings, and certain details about how the police intercepted the parcel containing cocaine and replaced it with a similar substance.¹²⁴ An open trial could expose the police’s operational methods to the general public, which would negatively affect future police operations.¹²⁵

Under the national security priority model, the Court does not proceed with a rigorous scrutiny on governing authorities’ assessment of the danger to national security. Nevertheless, by declining those generally formulated claims, the Court tries to ensure that the interference is based on the government’s elaborate contemplation of the likelihood of the danger to national security, instead of any one-size-fits-all account.

As a conclusion, a relatively higher degree of discretion granted to the state under the national security priority model is counterbalanced by an assessment of specific merits and particular circumstances in each given case. A low intensity of scrutiny does not necessarily mean that the government’s interference for protecting national security always survives the suitability test. It may fail the test in the following circumstances: the alleged danger has been considerably mitigated, the impugned measure lacks efficacy, or in connection with certain limited rights, the danger has been illustrated only in a general way. In short, under the national security priority model, the bias in favour of protecting national security is not necessarily absolute when the Court takes concrete circumstances into considerations.

6. Conclusion

In this article, I answer how the suitability test, part and parcel of the application of the principle of proportionality, is conducted by the European Court of Human Rights in cases concerning national security. I argue that the Court reviews the suitability of the government’s interference with human rights in a consistent and predictable way, when we read its decision-making progress from two distinct but closely related perspectives: the broad categorical perspective and the case-specific perspective.

From the broad categorical perspective, I conclude that there are two models in accordance with the nature of the rights under question, which are corresponding to the margin of appreciation enjoyed by the state. Between the two models, the Court shows different extent of intensity of its scrutiny and applies different approaches to redressing the model’s established bias towards human rights or national security. The first model, the human right priority model is visible in cases concerning rights underlying the democratic values of ‘pluralism, tolerance, and broadmindedness’. The cases examined in this article frequently raise concerns over the freedom of speech and freedom of assembly and association. The persuasive reasoning is that the values of human rights ‘outweigh’ those of national security. It thereby requires the Court to adopt an intense scrutiny of the contested government’s interference. The other model is the national security priority model, which is visible in the cases concerning other rights that are not closely linked with democratic values. The importance of protecting national security has been highlighted, and the scrutiny on suitability of the contested government action is therefore less intense than that in the first model.

To assess whether the government’s interference is suitable, the Court looks into the three main elements: the danger to national security, the effectiveness of the impugned measure, and the evidence or argument to substantiate the former two. That is where the two models of scrutiny come in: the human rights priority model requires the state to prove the existence of a real danger, and the measure to be highly capable of achieving its purpose. The government’s argument and evidence will be reviewed closely by the Court. By proceeding with such strict review standards, at the outset, it will be arguably easier for the Court to find a violation on human rights, rather than for it to find that the interference is suitable. In this regard, the approach is named the ‘human rights priority model’, as it offers a quite strong protection of rights and freedoms. By contrast, under the the ‘national security priority model’, as the Court assesses the suitability of the impugned measure of the state, the alleged danger can be merely potential, cumulative, or remote, and the measure is deemed ineffective only if there is a serious defect.¹²⁶. In addition, the Court is often prepared to accept the state’s argument.

Then from the case-specific perspective, the Court tries to redress the model’s embedded bias towards human rights or national security. In this regard, concrete facts and specific circumstances of each case matter. For the human rights priority model, apart from confining its application to several given types of cases, the Court adopts a proactive strategy to evaluate whether the alleged danger is real, by taking account of the context and specific circumstances of a case. After all, the Court does not expect or require the state to wait until it is too late to take actions against national security threats. When it comes to the national security priority model, the government’s interference is held unsuitable if some major defects are identified by the Court, such as: ex post disappearance of danger, a lack of efficacy of the impugned measure, or in cases related to certain limited rights, a lack of specific description of the danger.

Findings presented above contribute to an interpretive framework in which the Court’s decisions can be understood more consistently and concretely in regard to its suitability test in cases concerning national security vis-à-vis protection of rights. How should human rights reconcile with national security? Each party has a good reason to be prioritized: while the security of the state constitutes the foundation for safeguarding human rights, the protection of human rights constitutes the legitimate basis of the state. That long-existing dilemma may explain why the Court’s reasoning and decisions seem to be inconsistent from one case to another. In this light, the two models which I conclude help to explain how the Court applies the doctrine of margin of appreciation and conducts the suitability test, which has not been sufficiently addressed in the current literature.

At last, although it is not within the research scope of this article, I need to point out that the Court’s review on the necessity test (less-intrusive-means test) can also be made sense of under these two models. Under the two models, the Court, in different manners and with different major concerns, makes the decision on whether the government’s measure is specifically and narrowly framed to accomplish its alleged purpose.¹²⁷ A more systematic account certainly requires another research effort, as my point here is to indicate that the models proffered in this article are not confined to assessing merely the suitability test, but also other core elements under the principle of proportionality.