Exclusion of evidence in times of mass surveillance. In search of a principled approach to exclusion of illegally obtained evidence in criminal cases in the European Union

Starting point: Issues of admissibility and exclusion are matters for the member states

The ECJ is rather equivocal about the use in criminal proceedings of traffic and location data obtained—and accessed—on consequence of their retention. The Court's starting point is that issues of admissibility and exclusion of evidence are distinctly a matter of national law:

[A]s EU law currently stands, it is, in principle, for national law alone to determine the rules relating to the admissibility and assessment, in criminal proceedings against persons suspected of having committed [serious] ¹⁵ criminal offences, of information and evidence obtained by such retention of data contrary to EU law ¹⁶ [or by access of the national authorities thereto contrary to EU law]. ¹⁷

In La Quadrature du Net, the ECJ refers in the above statement to ‘persons suspected of having committed serious criminal offences’. It is striking that the Court initially emphasised the member states’ procedural autonomy in respect of admissibility issues only in relation to serious criminal offences. This could have created the impression that issues of admissibility and exclusion may not be a matter for the member states when relatively minor offences are concerned. There is no satisfactory principled explanation for this. The gravity of the offence is an arbitrary criterion on the basis of which to differentiate between the EU and the member states in terms of their competence to decide on admissibility issues. This is even more so in light of the large discrepancy in sentencing regimes between member states and the vast difference in the sensitivity around what offences can be considered serious. Such criterion raises a host of questions in terms of where the dividing line would lie between offences that are sufficiently serious as to fall under the competence of the member states on the one hand, and offences of lesser gravity that would be a matter of national law.

However, the term ‘serious’ was later omitted in Prokuratuur. This suggests that presumably the CJEU did not intend to establish a European admissibility rule based on the criterion of the gravity of the offence. It is more likely that the reference to ‘serious criminal offences’ in La Quadrature du Net was a remnant of references to serious criminal offences throughout other parts of the judgment—where the ECJ indicates that retention of traffic and location data may be justified under very specific circumstances, and in any event only in order to combat serious crime or to prevent a serious risk to public security or a risk to national security. ¹⁸

Qualification: National criminal courts must exclude illegally obtained evidence if two cumulative conditions are fulfilled

While the ECJ starts by saying that admissibility and exclusion are a matter for the member states, it qualifies this statement later in its judgments. National evidence rules are not entirely exempted from EU law, as member states are obliged to respect the EU principles of equivalence and effectiveness. It is this second principle which compels member states to ‘prevent information and evidence obtained unlawfully from unduly prejudicing a person who is suspected of having committed criminal offences.’ ¹⁹ Such effective judicial protection could be achieved in different ways: for instance, the unlawful information could be given less probative weight (for instance, by requiring that it be corroborated by other evidence, or by excluding that it plays a main or predominant role in the decision). The Court also hints to the possibility that the sentence be reduced when the information has been collected unduly. There are, however, circumstances where exclusion of the evidence is the requisite option. On this, the Court first notes that, in their decision whether to exclude evidence obtained in breach of EU law, national courts must have regard to ‘the risk of breach of the adversarial principle and, therefore, the right to a fair trial entailed by the admissibility of such information and evidence.’ ²⁰ The ECJ imposes an obligation on national criminal courts to exclude evidence obtained through the general and indiscriminate retention of data and access thereto in violation of EU law if the following conditions are fulfilled:

[T]he principle of effectiveness requires national criminal courts to disregard information and evidence obtained by means of the general and indiscriminate retention of traffic and location data in breach of EU law [or by means of access of the competent authority thereto in breach of EU law], ²¹ in the context of criminal proceedings against persons suspected of having committed criminal offences, where those persons are not in a position to comment effectively on that information and that evidence and they pertain to a field of which the judges have no knowledge and are likely to have a preponderant influence on the findings of fact. ²²

The assertiveness of the ECJ on the exclusion of evidence is quite surprising when compared to the caution of the European Court of Human Rights (ECtHR) to take a sharp position on this matter. Save for cases of evidence obtained through inhumane and degrading treatments and a few other instances, the ECtHR does not endorse a sharp obligation to exclude evidence (Pinar Ölçer, 2013: 397). This is particularly the case with regard to evidence obtained by means of breaches of privacy. The ECtHR has emphasized more the procedural fairness of the procedure, requiring that parties be allowed to challenge the evidence and to obtain an appropriate response by the court. ²³

Another striking aspect of the ECJ decision are the two conditions that trigger the obligation to exclude evidence: (i) that the suspect has the possibility to comment effectively on that information and that evidence; and (ii) that evidence/information pertain to a field of which the judges have no knowledge and are likely to have a preponderant influence on the findings of fact. In neither of the judgments does the ECJ clarify what these conditions for mandatory exclusion actually entail, nor does it offer any arguments for why these particular circumstances justify mandatory exclusion. In order to make sense of these conditions, it is necessary to revisit the case in which the ECJ initially introduced them. Steffensen was a case stemming from an administrative litigation. It concerned analyses conducted by national authorities which indicated that certain foodstuffs failed to comply with domestic rules on labelling. ²⁴ The EU directive on quality control of foodstuffs, which was in force at the time, obliged member states to ensure that manufacturers who are subject to inspection of foodstuffs for quality control ‘may apply for a second opinion’. ²⁵ In the underlying dispute, Steffensen had allegedly been deprived of the possibility to obtain such a second opinion in the form of an independent scientific evaluation of the quality of a product. One of the questions before the ECJ asked essentially whether the results of analyses of samples of a manufacturer's product could be used in domestic administrative proceedings against the person who allegedly violated the food quality standards, if the manufacturer's right to a second opinion as guaranteed by EU law was breached. In response, the ECJ held that issues pertaining to admissibility of evidence are not governed by EU law, and are in principle matters for the Member States, who must respect the principles of equivalence and effectiveness. But since respect for the right to a second opinion is guaranteed by EU law and the breach of that right may have consequences for the admissibility of evidence, the national rules on the taking of evidence fall within the scope of EU law. Accordingly, those rules must comply with the requirements arising from fundamental rights, in this case particularly the right to fair trial as enshrined in Article 6 ECHR. ²⁶ An essential aspect of the right to a fair trial is the right to adversarial proceedings, which requires that the parties must be afforded a real opportunity to have knowledge of and comment on all evidence adduced or observations filed with a view to influencing the Court's decision. ²⁷ The ECJ held that, in its determination of whether evidence obtained in breach of the right to a second opinion may be admitted at trial, the national court must assess whether the adversarial principle and thus the right to a fair hearing was respected. In the context of that assessment, the national court will have to examine specifically ‘whether the evidence at issue in the main proceedings pertains to a technical field of which the judges have no knowledge and is likely to have a preponderant influence on its assessment of the facts’. ²⁸ If admission of the evidence obtained in breach of EU law is likely to violate the adversarial principle and thus the right to a fair trial, the national court must exclude the evidence. ²⁹

In specifying the condition of respect for the adversarial principle and emphasising in particular that it should be examined whether the evidence pertains to a technical field of which the judges have no knowledge and is likely to have a preponderant influence on its assessment of the facts, the ECJ drew inspiration from the ECtHR case Mantovanelli v. France. ³⁰ This case concerned the death of the applicants’ daughter by alleged medical negligence through the excessive administration of an anaesthetic. The applicants applied to an administrative court for the appointment of an expert to investigate the circumstances surrounding their daughter's death. The expert's report concluded that there was no evidence linking the death to the use of the anaesthetic. The applicants claimed that they had not been informed of the dates of the steps taken by the expert, nor did they have access to the documents referred to in his report. They alleged that they did not receive a fair hearing as required by Article 6 ECHR since the expert report had not been prepared in accordance with the principle of adversarial proceedings. The ECtHR held that, while the applicants could have made submissions to the administrative court on the content and findings of the expert report after receiving it, this did not afford them a real opportunity to comment effectively on it. The question the expert was instructed to answer was identical to the one the administrative court had to determine, namely whether the circumstances in which the anaesthetic had been administered demonstrated the hospital's negligence. This issue pertained ‘to a technical field that was not within the judges’ knowledge’. Although the administrative court was not in law bound by the expert's findings, ‘his report was likely to have a preponderant influence on the assessment of the facts by that court’. The ECtHR concluded that the applicants could only have expressed their views effectively before the expert report was lodged. Since they had not been afforded that opportunity, their right to a fair trial as required by Article 6 ECHR had been breached. ³¹

This clarifies where the ECJ got its inspiration for the conditions for mandatory exclusion of evidence obtained in violation of EU law. The case Mantovanelli was about evidence that was not only obtained in breach of the adversarial principle, but that concerned a technical field of which a judge would have no knowledge, namely highly specialised medical questions. In that context, it would make sense for a judge to rely heavily on expert evidence to reach a conclusion on the facts. It is logical that the more technical, and often consequently the more decisive the evidence may be, the more important it is that parties are given a fair opportunity to express their views during the preparation of the expert report. The ECJ then adopted these conditions in Steffensen, a case concerning the use of evidence obtained in violation of the right to a second opinion in the form of an independent scientific evaluation of the quality of a product, which is equally a highly technical matter.

It appears that the ECJ is following—almost pasting—the above precedents when it re-affirms those conditions for mandatory exclusion in La Quadrature du Net and Prokuratuur. It is questionable whether these conditions are appropriate in the context of the new case. First, both Mantovanelli and Steffensen are cases of administrative proceedings, which rules might differ from criminal proceedings. Secondly, the use of metadata retained by electronic communications providers in criminal proceedings can raise technical issues that may prompt the court to rely on expert evidence to interpret the data correctly and verify their reliability. It is nonetheless highly dubious that they concern a technical field of which a judge has no knowledge.

It is somewhat odd that the ECJ ties the procedural condition that the suspect did not get the opportunity to challenge the evidence (‘comment effectively on the information’) to the mandatory exclusion of illegally obtained evidence. It is difficult to see how this procedural condition could apply to traffic and location data obtained by means of undue retention. The principle that parties should be afforded a real opportunity to comment effectively on the evidence is an essential aspect of the right to fair trial and should be guaranteed in all cases, regardless of whether or not the evidence was illegally obtained. ³² Moreover, the application of this condition risks defeating the purpose of exclusion. In a criminal trial it is normally the case that parties are given an opportunity to comment on the evidence. Hence, every time the metadata has been disclosed to the parties and the parties are given the possibility to express their views upon it and challenge its admissibility, the condition would be fulfilled. In other words, since the protection of the adversarial principle is a mainstay of criminal proceedings, the procedural condition would normally lead to the conclusion that the evidence should not be excluded. The ECJ fails to explain how illegalities in the gathering of evidence could compromise the adversarial principle at a subsequent trial, or be remedied by it. If the ECJ will go as far as to specify the circumstances in which exclusion is mandatory, it could adopt a more coherent approach by clarifying the relationship between evidence so obtained, the prohibition on its use in domestic proceedings and the impact on the rights of the suspect to challenge the evidence against him.

In sum, the interpretation of what these conditions for mandatory exclusion actually entail is difficult without the background of cases from which the conditions originated, namely the ECJ and ECtHR case law which narrowly concerned evidence obtained in breach of the adversarial principle and pertaining to highly technical issues. The conditions are perhaps best viewed as general conditions to ensure the right to a fair trial, independent of whether or not the evidence was obtained in breach of EU law.

The endorsement of a particular rationale for exclusion of illegally obtained evidence

What is perhaps most striking about the sections of the judgment in La Quadrature du Net pertaining to exclusion of illegally obtained evidence is that for the first time the ECJ includes a statement on the normative principle that should underpin exclusionary rules, which it then confirms in Prokuratuur:

As for the principle of effectiveness, it should be noted that the objective of national rules on the admissibility and use of information and evidence is, in accordance with the choices made by national law, to prevent information and evidence obtained unlawfully from unduly prejudicing a person who is suspected of having committed criminal offences. ³³

As mentioned, the ECJ further specifies that the objective of preventing illegally obtained evidence from unduly prejudicing the suspect may be achieved under national law not only by excluding such evidence, but also by means of national rules and practices governing the assessment and weighting of such material, or by factoring in whether that material is unlawful when determining the sentence. ³⁴ The Netherlands provides an illustration of the latter option. Article 359a of the Dutch Code of Criminal Procedure lists the three possible sanctions which criminal courts may impose at their discretion when faced with illegally obtained evidence, namely excluding the evidence, declaring the criminal proceedings as a whole inadmissible or admitting the evidence and offering the defendant a sentence discount in proportion to the gravity of the illegality. ³⁵

While the ECJ leaves open the possibility for member states to achieve the objective of not unduly prejudicing the suspect in ways other than through exclusion of evidence, it certainly also ties this objective to exclusionary rules. By emphasising that rules on admissibility and exclusion of illegally obtained evidence aim to avoid ‘unduly prejudicing’ the suspect, the ECJ seems to endorse a particular rationale for exclusion of illegally obtained evidence, namely the protective or remedial principle. In short, the protective principle posits that exclusionary rules are essentially concerned with protecting the suspects’ rights and that prohibiting the use of illegally obtained evidence is necessary to vindicate these rights (see ‘Rationales for exclusion of illegally obtained evidence’ for more detail). ³⁶ Next to the many objections that this position raises, it is particularly problematic that the Court's endorsement of this rationale is under-theorised. An analysis of what the rationale entails, why it is most appropriate for underpinning an exclusionary rule pertaining to the collection of evidence by general and indiscriminate data retention and access thereto in breach of EU law, and a discussion of other relevant principles for exclusion is notoriously absent in the judgment. The ECJ also demonstrates no awareness of how the choice for this particular principle would influence how national judicial discretion is exercised and which factors are important in the exercise of discretion.

La Quadrature du Net and Prokuratuur are not the first cases in which the ECJ ruled on admissibility of evidence obtained in violation of EU law in domestic proceedings. As mentioned, Steffensen concerned evidence obtained improperly in an administrative context and its use in subsequent appeal proceedings. In WebMindLicenses, the ECJ was asked to give a preliminary ruling regarding the interpretation of the VAT Directive and whether licensing agreements were abusive for VAT purposes. ³⁷ In the underlying dispute, the tax authorities, in order to establish the existence of an abusive practice concerning VAT, had used telecommunications that were intercepted and emails that were seized without prior judicial authorisation in the context of a parallel criminal procedure which had not yet been concluded. This raised questions as to the compatibility of the collection and use of such evidence with inter alia Article 8 ECHR and Article 7 CFEU, which both protect the right to respect for private life. The ECJ ruled that EU law does not preclude tax authorities from using evidence obtained without the taxable person's knowledge in the context of a parallel criminal procedure that had not yet been concluded, provided that the rights guaranteed by EU law, especially the CFEU, are respected. ³⁸ It held that, before admitting such evidence (in this case the intercepted communication and seized emails), the national court must verify that the manner in which the evidence was gathered and whether the use of the evidence by the tax authorities was authorised by law and was necessary. The national court also needed to establish that the taxable person had the opportunity, in the context of the administrative procedure, of gaining access to the evidence and being heard concerning it. If these requirements were not met or if the court was not empowered to carry out such a judicial review, the evidence obtained in the criminal procedure should be disregarded in the administrative procedure. ³⁹ The ECJ position in WebMindLicenses seem to be less prescriptive with regard to exclusion of unlawful evidence, leaving greater discretion to national courts.

The ECJ went a step further in La Quadrature du Net and Prokuratuur compared to its earlier case law on illegally obtained evidence by explicitly endorsing a particular rationale for exclusion, which it made no reference to in Steffensen or WebMindLicenses. The new endorsement of a rationale for exclusion raises the question of whether the ECJ is developing a more elaborate doctrine on exclusion of evidence obtained in violation of EU law. If this would be the case, it would be all the more important that the Court's choice for a particular rationale for exclusion of illegally obtained evidence be supported by (adequate) arguments. As will be explained in the following sections, the choice for a particular principle is not without consequences.

Reliability principle

According to the reliability principle determining the truth of the criminal charges is the sole purpose of the criminal trial, and evidence should be admitted or excluded only based on its reliability (Ashworth, 1977: 723; Duff, 2004: 152, 154–155). Bentham (1825: 2–3) is often cited as arguing that the rectitude or accuracy of the decision is paramount in the law of evidence. Consequently, all evidence that is probative of facts at issue should be admissible, subject to only a few exceptions in which exclusion is warranted such as delay, vexation or expense (Mirfield, 1997: 9). This approach reached in early applications the extremes of considering irrelevant whether the evidence had been stolen, or unlawfully procured, as long as this did not affect its reliability. ⁴² Adoption of the reliability principle does not entail that the way in which evidence was obtained is entirely irrelevant; indeed, the manner in which the evidence was gathered may affect the degree of its reliability, particularly where confessional evidence as opposed to real evidence is concerned (Ashworth, 2003: 115). It is important to bear in mind that the reliability principle is also an unreliability principle (Mirfield, 1997: 31): while courts may have some discretion in determining whether the evidence is in fact reliable, once they conclude that the evidence is unreliable they have no other choice but to exclude it. Reliability is directly connected with the accuracy of fact finding. Excluding unreliable evidence prevents miscarriages in the finding of past events, it ensures the discovery of the truth (or the best truth) about the alleged/charged facts. As such it seems difficult to reject this rationale, without doing away with the very basic essence of a fair trial.

A problem of the reliability principle may be that it fails to explain the current law in many jurisdictions, where illegally obtained evidence is often excluded despite it having a high degree of reliability (Jackson and Summers, 2012: 154). However, this is not a weakness of the reliability principle as a rationale underpinning exclusion of illegally obtained evidence as such; rather, it illustrates a problem with adopting reliability as the sole or exclusive concern in the law of evidence. Indeed, relying solely on accuracy of the verdict in terms of the substantive truth to justify exclusion of evidence is to ignore competing criminal justice values which may also be at stake, notably respect for notions of rights, such as trial fairness, human dignity or privacy.

Should the ECJ have invoked the reliability principle to underpin exclusion of evidence obtained by general and indiscriminate data retention and access in breach of EU law? It is undoubtedly possible that the reliability of telecommunication data could be tainted. Software is required to convert raw data from phone masts into usable evidence. Problems with this software may affect the reliability of the data obtained. In 2019, courts in Denmark imposed a two-month moratorium on the use in criminal trials of geolocation data obtained from telecommunication operators. It was found that, due to a malfunctioning of the software, the system tended to omit certain data during the conversion process and linked phones to the wrong masts, connected them to several masts simultaneously and erred in identifying the location of specific masts. The errors in the software meant that innocent people may have been located at crime scenes and that factually guilty people could have been wrongly excluded from the criminal investigation (Henley, 2019). Disregarding such evidence in a criminal trial would be the only correct approach, as relying on potentially unreliable evidence would entail the risk of wrongful convictions. Yet the assessment of the reliability of this evidence would be part of the trial judge's general assessment of the probative value of the evidence, rather than it being part of the assessment on the admissibility of illegally obtained evidence. In the Danish case the unreliability of the evidence was not brought about by any illegal gathering of the data. By contrast, in La Quadrature du Net the ECJ explicitly said that as a rule general and indiscriminate retention of traffic and location data as a preventive measure is illegal under EU law, and it added in Prokuratuur that the same is true for access to those data by public authorities for the purpose of prevention, investigation, detection and prosecution of criminal offences (when such access does not remain confined to proceedings to combat serious crime or prevent serious threats to public security). However, it is not the general and indiscriminate retention of data or access thereto in breach of EU law as such that will taint the data's reliability. Consequently, according to the reliability principle, any evidence so obtained would be admissible. Hence, it comes as no surprise that the ECJ did not consider or adopt this principle. Presumably the Court did not wish to establish first that general and indiscriminate data retention and access can be in breach of EU law, to then subsequently endorse a rationale for exclusion that would in effect still allow those data to be used in criminal proceedings. This begs the question: to what extent can reliable information be excluded from criminal adjudication?

Deterrence or disciplinary principle

The deterrence or disciplinary principle ⁴³ posits that courts should exclude illegally obtained evidence in order to discourage law enforcement officers (police, prosecutors, investigating judges) from committing improprieties and illicit acts in the investigation of crime and thus protect citizens’ civil liberties (Duff, 2004: 160–161; Zuckerman, 1989: 349). The idea is that, if judges routinely exclude improperly obtained evidence, this sends a message to state officials that there is nothing to be gained from breaking the law and consequently they will refrain from doing so (Roberts and Zuckerman, 2010: 151). This is the dominant theoretical justification for exclusion of illegally obtained evidence in the United States. ⁴⁴

A common practical objection to this principle is that it is highly questionable whether the exclusionary rule does in fact deter law enforcement officers (Alschuler, 2008: 1365; Duff, 2004: 163–164; Roberts and Zuckerman, 2010: 156–157; Sklanski, 2008: 567, 582; Slobogin, 1999: 363, 368; Zuckerman, 1989: 349–350). A range of considerations, other than the prospect of exclusion of evidence, may be on law enforcement officers' minds as they are gathering evidence in an improper way. Ashworth notes that ‘the police may not be deterred if they are unaware of the relevant rules, if they think that the misconduct will not come to light, if they think that the suspect will plead guilty, or if for any other reason gathering evidence for use at trial is not what is motivating their conduct…’ (Ashworth and Redmayne, 2010; 344; see also Ashworth, 2003: 112). It is unsurprising that in jurisdictions where this principle is endorsed, it has its limitations, notably in the United States (Cammack, 2013: 15–20; Thaman, 2010: 335). For instance, the principle is applicable only for intentional (mala fide) violations. Additionally, the principle ceases to apply in the context of the ‘inevitable discovery’ doctrine, which permits admission of illegally obtained evidence if it would have been gathered in any event (i.e. ‘inevitably’) in a legal manner. ⁴⁵

There are also issues of principle with the disciplinary approach. Some authors question whether it is justified to acquit a factually guilty person in order to ‘protect’ the public against police misconduct (Zuckerman, 1989: 349). This would effectively entail punishment of the public, whereas the responsible police officer may not suffer any direct sanction (Calabresi, 2003: 112; Duff, 2004: 161). Moreover, endorsing the principle requires approving two of its logical corollaries. First, if the disciplinary principle is adopted, the motives of the police become relevant (Ashworth, 1977: 731, 2003: 119). Precisely in cases in which law enforcement officers have deliberately violated procedural rules, there is more reason to exclude the resulting evidence in an attempt to deter them from flouting the law in the future, compared to where the breach is a result of mere ignorance or a mistake (Mirfield, 1997: 31). While emphasis on the law enforcement officer's motives is consistent with only endorsing the disciplinary principle, it becomes less defensible if also the protection of the suspect is considered a relevant factor in deciding on admission of illegally obtained evidence, as will be argued below (Duff, 2004: 161). Secondly, according to the disciplinary principle, evidence would need to be excluded only if it was gathered illegally by a state official. If the argument underpinning the principle is that exclusion is required to deter and punish law enforcement officers, exclusion would be pointless if the illegality had been committed by a private citizen (Duff, 2004: 162).

Difficulties arise when such exclusionary principle is applied to the case decided by the ECJ (concerning the indiscriminate retention by providers of electronic communication services of traffic data and location data ⁴⁶ and the subsequent access thereto by national public authorities ⁴⁷ ). First, from the perspective of the disciplinary principle, the focus would be on the actions of the law enforcement authorities as opposed to the actions of telecom providers, who are private companies. Secondly, it is more difficult to speak of deterrence when retention of data—and access—was made possible by a national law, although in breach of EU law. The disciplinary principle could at most justify exclusion, as a means to dissuade law enforcement officers if they intentionally accessed metadata despite being aware that the national rules were in breach of EU law—and, to make things even more complicated, it could also be argued that before the two judgements were delivered, the exact scope of the EU rule prohibiting retention and access was not (yet) too clear. But even if mass retention was illegal under national law, one could wonder if excluding the evidence from trial would have a deterrent effect, given that there may be many other advantages obtained from that knowledge—for instance, state authorities could discover crimes and then search for lawful evidence to prosecute and punish them. It should not be forgotten that excluding reliable evidence for disciplinary purposes entails losing valuable information: it makes sense to do so only if the deterrence effect is readily visible.

Protective, remedial or vindicatory principle

According to the protective, remedial or vindicatory rationale the holder of a right should be protected against the consequences of a state official's breach of that right. ⁴⁸ If a legal system declares certain standards for the conduct of criminal investigations, then citizens have corresponding rights. If the legal system is to respect those rights, a suspect whose rights have been violated should not thereby be placed at a disadvantage; the evidence obtained as a result of the rights violation should not be used against the suspect (Ashworth, 1973: 725). Exclusion is seen as the most appropriate remedy for violation of defendants’ rights committed by the investigating authorities while collecting evidence. Criminal courts should use their authority and rule in such a way that the ‘state and the citizen are placed in the positions they would have been in’ had their rights not been violated (Ashworth, 2003: 111–112). In contrast to the disciplinary principle, the protective principle's distinct focus on the rights of the suspect explains why the intentional or inadvertent nature of the police's conduct is irrelevant: what matters is whether there was an actual violation of the suspect's rights, not the intentions of the police in committing the violation (Ashworth, 1977: 731, 2003: 119; Mirfield, 1997: 30).

A range of objections have been formulated against the protective principle. According to a ‘pure’ protective principle, exclusion would be the appropriate remedy in any and every case where the rights of the accused have been violated. The problem is that a ‘pure’ protective rationale corresponds with an exclusionary rule that is a very blunt instrument to vindicate the rights of the suspect, as it gives courts no discretion to vary the remedy depending on the harm done to the suspect (Duff, 2004: 152, 166). Critics argue that the protective principle is blind to ‘comparative reprehensibility’: exclusion as a remedy would be disproportionate where there is a breach of a minor right and there is strongly probative evidence of a serious offence (Duff, 2004: 169; Kamisar, 1987: 1; Zuckerman, 1989: 349). Some proponents of this rationale acknowledge this weakness of the protective principle and adopt a qualified version thereof. Ashworth (1977: 725–726, 729) states that exclusionary discretion remains the most appropriate means of protection of the suspect's rights, albeit with clear principles to guide or structure this discretion, accepting that ‘the prima facie justification for exclusion supplied by the protective principle may occasionally be outweighed by other considerations’. In his later work, he also argues that the case for applying this principle is strongest where fundamental or human rights, as enshrined in the ECHR or a constitution, were violated, but that it should probably not be applied in cases where the breach of a right is merely small or ‘technical’ (Ashworth, 2003: 112).

The issues presented by adoption of the ‘pure’ protective principle raise questions about the ECJ's endorsement of this principle. Recall that, while the ECJ stated that the aim of exclusionary rules is to prevent undue prejudice to the suspect, it added that this objective may also be achieved in other ways, namely by adopting rules regarding the assessment and weighting of evidence or by mitigating the sentence. ⁴⁹ The latter two approaches may enable courts to vary the remedy in proportion to the seriousness of the breach of procedural rules and the disadvantage caused to the victim of the rights violation. For instance, if the illegality is relatively minor, member states could still allow the illegally obtained evidence to be used but require corroborative evidence to support the illegally obtained evidence before convicting the suspect, or they could choose to apply a modest sentence discount. Yet alternative remedies like these are inconsistent with the adoption of a pure protective principle, according to which exclusion of evidence is the sole appropriate remedy to vindicate the suspect's rights breached during the investigation. The fact that the ECJ left the door open for member states to adopt alternative remedies that allow national courts to differentiate depending on the degree of disadvantage to the suspect may imply that the ECJ wished to endorse a qualified version of the protective principle. However, rather than just listing the possible remedies, the ECJ could have adopted a more principled approach by considering how each remedy might correspond to the degree of disadvantage to the suspect. Where the degree of disadvantage to the suspect is particularly high, exclusion of evidence or even staying the proceedings may be appropriate. Conversely, if the disadvantage is minor, mitigation of sentence may suffice (Mirfield, 1997: 18).

A further issue with the protective principle is that, while it takes the rights of the accused seriously, it does not provide a basis for excluding evidence obtained by violation of the rights of a third party (Duff et al., 2007: 232). Mass collection of traffic data breaches much more than just the suspect's rights. Hence, it is surprising that the ECJ endorsed the protective principle specifically in cases concerning the general and indiscriminate retention of data and access thereto. The most straightforward application of the protective principle would be in a case where only the suspect's rights were violated in the course of the investigation. If an illegal home search is carried out in the private premises of a person suspected of murder and the dead body is found, the pure protective principle would require that any evidence relating to the dead body is excluded. The scenario is somewhat different if, in the context of a criminal investigation against a particular suspect, telecommunication data are accessed that were gathered and retained in a general and indiscriminate manner in breach of EU law. The general and indiscriminate retention of telecommunication data violates the rights of a large group of citizens, even though those persons are neither directly nor indirectly involved with the crime that is being investigated. The suspect is but one individual member of this group (provided that their traffic data were also retained). The application of the protective principle would still require that the data be excluded since the rights of the suspect were violated, regardless of whether the rights of third parties were also infringed. Yet the illegality would not exist but for the violation of third-party rights; if there had been targeted retention of and access to data of a specific suspect, there would be no violation of EU law. EU law does not preclude member states from adopting legislation that permits telecommunications providers to retain the data of a specific suspect or even categories of persons, for instance, and allows public authorities to access this. In that sense, the suspect would have an advantage by virtue of the violation of the rights of third parties, since it is precisely the violation of rights on a larger scale that would lead to exclusion in the criminal case against the suspect. Moreover, it could also be argued that once the privacy of people has (unduly) been pierced, the exclusion of evidence does in no way redress or alleviate such rupture, as the private information gathered remain disclosed (at least to those who can lawfully access the materials of the case, and unless exclusion is intended as destruction).

Relatedly, the pure protective principle risks ignoring the wider public interest since no elements other than the violation of the suspect's rights are considered relevant. The public interest could be simply to convict the guilty and ensure public safety (Duff, 2004: 166). But the public interest may equally be to avoid large-scale violation of citizens’ rights, which is relevant in a case concerning mass data gathering. The protective principle, with its distinct focus on the suspect, is not well-suited to take account of the wider citizens’ interests at stake. In La Quadrature du Net and Prokuratuur there may be a tension between what makes certain investigative action illegal under EU law on the one hand, which requires the large-scale violation of rights of a group of citizens, and the requirements for exclusion of the illegally obtained evidence on the other, which according to the ECJ is based on the protective principle and thus strictly only requires violation of the rights of the accused. In fact, exclusion of evidence does not protect third parties, unless it is intended in a deterrent rationale (see ‘Deterrence or disciplinary principle’ above). The protective principle is not entirely flawed or obsolete in an era of mass data gathering, but its function remains confined within narrow boundaries. A pure application of the principle to the case of evidence by means of general and indiscriminate data gathering would still lead to exclusion of evidence, but limited to the sole information obtained by breaching the suspect's right to privacy. In any case, the fact that the ECJ did not acknowledge or explore this aforementioned tension suggests that the Court's endorsement of the principle may be insufficiently considered.

Integrity principle

A final justificatory theory for the exclusion of illegally obtained evidence centres on integrity or legitimacy. ⁵⁰ Its strongest proponents are theorists who take issue with the three traditional rationales discussed in the previous sections. Zuckerman argues that the criminal trial has ‘a special moral dimension’ and that the willingness of the public to accept the authority of the criminal court depends on the extent to which the public believes in the moral legitimacy of the system. He suggests that a judicial community that is perceived as condoning or even encouraging violations of the law will undermine the morality of the adjudication system (Zuckerman, 1989: 344). Dennis equally acknowledges that the criminal trial has ‘moral authority and expressive value’—in the sense that it is a public process which serves as a medium of communication with the public at large, it may carry a message of moral condemnation of the defendant, and it is an expression of the norms of the criminal law and the consequences of breach of those norms (Dennis, 2020: 52–53)—which may be jeopardised by the admission of illegally obtained evidence (Dennis, 2020: 109).

Different ways of theorising the integrity principle have been put forward (Ashworth, 2003: 108; Ho, 2019: 829–833; Mirfield, 1997: 23–25). Two prominent variations of the principle are distinguished here. ⁵¹ According to the ‘public attitude integrity’ variation, judgments need to appear legitimate in the eyes of the public. The idea is that citizens will lose their faith in the administration of justice if judges too easily condone wrongdoing by state authorities. Equally, the legitimacy of the administration of justice will suffer if citizens see that people who have committed serious offences go free because of relatively minor wrongdoing (Ashworth and Redmayne, 2010: 346). This variation is adopted in s. 24(2) of the Canadian Charter of Rights and Freedoms, which essentially centres on whether the admission of the evidence would bring the administration of justice into disrepute. ⁵² Another variation is ‘court-centred integrity’, according to which the use of illegally obtained evidence jeopardises the moral and expressive authority of the verdict and of the court delivering it. In deciding whether exclusion of evidence is required, the court must apply its own standards of decency and propriety (Mirfield, 1997: 24). ⁵³ Exclusion of the tainted evidence is considered a way of renouncing the impropriety and preserving the integrity of the court specifically and of the criminal justice system generally.

Adopting the integrity principle—in one of the above variants—may provide courts with a theoretical framework within which they can navigate the fundamental tension that is at the heart of the decision whether to admit illegally obtained evidence. On one end of the spectrum, states would emphasise truth finding as the main aim of the criminal process but routinely condone malpractice by law enforcement agencies. On the other end of the spectrum, states would uphold due process values and demonstrate respect for suspects’ procedural rights, whilst possibly letting criminals go free on the basis of a mere technicality and thus failing to protect citizens from (future) criminality. In both extremes, the moral integrity and procedural legitimacy of the criminal justice system may be compromised (Roberts and Zuckerman, 2010: 158–159). A merit of the integrity principle is that it enables courts to perform a balancing exercise, weighing competing public interests, in which a variety of factors can be considered in determining whether to admit illegally obtained evidence. These factors may include whether the investigating officers’ misconduct was intentional or a mere mistake (which is also a relevant consideration under the disciplinary rationale), the need to protect citizens’ civil liberties (which is central to the protective principle), the seriousness of the offence, the seriousness of the impropriety, the fact that the illegality was committed by a private citizen as opposed to a state official, the probative value of the evidence, and the importance of the evidence (Dennis, 2020: 108; Duff, 2004: 172; Zuckerman, 1989: 350).

However, the integrity principle also has its downsides. Among the various objections that can be made against the integrity principle, ⁵⁴ the strongest is that the principle is rather vague (Duff, 2004: 172). If courts would be allowed to pick and choose which factors they take into account and what weight they attach to them in the exercise of their discretion, this may lead to arbitrary and inconsistent decision making (Re, 2014: 1904). Duff argues in response that the guiding criterion for courts should be that of the moral legitimacy of the admission of evidence (Duff, 2004: 173), but it is questionable whether this criterion is sufficiently clear and precise to ensure consistent case law. It seems proponents of the principle would counter this criticism by re-emphasising the strengths of the principle. While the principle may lack clarity, it at least allows courts to resolve the tension between competing imperatives in determining whether to admit illegally obtained evidence in a way that none of the traditional principles do (Dennis, 2020: 50–51, 109; Duff, 2004: 172), and requires courts to make their reasoning in this regard transparent for the public (Zuckerman, 1989: 350).

Although there is merit to the objection that the integrity principle opens the door to inconsistent and arbitrary decision-making, the fact that it enables courts to take into account the public interest and weigh up various considerations may make it a suitable principle for exclusion of evidence in an era of mass data gathering. The fact that evidence was obtained in breach of not only the individual suspect's rights but also the rights of an entire population using electronic communications services could be a relevant consideration in determining whether to exclude the evidence. Consideration of this factor would not entail that any evidence so obtained would immediately lead to exclusion. Whether or not exclusion is warranted is dependent on which variation of the integrity principle is adopted and on the other factors taken into consideration in the balancing exercise. For instance, one factor is whether the breach is minor or ‘technical’. Judicial integrity would be less compromised if evidence is admitted that was obtained by a minor or technical breach as opposed to a substantive breach of rights (Ashworth, 2003: 119). ⁵⁵

From the perspective of the integrity of the justice system there is a clear difference between a police officer who is mistaken or ignorant and one who deliberately flouts the rules (Ho, 2019: 838). This distinction is clear regardless of whether one adopts the public attitude or the court-centred variety of the integrity principle (Ashworth, 2003: 119). If the public attitude variant is adopted, it could be argued that the public is willing to draw the aforementioned moral distinction between different behaviours and intention of the officers and might find it relevant to what it thinks of the action taken by the court. Under the court-centred variety, the more serious the breach, the more likely the court will exclude the evidence, as it is more damaging to the integrity of the justice system for the court to condone more serious or significant breached than to condone lesser ones (Mirfield, 1997: 31).

It is not straightforward whether the factor of the seriousness of the offence charged is relevant under the integrity principle. Under the public attitude variation of this principle, on the one hand, the more serious the offence, the more difficult it is to justify to the public that evidence is excluded and that a factually guilty person may be acquitted. On the other hand, the more serious the offence, the more serious the consequences of conviction will be and so ‘the higher should be the moral rectitude of the means by which it is achieved’ (Zuckerman, 1987: 62–63, 1989: 356; see also Mirfield, 1997: 364). Under the court-centred variety, it may be appropriate for the court to condone the investigating officer's conduct, who committed the breach in his own knowledge or belief regarding the seriousness of the offence, but that would depend on the particular moral theory the judge would adopt (Mirfield, 1997: 32–33).

In a case where general and indiscriminate retention of data and access thereto is concerned, the breach of rights may be considered minor if the public attitude variation of the integrity principle is adopted. Many citizens may suffer no—or limited—direct harm and may not actually be disadvantaged by the breach. Indeed, many may never even be aware that their rights were breached. When the offence charged is then also a particularly serious one, the public in general may not find it objectionable to admit the evidence. However, a complicating factor here is how the public attitude should be conceptualised. As Mirfield (1997: 32–33) puts it:

Does every citizen, no matter how miserable his own morality, get a vote, so to speak? Or do we view matters from the perspective of the reasonable citizen with a responsible and thoughtful character? Or do we see things from the point of view of the judge himself, albeit a judge taking proper account of community values?

Depending on which of the conceptions of the public attitude is adopted, views may differ as to how heinous a mass violation of privacy rights is considered to be.

According to the court-centred variation of the principle, the court would need to assess on the basis of its own standards of morality what the relationship should be between the mass violation of privacy rights and the consequences of the remedy of exclusion of evidence. If a court applies its own standards of morality, it may be more inclined to focus on the fact that violation of privacy rights occurred on a large scale, and thus consider the breach of rights significant.

This section has demonstrated that all the rationales for exclusion of illegally obtained evidence have both strengths and weakness. It has also argued that the choice for a particular rationale is not without consequences; a change of perspective will affect how exclusionary rules are applied and which factors are relevant in the exercise of discretion. While the reliability principle, disciplinary principle and protective principle on their own are problematic as a ‘self-contained account’, all three of them offer relevant considerations on the issue of whether or not it is opportune to exclude illegally obtained evidence (Dennis, 2020: 104). The integrity principle allows courts to take into account factors that are relevant on all three accounts, and other factors as well. And while the latter is the principle that best captures the complexity of the reasoning around exclusion, it is also the one that offers less guidance to courts.

All principles have pros and cons. How can a choice be made among these different principles? The next section will attempt to give some indications, suggesting that a structured discretion combining the different principles might be an apt solution for legal systems grappling with admission and exclusion of illegally obtained evidence, both at national and supranational level.