DISSENTING OPINION OF JUDGES MATSCHER AND VALTICOS

(Translation)

This is a borderline case. On the one hand, the conditions in which the trial took place and the witnesses were examined were certainly not entirely satisfactory and no doubt they could have been improved, although it has to be acknowledged that efforts had been made in Netherlands law to adapt the procedure for hearing anonymous witnesses to the requirements of Article 6 of the Convention (art. 6) as set out in the judgment of Kostovski v. the Netherlands (20 November 1989, Series A no. 166). On the other hand, this was a case of armed robbery and it is understandable that the witnesses - even though they were police officers - should be in fear of reprisals from trigger-happy criminals. Were a similar situation to arise in the future, it would certainly be desirable for even more attention to be paid to the requirements of Article 6 of the Convention (art. 6) when measures were taken.

In the instant case, however, having regard to all the circumstances, we are unable to find a violation of Article 6 of the Convention (art. 6) and concur on the whole with the opinion of Judge van Dijk .

DISSENTING OPINION OF JUDGE VAN DIJK

1.   To my regret I am unable to agree with the conclusion of the majority that there has been a violation of Article 6 para . 1 in conjunction with Article 6 para . 3 (d) of the Convention (art. 6-1+6-3-d). Nor can I follow the majority as regards the essence of the reasoning supporting that conclusion.

2.   Although the Court ’ s case-law has not yet been fully developed as to the conditions subject to which a conviction in criminal proceedings may be based partly on statements of anonymous witnesses, the Court has drawn certain lines. I am of the opinion that the present judgment is neither within these lines nor a logical continuation thereof, while on the other hand the facts of the case are not specific to such a degree as to justify distinguishing the present case from, especially, the Doorson case ( Doorson v. the Netherlands judgment of 26 March 1996, Reports of Judgments and Decisions 1996-II, in which the Netherlands were found not to have acted in violation of Article 6 para . 1 taken together with Article 6 para . 3 (d) of the Convention (art. 6-1+6-3-d)). Even though the Court is not bound by precedent, legal certainty and legal equality require that the Court ’ s case-law be both consistent and transparent as well as reasonably predictable in so far as the facts of the case are comparable to those of earlier cases.

3.   Since "the admissibility of evidence is primarily a matter for regulation by national law and as a general rule it is for the national courts to assess the evidence before them" (see the previously cited Doorson judgment, p. 470, para . 67), the applicable national legislation and case-law and the practice followed by the domestic courts are to some extent relevant also for the Court.

The Netherlands Supreme Court revised its case-law concerning the conditions under which a conviction may be based on statements of anonymous witnesses in view of the judgment of the Court in the Kostovski case (see paragraph 40 of the present judgment). Moreover, taking that revised case-law as a starting-point and basing itself, inter alia, on an analysis of the Court ’ s case-law on the admissibility of statements of anonymous witnesses in criminal proceedings and the implications of the Court ’ s case-law for the relevant domestic law and legal practice in the Netherlands, the Netherlands Government proposed, and the legislature adopted, several amendments to the Code of Criminal Procedure (see paragraph 42 of the judgment, and the explanatory memorandum to the Act of 1993 cited there).

The Act of 1993 was not yet in force when the judgments of the domestic courts in the present case were given. However, had it been in force, the procedure followed by the Court of Appeal would have been in conformity with the rules thereby introduced. It is true that the reasons listed in the Act for keeping the identity of a witness secret do not include the desirability, for tactical reasons, of not disclosing the identity of a police officer in order not to impair his future effectiveness. The explanatory memorandum of the Act states in so many words that in the opinion of the Government the public interest in investigating serious crimes cannot alone justify guaranteeing complete anonymity (see paragraphs 41 and 42 of the present judgment). However, the Court of Appeal did not base its acceptance of the wish of the witnesses to remain anonymous on this reason, but on the fear of the witnesses for their lives and safety and those of their families (see paragraph 26 of the present judgment), which is a ground provided for in Article 226a of the Code of Criminal Procedure.

The foregoing does not, of course, guarantee per se that the revised case-law of the Supreme Court and/or the relevant provisions of the amended Code of Criminal Procedure will in all circumstances be found to be in conformity with the Convention. However, as noted above, in this matter the domestic case-law and legislation have a relevance of their own. Moreover, in the present case, in view of the legal background of the relevant Netherlands case-law and the drafting history of the new legislation, in both of which Strasbourg case-law was expressly taken into account, there would seem to be good cause for a presumption of conformity, at least in so far as the issues dealt with have also been considered in the Strasbourg case-law.

4.   Consideration of the various issues involved ultimately led me to the conclusion that Article 6 paras . 1 and 3 (d) (art. 6-1, art. 6-3-d) have not been violated in the present case. It might have been preferable for the Court of Appeal or the investigating judge to have interrogated the witnesses in the presence of counsel and the Procurator General, the accused being able to follow the proceedings in a separate room. I cannot discover from the case file that was before the Court whether this possibility was considered at all. Be that as it may, taking all the facts and circumstances into account, I am of the opinion that the right of the defence to examine the witnesses was not limited to such an extent that the defence was not given an adequate and proper opportunity to challenge and question the witnesses, as required by Article 6 paras . 1 and 3 (d) (art. 6-1, art. 6-3-d) (see the Lüdi v. Switzerland judgment of 15 June 1992, Series A no. 238, p. 21, para . 47). I therefore consider the trials to have been fair, taking into account also the compensating elements of the procedure decided on by the Court of Appeal and follow ed by the investigating judge. In reaching this conclusion I would stress the following aspects:

(a) The anonymous witnesses were not interrogated only by a prosecuting authority, but also by an independent and impartial judge who, judging from his official report of his findings to the Court of Appeal, took great care to compensate the defence for the handicap resulting from the lack of a face-to-face confrontation. The applicants and their counsel were able to hear the interrogation by the investigating judge and to ask questions of their own. In these respects, as was also observed by the Commission, the present case differs from the cases of Kostovski ( Kostovski v. the Netherlands judgment of 20 November 1989, Series A no. 166, p. 20, para . 42), Windisch ( Windisch v. Austria judgment of 27 September 1990, Series A no. 186, p. 10, para . 27), Lüdi (loc. cit., p. 21, para . 49), and Saïdi ( Saïdi v. France judgment of 20 September 1993, Series A no . 261-C, pp. 56-57, para . 44). The practice in cases such as the present of having witnesses heard by an investigating judge instead of by the trial court itself was accepted by the Court in its Doorson judgment as being in conformity with the Convention (loc. cit., p. 471, para . 73).

(b) The Court of Appeal has given reasons for delegating the hearing of the witnesses to the investigating judge. The weight which the minority of the Commission laid on the fact that the Court of Appeal, being the trial court, did not avail itself of the possibility to assess for itself the reliability of the witnesses, is in my opinion not conclusive; there is no good reason why the court could not rely for this on the assessment of the equally independent and impartial investigating judge. In that respect, it is also of relevance that the interrogations before the investigating judge did not take place in a pre-trial phase but during a suspension of the trial before the Court of Appeal and pursuant to an order of the Court of Appeal; they formed part of the trial. In the Kostovski case, where the Court emphasised the importance of the possibility for the trial judge to observe the witness, only one of the witnesses was heard by a judge, who was, however, unaware of the identity of the person concerned (loc. cit., p. 21, para . 43).

(c) The statements made before the investigating judge were statements by witnesses who had been identified by the investigating judge as having been, at the relevant moment, sworn police officers who were authorised to perform prosecuting duties invested with investigative competence and were under oath in relation to any statement made in that context (see the Lüdi judgment previously cited p. 21, para . 49).

(d) The investigating judge, who had observed the witnesses during the interrogations, gave a reasoned opinion as to their reliability; this was also intended to compensate the defence for being deprived of the visual information which would have permitted them to test the witnesses ’ reliability (see the Windisch judgment previously cited, pp. 10-11, paras . 28-29).

(e) The investigating judge gave his reasoned opinion as to whether the wish of the police officers to remain anonymous was justified; these reasons were also found to be justified by the Court of Appeal on the grounds set out in its judgment. Their shared opinion that the accusations and events were such that fear of violent repercussions was not unsubstantiated cannot be considered unreasonable. Article 6 (art. 6) does not guarantee an unlimited right to question witnesses. It is necessary not only to recognise the discretion of the competent domestic court in maintaining conformity with the exigencies of the proper administration of justice, but also to balance the interests of the defence under Article 6 (art. 6) against the interests of witnesses protected by other substantive provisions of the Convention (see the Doorson judgment previo usly cited, p. 470, para . 70). Although in the Lüdi judgment (loc. cit., p. 21, para . 49) the Court considered the interest of the police authorities in preserving the anonymity of their agents "legitimate", greater weight should be given in this case, as was done by the Court of Appeal, to the agents ’ interest in the protection of their lives and safety and those of their families (Articles 2, 3, 5 and 8 of the Convention) (art. 2, art. 3, art. 5, art. 8).

(f) The defence were given ample opportunity to hear and question the witnesses, and to comment on the recording of their answers, and in fact made extensive use of that opportunity; the technical deficiencies complained of were inconvenient and might perhaps have been avoided, but, given especially the extensive time reserved for the hearings and the detailed way in which the statements were recorded, these deficiencies were not such as to hamper the defence to a significant extent.

(g) The Court of Appeal did not exclude beforehand the possibility that additional questions might be put to witnesses at the trial, but was of the opinion that the defence had insufficiently substantiated their wish to do so. Moreover, the defence were offered the possibility to challenge the statements and their use as evidence in open court before the Court of Appeal.

(h) The convictions were not based solely upon the statements of anonymous witnesses. Although these were undoubtedly the core of the evidence, there were also statements by identified witnesses, there was some technical evidence and there was the recording of the telephone conversation. In this respect, too, the Court should recognise that "as a general rule it is for the national courts to assess the evidence before them".

In the Doorson judgment, in which the criterion of "decisive extent" was developed and applied (loc. cit., p. 472, para . 76), the Court found that this criterion had been met in a situation where the conviction was based, in addition to statements of anonymous witnesses, on a statement made by an identified witness to the police but retracted during the trial and a statement by an identified witness who disappeared before the defence had had the opportunity to question him (loc. cit., p. 472, para . 76, in conjunction with pp. 458-59, para . 34).

In view of all these aspects of the case I come to the conclusion that the trial which led to the applicants ’ conviction was "fair" in the sense of Article 6 of the Convention (art. 6) as construed in the Court ’ s previous case-law.

Having stated my conclusion I wish, with due respect, to make the following observations with regard to the reasoning on which the majority bases its conclusion.

5.   Like the majority, and in accordance with the case-law of the Court, I take as my starting-point that evidence must normally be produced at a public hearing. Therefore, I find in the abstract that interrogation of the police officers at the trial before the Court of Appeal, in disguise if necessary to protect their anonymity, would have been preferable. However, I also take note of the opinion of the Court of Appeal that this would have been too risky because disclosure of the identity of the witnesses could not have been excluded. I would have preferred it if the Court of Appeal had given concrete arguments for that opinion. On the other hand, I lack the expertise - as, I assume, do my colleagues in the Court - to judge whether its fear was justified or not. Then again, one may well wonder whether the defence would have been in a better position to observe the witnesses ’ demeanour and test their reliability had they appeared in disguise, given the fact that an effective disguise may also substantially disguise the sound and intonation of the voice and the body language of the person concerned. A national court may in general be considered to be in a better position to judge such a complex and factual issue than is our Court, which should substitute its judgment for that of the national court only if the latter ’ s judgment is unreasonable. From the wording of the Doorson judgment it is clear that in that case the Court recognised this primary responsibility of the national court.

6.   The fact that police officers were fired at in pursuit does not necessarily mean that at a later stage, during or after the trial, their lives and safety and/or those of their families were in danger. However, on both points the national authorities - in this case the national courts - should be left some latitude to balance the interests of the defence against those of the witnesses (see the Doorson judgment previously cited, p. 470, para . 71, where a reasonableness test was applied). In my opinion the Court of Appeal has not overstepped the mark in finding that the risk of disclosure of the identity of the witnesses was present and that their fear for their lives and safety or those of their families was justified in view of the seriousness of the crimes committed and the violence used.

I disagree with the majority that the Court of Appeal failed to make any real effort to assess the threat of reprisals. The Court of Appeal had at its disposal the report of findings of the investigating judge, in which the latter referred not only to the statements made on this subject by the police officers concerned - which were in some cases supported by previous experience (paragraphs 17 and 21 of the judgment) - but also gave his own evaluation, based upon the seriousness of the crimes and the violence used by the perpetrators. Given these reasoned opinions of both the investigating judge and the Court of Appeal, the latter of which was reviewed by the Supreme Court, the Government were not called upon to give an explanation of their own, as held by the majority (see paragraph 60 of the judgment); nor in my opinion would the Government have been in a position to do so.

In the Doorson judgment (loc. cit., pp. 470-71, para . 71) the Court found that an actual threat against the witnesses was not required for the decision to maintain their anonymity to be reasonable, and that previous experience might be relevant. In the present case it should also be taken into consideration that some of the witnesses had been wounded while pursuing the robbers. Even if one were to take the position that a certain risk is implicit in the profession of police officer, that should not mean that the latter has to take unnecessary risks and it certainly does not mean that the lives and safety of his family deserve any less protection than that of other persons. The "general duty of obedience to the State ’ s executive authorities" (see paragraph 56 of the judgment) owed by policemen cannot imply that their lives and safety and those of their families are any less worthy of protection. Therefore I cannot agree with the majority that, for the sole reason that in the present case members of the police force were involved, the case has to be distinguished from that of Doorson as far as the balancing of the interests of the defence against those of witnesses is concerned.

The fact that Mr Engelen , who had originally made a highly incriminating statement, was not granted anonymity by the police and nevertheless did not suffer any harm at the hands of the applicants cannot be decisive in this context. The police may have made a mistake in his case by revealing his identity, but in any case later events do not automatically and retrospectively invalidate a reasonable assessment of a risk. One can hardly blame a witness who expresses serious fears for not waiting until something serious happens to him or any other witness; the only criterion for the national court to consider is the reasonableness of the fear.

7.   I fail to see why policemen should be under a special duty to give evidence in open court (see paragraph 56 of the judgment) since this is a general civic duty prescribed by law. And even if one may agree that the use of policemen as anonymous witnesses "should be resorted to only in exceptional circumstances" (ibid.), one may argue on the other hand that their anonymity should meet with fewer objections from the point of view of the defence, because their statements are statements by sworn professionals, whose identity and investigative competence can easily be checked by the investigating judge.

8.   The majority also deal with the question whether "operational needs" provided sufficient justification for preserving the anonymity of the police officers. Reference is made in that context to the explanatory memorandum of the Act of 11 November 1993 in which the Netherlands Government indicate that in their opinion that interest can be sufficiently protected by less far-reaching restrictions on the rights of the defence. However, in my opinion that issue is of only minor importance in the present case, since the Court of Appeal stated in its judgment that, of the arguments advanced by the witnesses for claiming anonymity, it considered the argument concerning the personal safety of these witnesses and their families to be "decisive" (see paragraph 26 of the judgment).

9.   It cannot be denied - and indeed it has not been denied - that the possibilities for the defence to interrogate the witnesses were not without limitations. Nor are they always without limitations in normal situations, where the identity of the witnesses is known to the defence. The fact that certain questions were not answered by witnesses and that this was accepted by the investigating judge may be open to criticism but this would in all probability also have occurred had the witnesses been interrogated in open court in some form of disguise. It may therefore be questioned whether it was necessary and proportionate to refuse to answer certain questions put by the defence in order to protect the anonymity of the witnesses and the secrecy of the police tactics. However, this question has to be answered primarily by the competent domestic court, and indeed it was answered in the affirmative by the investigating judge, and after his decision was challenged by the defence before the Court of Appeal, by the latter as well. On the whole, in my opinion, neither the limitations necessitated by the situation nor those accepted by the investigating judge were such as to warrant the finding that there was no "adequate and proper opportunity to question the witnesses" as required by the Court in its Kostovski judgment (loc. cit., p. 20, para . 41). The handicaps under which the defence laboured were, therefore, sufficiently counterbalanced by the procedures followed by the judicial authorities (see the Kostovski judgment previously cited, p. 21, para . 43, and the Doorson judgment previo usly cited, p. 472, para . 76).

10.   Finally, although the statements of the anonymous witnesses were an essential part of the evidence, it cannot be said that the convictions were based solely on these statements. Whether they were based on these statements "to a decisive extent", as the majority holds (paragraph 63 of the judgment), is hard to say. I share the opinion expressed by the Delegate of the Commission at the Court ’ s hearing that this criterion, laid down in the Doorson judgment (loc. cit., p. 472, para . 76), is difficult to apply, because if the testimony of anonymous witnesses is used by the court as part of the evidence, that will always be because the court considers it a "decisive" part of that evidence, making the proof complete or at least sufficient. Here again, the Court ’ s finding that "as a general rule it is for the national courts to assess the evidence before them" (paragraph 50 of the present judgment) should prevail. Moreover, as already mentioned (see paragraph 3 (h) above), in this respect the facts of the case would not seem to differ substantially from those of the Doorson case.

[1] The case is numbered 55/1996/674/861-864.  The first number is the case's position on the list of cases referred to the Court in the relevant year (second number). The last two numbers indicate the case's position on the list of cases referred to the Court since its creation and on the list of the corresponding originating applications to the Commission.

[2] Rules of Court B, which came into force on 2 October 1994, apply to all cases concerning the States bound by Protocol No. 9 (P9).

[3] Note by the Registrar: f or practical reasons this annex will appear only with the printed version of the judgment (in Reports of Judgments and Decisions 1997-III), but a copy of the Commission's report is obtainable from the registry.