comment 0

Conceptualising Confrontation | Part 6 | Anonymous Witness Evidence in English Law

In English law, the reception of anonymous witness evidence is governed by the statutory regime contained in Part III of the Coroners and Justice Act 2009 (CJA 2009) which placed the temporary powers contained in the Criminal Evidence (Witness Anonymity) Act 2008 on a permanent legislative footing. Arguably, however, a power enabling courts to receive anonymous evidence can be traced back to the 1989 decision in R v Murphy, where it was held that criminal courts had a common law power to withhold the identity of witnesses from the accused, the press and the public. Even as early as 1992, Beldam LJ felt able to say that ‘it is now well established that there may be occasions upon which the interests of justice require that the identity of witnesses should be withheld.’ (R v Watford Magistrates’ Court, ex parte Lenman (unreported) 10 June 1992 (CA). Quoted in R v Liverpool Magistrates’ Court, ex parte DPP (1996) 161 JP 43, 47)

As David Ormerod recently observed in his commentary on the provisions of the 2009 Act, ‘the evidence of an anonymous witness engages issues which are numerous, complex and controversial.’ (D. Ormerod, A. Choo and R. Easter, ‘Coroners and Justice Act 2009: the “witness anonymity” and “investigation anonymity” provisions’ [2010] Crim LR 368)

In English law, matters are further complicated by the incorporation of the European Convention on Human Rights (ECHR) into domestic law by s 1 of the Human Rights Act 1998 and all of the resultant obligations this ushers in. At the heart of the matter however, is a tension between the interests of the accused on one hand, and the interest of witnesses and of the State on the other.

From the accused’s point of view, anonymous evidence presents a number of difficulties. Firstly, anonymous evidence directly impinges on the disclosure of identity component of the confrontational model, thereby encroaching upon the remaining three components of the right. It will not be possible for the accused to fully investigate the credibility of adverse witnesses out-of-court, nor will it be possible to investigate credibility in cross-examination. Where the witness is screened, face-to-face confrontation vapourises. It will not be possible for the accused to visually identity the witness, nor will it be possible to observe her demeanour. It is also highly likely that the witness will be shielded from the gaze of the public gallery, making the proceedings visually opaque. In short, the circumstances in which the confrontational model will not go into meltdown are likely to be rare. Secondly, the accused’s ability to advance her case and undermine that of the prosecution is ‘unequal to the Crown’s’, which potentially engages her rights under article 6(3)(d) ECHR. It becomes ever more crucial that prosecution faithfully honour their disclosure obligations under ss 3-8 of the Criminal Procedure and Investigations Act 1996. Finally, there is the distinct possibility that a jury, upon learning that a witness is to give evidence anonymously, will assume that ‘the defendant is a dangerous criminal capable of serious violence.’ (C. Coleman, ‘Can justice afford witness anonymity?’ The Times June 23, 2008)

Anonymity under the 2009 Act

The genesis of the 2009 Act is of itself of interest. In the landmark House of Lords decision in R v Davis [2008] UKHL 36; [2008] AC 1128 it was held the use of anonymous evidnce was irreconcilable with the common law right to confrontation and the European Court of Human Rights’ article 6(3)(d) jurisprudence. The effect of this decision was seismic because it effectively abolished a practice, quietly carried on by the courts, which permitted the Crown to lead material evidence from anonymous witnesses. Unsurprisingly, a frenzy followed soon afterwards in the British gutter-press. ‘Anarchy’ it was said, would be unleashed ‘by barring anonymous witnesses in court trials.’ (M. Sullivan, ‘Anarchy Unleashed’ The Sun June 25, 2008) ‘Chaos in court as loony Lords spike £6m trial’, (R. Kaniuk, ‘Chaos in court as loony Lords spike £6m trial’ Star June 25, 2008) boomed the Star newspaper. Within four weeks of the Davis-decision, following only one day of parliamentary debate, the Criminal Evidence (Witness Anonymity) Act 2008 entered into force, overruling Davis and establishing the first statutory framework for the reception of anonymous evidence since the days of the Star Chamber. Following the 2008 Act’s expiry on 31 December, 2009, its provisions were adopted wholesale by the CJA 2009.

Under s 86(1) of the 2009 Act, a court is empowered to order ‘such specified measures to be taken in relation to a witness…to ensure that the identity of the witness is not disclosed in or in connection with the proceedings.’ Section 86(2) goes on to set out a non-exhaustive list of measures for preventing disclosure of the witness’s identity. Such measures include withholding the witnesses name or other identifying details; permitting the witness to give evidence under a pseudonym; prohibiting questions that may lead to identification; screening the witness from the public and the accused; and subjecting the witness’s voice to modulation. The court’s discretion here is incredibly wide – there is no limit on the number of measures that can be applied nor are there any restrictions as to their combination. Section 86(2) equips the court with enough ammunition to neutralise each component of the confrontational model.

The granting of a witness anonymity order is dependent upon the satisfaction of three ‘conditions’ set out in s 88. Condition A (s 88(3)) stipulates that the order be necessary to (a) protect the safety of the witness or (b) in order to prevent harm to the public interest. Arguably, the former of these contingencies is directed at intimidation cases, whereas the latter will cater for situations where state agents are called to give evidence. Condition B (s 88(4)) requires that the effect of the proposed order would be consistent with the defendant receiving a fair trial. As Ormerod notes, ‘this is a provision designed to ensure ECHR compatibility.’ Finally, Condition C (s 88(5)) provides that,

… the importance of the witness’s testimony is such that in the interests of justice the witness ought to testify and –
(a) the witness would not testify if the proposed order were not made, or
(b) there would be real harm to the public interest if the witness were to testify with the proposed order being made.

Two observations can be made here. Firstly, from a drafting point of view, it would have been more logical to have made Condition C the first consideration and Crown applications for anonymity orders are governed by s 87(2). On such an application, the identity of the proposed witness can, and regularly will, be withheld from the defence, but must generally be disclosed to the court. This means that the accused will be reliant on the Crown thoroughly investigating the credibility of the witness and subsequent disclosure of anything untoward in a manner consistent with the order. The inequality between the Crown and the accused kicks-in when one looks to s 87(3) which governs anonymity applications made by the accused. Under this regime, the identity of the proposed witness must be revealed both to the court and to the Crown. As Ormerod observes, ‘the principle reason for compulsory defence disclosure…is…to enable the Crown to make inquiries regarding the witness’s credibility.’ The rub is that this ‘is precisely what the defendant is not entitled to do in the converse situation.’

Can anonymous evidence be the sole or decisive basis of a conviction?

The regime in the CJA 2009 presents a significant threat to the integrity of the confrontational model and places the accused at a distinct tactical disadvantage in relation to the Crown. Nevertheless, the 2009 Act is probably here to stay. However, the dust refuses settle over one crucial issue: can anonymous evidence be the sole or decisive basis of a conviction without infringing article 6(3)(d)?

Immediately after the decision in Davis, one would be safe in answering this question in the negative. Lord Bingham clearly stated in that case that ‘no conviction should be based solely or to a decisive extent upon the…testimony of anonymous witnesses.’ Here, Lord Bingham was merely applying the test that had been developed in the jurisprudence of the Strasbourg court since Doorson v Netherlands (1996) 22 EHRR 330 in which the court held that ‘even when [procedures are adopted to counterbalance the effects of anonymous witness evidence]…a conviction should not be based either solely or to a decisive extent on anonymous statements.’ The force behind the ‘sole and decisive’ test is obvious: it is unfair to adduce evidence that is solely or decisively supportive of the accused’s guilt in a manner that makes it impossible for the accused to challenge it.

Since the recent Supreme Court decision in R v Horncastle [2009] UKSC 14; [2010] 2 AC 373, however, the future of the ‘sole and decisive’ test is in doubt. Horncastle concerned the admission of out-of- court statements made by the complainants in each appeal. In one case, the complainant had died prior to trial: his statement was admitted via s 116(1) and 2(a) of the CJA 2003. In the other case, the complainant ran away before the trial through fear of giving evidence: her statement was admitted via s 116(1) and 2(e). The principal issue in that case was whether a conviction based solely or decisively on the statement of an absent witness infringed the accused’s rights under articles 6(1) and 6(3)(d) of the Convention.

Giving the judgment of the court, Lord Phillips held that in the circumstances ‘it would not be right for this court to hold that the sole or decisive test should have been applied rather than the provisions of the 2003 Act’ (see para 117 of the judgment). This decision is in direct conflict with the ECtHR’s earlier decision in Al-Khawaja v United Kingdom (2009) 49 EHRR 1 in which the ‘sole and decisive’ test was applied to very similar facts. In so holding, Lord Phillips appears to have based the court’s rejection of the test on two grounds. Firstly, Lord Phillips stated that ‘the sole or decisive test produces a paradox. It permits the court to have regard to evidence if [it is] peripheral, but not where it is decisive. The more cogent the evidence the less it can be relied upon.’ But is it right to assume that merely because an item of evidence is decisive of guilt it is also cogent? Even if the paradox exists, Lord Phillips ignores its flipside: the more decisive the evidence, the more necessary it is for the accused to be able to test it.

The second ground is one of practicality. In Lord Phillips’ view,

…a direction to the jury that they can have regard to a witness statement as supporting evidence but not as decisive evidence would involve them in mental gymnastics that few would be equipped to perform. [see para 90 of the judgment]

In the case of hearsay evidence, this ground of objection is more persuasive since traditionally questions of evidential weight have been left to the jury. However, it would be quite possible, in the case of anonymous witnesses, for the court to consider how a reasonable jury would treat the testimony of a proposed witness on the basis of the contents of their witness statement. If the court concluded that a reasonable jury could convict solely or decisively on the basis of anonymous testimony, the application for anonymity should be rejected.

Whilst primarily concerned with the admission of statements made by absent witnesses under the CJA 2003, the decision in Horncastle may well be taken to apply to anonymous witness evidence. In an important dictum not forming part of the ratio in Horncastle, Lord Phillips stated:

So far as a sole or decisive rule is concerned, I am not persuaded that there is a difference in principle between its existence in relation to absent witness and its existence in relation to anonymous witnesses. Each situation results in a potential disadvantage to the defendant…I cannot see why a sole or decisive test should apply in the case of anonymous evidence but not in the case of a witness statement. [see para 50 of the judgment]

At the time of writing [this piece was originally written in the summer of 2010 – see update below] the Grand Chamber of the ECtHR are considering an appeal by the UK government against the decision in Al-Khawaja. The outcome of the appeal is difficult to predict. Push come to shove, the Grand Chamber has three options. Firstly, they could agree with the Supreme Court’s analysis in Horncastle and simply hold that even where the evidence in question is sole and decisive a breach of Article 6 may not automatically follow.. This is unlikely, since the court will want to save face. Secondly, the court may seek to justify the test on the basis of an extrinsic, rather than instrumental, theory of the right to confrontation, rooting the rule within a societal or dignity-based rationale. Or thirdly, they may simply affirm the existing rule more forcefully.


The text in this blog post was originally drafted in the summer of 2010. Since that time,  Al-Khawaja v United Kingdom went from the Fourth Section Chamber decision cited in the post above and was heard by the Grand Chamber. The Grand Chamber’s decision can be found here.

In the original draft I set out three possible ways in which the Grand Chamber would deal with the issue (see the paragraph immediately above this update). In the event, the GC went for option one. The GC held that a conviction based solely or decisively on the statement of an absent witness would not automatically result in a breach of Article 6. However, the court went on to say that counterbalancing factors had to be in place, including strong procedural safeguards, to compensate for the difficulties caused to the defence. 

In other words, the GC swallowed the Supreme Court’s decision in Horncastle wholesale.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s