Public inquiries and the limits of the Attorney General’s undertaking

Public inquiries have become an increasingly common method of investigating major events or tragedies. As their popularity has increased, so too has their overlap with criminal investigations. Sometimes the inquiry delays its work to avoid running in parallel with the criminal investigation, but more often, to avoid the vice of the inquiry being delayed, the two run alongside each other. This inevitably results in both the inquiry and the criminal investigators seeking to question the same individuals.

The Inquiries 2005 Act recognised that this situation would arise and sought to provide a solution. It prioritised the rights of the suspect over the inquiry’s mandate to discover the truth. Section 22 of the Act entitles an inquiry witness to assert what lawyers call the privilege against self-incrimination and so refuse to answer questions if in doing so the witness would admit to committing criminal offences.

But this legislative compromise did not achieve the solution it promised. The lack of candour or co-operation from individuals who played a significant role in the relevant events was said to render the inquiry ineffective. The truth was being withheld. Widespread dissatisfaction led to resort to the common law to provide a real solution – an undertaking from the Attorney General that any evidence given to the inquiry will not be used against that witness in future criminal proceedings (“the AG’s undertaking”). The expedience of a politician’s promise of non-prosecution would thus be the artifice by which witnesses could be compelled to answer an inquiry’s questions.

The scope of the AG’s undertaking has varied from inquiry to inquiry, becoming broader in recent years to prohibit not only direct use of the evidence in a criminal trial, but also, on some occasions, derivative use, including (a) deciding whether or not a prosecution should be brought, and (b) opening further lines of investigation which produce evidence relied upon in criminal proceedings. (b) was first introduced in the AG’s undertaking to the Baha Mousa inquiry in 2006 and replicated in the Al Sweady inquiry of 2010. However, the wording of the AG’s undertaking to the Azelle Rodney inquiry in 2011 added an important exclusion – that “this undertaking does not preclude the use of information and/or evidence identified independently of the evidence provided by that person to the Inquiry.”

This exclusion prompts a question for those individuals (or corporate entities) who may be compelled to provide evidence to an inquiry: to what extent does the AG’s undertaking represent a complete and effective barrier against the use of evidence? Are there circumstances in which the evidence the witness gives to the inquiry, despite the undertaking, either becomes admissible against them, or is otherwise deployed to their prejudice, in criminal proceedings?

In considering when a criminal prosecution can circumvent the apparent protection of the AG’s undertaking, we can mention but otherwise ignore the standard caveat that the AG’s undertaking does not apply to circumstances where a witness is subsequently prosecuted for providing false evidence to the inquiry. The untruthful witness faces prosecution for perjury in the same way as any other who is considered to have lied under oath.

This article considers three ways in which the AG’s undertaking given to a public inquiry might, in practical terms, be circumvented. The first is where a witness “adopts” in criminal proceedings the evidence previously given to the inquiry; the second is the derivative use of inquiry evidence to identify adverse lines of investigation; and the third is the (more practical) prejudice caused by publication of the inquiry’s findings.

The retrospective “adoption” of inquiry evidence

Suspect A is under investigation for a suspicious death whereby the victim was stabbed in a fight between two rival gangs. The police cannot identify A as having been directly involved in the stabbing nor account for his movements on the day. A exercises his right to silence during his first police interview under caution. A is then compelled to give evidence to a public inquiry relating to incidents of gang violence in the area, protected by an AG’s undertaking which states that “no evidence a person may give before the Inquiry will be used in evidence against that person in any criminal proceedings or for the purpose of deciding whether to bring such proceedings.” A gives evidence to the inquiry about his movements on the day, including that he was present near the scene of the murder with the group involved in the fight, but denies being party to the stabbing. Following publication of the inquiry’s findings, the police interview other witnesses referred to during the course of A’s evidence. They attest to A’s movements near the scene of the crime but state that he was in fact involved in the fight. A is subsequently invited by the police to attend a second interview under caution. He is properly informed of his legal rights and entitlements, including the right to silence. He is asked whether the bystanders’ evidence is correct or if he instead stands by and adopts his account as provided to the inquiry. He confirms the latter. He is charged with murder on a joint enterprise basis.

The question for consideration is whether, by adopting the evidence provided to the inquiry in a subsequent non-compelled and fully PACE-compliant interview, the investigators are able to undermine the protection afforded by the AG’s undertaking, so that A’s evidence to the inquiry becomes admissible against him in a criminal trial. The terms of the undertaking protect only that evidence given “before the Inquiry”, not evidence adopted or repeated in a subsequent police interview.

This question is more than merely hypothetical. It was addressed recently during the prosecution of two soldiers (known as A and C) [1], accused of the murder of Joseph McCann in Belfast on 17 April 1972. This case reveals an attempt by the Public Prosecution Service of Northern Ireland (“the PPSNI”) to adduce evidence in precisely this way, albeit not in the context of a statutory public inquiry. As was common practice at the time, the two soldiers were interviewed by the Royal Military Police in 1972 following the discharge of their weapons. Those interviews were compelled by a superior officer, no legal rights were available to the soldiers and there was no caution administered. It was not in dispute that A and C’s 1972 interviews were inadmissible against them as evidence in a criminal trial.

However, in 2010, as part of an inquiry by the Historical Enquiries Team (established to re-examine the facts of Troubles-related deaths), the soldiers were invited to interview. These new interviews were conducted under caution with lawyers present, and the soldiers confirmed (or did not dispute) the accuracy of their 1972 statements. Both soldiers were subsequently charged. The PPSNI position was that the 2010 interviews (specifically the parts in which they adopted their 1972 statements) were admissible as evidence against the soldiers as they were conducted voluntarily and with all the benefits of legal representation and rights accorded under PACE.

In the Crown Court’s ruling on the admissibility of the 2010 interviews, the prosecution’s submissions were rejected and the interviews were judged to be inadmissible. But the Court’s reasoning reveals that this was a highly fact-specific decision. The Court referred to the well-known Strasbourg authority of Saunders[2] and Lord Hughes’ dicta in Beghal v DPP[3], which establish the general rule that compelled oral evidence is inadmissible in criminal proceedings. However, the main thrust of the Court’s reasoning was that the 2010 interviews were not part of a formal criminal investigation; the caution had not identified any criminal offence (a breach of the PACE Codes of Practice); and the interviewees (and their solicitors) did not know at the time of the 2010 interviews that the 1972 statements were inadmissible.[4] This decision is therefore no precedent for saying that evidence given to a public inquiry by a witness protected by an AG’s undertaking can never become admissible evidence in criminal proceedings if voluntarily adopted by the witness in a subsequent interview under caution.

By analogy, statutes conferring compulsory interview powers on criminal law enforcement bodies nearly always provide for circumstances in which the interview, which is inadmissible against the interviewee by virtue of its compulsion, can become admissible against the interviewee if evidence relating to the interview is adduced, or a question relating to it is asked, by or on behalf of the interviewee.[5] Similarly, if a person voluntarily affirms the accuracy of the evidence he gave to a public inquiry in an interview under caution, or in oral evidence during a criminal trial, the AG’s undertaking will not necessarily exclude the public inquiry evidence. Under sections 76 and 78 PACE, its admissibility will depend on factors including the actual prejudice caused by admitting the evidence.

The development of additional lines of investigation

The second way in which an individual’s evidence to a public inquiry might be used against them is by utilising it to pursue adverse lines of investigation. As noted above, the AG’s undertakings for the Baha Mousa and Al Sweady inquiries prevented this derivative use, but this is not a standard exclusion found in all AG’s undertakings. The undertaking to the Grenfell Tower inquiry does not, for example, preclude the criminal investigator from making derivative use of the evidence provided by witnesses. Doubtless the AG in that case did not want to hamper the progress of an ongoing, highly scrutinised, criminal investigation, or the prospect of bringing successful prosecutions by providing such an assurance.

For example, a witness may, during the course of their evidence to a public inquiry, refer to previously unidentified witnesses or third party material that may tend to incriminate them. Whilst the police would be unable to adduce that witness’s oral account as evidence against them in criminal proceedings, there is nothing to stop the police from contacting those witnesses, or obtaining the third party material, as independent evidence to support a prosecution, as long as they follow proper procedures in respect of the seizure and production of such evidence.

In a less direct manner, the evidence provided by witnesses as to their own involvement or culpability inevitably feeds into the ultimate findings by the Chair of an inquiry. Whilst under the 2005 Act a Chair is prohibited from making determinations as to guilt in law, there is nothing to stop them from making factual findings as to responsibility, and also as to the quality of the evidence provided by each witness. Where those witnesses are subsequently interviewed under caution as suspects, it is impossible to know the impact those findings may have on an investigation. Whilst investigators will often assure suspects who attend an interview that they have not considered the evidence given during an inquiry, it is, with respect, naïve to assume that they are not at least aware of its key findings. Those findings will invariably influence how their investigation is carried out, including identifying lines of investigation adverse to the interviewee.

Using Grenfell again as an example, the public inquiry and the criminal investigation are running in parallel, but have different mandates and powers. And yet it appears that the progression of the criminal investigation may in some way be contingent upon the findings of the inquiry. On 18 September 2020, the Mayor of London was asked at the London Assembly when the criminal case into Grenfell might be concluded with decisions as to charge. His response, ostensibly on behalf of the police and CPS, was that “the criminal investigation must take into account any findings or reports produced by the Grenfell Tower public inquiry”[6] [emphasis added]. Whilst this assertion could apply to the elimination of suspects from police enquiries, findings of fact as to responsibility or a witness’s credibility could be equally persuasive to the formulation of a case theory, even it is not directly relied upon in evidence.

To take a more historic example, Lord Saville published his findings into the events of Bloody Sunday in 2010. An AG’s undertaking had been provided a decade earlier. Various findings of fact as to the actions of British soldiers were made, including that several were likely responsible for the deaths of civilians, and that there was no evidence the civilians were armed. Specific soldiers were identified as having been likely to have killed or injured specific individuals. Fresh criminal investigations were subsequently launched, and the soldiers identified by Lord Saville were invited to attend voluntary interviews under caution, 45 years after the events in question. Given there were no significant evidential developments in that 45 year period, the inquiry must have been the vector for change.

Therefore, unless the AG’s undertaking includes a derivative use prohibition, there is no practical protection for suspects against the building of a criminal case which takes into account (although does not explicitly rely on) the evidence given to, or the factual findings made by, an inquiry.

Jury prejudice as a by-product of an inquiry

Thirdly, it is impossible to quantify the potential prejudice to a jury that may be caused by an inquiry’s adverse findings. The events examined at a public inquiry are usually those which have caused a particular affront to the public conscience. The findings of an inquiry are typically front page news, with individuals named and shamed across newspaper headlines and social media. If those individuals subsequently stand trial for the same matters, there is, in legal terms, no effective recourse for them to claim that their right to a fair trial would be prejudiced. The protections available have diminished over time, such that courts will now only stay proceedings in the most exceptional of circumstances. A significant degree of trust is placed in the ability of juries to simply put aside any knowledge they may have of pre-existing proceedings and to focus on the matter in hand. However, a defendant in the dock who has been vilified in the media on the basis of his previous evidence to a public inquiry may not feel the same degree of confidence in a jury’s powers of objectivity.

Conclusion

As we see more and more issues progressed to public inquiries alongside criminal investigations, it will be important not to take for granted the apparent safeguards provided by an AG’s undertaking, if one is given. Likewise, witnesses to public inquiries must always consider the possible future use of their evidence in criminal proceedings. A witness may, in extreme circumstances, need to weigh up the risks associated with giving evidence (direct and indirect use of the evidence in criminal proceedings, reputational damage caused by adverse findings and media attention) alongside the sanctions for non-compliance, which is currently summary conviction with a fine and/or maximum 51 weeks imprisonment under section 35 of the 2005 Act. Faced with these competing risks, statutory inquiries may find that suspects become increasingly less willing to answer the call for full and frank disclosure, without more concrete guarantees that they will not inadvertently assist their own prosecution.

Danielle Reece-Greenhalgh, Senior Associate   

[1] R v Soldier A & Soldier C [2021] NICC 3.

[2] [1997] 23 EHRR 313

[3] [2015] UKSC 49

[4] See para [34]

[5] See, for example, section 2(8AA) Criminal Justice Act 1987

[6] https://www.london.gov.uk/questions/2020/3082