Disclosure: time to cut the Gordian knot?
Non-disclosure in criminal proceedings has long been recognised as a “potent source of injustice” (per Glidewell J in R v Ward  1 WLR 619, the successful appeal against convictions relating to the 1973 Euston station and M62 bombings). The collapse of the Liam Allan trial has brought the issue back to national attention.
Regrettably, current evidence suggests that this is not an isolated incident. The 2017 joint review of disclosure by HM Chief Inspector of the CPS and HM Inspector of Constabulary found that scheduling of unused material is “routinely poor” and, alarmingly, “revelation by the police to the prosecutor of material that may undermine the prosecution case or assist the defence case is rare.” Non-disclosure also remains the most common reason for the Criminal Cases Review Commission referring cases back to the Court of Appeal.
The disclosure regime established by the Criminal Procedure and Investigations Act 1996 (‘CPIA’) was the statutory response to notorious miscarriages of justice in the 1970s and 1980s, but, 20 years on, it does not appear to be functioning as it always should. The recent explosion in the volume of electronic material is posing still further challenges for investigators and prosecutors seeking to comply with the regime by identifying disclosable material. The R v R & Others  EWCA Crim 1941 is a prime example. The unused material constituted 7 terabytes of data. At the point the case was stayed (the decision was later overturned on appeal) there had been no progress beyond primary disclosure for five years.
The SFO is turning to technology for answers. In the Rolls Royce case, artificial intelligence was used to sift, index and summarise documents. While that may provide the long-term solution, the software is not readily available beyond the SFO, not least because of cost.
Some commentators and academics suggest the solution lies in adopting the currently forbidden ‘keys to the warehouse’ approach whereby the defence are given access to all material obtained during the investigation (subject to limited public interest grounds for non-disclosure). That would undoubtedly represent a dramatic shift in the approach to disclosure in England and Wales. If we look to other jurisdictions, however, the proposal is not as radical as it might first seem.
In most civil law jurisdictions, the defence have access to all material obtained during the investigation. Generally, in inquisitorial systems, the prosecution prepares for the investigating judge a dossier that contains all the evidence and the full record of the investigation. The judge will then grant the defence access to the full file at some point in the proceedings; exactly when depends on the jurisdiction.
The approach in common law jurisdictions varies. Canada, for example, allows the defence full access to all relevant material obtained in the investigation. In the landmark ruling in R v Stinchcombe  3 SCR 326, the Canadian Supreme Court affirmed that all relevant information, including unused material, must be disclosed subject only to the reviewable discretion of the prosecution to withhold material irrelevant to the case or on public interest grounds (such as national security or witness safety).
The Canadian Supreme Court dismissed the suggestion that such a general disclosure duty was incompatible with an adversarial system as contrary to the burden of proof and the function of the prosecutor as a minister for justice. Sopinka J went further and observed that:
“…the fruits of the investigation which are in the possession of counsel for the Crown are not the property of the Crown for use in securing a conviction but the property of the public to be used to ensure that justice is done.”
Sopinka J also pointed to a further principled argument for full disclosure, namely, that the decision as to whether an unused witness statement was “…sufficiently useful to put into evidence should be made by the defence and not the prosecutor.”
Similarly, in England and Wales, it is arguable that notwithstanding the provision of a carefully-worded defence statement, a defendant remains in a better position to know what might assist her case than the most diligent prosecutor.
In the USA, the ‘Brady rule’ (so-called after the seminal Supreme Court case of Brady v Maryland 373 U.S. 83 (1963)) provides a constitutional right to exculpatory material. There is, however, a growing ‘open file’ trend at State-level, partly in response to restrictive Federal discovery rules. Some of those schemes make reciprocal defence disclosure a condition of full access to the case file to ensure equality of arms.
The proponents of such schemes argue that they improve efficiency by avoiding pre-trial disclosure arguments and appeals and allow defendants to make better-informed and timely plea decisions. Those points accord with the reasoning of the Canadian Supreme Court in Stinchcombe, which found that any increase in prosecutorial workloads because of full disclosure would be offset by the time saved in resolving disclosure disputes. The ‘open file’ approach would also arguably reduce the scope for police or prosecutorial incompetence or impropriety and improve confidence in the fairness of trials.
Such an approach is also not entirely without precedent in England and Wales. Before the CPIA regime, the common law had developed its own answers to non-disclosure. In 1993, Ward (supra) established a new disclosure standard: all evidence of help to the accused should be disclosed. The Court of Appeal emphasised that,
“…all relevant evidence of help to the accused” is not limited to evidence which will obviously advance the accused’s case. It is of help to the accused to have the opportunity of considering all the material evidence which the prosecution have gathered, and from which the prosecution have made their own selection of evidence to be led.”
That potentially broad obligation was clarified in the later cases (see most notably R v Keane  1 WLR 746 which held the test was one of materiality). Ultimately, the CPIA, which came into force a few years later, curtailed the prosecutorial disclosure obligation. It must be remembered, however, that the statutory scheme requires that a full schedule of material in the possession of the prosecution is always provided to the defence, save for material which attracts public interest immunity. The defence are, therefore, always made aware of the existence of material and can request copies of it. If the prosecution decline to provide access to the material then application can be made to a judge.
Shifting the burden of identifying material that might undermine the prosecution or assist the defence to the defendant does present its own difficulties. Arguably it is inefficient as the time and cost of examining vast quantities of unused material would fall on both sides, instead of the prosecution alone. To accommodate that, the current public funding arrangements for defence lawyers, which do not provide for specific remuneration for reviewing unused material, would have to change. Even then, it is unlikely to be workable in the most document-heavy cases as the defence would require enormous resources, in terms of both manpower and IT systems, to review all the material. In short, and perhaps unsurprisingly, there are practical difficulties with any disclosure regime.
In the final analysis, such a sweeping change is unlikely to be attractive to the legislature. There is after all broad consensus that the CPIA provides a principled, Article 6 ECHR compliant disclosure regime. The recent criticism relates to its practical implementation and the 2017 joint inspection report provides a series of recommendations from better training to improved strategic oversight which, if implemented properly, would go some way to remedying those concerns.