Efficiency without adversaries? Leveson,Fisher and the re-engineering of criminal procedure in England and Wales

Introduction

The criminal courts of England and Wales are in a state of abject crisis. In December 2025, the backlog of criminal cases waiting to be heard in the Crown Court of stood at approximately 78,000 cases. ¹ This is an increase of 11% on the previous year and is almost twice as high as that at the end of 2019. ² Magistrates’ courts also carry record volumes of outstanding cases, while leaking roofs, judicial vacancies, and practitioner attrition have become familiar features of the landscape of criminal justice in England and Wales. In this environment, William Gladstone's maxim of ‘justice delayed is justice denied’ ³ has ceased to be a rhetorical flourish and has hardened into the governing premise of reform. That premise, however, did not arrive with either COVID-19 or the industrial action taken by the Criminal Bar. The Criminal Bar Association (CBA) notes that defence-bar industrial action in 2022 added to the backlog, but its own data shows the increase in backlog over that strike period was relatively modest (from 58,000 to ∼62,000 between April 2022 and October 2023) – indicating that the courts were already under pressure and that the strike contributed on top of an existing upward trend. ⁴ For three decades, major policy interventions have been framed around ‘efficiency’: the Royal Commission on Criminal Justice; the Auld Review; the Criminal Justice: Simple, Speedy, Summary initiative; Swift and Sure Justice; Sir Brian Leveson's 2015 Review of Efficiency in Criminal Proceedings; and the National Audit Office's (NAO's) landscape reviews. ⁵ Each has promised to cut delays, rationalise procedures and improve throughput. The 2025 Independent Review of the Criminal Courts, ⁶ again chaired by Sir Brian Leveson, represents the most comprehensive step in this ongoing trend. Its terms of reference instruct the Review to advise, firstly, on how courts might handle cases proportionately given the pressures on the Crown Court, and secondly, on how they and their partner agencies can operate ‘as efficiently as possible’. Part 1 of the Review offers a structural plan: mainstreaming out-of-court disposals (OOCDs) and diversion at the charging and review stages; reclassifying either-way offences and limiting the right to elect jury trial; establishing a new Crown Court Bench Division (CCBD), comprising a judge and magistrates, for mid-range cases; and increasing judge-alone trials in fraud and other complex cases. Each proposal is framed with language of proportionality, access and responsiveness to vulnerability. However, the underlying logic is clear: conserve limited jury-trial capacity and channel as many cases as possible into leaner, more tightly managed processes. Meanwhile, Jonathan Fisher KC's Disclosure in the Digital Age: Independent Review of Disclosure and Fraud Offences ⁷ tackles a different challenge within the same efficiency drive: the explosion of digital material in criminal investigations and the consequent pressure on the disclosure regime under the Criminal Procedure and Investigations Act 1996 (CPIA). Fisher's assessment is stark. Phones, social media, cloud storage, and widespread CCTV have rendered disclosure increasingly time and resource-intensive, ⁸ impacting both timeliness and the State's ability to fulfil its duty to disclose material that undermines the prosecution or supports the defence. His recommendations aim to create a more ‘viable and efficacious’ ⁹ system: a cross-system Digital Disclosure Working Group; increased utilisation of advanced technologies, including AI-assisted review; revised CPIA Code and Attorney General's Guidelines (AGG) to clarify ‘reasonable lines of Enquiry’; and more assertive judicial case management of disclosure. ¹⁰

Read together, Leveson and Fisher do not simply propose technocratic adjustments to a beleaguered system. Leveson's proposals reorganise where and how cases are adjudicated, intentionally shifting work away from traditional, jury-centred trials towards out-of-court, summary or hybrid venues. Fisher's settlement addresses the evidential infrastructure of that system: how investigations and disclosure can be made sustainable in the digital age through technological triage and managed engagement. In both, efficiency, proportionality and resource optimisation are central; fairness and due process are referenced, but increasingly as values to be balanced against efficiency rather than as its initial foundation. This article examines those settlements through the lens of traditional adversarial justice. On the standard account, the adversarial model is defined by party-led investigation and proof presentation, a relatively passive judge, a strong presumption of innocence, a burden of proof which rests on the prosecution, a meaningful equality of arms in access to evidence, and – at least for serious offences – trial by jury as the default adjudicative forum. ¹¹ These features underpin the Article 6 fair-trial guarantees of the European Convention on Human Rights and shape ethical and procedural expectations regarding disclosure, cross-examination and effective participation. ¹² By contrast, managerial criminal justice treats criminal process as a system to be steered: risk is managed, throughput maximised and professional roles recast around administrative performance metrics rather than adversarial contest. ¹³ The central question is whether Leveson and Fisher's proposals can plausibly be reconciled with the adversarial commitments they formally retain, or whether, together, they re-engineer the defence role and the trial forum in ways that hollow out the traditional model in practice. The article's central contribution is to recast Leveson and Fisher as components of a single digital–managerial settlement and to show how that settlement restructures defence work, digital disclosure practice and the Article 6 equality-of-arms guarantee. The article is divided into five substantive parts. Section ‘Traditional adversarial justice and the managerial turn’ outlines the traditional adversarial model and examines the rise of managerial criminal justice. Sections ‘Varieties of managerialism in criminal justice’ and ‘Managerialism in English criminal justice’ explore Leveson's Independent Review of the Criminal Courts and Fisher's Disclosure in the Digital Age. The section ‘Managerialism, disclosure and early engagement’ juxtaposes the continuing drive for efficiency with the classic notion of adversarialism. The final substantive section proposes an adversarial-grounded reform agenda, identifying structural conditions – especially defence funding, disclosure and jury trial – where efficiency can be aligned with fair-trial protections. It concludes by considering the implications of treating efficiency as a new constitutional value in criminal justice.

Traditional adversarial justice and the managerial turn

Any evaluation of Leveson and Fisher has to start from the traditional adversarial model that still formally structures criminal procedure in England and Wales. Adversarial justice rests on a cluster of familiar commitments: the parties, rather than the court, control the investigation and presentation of proof; the judge acts as an impartial arbiter rather than an inquisitorial fact-finder; the prosecution carries the burden of proving guilt beyond reasonable doubt; and defence counsel's primary function is to test the prosecution case through rigorous challenge, not to assist the court in meeting administrative goals. ¹⁴ Jury trials are widely regarded as the institutional expression of this model in serious cases, embodying a lay community's verdict on contested facts and providing an additional buffer between the individual and state power. ¹⁵ Within this paradigm, procedural safeguards – disclosure, cross-examination, rules of evidence, the privilege against self-incrimination – are primarily justified because they sustain a robust contest between structurally unequal parties. ¹⁶ The prosecution's investigative advantages are tempered by positive disclosure duties and by the opportunity for the defence to probe the reliability, completeness and interpretation of the State's case at trial. Defence counsel are cast as partisan advocates, obliged (within ethical limits) to put the prosecution to proof and to exploit evidential gaps wherever they arise. ¹⁷ Case management, where it features at all, is conceived as a relatively light-touch, reactive function: the judge responds to applications made by the parties and polices compliance with rules framed in rights-protective terms. Article 6 ECHR frames these concerns. Strasbourg has long held that the right to a fair trial includes both equality of arms and adequate disclosure of material capable of undermining the prosecution or assisting the defence. In Edwards v United Kingdom, ¹⁸ the Court found a violation where undisclosed information about a police informant prevented effective challenge to the Crown's case, while Rowe and Davis v United Kingdom ¹⁹ and Dowsett v United Kingdom ²⁰ confirm that the prosecution cannot be the final arbiter of what is revealed to the defence when public interest immunity is claimed. In Nideröst-Huber v Switzerland, ²¹ the Court stressed that aims of saving time or expediting proceedings cannot justify undermining the adversarial character of the process. Domestically, R v H; R v C ²² reads the CPIA disclosure regime through the same lens, insisting that fairness ordinarily requires disclosure of any material that significantly weakens the prosecution or strengthens the case of the defence, and that departures from this must be exceptional. On this idealised view, efficiency is not irrelevant – long delays and wasteful hearings can obviously harm defendants, victims and witnesses – but it is not the system's primary organising value. Delay is objectionable because it undermines accuracy, prolongs anxiety and erodes legitimacy, rather than because it impacts throughput or key performance indicators. The vocabulary is one of fairness and due process, not flow, productivity or failure demand. It is this ordering of values that is disturbed by the various forms of managerialism that have taken hold in the criminal courts over the last two decades.

Varieties of managerialism in criminal justice

Managerialism is not a single, unified theory but an umbrella under which several related developments sit. From the late 1980s onwards, public services in England and Wales were increasingly governed through targets, performance indicators and audit mechanisms. ²³ Courts and criminal justice agencies were drawn into this orbit through the creation of Her Majesty's Courts and Tribunals Service (HMCTS), successive Public Service Agreements, and close interest from the NAO in value for money and efficiency in criminal justice. ²⁴ NAO reviews and Ministry of Justice policy papers on Criminal Justice: Simple, Speedy, Summary, Swift and Sure Justice and Transforming the CJS framed the criminal process as a production line to be streamlined: fewer hearings per case, fewer cracked trials and shorter waiting times. ²⁵ The language of audit and performance permeated how problems were defined, and solutions were evaluated. Second, procedural managerialism and managerial judging. At a more granular level, academic commentators have charted the rise of active judicial case management and procedural steering of criminal trials. ²⁶ The Criminal Procedure Rules (CrimPR), first introduced in 2005 and now consolidated in the 2025 rules, are central here. The overriding objective in r 1.1 requires courts to deal with cases ‘justly’, but then defines that in terms that marry classic fairness concerns (acquitting the innocent, convicting the guilty) with explicitly managerial aims: dealing with cases efficiently and expeditiously and allotting to each case an appropriate share of the court's resources. ²⁷ Practice Directions and judicial guidance encourage robust intervention to identify ‘real issues’, limit evidence, enforce timetables and control the length and number of hearings. The trial and pre-trial processes are no longer simply neutral stages at which parties present and contest evidence, but points in a managed process whose success is measured partly in terms of time taken, number of adjournments avoided and compliance with standardised timetables. Judges are expected to challenge both prosecution and defence when their conduct is perceived to cause delay or waste court resources. Even before the inception of the CrimPR, appellate courts had begun to recast the CPIA-era defence case statement as a tool for proactive case management. In R v Tibbs, ²⁸ the Court of Appeal held that a defence statement should set out the nature of the defence, identify the prosecution facts in issue and explain in outline why they are disputed, rejecting bare denials as insufficient. In R v Gleeson, the court condemned deliberate defence ‘ambush’ on a purely legal point and insisted that material issues of law and fact should be raised at the plea and directions hearing rather than withheld in the hope of an acquittal on a technicality. ²⁹ Jisl ³⁰ then articulated a broader philosophy of judicial case management, stressing that trials must be controlled so as to focus on genuinely disputed issues in light of limited court resources. Once the Criminal Procedure Rules were in force, this trajectory crystallised: DPP v Chorley Justices ³¹ Thomas LJ described the Rules as effecting a ‘sea change’ ³² in criminal process and made clear that parties were under a duty to assist the court by identifying and remedying technical defects in advance; Malcolm v DPP ³³ and Penner ³⁴ confirm that courts may deploy their powers to neutralise late tactical points where the defence has failed to engage with case-management obligations, so long as no unfair prejudice is caused to the accused.

These managerial imperatives have also reshaped how professional roles are conceived within the system. The Crown Prosecution Service has been studied as a paradigmatic managerial organisation, balancing legal judgment with targets, Key Performance Indicators and demands for consistency and throughput. ³⁵ Defence practice, though structurally less powerful and resourced, has been affected in parallel. Johnston argues that the defence advocate has been progressively reimagined as a case-progression actor: obliged to cooperate in narrowing issues, to facilitate early guilty pleas where appropriate, and to assist the court in meeting efficiency objectives. ³⁶ This sits alongside legal aid contracting models that remunerate fixed events (police station attendance, plea hearing, trial), thereby aligning economic incentives with rapid resolution rather than protracted contest. Across these strands, managerialism does not necessarily present itself as anti-fairness. Policy documents routinely insist that reforms are intended to make justice ‘swifter and surer’, not merely cheaper. ³⁷ But the techniques and metrics employed – targets for guilty plea rates, cracked trial statistics, time-to-disposal dashboards – mean that delay, contest and complexity are routinely treated as pathologies to be reduced, even when they reflect legitimate efforts to test the prosecution case.

Managerialism in English criminal justice

The Plea and Trial Preparation Hearing (PTPH), introduced in 2015, was designed as a working hearing at which the parties are expected not only to enter pleas but to identify the issues in dispute, agree as much evidence as possible, resolve admissibility questions and set a tight timetable for any trial. ³⁸ Linked initiatives such as Better Case Management in the Crown Court and Transforming Summary Justice in the magistrates’ courts seek to ensure that cases are ‘right first time’. ³⁹ In principle, this should promote both fairness and efficiency: properly prepared cases are less likely to collapse late or to require repeated adjournments. In practice, empirical work and practitioner commentary suggest that the relentless emphasis on avoiding ineffective hearings and on keeping cracked-trial rates down can incentivise early guilty pleas and discourage robust late-stage challenge. ⁴⁰ Applications for further disclosure, expert evidence or witness summonses may be met with scepticism if they are seen as disrupting a carefully managed timetable. Sentencing policy has also been shaped by managerial thinking. The now entrenched ‘one-third’ reduction for a guilty plea at the first reasonable opportunity, with smaller discounts thereafter, functions as a tool of case-flow management: it reallocates the cost of insisting on trial onto defendants, who face a significant sentencing premium if they do not plead early. ⁴¹ From a managerial perspective, this is rational; from an adversarial perspective, it alters the baseline against which the exercise of the right to trial is experienced and evaluated. Choosing to put the prosecution to proof becomes, in effect, an expensive option.

Commentary on the adversarial defence lawyer argues that these developments, taken together, have altered both the expectations placed on defence counsel and the normative environment in which they operate. ⁴² Defence practitioners are increasingly expected to advise on and encourage early guilty pleas where the evidence appears strong; to co-operate in pruning witness lists; to agree the scope of cross-examination; and to accept judicial indications about the length and shape of trials. Their performance is evaluated not only by clients, appellate courts and professional regulators, but informally through judicial comment and through the system's own performance metrics (e.g. cracked-trial statistics attributed to particular courts or listing teams). This does not mean that adversarial advocacy has disappeared. Defence lawyers still challenge disclosure failings, cross-examine witnesses and run trials. But they are doing so in an environment in which time and contest are increasingly constructed as deviations from the norm that must be justified.

Managerialism, disclosure and early engagement

Disclosure and early engagement schemes provide particularly stark examples of how managerial and adversarial logics can clash. The CPIA 1996 and its accompanying Code of Practice created a disclosure regime that, on paper, is strongly prosecution-focused: investigators must pursue ‘all reasonable lines of Enquiry’, whether they point towards or away from the suspect, and the prosecution must disclose any material that might reasonably be considered capable of undermining its case or assisting the defence. ⁴³ This was intended as a structural safeguard against the inherent imbalance in investigatory resources. Over time, however, disclosure practice has become a key site of procedural managerialism. Judicial case management now extends explicitly to disclosure: courts require timetables, progress updates and assurances that ‘reasonable lines of Enquiry’ have been identified and pursued. ⁴⁴ Judges are urged to discourage broad, unspecific defence requests, to focus attention on genuinely contested issues and to prevent disclosure from becoming an open-ended trawl through digital material. Fisher's review records that this development has been widely described as a shift to more ‘managed’ disclosure, with courts playing a central role in steering both sides towards proportionate handling of material. ⁴⁵ Early engagement schemes crystallise many of these tensions. Pre-charge engagement (PCE), introduced in 2020, is predicated on the idea that suspects and/or their lawyers might voluntarily engage with investigators after interview but before charge, identifying lines of Enquiry, disclosure issues and possible bases for diversion. ⁴⁶ In theory, this is a paradigmatic managerial reform: it promises to narrow issues, divert weak cases without charge, and avoid the cost and stress of unnecessary prosecutions. In practice, the scant empirical research on the scheme suggests that it has been rarely used, poorly understood and structurally unfunded. ⁴⁷ Respondents reported that meaningful participation often requires 15–20 h of work for reviewing material, taking instructions, drafting representations and negotiating with investigators, that the standard police station and litigation fees do not remunerate. The result is an ‘empty gesture’: a process that assumes defence capacity that does not exist in most legally aid cases, and thus risks operating as a privilege of the privately funded few. Fisher's own recommendations on digital disclosure reinforce this picture. His proposal for a cross-system Digital Disclosure Working Group and his endorsement of AI-assisted review are aimed at making disclosure ‘viable and efficacious’ amid digital volume. ⁴⁸ But they also presuppose a defence profession able to engage with, and to challenge, technologically mediated review processes in a meaningful way. Without serious attention to defence-side capacity – time, training, digital tools – there is a risk that managerial innovations will further entrench the asymmetry between a digitally equipped State and a chronically under-resourced defence. What emerges from these different strands is not that managerialism has simply displaced adversarial justice, but that it has overlaid and reoriented it. The formal structures of adversarialism – party presentation of proof, burdens of evidence, the possibility of jury trial – remain in place. Yet they are increasingly embedded within a system governed by performance metrics, case-flow targets and expectations of cooperative, efficiency-minded behaviour from all participants, including the defence. It is against this background that Leveson's 2025 proposals for restructuring the criminal courts and Fisher's digital disclosure settlement must be read. Both assume, and deepen, a managerial environment in which efficiency is not merely a desirable side-effect of fair procedures but a constitutional value in its own right.

Leveson's efficiency blueprint for the criminal courts

Part I of Leveson’s Independent Review of the Criminal Courts is framed explicitly as a response to crisis: an ‘open caseload’ of over 78,000 Crown Court cases, trials listed into 2029, and the spectre of ‘system collapse’ if delay remains unchecked. ⁴⁹ Efficiency is not presented as a marginal value, but as the organising principle through which proportionality, timeliness and – only derivatively – fairness are to be delivered. ⁵⁰ The package builds self-consciously on the 2015 Review of Efficiency in Criminal Proceedings, which celebrated ‘effective and consistent judicial case management’ as a central safeguard and sought to align professional obligations with the Criminal Procedure Rules’ overriding objective. ⁵¹ In Part I of the Review, that logic is deepened and constitutionalised: procedural design, judicial case-management powers and financial incentives are calibrated to maximise throughput, with the adversarial trial increasingly treated as a scarce and expensive resource. Although many individual recommendations address long-recognised problems – such as the lack of recording in the magistrates’ courts or the dysfunctional appeals structure – taken together they recast the criminal process as a multi-level triage system.

Diversion, OOCDs and deferred prosecutions

Recommendations 1–10 seek to reduce ‘demand’ on the courts by mainstreaming Out of Court Disposals (OOCDs), rehabilitation pathways and deferred prosecutions. Leveson urges that in ‘all appropriate cases’ prosecutors and police must consider OOCDs at the charging stage, and that the open caseload should be reviewed to identify cases suitable for diversion. ⁵² He couples this with a call for a standardised national diversion framework, expanded treatment and deferred prosecution options, and simplified rehabilitation rules, all presented as proportionate ways to keep low-tariff cases out of court. Professional responses acknowledge the potential of this diversionary turn. The Bar Council, for example, welcomes Recommendations 1–10 and stresses that reducing the inflow of cases is one of the few measures capable of generating immediate impact on the backlog. ⁵³ Restorative justice advocates have similarly endorsed Recommendation 8, which explicitly links OOCDs with restorative processes for low-tier offences. ⁵⁴ Through the managerial lens, however, the same measures raise familiar questions. Redirecting resolution to informal or administrative settings risks further hollowing out the adversarial trial as the core site of fact-finding and challenge. Put bluntly, the more that charge decisions, evidential concessions and admission of guilt are negotiated in opaque pre-court forums, the more the protections attached to open, contested hearings become a residual safeguard for the few. The Review contains little engagement with the structural resource constraints that shape defence participation in these diversionary regimes – an issue sharply exposed in recent research on PCE, which suggests that early resolution schemes heavily depend on unfunded defence labour. ⁵⁵

Rebalancing work between magistrates’ courts and the Crown Court

A second strand of the blueprint re-engineers the boundary between the magistrates’ courts and the Crown Court. Leveson proposes:

Removing the right to elect Crown Court trial for either-way offences with a maximum of two years’ imprisonment (and for other specified offences); ⁵⁶

These recommendations are underpinned by modelling, which suggests that the combined effect of reclassification and altered allocation would save around 4000 Crown Court sitting days per year and facilitate ‘efficient and timely’ distribution of cases. ⁶⁰ The stated principles are based in proportionality – reserving the Crown Court for more serious offences – and fairness, with Leveson claiming that his proposals do not compromise defendants’ rights. ⁶¹ However, critics contest that assessment. For APPEAL and others, restricting election for a swathe of either-way offences (including protest-related public order offences and assault occasioning actual bodily harm) represents a significant encroachment on the right to jury trial, particularly for defendants whose alleged conduct is politically or socially contentious. ⁶² Survey data from the CBA indicates overwhelming opposition to these measures among frontline barristers, who stress the importance of jury trials for public confidence and as a guard against bias, and suggest that jury trials are being scapegoated for problems in the justice system. ⁶³ From a managerial perspective, the rebalancing exercise exemplifies what Johnston and Pivaty describe as the ‘downward displacement of complexity’: the system's structural pressures are addressed by pushing more work onto lower courts and cheaper forums, supported by modest procedural tweaks (audio recording, modest legal aid reforms) intended to sustain legitimacy. ⁶⁴ The Review recognises that retaining more serious work in the magistrates’ court will increase demands on defence practitioners, and recommends removing ‘perverse incentives’ in legal aid remuneration that encourage election, but does not confront the deeper problem of long-term underfunding in the summary jurisdiction. ⁶⁵

The practical effect is to narrow defendant control over the trial forum in the name of resource management. Where the adversarial tradition imagined the choice of jury trial as a critical safeguard for those facing criminal stigma, Leveson's model treats that choice as a variable that can be re-tuned to optimise throughput, subject only to broad assertions that proportionality and fairness are maintained.

Crown Court (bench division) and judge-alone trials

The most radical structural innovation is the proposed Crown Court (Bench Division) (CCBD). Leveson rejects the long-mooted ‘Intermediate Court’ in favour of a new division of the Crown Court, staffed by a judge sitting with two magistrates and empowered to try all either-way offences without a jury. ⁶⁶ Allocation decisions would be made at a PTPH, with a presumption that cases likely to result in a sentence of three years or fewer should be heard in the CCBD. ⁶⁷ The CCBD is presented as a means to ‘efficient and proportionate use of resources’ ⁶⁸ while preserving Crown Court sentencing powers and allowing appeals to proceed as now. In modelling, Leveson assumes that reallocating cases to a bench-trial structure would significantly reduce average trial length; he notes that replacing the jury with a judge alone would save even more time than the judge-plus-magistrates model used for the CCBD estimates. ⁶⁹ This structural reform is reinforced by Chapter 9's recommendations on trial by judge alone. Recommendation 43 proposes that defendants should be permitted to elect judge-only trial, subject to the trial judge's consent, with allocation decisions at PTPH and no appeal against the judge's decision either way. ⁷⁰ Leveson further recommends extending existing powers for non-jury trials in cases of jury tampering so that judges may direct judge-alone trials where cases turn predominantly on expert evidence beyond the ‘common understanding’ of jurors. ⁷¹

The combined effect is a systematic thinning of jury involvement. For either-way cases, the default arbiter of fact becomes a professional lay bench; for complex fraud and similar cases, the Review envisages judge-only trials becoming routine. Commentators have repeatedly identified this as the most constitutionally sensitive aspect of the blueprint. Benn argues that Leveson not only rejects the idea of a constitutional right to jury trial, but appears to deny any right at all – contending that nothing in the legal order would prevent Parliament from legislating for non-jury trials even in murder cases. ⁷² APPEAL's analysis emphasises the risk that displacing juries will exacerbate racial disparities, given evidence of bias in judicial decision-making and the relative ‘colour-blindness’ of jury verdicts. ⁷³ Professional responses from the CBA echo these concerns. In its position paper, based on survey responses from over 2000 criminal barristers, 88.5% opposed the creation of the CCBD. ⁷⁴ The CBA stresses that the Crown Court is defined not by architecture or judicial title but by the presence of a jury, and warns that the CCBD would, in substance, be a summary court ‘with all power vested in too few hands’. ⁷⁵ The Bar Council similarly records broad opposition to measures reducing the scope of jury trial and notes the absence of robust empirical evidence that jury trials in complex fraud are a principal driver of delay. ⁷⁶

Viewed through the lens of managerialism, the CCBD and judge-alone reforms perform a double move. They reconfigure trial structures to facilitate ‘case-progression’ and sitting-day savings, while simultaneously re-describing the jury as a discretionary instrument – available where system capacity permits and where Parliament chooses. It is no longer a presumptive baseline for serious criminal allegations. Efficiency is thus embedded not only in how cases are processed, but in who is permitted to decide contested facts. ⁷⁷

Plea incentives, appeals and routinised early resolution

A third set of recommendations adjusts plea incentives and appeal rights to push the system towards earlier, less resource-intensive resolution. In Chapter 7, Leveson positions the PTPH as the critical decision point in Crown Court cases and laments the persistence of ‘cracked trials’ where guilty pleas are entered on the day of trial, wasting court time. ⁷⁸ The response is to:

Normalise Goodyear indications, ⁷⁹ with a Criminal Practice Direction encouraging judges to provide advance indications of sentence in all trials unless a good reason is given otherwise; ⁸⁰

The appeals chapter pursues parallel aims with Recommendations 21 and 22 seeking to abolish the automatic right to appeal conviction or sentence from the magistrates’ court to the Crown Court, replacing it with a permission filter that limits appeals to issues for which leave has been granted. ⁸⁵ Leveson identifies the resource cost and emotional toll of de novo re-hearings as justification for moving to a more streamlined model focused on error correction rather than further hearings. Again, these changes are explicitly efficiency-driven but normatively defended in terms of finality and fairness to victims. The Bar Council strongly opposes limiting appeals to point-of-law style hearings, warning that the current model serves as a vital safeguard against the well-documented weaknesses of summary justice. ⁸⁶ Concerns have also been raised that further increasing plea discounts, when combined with widespread Goodyear use, risks deepening the ‘plea-bargaining’ dynamics already identified by scholars and practitioners as problematic in the English system. ⁸⁷ Marsh notes the inherent dangers of defendants entering inappropriate pleas as the defence profession has suffered an ‘emasculating process [which] has occurred through judicial interventions and policy changes’. ⁸⁸ Appellate decisions underline that these plea-structuring devices do not simply recalibrate risk; they can generate unsafe convictions where ‘efficiency’ pressures overwhelm free choice. In R v Nightingale (Danny), ⁸⁹ the Court Martial Appeal Court held that a judge's unsolicited and strongly framed sentence indication had so narrowed the defendant's freedom of choice that his guilty plea was a nullity and the conviction had to be quashed. In R v Inkster, ⁹⁰ the Court of Appeal reached the same conclusion where an unrepresented defendant was told by the judge, incorrectly, that he had ‘no defence’ and that a guilty plea was his only realistic option. Moore ⁹¹ goes further, treating repeated, concrete sentencing indications in the Crown Court as impermissible judicial pressure rendering guilty pleas unsafe. The Horizon appeals in Hamilton and others v Post Office Ltd ⁹² show the systemic dimension: dozens of sub-postmasters entered guilty pleas in the shadow of over-charging, non-disclosure and financial leverage, yet the Court nevertheless treated their convictions as unsafe because the underlying process was an abuse. For managerialism scholars, the pattern is familiar: procedural rights (appeal, jury trial, the ability to maintain a not-guilty plea without excessive penal risk) are recalibrated so that the path of least resistance aligns with system goals of timeliness and disposal. ⁹³ Early guilty pleas become the rational choice for risk-averse defendants, while appellate review is recast as a narrow, permission-controlled safety valve rather than a routine second hearing.

Interim assessment: Leveson

Taken together, Leveson's recommendations are more than a menu of technical reforms, they amount to a re-specification of what it means for the criminal courts to function ‘justly’ under the Criminal Procedure Rules 2025. ⁹⁴ The system is reconceived as an efficiency-oriented network in which: police and prosecutors are expected to divert cases wherever possible; magistrates’ courts absorb a larger share of serious work; jury trial is reserved for a shrinking subset of high-tariff cases; and strong financial and procedural pressures channel defendants towards early guilty pleas, with limited late-stage opportunities for contest or correction. The suggested reforms from Leveson do nothing to reverse this reconfiguration of the defence role. In fact, the approach is entirely consistent with broader trends identified in the literature on efficiency and bureaucratisation of criminal justice. Johnston's earlier work on the adversarial defence lawyer argued that piecemeal reforms since the turn of the millennium had already repositioned defence practitioners as co-managers of case progression, tasked with identifying ‘real issues’ and facilitating early pleas in service of the CrimPR's overriding objective. ⁹⁵ Leveson's Review can be read as the culmination of this trajectory. Its explicit crisis framing, its use of modelling to justify shifts in forum and structure, and its readiness to reconceive the role of juries and appeals all signal that efficiency has moved from being a managerial desideratum to a quasi-constitutional value. In that sense, the Review crystallises the tension at the heart of this article – whether an efficiency-first blueprint for the criminal courts can be squared with a traditional adversarial conception that treats public, contested jury trial and robust appellate review as the central mechanisms for constraining state power.

Fisher's Disclosure in the Digital Age and the digital–managerial settlement

Jonathan Fisher KC's Disclosure Review is explicitly framed as a response to the rising tide of digital material in criminal investigations and the corresponding strain on the CPIA 1996 regime. ⁹⁶ The terms of reference tasked Fisher with reviewing the operation and effectiveness of the disclosure framework in the context of digital material, assessing whether the Attorney General’s Gudielines on Disclosure and CPIA Code of Practice remain fit for purpose. The central diagnosis is stark. Digital proliferation – phones, messaging apps, social media, cloud storage and ubiquitous CCTV has rendered traditional, document-based conceptions of unused material obsolete. Fisher warns that, unless adapted, the current regime will ‘severely hinder’ the Crown's ability to investigate and prosecute ‘disclosure-heavy’ cases such as fraud, organised crime and Rape and Serious Sexual Offences (RASSO). ⁹⁷ Commentators from practice echo this, describing disclosure as ‘pivotal’ to timely case progression and emphasising that digital review can take months or years. ⁹⁸ Like Leveson's Review, Fisher casts efficiency as an existential concern: the problem is not simply that failures of disclosure cause miscarriages of justice, but that digital burdens threaten the basic viability of prosecutions. The Review's 45 recommendations accordingly seek to build a ‘modern disclosure regime’ that ‘embraces technology’ to minimise ‘needless administrative burdens’ on investigators and prosecutors while maintaining fairness. ⁹⁹ In that sense, the Review is the evidential complement to Leveson's structural blueprint: where Leveson redistributes and redesigns trial forums, Fisher re-engineers the evidential plumbing that feeds them. This section argues that, although many recommendations are normatively attractive in isolation, the overall settlement further normalises a managerial conception of disclosure – one in which proportionality, technological triage and inter-agency coordination take centre stage, with adversarial challenge and equality of arms at risk of being squeezed to the margins.

Diagnosing digital overload

Fisher's starting point is a detailed account of digital overload in criminal investigations. Drawing on evidence from police, CPS, defence and judiciary, the Review notes that vast quantities of digital material are now routinely seized, particularly in fraud, organised crime and Rape and Serious Sexual Offences (RASSO), and that existing processes for capture, review and disclosure are ‘increasingly time and resource-intensive’. ¹⁰⁰ In many cases, multiple devices are seized per suspect and per complainant; communications data may span several platforms over years; and cloud-stored content is protected by complex legal and technical access barriers. The Review emphasises three systemic consequences. First, digital volume can delay charging decisions and trial readiness, as investigators and prosecutors struggle to review material within realistic timescales. Secondly, the sheer scale of material increases the risk of disclosure failings: relevant items may be missed, mis-categorised or omitted from schedules, particularly where disclosure officers lack specialist digital training. Thirdly, these pressures interact with the CPIA ‘reasonable lines of Enquiry’ test and with Article 8 privacy concerns, often producing highly intrusive downloads and data-sharing practices in RASSO and other sensitive cases. ¹⁰¹

Fisher is careful to affirm that disclosure serves fundamental fairness interests. The Review acknowledges that unused material has historically been ‘the lifeblood’ of successful defence challenges and that failures have caused high-profile miscarriages, which are ‘lamentable and erode the public's trust in the criminal justice system’. ¹⁰² The risk of a miscarriage of justice is not abstract. In R v Kay, ¹⁰³ the conviction for rape was quashed after the Court of Appeal accepted that the jury had been presented with an incomplete and misleading set of Facebook messages between the complainant and the appellant; only post-conviction did the defence obtain the whole archived thread, which painted a very different picture of the relationship and supported the account of consensual sex. The now-notorious near-miscarriages in R v Allan, R v Itiary and R v Makele likewise turned on the mishandling of digital material. In each, phones had been seized and voluminous text, image and social-media data obtained, yet exculpatory messages and photographs were not reviewed or disclosed until the eve of trial, when independent or defence scrutiny exposed material which undermined the prosecution case and led to abandonment of the proceedings. ¹⁰⁴ Taken together, these cases show how the combination of digital volume, ambiguous ‘reasonable lines of Enquiry’ and resourcing pressures can produce wrongful convictions and near-catastrophic unfairness even under the ostensibly strengthened CPIA/AGG regime. Far from being outliers, those examples sit against a wider pattern emerging in post-conviction practice. A recent Criminal Cases Review Commission study of rape-case applications in which digital communications were central identified recurring problems with the scope of device seizure, the quality and recording of review, and the disclosure of that review to the defence. ¹⁰⁵ The Commission's sample suggests that digital disclosure failures are now a structural feature of serious sexual offence litigation, rather than one-off lapses. Fisher's endorsement of targeted searches, triage tools and structured ‘reasonable lines of Enquiry’ must therefore be read not as technocratic refinements to an otherwise functioning system, but as interventions into an environment where digital evidence has already generated concrete miscarriages and a steady flow of alleged wrongful convictions.

The main rhetorical thrust is that the current configuration of duties and practices is unsustainable: without reform, the State's capacity to prosecute will be ‘severely hindered’, with knock-on effects for public confidence. ¹⁰⁶ Defence lawyers note that prosecutors’ ability to prosecute ‘swiftly’ is ‘hampered by vast quantities of digital material’, and place Fisher's work alongside the Attorney General’s Guidelines on Disclosure 2022 and subsequent 2024 update as part of a broader recalibration of expectations around digital review. ¹⁰⁷ Viewed through the managerial prism, this diagnosis sets up a familiar move: digital volume is constructed as an external, quasi-natural pressure on the system, to which disclosure rules and practices must adapt. The central question becomes how to ration attention and time across cases while preserving a baseline of fairness, rather than how to re-engineer structural conditions (funding, staffing, training) so that the original adversarial ideal – broad defence access to material capable of undermining the prosecution or assisting the defence – can be realised.

Technological solutions and institutional re-engineering

Fisher's core recommendations fall into three clusters: institutional architecture, technological tools and procedural standardisation. The headline proposal is a cross-system Digital Disclosure Board or Working Group, bringing together senior representatives from police, the CPS, the defence, the judiciary and relevant oversight bodies to set standards, monitor performance and drive implementation. ¹⁰⁸ The Board would oversee national policy on digital capture, retention and review, develop consistent triage models and promote good practice across forces and prosecuting agencies. Fisher urges the system to ‘fight fire with fire’: ¹⁰⁹ the same technologies that generate digital overload – powerful search, analytics and AI – should be harnessed to manage it. Recommendations include:

Investment in advanced e-disclosure platforms and AI-assisted review tools to filter and cluster material;

Private-sector commentators, particularly in the areas of fraud and corporate crime, have largely welcomed these proposals as aligning criminal disclosure with practices long used in complex civil litigation and regulatory investigations, where Technology-Assisted Review (TAR) is familiar. ¹¹¹ Fisher also recommends clearer, more prescriptive guidance on digital capture and review, both through revisions to the CPIA Code and through updated AGG. This will provide clarity for both investigators and prosecutors, and ensure firmer parameters for what must be seized, what may be sampled, and how ‘reasonable lines of Enquiry’ should be operationalised with digital material. ¹¹² From a managerial perspective, these measures are archetypal: they seek to embed governance structures, tools and standards that will allow disclosure to be planned, monitored and audited as a process. Digital material is reconceptualised as a flow to be triaged and processed efficiently, with technologies and central oversight bodies deployed to measure and optimise that flow. The risk, from an adversarial standpoint, is three-fold. First, where AI-assisted tools and standardised search methodologies are deployed primarily by the State, there is a danger that untested technological filters will become de facto gatekeepers of what the defence ever sees. Unless defence practitioners have meaningful input into search strategies – and access to their own tools – the ‘black box’ risk is substantial. Secondly, centralised governance can entrench ‘proportionate’ practices that, while rational at system level, normalise more limited review in individual cases, especially where material is seen ex ante as peripheral. Thirdly, the emerging use of AI tools elsewhere in the criminal justice system illustrates why ‘technological efficiency’ cannot be treated as normatively neutral. Durham Constabulary's Harm Assessment Risk Tool (HART), a machine-learning model trained on custody and offending data, has been used to classify arrestees as high, medium or low risk of reoffending, informing decisions about charge versus diversion into the Checkpoint scheme. ¹¹³ Academic analysis of HART highlights the opacity of its modelling, the inclusion of variables such as postcode that track deprivation, and the risk that apparently neutral outputs mask and reproduce existing patterns of over-policing. ¹¹⁴ Civil-society scrutiny via Freedom of Information requests suggests that more than 12,000 people were profiled using the tool between 2016 and 2021, with thousands deemed ‘high risk’ and therefore likely to have been charged rather than offered rehabilitation, yet defendants had no meaningful way of understanding, let alone contesting, how their risk scores were produced. ¹¹⁵ In fair-trial terms, this is a classic equality-of-arms concern: AI-inflected decisions at the charging and diversion stage alter the stakes of prosecution without any corresponding defence visibility of the underlying model, training data or error rates. ¹¹⁶

A similar pattern is visible in the deployment of facial-recognition and risk-scoring systems. In Bridges v Chief Constable of South Wales Police, ¹¹⁷ the Court of Appeal held that the force's use of live automated facial recognition technology was unlawful, citing ‘fundamental deficiencies’ in the legal framework, inadequate safeguards and a failure properly to assess equalities impacts. ¹¹⁸ Commentators have treated Bridges as emblematic of a wider governance gap: powerful biometric AI adopted in the name of efficient targeting, with privacy, discrimination and oversight concerns addressed only ex post. ¹¹⁹ Comparative experience with algorithmic risk scores reinforces the point. ProPublica's analysis of the COMPAS tool in US sentencing suggested that Black defendants who did not reoffend were nearly twice as likely as white non-recidivists to be falsely labelled ‘high risk’, prompting a technical literature which concedes that different statistical fairness criteria cannot be simultaneously satisfied and that trade-offs are inherently value-laden. ¹²⁰ For present purposes, what matters is not the resolution of that debate, but its procedural implication; when opaque, proprietary systems influence policing, diversion, surveillance or sentencing, defence lawyers face profound difficulties in obtaining disclosure about model design and performance, let alone mounting an informed challenge to their use. ¹²¹

Reasonable Lines of Enquiry: The Attorney General's Guidelines and the narrowing debate

A key axis of the report is the concept of ‘reasonable lines of Enquiry’ under s.23 CPIA 1996 and the AG Guidelines. The 2022 Guidelines, updated again in 2024 to address digital material, set out factors investigators must consider when deciding what digital material to review, emphasising lawfulness, necessity and proportionality, and reiterating that all reasonable lines should be pursued. They explicitly acknowledge tensions between privacy and fair-trial rights, but insist that where the two conflict, the fair-trial right is ‘absolute’ and must prevail. ¹²² Academic commentary has already raised concerns that the AG Guidelines, while formally restating fairness principles, in practice give investigators substantial discretion to construe digital reasonableness narrowly, particularly in resource-strained contexts. Griffiths argues that the Guidelines signal a further step away from a rights-based conception of disclosure towards a managerial one, by framing decisions in terms of proportionality and impact on victims and witnesses. ¹²³ Fisher's review sits squarely within this trajectory. On one hand, he is emphatic that the core CPIA test should not be diluted: investigators must still identify and follow all reasonable lines of Enquiry, including those pointing away from the suspect, and the prosecution must disclose anything capable of undermining its case or assisting the defence. He rejects calls for a fundamentally different statutory test for digital material and suggests that additional guidance be created that encompasses best practices in the digital age. ¹²⁴ On the other hand, he endorses the AG Guidelines’ proportionality-driven approach and urges further clarification of what ‘reasonable’ means in digital contexts. This includes recognising that it may be reasonable not to download or review entire devices where targeted enquiries suffice; that investigators can legitimately sample or use keyword searches rather than exhaustive review; and that privacy and resource considerations can shape what is ‘reasonably’ pursued. ¹²⁵ The managerial character of this move is clear: ‘reasonableness’ is recast as a balancing exercise between fairness, privacy and resource constraints, with investigators and prosecutors guided by high-level principles and local protocols. The adversarial worry is that, in practice, resource and workload pressures will dominate, leading to routine under-sampling and to a culture in which broad defence requests for disclosure are reframed as unreasonable or ‘speculative’.

The Court of Appeal has grappled directly with these digital pressures in an attempt to supply workable standards. In R v R and Others, ¹²⁶ a high-revenue fraud prosecution involving terabytes of seized material, the Court of Appeal stressed that the CPIA duty does not require exhaustive review of every file on every device; rather, investigators must adopt proportionate search strategies – using keywords, sampling and iterative defence engagement – to identify material which might reasonably undermine the prosecution or assist the defence. Later decisions in McPartland and Grant ¹²⁷ and Bater-James ¹²⁸ and Sultan Mohammed ¹²⁹ refined this approach in the context of complainants’ mobile phones and social-media accounts, insisting that there is no blanket obligation to seize and interrogate digital devices in every sexual offence case, and that any intrusion must rest on a properly articulated ‘reasonable line of enquiry’ grounded in the issues in dispute rather than speculative trawling. These cases, when read together, embody the same managerial logic that underpins Fisher. They treat digital disclosure as a problem of project management and proportionality – search terms, scoping exercises, triage tools, and negotiated plans – while leaving largely unaddressed the asymmetry in capacity between state agencies and typically under-resourced defence teams confronted with the output of those processes. The jurisprudence offers a vocabulary of ‘reasonable lines of enquiry’ and ‘targeted’ review, but it is largely silent on how equality of arms is to be preserved once terabytes of material have been gathered, sifted and presented under conditions where the defence may have only a few funded hours to respond. In that sense, appellate guidance has helped normalise a digital disclosure settlement that is formally compliant with Article 6, but substantively fragile when measured against the adversarial ideal of meaningful, properly resourced defence participation.

This intersects directly with the reconfigured role of the defence under the AG Guidelines and CrimPR. Defence practitioners are increasingly expected to identify, at an early stage, specific lines of Enquiry and categories of material they say should be reviewed or disclosed, often before they have seen much unused material at all. ¹³⁰ In theory, this enhances focus and cooperation. In practice, it burdens the defence with a quasi-managerial role in triaging digital material, without commensurate resources.

The Intensive Disclosure Regime: a Bespoke Pathway for the hardest cases

A headline structural reform is Fisher's Intensive Disclosure Regime (IDR) for ‘serious, complex or otherwise voluminous’ cases. ¹³¹ The premise is simple: exceptionally data-heavy prosecutions overwhelm ‘business as usual’ case management, so disclosure must be handled as a discrete, judge-managed issue with continuity of oversight. The IDR is not an automatic label; it is a discretionary designation a judge can make at the PTPH after hearing representations from both the prosecution and defence. ¹³² If a case is designated as IDR, several consequences follow. Firstly, the prosecution must file a revised Disclosure Management Document (DMD) that includes full particulars of any advanced technology (configuration and operation) used or proposed for data management and review. ¹³³ Second, the defence is obliged to file a Response to the DMD (RDMD) commenting on trial issues, agreeing or contesting the approach of prospection, where they can propose additional categories for review, and critically, addressing search methods (e.g., search terms, filters, third-party material). ¹³⁴ Third, there will be a Disclosure Management Hearing (DMH), this heading has the ultimate goal of streamlining the process, albeit by adding an additional hearing. At the PTPH the court fixes a timetable for DMD/RDMD service and sets a DMH to resolve disclosure issues, which can be conducted in person or remotely. ¹³⁵ These are not bureaucratic niceties. By forcing both parties to articulate their approach to finding exculpatory or undermining material, the IDR creates an audit trail against which fairness can later be tested. That in turn allows proportionate use of technology to be explained and, if necessary, revised – arguably, a pragmatic response to past episodes in which misunderstanding of software features contributed to disclosure error. ¹³⁶

Fisher envisages the court as an active guarantor of the IDR's integrity. Judges should be able to order further particulars of the disclosure strategy, direct revisions to a DMD, and enforce the timetable to keep issues moving. ¹³⁷ The Review also recommends consolidating the currently scattered guidance (AGG; judicial protocols) into a single Consolidated Guidance document – referenced in legislation and expressly covering the IDR pathway, thus meaning that everyone is reading from the same script. ¹³⁸ A new central GOV.UK landing page and archiving of outdated versions would remove avoidable confusion. ¹³⁹ Finally, because IDR asks more of everyone, Fisher links it to funding (Rec 41) and to oversight/evaluation. The proposed Disclosure Scrutiny Joint Committee should monitor implementation, while Fisher also floats appointing a single figure to keep the system's ‘disclosure performance’ under regular review. ¹⁴⁰ He also calls for quantitative data on IDR progression to test whether the pathway delivers the promised benefits. ¹⁴¹

Defence capacity, equality of arms and PCE

Fisher's Review acknowledges, but does not dwell on, the resource asymmetry between prosecution and defence in managing digital disclosure. The Home Office announcement and subsequent progress updates emphasise consultation with ‘defence professionals’, but the bulk of recommendations are addressed to investigators, prosecutors and central agencies. For a system already operating with a hollowed-out defence legal aid scheme, this is a glaring omission. Pre-Charge Engagement envisages a structured channel through which defence practitioners can engage with investigators before charge, suggesting lines of Enquiry, commenting on digital material and making representations about charge and disposal. ¹⁴² In principle, it is precisely the kind of early, cooperative mechanism Fisher and the AG Guidelines favour.

In practice, survey evidence from 40 defence practitioners indicates that PCE is ‘rarely used, poorly understood and structurally unfunded’. ¹⁴³ Respondents reported that meaningful participation often requires in excess of 15 hours work – reviewing material, taking instructions, drafting representations and negotiating with investigators – which would exceed the level of remuneration of offered by the scheme. ¹⁴⁴ The result is a pattern of ‘empty gesture’ proceduralism: a formally available mechanism that, in reality, is confined to a minority of cases where clients can pay privately or where particularly motivated practitioners absorb the cost. Against that background, Fisher's emphasis on more detailed early discussions of ‘reasonable lines of Enquiry’, search parameters and digital scope risks reinforcing these structural inequalities. If ‘good practice’ involves early, cooperative agreement between investigators and defence on digital strategy, but only the better-resourced defence teams can provide that input, the system risks hardwiring in a two-tier disclosure culture: one for complex, well-funded corporate and serious-crime defendants, another for ordinary legal aid suspects. Commentaries from defence firms suggest that some would welcome the opportunity for defence teams to engage closely with disclosure strategies but implicitly assume that parties have access to specialist e-disclosure tools and technical expertise. ¹⁴⁵ For the bulk of criminal defendants those assumptions do not hold. The equality of arms issue is thus less about the formal tests than about capacity. A digitally enabled Crown, supported by central boards, AI tools and updated guidance, faces a defence profession operating on thin margins, with limited access to comparable technology and with legal aid structures that incentivise swift resolution rather than deep engagement with unused material. In that setting, managerial disclosure reforms risk consolidating a model in which defence input is formally welcomed but practically truncated.

Interim assessment: Fisher

The Fisher Review is, in many respects, a thoughtful and sophisticated attempt to confront the problem of digital disclosure in a system already under strain. It rejects calls to dilute the ‘golden rule’ that the prosecution must disclose material which might reasonably be considered capable of undermining its case or assisting the defence and reiterates that ‘reasonable lines of Enquiry’ under the CPIA remain the statutory anchor rather than a discretionary gloss. By recognising that technology can enhance fairness as well as threaten it, and by explicitly linking disclosure failures to miscarriages of justice, the Review avoids the more overtly cost-cutting tone that has characterised some earlier managerial reforms. At the same time, the Review is unmistakably written within a managerial register. Its core solutions are institutional and procedural: a system of digital governance boards; revised CPIA Code and AGG to re-specify ‘reasonableness’ in proportionality terms; more assertive judicial case management of disclosure; and, for the most complex cases, an IDR built around DMDs, RDMDs and a dedicated DMH. ¹⁴⁶ These devices mirror, in the disclosure sphere, the broader move under the Criminal Procedure Rules towards a case-managed, ‘right-first-time’ process in which timeliness, throughput and the avoidance of failure demand are treated as central performance metrics. Fairness remains an important value, but it is increasingly framed as something to be reconciled with resource constraints, rather than as the starting point from which constraints must be justified. Viewed in isolation, the IDR has real potential as a mechanism for transparency and ex post accountability in high-volume digital cases. Requiring the prosecution to set out, in a revised DMD, the configuration and operation of any advanced technology used for data capture and review forces articulation of choices that would otherwise remain opaque: search terms, filters, sampling strategies and third-party coverage. Obligations on the defence to respond substantively, and on the court to hold a DMH where strategies can be probed and, if necessary, revised, create an audit trail against which equality of arms can later be assessed. ¹⁴⁷ Read alongside the Attoney General's Guidelines on Disclosure and the consolidated guidance structure Fisher envisages, the scheme offers a plausible institutional framework within which judges can move beyond general invocations of proportionality and engage with the concrete detail of digital search methodology.

However, once the IDR is located within the wider settlement traced in this article, its normative attractiveness looks more contingent. From a traditional adversarial perspective, robust digital disclosure is valuable not because it produces tidy management information, but because it makes possible a meaningful challenge to the State's account of the evidence. That, in turn, presupposes defence capacity: time to review unused material, technical competence to understand and critique search processes, and access to at least basic digital tools. As the empirical experience of PCE suggests, formal opportunities for early engagement can easily become ‘empty gesture’ ¹⁴⁸ proceduralism where they are not matched by legal aid structures that fund the work required. In legally aided cases, there is a serious risk that the IDR replicates that pattern: designation triggers additional defence ‘obligations’ on paper, but limited resources mean that responses are minimal, and ‘agreement’ with prosecution search strategies becomes a form of compliance theatre rather than an informed, adversarial choice. Fisher is not blind to these concerns. He links the IDR expressly to legal aid and funding, recommending that the Ministry of Justice and the Criminal Legal Aid Advisory Board review whether current arrangements adequately support early disclosure engagement and the new pathway, and that the PCE funding process be simplified. ¹⁴⁹ He also proposes mechanisms of oversight and evaluation, including a Disclosure Scrutiny Joint Committee and improved quantitative data – specifically, data on IDR progression – to test whether the pathway delivers ‘swift and fair’ resolution rather than simply re-labelling delay. ¹⁵⁰ Yet those safeguards remain, at this stage, programmatic. Their effectiveness will depend not only on ministerial willingness to invest in defence capacity, but also on how courts interpret their case-management powers under the Criminal Procedure Rules and the consolidated guidance when faced with defence arguments that digital proportionality has been pushed too far.

The European Court of Human Rights has repeatedly emphasised that equality of arms in disclosure is integral to a fair trial, condemning regimes in which non-disclosure or one-sided control over sensitive material prevents the defence from testing the prosecution case. ¹⁵¹ In the domestic context of England and Wales, domestic authority has likewise stressed that the ‘golden rule’ of disclosure is a constitutional safeguard, not a dispensable luxury. ¹⁵² The interim conclusion, therefore, is that Fisher's settlement is normatively attractive only if its managerial machinery is matched by adversarial safeguards: ring-fenced funding that enables defence participation in IDR cases; a judicial culture willing to treat digital disclosure as a site of genuine contest rather than as a box-ticking exercise; and oversight arrangements with the capacity to call out practices that erode equality of arms. If the IDR becomes a vehicle for transparency, rigorous scrutiny and properly resourced defence involvement, it could improve both efficiency and fairness in complex digital cases. If not, it risks hardwiring the managerial settlement: a modernised process that protects throughput by shifting work and risk onto a structurally constrained defence profession, leaving the ‘golden rule’ of disclosure as a principle proclaimed but unevenly realised in practice.

Efficiency versus adversarialism

Set against the traditional adversarial model outlined earlier, the Leveson–Fisher settlement reorders the values that structure criminal procedure. On the ideal-type account, adversarial justice is organised around party control of proof, a relatively passive judge, a robust presumption of innocence and equality of arms in access to evidence, with contested jury trial as the presumptive forum for serious crime. ¹⁵³ Delay matters because it undermines accuracy and legitimacy, not because it impairs throughput. Within that framework, the defence lawyer's core duty is to subject the prosecution case to rigorous forensic challenge; cooperation with the court is instrumental to that end, not an independent objective. By contrast, Leveson and Fisher embed efficiency – as timeliness, throughput and the avoidance of ‘failure demand’ – as a quasi-constitutional value. Leveson treats the criminal courts as a triage system: cases should be diverted out of court wherever ‘appropriate’; mid-range work should be absorbed by strengthened magistrates’ courts and a new CCBD; jury trials should be rationed; appeals from the magistrates’ court should be subject to permission filters; and guilty-plea discounts should be sharpened through more aggressive Goodyear practice and a higher maximum discount at the first opportunity. ¹⁵⁴ Fisher, for his part, reconstructs disclosure as a managed, technologically assisted workflow, overseen by boards and IDR hearings, in which proportionality and resource optimisation are central. ¹⁵⁵

In both Reviews, the defence lawyer is recast as an efficiency partner. Advising on OOCDs, restorative pathways, deferred prosecutions and rehabilitative interventions becomes part of the ordinary repertoire of defence practice. So too does participation in early digital scoping, DMD/RDMD cycles and IDR hearings. Refusal to engage, or insistence on a more traditional adversarial stance (e.g. declining to commit to trial issues before sufficient disclosure, or resisting pressure for early pleas), risks being framed as professionally unreasonable or contrary to the CrimPR's overriding objective. ¹⁵⁶ From a traditional adversarial perspective, this is a significant normative shift. The price of efficiency is not merely a change in forums and procedures, but a re-orientation of defence ethics and expectations. As McConville and Marsh argue the cumulative effect of sentencing discounts, case-management pressures and structural incentives can be to render insistence on trial an ‘expensive’ and culturally abnormal choice. ¹⁵⁷ The Leveson–Fisher settlement deepens that dynamic: the rational, system-conforming defence lawyer is now one who steers clients towards early resolution and participates constructively in digital triage, even where that may limit opportunities for later challenge.

Disparity and the uneven defence burden

The managerial defence role also has distributional consequences. The burdens generated by diversionary programmes, digital disclosure governance and IDR participation do not fall evenly across the profession or across defendants. Large firms, CPS units and specialist agencies are better placed to absorb the demands of dashboards, AI-assisted review and intensive early engagement: they can invest in dedicated disclosure teams, case management software and training. High-street legal-aid firms, duty solicitors and small chambers cannot. The existing work on PCE already points to a two-tier pattern: meaningful early engagement is feasible where clients can pay privately, or where serious or complex cases justify exceptional investment; in routine legally aided work, the scheme is either ignored or engaged with at a minimal level. ¹⁵⁸ Conceptually, the IDR promises transparency and auditability: the defence can scrutinise and contest search parameters, challenge over-reliance on sampling or keyword searches and insist on revisions where digital strategy is deficient. But those benefits presuppose a defence with time, technical competence and access to digital tools. Without ring-fenced funding and support, IDR hearings risk becoming another environment in which well-resourced corporate or serious-crime defendants can exploit the regime. In contrast, ordinary defendants experience only a formalised, court-endorsed version of the prosecution's preferred digital strategy.

When combined with Leveson's proposals to shift more work into magistrates’ courts and hybrid bench-trial forums, and with documented concerns about racial and socio-economic disparities in non-jury settings, the effect is to skew the playing field further away from the adversarial ideal of equality of arms. The system that emerges is stratified: at the top, a small subset of cases in which resource-rich defence teams can participate fully in diversion negotiations and digital governance; at the bottom, a mass of volume cases in which managerial expectations are nominally applied but practically hollow, with defendants channelled through streamlined forums and processes in which adversarial challenge is structurally weak.

Injustice Risk and the Hallowing out of the trial

Finally, the Leveson–Fisher settlement has implications for miscarriage risk and the future of adversarial trial. Fisher's review is explicit that disclosure failures have caused serious miscarriages of justice and that unused material has historically been ‘the lifeblood’ of effective defence challenge. ¹⁵⁹ Yet his solutions remain predominantly managerial: consolidating guidance, clarifying ‘reasonable lines of Enquiry’ in proportionality terms, institutionalising digital governance and introducing the IDR for the most complex cases. Leveson, meanwhile, proposes reforms that make fully contested jury trials rarer by design: more diversion, more summary and bench-trial disposal, steeper trial taxes through enhanced plea discounts and reduced appellate opportunities. From a traditional adversarial standpoint, this is a worrying combination. Digital complexity increases the need for robust adversarial testing – of software, search strategies, data integrity and interpretive assumptions – at exactly the moment when trials are being thinned out and displaced into forums less conducive to searching challenge. The risk is not only that individual errors will go undetected, but that a culture of ‘managed’ justice will take hold in which the legitimacy of outcomes rests more on the appearance of procedural compliance (boxes ticked, dashboards green, guidance followed) than on substantive scrutiny of the prosecution case. The PCE experience again provides a cautionary note. A scheme intended to prevent weak or abusive prosecutions has become, for many defendants, an ‘empty gesture’ whose availability on paper is not matched by real adversarial protection in practice. ¹⁶⁰ There is no inherent reason to assume that diversion programmes, IDR pathways or CCBD trials will fare differently if they are built on the same under-resourced defence foundation. Without serious reinvestment in defence legal aid, digital capacity and time for pre-trial investigation, the managerial defence lawyer imagined by Leveson and Fisher risks becoming a constitutional fiction – invoked to legitimate efficiency-driven reforms but largely absent from the everyday reality of criminal practice. On that view, the settlement crystallised by the two reviews is not simply an evolution of adversarial justice but a further slide toward a managed, administrative model of criminal process in which adversarial rights remain formally available yet are increasingly marginal in practice.

Towards an adversarial coherent reform agenda

Leveson and Fisher crystallise an efficiency-driven, managerial settlement that sits uneasily with traditional adversarial commitments. The aim here is not to roll back every efficiency measure, but to sketch an ‘adversarial coherent’ version of efficiency – one that respects the presumption of innocence, equality of arms and the centrality of public, contested adjudication in serious cases. Leveson's out-of-court agenda and Fisher's endorsement of early engagement and ‘reasonable lines of Enquiry’ are not doomed from an adversarial perspective. Diversion, triage and managerial steering can be legitimate, but only if certain pre-conditions are met.

First, there must be serious reinvestment in defence legal aid, ring-fenced for early engagement and diversion work. Existing police-station and Pre-Charge funding structures remunerate only a small fraction of the work required. Secondly, minimum disclosure standards must apply before any out-of-court or pre-charge disposal. From an adversarial standpoint, it is intolerable for decisions with long-term consequences – where conditional cautions, OOCDs or ‘no further action’ offers are contingent on admissions – to be taken on the basis of bare allegation or skeletal summaries. Thirdly, there must be robust safeguards around voluntariness and the trial tax. Leveson's suggestion of enhancing that discount to 40% and intensifying Goodyear practice risks further increasing the ‘trial tax’ for those who insist on contesting allegations. ¹⁶¹ An adversarial coherent approach would resist any upward ratcheting of plea discounts. McConville and Marsh's work on guilty pleas underscores the danger of treating ‘efficiency’ gains from early resolution as normatively neutral. ¹⁶² Finally, there must be systematic monitoring for disparity would be necesssary – critics of OOCDs and community resolutions have repeatedly pointed to disparities for marginalised suspects in their use. ¹⁶³ If diversion is to be a central plank of an efficiency agenda, data on OOCDs, PCE-based decisions and pre-charge disposals must be collected and published, disaggregated by offence type, race, gender, age and representation status, with a remit for the CPS Inspectorate or an independent body to report on patterns. Absent such transparency, an ostensibly benevolent diversionary regime risks becoming another site where the articulate and resourced secure quasi-non-criminal outcomes, while marginalised defendants remain in the fast lane to conviction.

Re-centring the jury and transparent adjudication

The second pillar of an adversarial coherent reform agenda concerns the adjudicative forum. Leveson's proposals to expand magistrates’ jurisdiction, create a CCBD and normalise judge-alone trials in complex fraud are explicitly justified in efficiency and proportionality terms. ¹⁶⁴ From an adversarial perspective, they must be reframed as exceptional deviations from the baseline expectation that serious criminal allegations are tried by a jury of lay citizens. ¹⁶⁵ An adversarial consistent settlement would adopt three principles. First, a jury trial as the normative default for serious crime. For offences carrying substantial custodial penalties or significant stigma – serious violence, sexual offences, protest-related public order and serious dishonesty – the starting point should remain trial by jury. ¹⁶⁶ Any departure (CCBD or judge-alone) should be treated as exceptional and require a compelling justification on the facts of the case, not merely the class of offence. This would align with with Denning’s classic characterisation of the jury as ‘the lamp that shows that freedom lives’, and with Campbell's account of its constitutional role in constraining state power. ¹⁶⁷ Second, strict temporal and substantive limits on non-jury forums. If CCBD or judge-alone trials are to be retained at all, they should operate as temporary backlog-reduction measures subject to sunset clauses and periodic parliamentary renewal, rather than as permanent structural reconfigurations. ¹⁶⁸ Their jurisdiction should be tightly confined to specified categories where there is strong evidence that a jury trial adds no meaningful value (e.g. certain high-volume, low-dispute regulatory offences), and expressly excluded for politically sensitive or pattern-based crimes where community legitimacy and collective judgment are particularly salient. Concerns raised by APPEAL and others about racial disparities in non-jury settings underline the need for caution. ¹⁶⁹ Thirdly, enhanced transparency and appeal safeguards in non-jury adjudication. Where CCBD or judge-alone trials do occur, adversarial honesty requires more than assurances of equivalent fairness. Written, reasoned judgments should be mandatory, and appeal routes should remain as broad as for jury trials. Robust collection and publication of data on conviction rates, sentencing patterns and disparity by race and gender in CCBD and judge-alone cases are essential if these forums are not to become opaque, administrative spaces. ¹⁷⁰ Re-centring the jury in this way does not deny the reality of backlog or the need for proportionate management of court time. It insists, rather, that efficiency gains must not be bought at the price of hollowing out a core adversarial institution whose value is only fully visible when it is lost. ¹⁷¹

Re-designing disclosure around adversarial equality

The third strand of an adversarial coherent agenda is redesigning disclosure to ensure equality of arms in a digital age. Fisher is right that the CPIA 1996 regime is under acute strain from digital proliferation, and right to resist wholesale dilution of the ‘reasonable lines of Enquiry’ standard. But a sustainable settlement cannot be built solely on managerial triage directed by the State. It must address the structural asymmetry between a digitally equipped prosecution and a chronically under-resourced defence. This article contends that there are three potential avenues that ought to be explored:

Defence lawyers should have greater direct access to seized digital material, not merely to prosecution-selected extracts. Griffiths’ critique of the Attorney General's Guidelines as embedding a managerial, proportionality-driven conception of reasonableness underlines the importance of ensuring that defence-proposed lines of Enquiry are not dismissed as ‘fishing expeditions’ simply because they are inconvenient. However, Fisher resisted any temptation to offer the defence the ‘keys to the warehouse’. ¹⁷²

Conclusion

This article has treated the Leveson and Fisher Reviews as a linked moment rather than two discrete technocratic exercises. Read together, they embody an efficiency-first settlement for criminal process in England and Wales. Diversion, forum re-allocation, intensified plea incentives, digital triage and tiered disclosure standards are presented as necessary responses to crisis. Yet, when mapped against a traditional adversarial model built around party-led contest, robust disclosure, lay adjudication and equality of arms, the settlement sits uneasily. The formal right to zealous defence representation remains, but the defence lawyer is operationalised within managerial frameworks that treat time, cost and throughput as governing constraints. Instead of a predominantly partisan advocate whose primary obligation is to put the prosecution to proof and advance the client's interests, Leveson and Fisher presuppose a case-progression actor who priortises throughput over open-ended contest.

The likely distribution of burdens and benefits should also trouble those concerned with miscarriages of justice. In a system that tiers both forum (magistrates, CCBD, judge-alone) and disclosure (ordinary vs ‘intensive’ regimes), it is easy to imagine a familiar hierarchy: complex, well-resourced cases enjoying structured digital governance and genuine adversarial participation; high-volume, legally aided work experiencing only the pressures – earlier pleas, thinner disclosure, fewer jury trials – without the promised safeguards. Efficiency gains will not be experienced evenly. There is a real danger that managerial reforms will harden into a two-track criminal process in which the most marginal defendants bear the greatest procedural austerity. None of this suggests that delay, digital overload or the human cost of backlogs can be ignored. The question, rather, is what must be non-negotiable in any reform: adequately funded defence legal aid, particularly for early engagement; disclosure rules and digital practices designed around equality of arms rather than prosecution viability; and a renewed commitment to jury trial as the normative default for serious crime. Efficiency measures that survive those filters may be compatible with adversarial justice. Those that do not are better understood as attempts to manage decline under conditions of austerity than as principled reforms.

Ultimately, the choice is not between romanticising an adversarial past and embracing an inevitable managerial future. It is between allowing efficiency to crystallise, by default, into a new constitutional value that silently reorders rights, or confronting openly the trade-offs involved. By placing Leveson and Fisher in dialogue with doctrinal, empirical and theoretical work on managerialism, this article has tried to make those trade-offs visible. The next phase of reform will show whether policymakers are prepared to fund and design a criminal process that remains recognisably adversarial – or whether ‘efficiency without adversaries’ becomes the new normal.