UK Reviews

MANAGEMENT BRIEF – Produced using The Weavers

UK Major Reviews and Public Inquiries:

What the current process produces, what it misses, and how a systems-based approach would change both the question and the answer


How this brief was produced This brief applies a systems-based analytical framework to the question of how the UK conducts major reviews and public inquiries. The framework was applied in its entirety — every element, every insight accumulated from previous applications — to produce analysis that the conventional framing of this question does not reach. The brief follows the same pattern established in a previous application of this approach to organisational learning: the conventional answer is set out first, in full and fairly; then the systems analysis is applied; then the combined answer is drawn together into a set of structural enhancements. The question was not set externally. It emerged from the observation — consistent across every major UK inquiry since 2000 — that the same structural failures recur in different towers, and the reviews that examine them are themselves subject to the same structural failures they are investigating. A system cannot examine its own blind spots using only its own vocabulary.

How this brief was produced

This brief applies a systems-based analytical framework to the question of how the UK conducts major reviews and public inquiries. The framework was applied in its entirety — every element, every insight accumulated from previous applications — to produce analysis that the conventional framing of this question does not reach. The brief follows the same pattern established in a previous application of this approach to organisational learning: the conventional answer is set out first, in full and fairly; then the systems analysis is applied; then the combined answer is drawn together into a set of structural enhancements.

The question was not set externally. It emerged from the observation — consistent across every major UK inquiry since 2000 — that the same structural failures recur in different towers, and the reviews that examine them are themselves subject to the same structural failures they are investigating. A system cannot examine its own blind spots using only its own vocabulary.

This brief is for senior leaders, policy practitioners, and anyone responsible for commissioning or contributing to reviews of serious institutional failure. No specialist background is assumed.

David Sutton CITP MBCS | April 2026

SECTION 1

How UK Major Reviews Currently Work

The UK has a well-developed apparatus for reviewing serious institutional failures. Public inquiries operate under the Inquiries Act 2005. Serious case reviews, rapid reviews, independent programme assurance exercises, parliamentary select committee investigations, and National Audit Office value-for-money studies each follow their own frameworks. The machinery is substantial, experienced, and staffed by people of considerable ability. The outputs are often detailed and analytically impressive.

The outcomes are, with notable exceptions, disappointing in a specific and consistent way: the failure patterns that major reviews identify recur in subsequent organisations, programmes, and institutions — often in almost identical form — without the benefit of what the previous review produced.

The standard process

A major public inquiry typically proceeds through five stages. Appointment of a chair (usually a senior judge or equivalent), publication of terms of reference, taking of evidence (written and oral, often over years), production of an interim report where appropriate, and a final report with recommendations. The government is then expected to respond to the recommendations, typically within a fixed timeframe.

The framing of terms of reference is the most consequential decision in any inquiry. It determines what the inquiry is authorised to examine, whose evidence is heard, and what kinds of conclusion the report is permitted to reach. Terms of reference are drafted before the inquiry has heard any evidence. They reflect what the commissioning body — typically the government — believes is the relevant scope. This is the most important structural feature of the current process: the question is set before the answer is sought, and the question is set by an institution that may itself be a party to the failure being examined.

The pattern the current process consistently produces

Across the major UK reviews of the past two decades — Grenfell Tower, Infected Blood, Mid Staffordshire NHS, Hillsborough, Post Office Horizon, Birmingham City Council — several consistent features appear in both the reviews and their outcomes:


6 years Infected Blood Inquiry (2018–2024) — took evidence from 2019 to 2023	The failure it examined began in the 1970s. The knowledge required to prevent it was available by the 1980s. The time between what was knowable and accountability being established spanned four decades and three inquiries.


£140m Estimated cost of the Infected Blood Inquiry to March 2024	Equivalent to a substantial fraction of the compensation fund it was convened to justify. The process itself consumed resources that the victims the inquiry was designed to serve could have received earlier.


5 Major institutional failures since 2000 sharing an identical root pattern	Post Office Horizon, NHS Mid Staffordshire, Grenfell Tower, Birmingham City Council, Infected Blood — in each case: accurate information about the failure was not reaching those with the authority and inclination to act on it, for years before the failure became undeniable.

The four consistent features of the current process

1. The attribution frame

Reviews that are convened to examine institutional failure typically begin with a frame that prioritises attribution — identifying the decisions, the decision-makers, and the points at which different choices could have produced different outcomes. This is not inherently wrong. Accountability for serious failures matters. But the attribution frame creates a systematic bias: the most visible cause of a failure is almost always an individual decision or a decision-maker’s action, and the review’s grammar makes this the natural resting place for the analysis. The deeper causes — the organisational practices, the methodological assumptions, the information flows, the cultural norms, the invisible barriers between those who knew and those who needed to know — require a different kind of investigation to surface, and are frequently not surfaced at all.

The media frame reinforces this. Public inquiries generate news through the examination of witnesses. The most newsworthy examinations are those in which senior individuals are questioned about their decisions. The public experience of an inquiry — and therefore the political experience — is an attribution exercise. Systemic causes are less visible, less narratively compelling, and less satisfying as accountability. They also produce less clear-cut recommendations, because they require structural rather than individual remedies.


The result: reviews frequently produce findings that accurately describe what went wrong at the level of individual decision-making, while leaving the structural conditions that made those decisions likely or inevitable largely unexamined. The same structural conditions then produce the same failure in a different institution in a different decade.

2. The retrospective frame

All major reviews are retrospective. They examine what happened, not what is about to happen. This is appropriate — accountability requires examining what occurred. But the retrospective frame creates a second systematic limitation: by the time a review is commissioned, the failure is complete and the consequences are visible. The transition boundary — the point at which governance needed to concentrate before trajectories committed — is already months or years in the past. The review examines a fait accompli.

This means that the most important moment in the failure — the point at which small choices were compounding into large consequences, at which intervention was still cheap, at which the design decisions were still uncommitted — is precisely the moment that the review process is least equipped to reach. Reviews reach it only through backward reconstruction, which depends on documentary evidence that survivors have had the opportunity to manage, and testimony from participants whose recollections are inevitably shaped by knowledge of what followed.


The result: reviews produce accounts of how failures became catastrophic. They are structurally unable to produce accounts of how the same failures could have been caught and corrected at the point when correction was still straightforward. The learning that would prevent recurrence is available at the development stage. Reviews arrive at the post-catastrophe stage.

3. The evidence filter

The evidence that reaches an inquiry is shaped by the same barriers that shaped the information flows during the failure itself. Formal testimony through legal processes favours those with access to legal representation, those whose accounts fit the inquiry’s terms of reference, and those whose organisational positions give them the authority to produce documentary evidence. Those closest to the failure — the frontline workers, the practitioners, the members of the public who experienced the failure directly — are heard as witnesses, not as investigators. Their knowledge of the actual texture of the failure, the workarounds, the things that were technically compliant but practically useless, the signals that were sent but not received — this knowledge is filtered by the same processes that filtered it during the failure itself.

The Infected Blood Inquiry explicitly acknowledged this — its terms of reference stated that those infected and affected would be at the heart of the inquiry, and considerable effort was made to ensure their evidence was heard. Yet the compensation scheme that followed was designed without their participation, and within days of the final report’s publication, the community was expressing ‘anxiety, confusion and distress’ at a scheme that had not involved the people it was designed to serve. The vine had regrown between the inquiry and its implementation.


The result: the evidence base of a review is systematically weighted toward those whose institutional positions gave them visibility during the failure, and away from those whose ground-level knowledge was most precise and least heard. The review reproduces, in its evidence-gathering, the information barriers that produced the failure it is investigating.

4. The implementation gap

Reviews produce recommendations. Recommendations require government responses. Government responses describe intentions. Implementation requires sustained effort across multiple departments, agencies, and organisations, often across political cycles. The Grenfell Tower fire occurred in 2017. The Phase 2 Inquiry report was published in September 2024. The implementation of its recommendations — concerning building safety across the entire country — has barely begun. The Infected Blood Inquiry recommended compensation in May 2024 and by July 2025 was publishing an Additional Report because the compensation scheme had been designed without the community’s involvement and was taking too long to reach people who were dying.

The implementation gap is not a failure of intention. It is structural. The inquiry produces findings in a concentrated moment of public and political attention. Implementation happens in the subsequent years, when attention has moved elsewhere, when the political cycle may have turned, when the officials who understood the inquiry’s nuances have moved to different roles, and when the organisations responsible for implementation are the same organisations whose structural features produced the failure in the first place.


The result: the knowledge produced by a review is sealed inside the tower that produced it. The subsequent towers — the next government department, the next organisation in the same sector, the next political administration — start from the same point. The pattern repeats because the network that would carry learning to where it could prevent repetition does not exist.

SECTION 2

What a Systems-Based Analysis Adds

The conventional critique of UK major reviews — that they take too long, cost too much, and produce recommendations that are not implemented — is accurate but incomplete. The systems analysis identifies the structural reasons why each of these problems persists despite genuine effort and considerable capability, and what would need to change for each to be resolved.

The analysis draws on the full range of insights accumulated through applying a cross-domain systems framework to institutional failure, AI governance, NHS strategy, public sector transformation, and democratic resilience. Six structural additions emerge that the conventional critique cannot reach.

Addition 1: The inquiry itself is a tower

The most important structural insight the systems analysis produces is this: a public inquiry is itself an example of the closed, vertically optimised architecture that it is investigating. It is convened, staffed, and concluded within its own boundary. It is I-aligned — internally coherent, optimised for the production of its own output, and disconnected from the network of organisations, practitioners, and community knowledge that holds the most precise understanding of the failure it is examining.

This is not a criticism of the people conducting inquiries. It is a structural observation. The inquiry operates under a terms of reference that closes around the question before the evidence is heard. It gathers evidence through processes designed for legal rather than systemic truth-finding. It produces a report that is internally consistent with its terms of reference. It then closes. The knowledge it has accumulated — the testimony, the documents, the understanding of the failure that has built up across years of examination — does not transfer to a network. It transfers to the public record, which is searchable but not systematically connected to the institutions whose next decisions will determine whether the same failure recurs.

The inquiry that concluded in 2024 does not know what the planning department is deciding in 2025. The building safety regulations emerging from Grenfell do not automatically reach the housing association whose procurement practices are now repeating the same pattern with different materials. The thread between the inquiry’s findings and the organisations whose decisions matter most is broken — not by malice, but by the same invisible barriers that produced the failure in the first place.


The structural enhancement: Redesign inquiries as network-generating exercises, not tower-completing ones. The inquiry’s findings should return to a living, sector-wide network that carries them to the organisations whose decisions matter most — before those decisions are made. This requires an active, structurally funded knowledge transfer mechanism that operates after the inquiry closes, not the passive publication of a report that organisations may or may not read.

The structural enhancement:

Redesign inquiries as network-generating exercises, not tower-completing ones. The inquiry’s findings should return to a living, sector-wide network that carries them to the organisations whose decisions matter most — before those decisions are made. This requires an active, structurally funded knowledge transfer mechanism that operates after the inquiry closes, not the passive publication of a report that organisations may or may not read.

Addition 2: The question is set by the wrong people at the wrong time

Terms of reference are the most consequential decision in any inquiry. They are drafted before the inquiry has heard any evidence, by an institution — typically the government — that may itself be party to the failure. The structural problem is not bad faith. It is that the grammar of the terms of reference — what the inquiry is authorised to examine — determines what the inquiry can find. A terms-of-reference exercise conducted within the existing institutional framework will produce questions that the existing framework can formulate. It will not produce questions that the existing framework cannot formulate, because those questions are invisible from within the frame.

This is the inversion principle applied to public inquiries. Before the terms of reference close around a set of questions, an instrument must be used that can surface what those terms of reference will systematically exclude. That instrument cannot be another consultation within the same institutional framework. It must operate outside the existing grammar — engaging with the frontline practitioners, the community members, the people whose experience of the failure is most direct and least institutionally mediated — before the question is set.

The Infected Blood Inquiry’s Additional Report of July 2025 is a precise illustration of this failure. The compensation scheme — which was the implementation of the inquiry’s central recommendation — was designed without the participation of the people it was designed to serve. The community experienced it as something done to them, not with them. The inquiry had spent six years hearing from people who were infected and affected. The implementation process then proceeded in the same closed architectural mode that had produced the original failure: experts designing systems for the people they were serving, without asking those people what the system needed to do.


The structural enhancement: Add a pre-terms-of-reference stage to every major review: a structured engagement with the people closest to the failure — frontline workers, community members, practitioners — before the institutional framing closes around the questions. The engagement must ask not only what happened but what questions the current institutional framing cannot ask. Record those questions. Require that the final terms of reference demonstrate how each of them is addressed, or explicitly account for why it is not. The questions the institution cannot formulate are the ones most likely to prevent recurrence.

The structural enhancement:

Add a pre-terms-of-reference stage to every major review: a structured engagement with the people closest to the failure — frontline workers, community members, practitioners — before the institutional framing closes around the questions. The engagement must ask not only what happened but what questions the current institutional framing cannot ask. Record those questions. Require that the final terms of reference demonstrate how each of them is addressed, or explicitly account for why it is not. The questions the institution cannot formulate are the ones most likely to prevent recurrence.

Addition 3: The broken thread between evidence and implementation

The golden thread in any governance system runs from every major decision back to the principles it is designed to serve. In a public inquiry, the thread should run from implementation back through the recommendations, back through the evidence, back through the experience of those who suffered the failure, back to the founding principle that the system failed to honour. That thread is broken in the same place in every major UK review: between the publication of the report and the implementation of its recommendations.

The thread breaks here because it is cut by the same barriers that produced the failure. The organisations responsible for implementation are the same organisations whose structural features — their information barriers, their governance assumptions, their institutional cultures — were part of what produced the failure. The inquiry produces findings about those features. The organisations then implement those findings using the same features the inquiry identified as the problem. Implementation is assigned to the tower. The tower does what towers do: it processes the recommendation through its existing architecture and produces an output that is compliant with the recommendation in form and does not change the structural conditions that made the recommendation necessary.

This is not cynicism. It is a structural prediction. When an organisation with a systematic bias toward governance attribution is tasked with implementing a recommendation that requires it to examine its own practices, methodologies, and culture, it will attribute the cause of the failure to governance — and address governance. The practices, methodologies, and culture will remain. The vine grows back. The symptom returns in a different part of the garden because the root was never reached.


The structural enhancement: Separate the implementation authority from the organisations identified as party to the failure. Create an independent implementation body with a statutory mandate and a fixed lifespan — not a monitoring body that reports on what is being done, but an implementation body with authority to direct and resource the structural changes the review identifies. Require that the thread between each recommendation and the specific structural condition it addresses is traced, named, and made publicly visible. Where implementation cannot demonstrate how it addresses the named structural condition, require a public account of why not.

The structural enhancement:

Separate the implementation authority from the organisations identified as party to the failure. Create an independent implementation body with a statutory mandate and a fixed lifespan — not a monitoring body that reports on what is being done, but an implementation body with authority to direct and resource the structural changes the review identifies. Require that the thread between each recommendation and the specific structural condition it addresses is traced, named, and made publicly visible. Where implementation cannot demonstrate how it addresses the named structural condition, require a public account of why not.

Addition 4: Root cause analysis stops too early

Every major UK review identifies root causes. The root causes consistently named are: leadership failure, cultural failure, governance failure, and inadequate regulation. These are real. They are also incomplete in a specific and structural way: they are the root causes that are visible from within the institutional framework that commissioned the review. The root causes that are not visible from within that framework — the organisational practices and methodological assumptions that made leadership failure likely, the information barriers that made cultural failure invisible to those with the authority to address it, the specific mechanisms through which what the frontline knew failed to reach those making decisions — these require a different kind of investigation to find.

The soil beneath every major UK institutional failure contains the same set of objects: processes that were designed without understanding how the information they created would be used; definitions that drifted between functions without anyone noticing; practices that were technically compliant but operationally dysfunctional; and cultural norms in which recording became a compliance exercise rather than an honest account. These are not visible in the testimony of senior witnesses. They are visible in the testimony of people doing the actual work — and in the documentary record of what the system was actually producing, as distinct from what it was reporting that it was producing.

The Infected Blood Inquiry found systematic failures of leadership and governance. What it also found — buried in the detail — was that clinicians treating patients with infected blood were working within a knowledge environment that had been shaped by the same institutional assumptions that prevented accurate information about risk from reaching them. The root cause was not only that leaders made wrong decisions. It was that the soil through which information had to travel had been contaminated by decades of institutional practice that privileged reassurance over accuracy. Cutting that vine requires examining the soil — not only the decisions of the people at the top of the tower.


The structural enhancement: Require that every major review includes an Independent Programme Assurance component with a specific mandate to examine the practices, methodologies, and information culture that produced the failure — not only the governance decisions made on the basis of filtered information. This component should be conducted independently of the formal inquiry process, with access to the full documentary record and the testimony of frontline workers, and should report to the inquiry rather than to the government. Its findings should be required reading for those implementing the inquiry’s recommendations.

The structural enhancement:

Require that every major review includes an Independent Programme Assurance component with a specific mandate to examine the practices, methodologies, and information culture that produced the failure — not only the governance decisions made on the basis of filtered information. This component should be conducted independently of the formal inquiry process, with access to the full documentary record and the testimony of frontline workers, and should report to the inquiry rather than to the government. Its findings should be required reading for those implementing the inquiry’s recommendations.

Addition 5: The transition boundary before the failure is never examined

Every major institutional failure had a development phase during which the conditions that produced it were being established. This is the period during which the assumptions were embedded, the processes were designed, the information barriers were first created, and the governance structures were first set up in ways that would later prove inadequate. At this stage, the failure was not a failure — it was a design choice, a policy decision, a procurement decision, or an operational assumption that seemed reasonable at the time and would not appear unreasonable for years.

The contaminated blood used to treat haemophilia patients in the 1970s and 1980s was supplied through procurement processes that prioritised supply security over risk management, in a knowledge environment where the risks were contested and the institutional incentives ran strongly in the direction of continued supply. The design choices that made the failure possible were not made by malign actors. They were made by people operating rationally within an institutional framework that could not see what it could not see. The framework’s broken clock was still set to the time at which it was designed. It could not update.

No major UK review has yet developed a methodology for examining the development phase of an institutional failure with the same rigour it applies to the failure event and its immediate antecedents. The reason is structural: the development phase is far in the past, the documentary evidence is partial, and the people who made the relevant decisions are no longer in the roles they held. Examining the development phase requires a different investigative methodology — one that works backwards from the failure to the design choices that made it likely, and then asks what governance instrument at the design stage would have caught what the designers could not see.


The structural enhancement: Require that every major review includes a development-phase audit: a structured retrospective examination of the design choices, procurement decisions, and governance assumptions made during the period when the conditions for failure were first established. The audit should ask: at what point was this failure still a small, reversible design choice? Who had the authority and the information needed to make a different choice at that point? What governance instrument would have made the information available to them? The answers produce the most actionable recommendations — because they name what needs to change before the next version of the same failure is being designed, not after it has become irreversible.

The structural enhancement:

Require that every major review includes a development-phase audit: a structured retrospective examination of the design choices, procurement decisions, and governance assumptions made during the period when the conditions for failure were first established. The audit should ask: at what point was this failure still a small, reversible design choice? Who had the authority and the information needed to make a different choice at that point? What governance instrument would have made the information available to them? The answers produce the most actionable recommendations — because they name what needs to change before the next version of the same failure is being designed, not after it has become irreversible.

Addition 6: The democratic resilience dimension

The most dangerous failure mode the systems analysis identifies is not the institutional failure itself. It is the failure of the feedback loop between reality and governance — the mechanism by which those in authority come to know that something is wrong, well enough and soon enough to act.

In every major UK institutional failure, this feedback loop broke down long before the failure became undeniable. The people in authority were receiving information about the performance of the institution. The information was accurate in the sense that it accurately reflected what the institution’s reporting systems were producing. It was systematically inaccurate in the sense that the institution’s reporting systems had been shaped by the same cultural practices, information barriers, and incentive structures that were producing the failure. The information that leaders received made the institution look better than it was, because the processes that created the information were designed by people who understood, consciously or unconsciously, that the alternative was to surface things that nobody in the institution had the authority or the will to address.

This dynamic is now being compounded by the increasing use of AI tools in institutional performance reporting. An AI system trained on the filtered, curated information that institutions produce will generate outputs that amplify the institutional narrative — making deteriorating conditions appear as normal variation, flagging the same things that the institution’s existing monitoring flags, and failing to surface what the institution’s monitoring was designed not to surface. The vine between reality and governance has always been present. AI introduces it at network speed and scale.


The structural enhancement: Establish independent real-time monitoring systems for the institutions most likely to be subject to major review — designed from the outside in, not the inside out. The monitoring should be designed specifically to surface what the institution’s own reporting does not surface: the gap between what the frontline experiences and what the management information reports, the discrepancy between what service users encounter and what the performance framework measures, the slow drift between the founding principle of the institution and the decisions being made in its name. This monitoring should be independent of the institution, transparent to parliament and the public, and specifically designed to provide early warning before the conditions for a major failure are fully established.

The structural enhancement:

Establish independent real-time monitoring systems for the institutions most likely to be subject to major review — designed from the outside in, not the inside out. The monitoring should be designed specifically to surface what the institution’s own reporting does not surface: the gap between what the frontline experiences and what the management information reports, the discrepancy between what service users encounter and what the performance framework measures, the slow drift between the founding principle of the institution and the decisions being made in its name. This monitoring should be independent of the institution, transparent to parliament and the public, and specifically designed to provide early warning before the conditions for a major failure are fully established.

SECTION 3

The Combined Analysis: Current Process and Structural Enhancements

The table below sets out the four consistent features of the current process alongside the structural enhancements the systems analysis produces. The enhancements do not abolish the existing process. They add the structural conditions under which each element of the process produces the result it is designed for.


Feature of current process	What it currently produces	With structural enhancement


Attribution frame	Findings accurate at the level of individual decision-making. Structural conditions that made those decisions likely — practices, methodologies, information culture — largely unexamined. Recurrence explained as a new failure rather than the same structural conditions.	Independent Programme Assurance examining practices, methodologies and information culture — reporting to the inquiry, not to government. Root cause analysis required to reach the soil, not only the governance decisions made on the basis of filtered information.


Retrospective frame	Accurate account of how failures became catastrophic. Structurally unable to reach the development-phase design choices that made catastrophe likely. The most important learning — what to do differently before the next version of the failure is designed — is not produced.	Development-phase audit: working backwards from the failure to the design choices that made it likely. Asks what governance instrument at the design stage would have caught what the designers could not see. Produces the most actionable recommendations for prevention.


Evidence filter	Evidence weighted toward institutional witnesses. Frontline knowledge heard as testimony, not investigation. The barriers that produced the failure shape the evidence that examines it. Ground-level knowledge of what the system was actually doing, as distinct from what it reported, does not reach the inquiry systematically.	Pre-terms-of-reference engagement with frontline workers and community members, before the institutional frame closes. Structured to surface the questions the existing terms of reference cannot formulate. Those questions become required components of the terms of reference, or the inquiry must publicly account for why they are excluded.


Implementation gap	Implementation assigned to the organisations identified as party to the failure. Systematic bias toward governance-level remedies that leave structural conditions intact. The vine grows back. The pattern recurs.	Independent implementation body with statutory mandate, separate from organisations party to the failure. Visible traceability between each recommendation and the specific structural condition it addresses. Network-generating knowledge transfer that carries findings to the organisations whose next decisions matter — before those decisions are made.


[Not addressed]	The democratic resilience dimension — the slow drift between what the institution is actually doing and what its reporting systems say it is doing — is not examined until the drift has produced a catastrophe visible enough to commission an inquiry.	Independent real-time monitoring designed from the outside in. Surfaces what the institution’s own reporting does not surface. Provides early warning before the conditions for a major failure are fully established. Designed to be AI-resistant — so that the AI tools used to process the monitoring data have been shown honest information, not the filtered version.

SECTION 4

Summary


The pattern in one sentence UK major reviews consistently produce accurate accounts of what went wrong at the level of individual and governance decisions, while leaving the structural conditions that made those decisions likely — the information barriers, the broken feedback loops, the practices and methodologies that shaped what leaders could see — largely unexamined; and the knowledge they do produce is sealed inside the inquiry that produced it, rather than carried to the network of organisations whose next decisions will determine whether the same pattern recurs.

The pattern in one sentence

UK major reviews consistently produce accurate accounts of what went wrong at the level of individual and governance decisions, while leaving the structural conditions that made those decisions likely — the information barriers, the broken feedback loops, the practices and methodologies that shaped what leaders could see — largely unexamined; and the knowledge they do produce is sealed inside the inquiry that produced it, rather than carried to the network of organisations whose next decisions will determine whether the same pattern recurs.

The conventional critique of major UK reviews is that they take too long, cost too much, and produce recommendations that are not implemented. This is true. The systems analysis explains why: each of these problems is a symptom of the same structural condition. The review process is itself organised as a closed, vertically optimised architecture — examining failures produced by closed, vertically optimised architectures, using tools and processes that reproduce the same structural features they are investigating.

The five major institutional failures since 2000 that share an identical root pattern are not evidence of collective incompetence or collective malice. They are evidence that the same structural conditions — broken information flows between what the frontline knows and what leaders decide on; governance frameworks designed for the institution as it was understood when they were built rather than the institution as it actually operates; and implementation processes assigned to the organisations that need to change — produce the same outcomes regardless of the individuals involved.

The systems analysis does not produce simple recommendations. It produces a set of structural changes that, taken together, would change the architecture of the review process itself — from a tower that examines failures and closes, to a network that generates learning and carries it to where it can prevent recurrence. The changes are structural rather than personnel changes. They require different institutions, different processes, different mandates. They do not require different people.

Three diagnostic questions before the next major review

1. Has the institution responsible for implementation been separated from the institutions identified as party to the failure? If not, the review’s structural findings will be processed through the same architecture that produced them, and governance-level remedies will substitute for structural ones.

2. Does the review process include a mechanism for surfacing the questions that its own terms of reference cannot formulate — specifically, the questions most likely to be asked by the people closest to the failure rather than the people most familiar with the institutional framework? If not, the most important learning will remain outside the inquiry’s scope.

3. Has the development-phase design been examined — the period when the conditions for failure were first established and the transition boundary had not yet committed to a trajectory? If not, the most actionable recommendations — the ones that would prevent the next version of the same failure before it is fully designed — will not be produced.

Where the honest answer to any of these questions reveals a gap, the intervention required is structural — in the design of the review itself, in the mandate of the implementation body, in the governance of information between what institutions report and what they actually do. Process improvements applied to structural problems produce process changes at the level where the intervention lands and structural continuity everywhere else.


The most important single change: Before the terms of reference of any major review are finalised, conduct a structured engagement with the people whose experience of the failure is most direct — asking not only what happened but what questions the institutional framing of the review cannot reach. Record those questions. Require the terms of reference to address them. The questions the institution cannot formulate are the ones that, had they been asked at the development stage, would have prevented the failure. They are also the ones most likely to prevent the next one.

The most important single change:

Before the terms of reference of any major review are finalised, conduct a structured engagement with the people whose experience of the failure is most direct — asking not only what happened but what questions the institutional framing of the review cannot reach. Record those questions. Require the terms of reference to address them. The questions the institution cannot formulate are the ones that, had they been asked at the development stage, would have prevented the failure. They are also the ones most likely to prevent the next one.

David Sutton CITP MBCS | April 2026