When the Specialists Mark Their Own Homework

RAIB Investigations, Coroners’ Constrained Powers, and Why Families Must Still Press

By James Byrne and Giles Mooney KC

Introduction

The recent collision between two trains at Bedford is a sober reminder of how suddenly lives can be lost on the railway, and of the machinery that swings into action when they are.  Whenever there is a serious railway accident, the Rail Accident Investigation Branch investigates and, in time, publishes a report on what it considers to have caused the accident and how a recurrence might be avoided. That report will be the work of acknowledged experts, and it will rightly carry great weight. But it will also, if recent history is any guide, come to dominate any inquest that follows and to narrow what a coroner is permitted to examine.

This article is about that consequence: about the limited powers the courts have so far held that coroners have to question or reject the conclusions of the RAIB and its sister body the Air Accidents Investigation Branch; about why, in our view, those conclusions should still be tested rather than taken on trust; and about why the inquest remains vital ground for the families who will later bring civil claims, where the deference shown to these reports falls away. We write not in the abstract, but from experience.

Beginning at Sandilands

We come to this subject not as theorists but as the barristers who sat with bereaved families through the inquest into the worst tram disaster in this country for more than a century.  On the morning of 9 November 2016, a tram derailed at the Sandilands curve in Croydon, having entered a sharp bend at roughly three-and-a-half times the permitted speed.  Seven people died and more than fifty were injured, several seriously. Representing the families, we watched an inquest that, for all its length, was hemmed in from the outset.

The Rail Accident Investigation Branch had reported in 2017 that the most likely cause of the accident was a temporary loss of awareness, a possible “micro-sleep”, on the part of a driver who had an otherwise unblemished record and had driven that same stretch hundreds of times.  The RAIB’s chief inspector told the inquest jury the crash was “absolutely” the product of excessive speed.  That much was not seriously in dispute.  What was in dispute was everything that sat behind the moment of derailment: the systemic failures, the inadequacy of the operator’s risk assessments against known human factors, and, most pointedly, an incident ten days earlier in which a tram had taken the very same bend at dangerous speed and very nearly overturned, an incident that had not been properly investigated.

The senior coroner declined to hear oral evidence from the management of Transport for London or Tram Operations Limited.  Her reasoning rested squarely on the line of authority that began with the Norfolk helicopter case: because there was, in her judgment, no credible evidence that the RAIB investigation was incomplete, flawed or deficient, she considered herself “not permitted to call further evidence.”  The jury returned a conclusion of accidental death.  The families called it a farce and said justice had been suffocated.  We understood why.  People who were, on any view, well placed to speak to the systemic causes of seven deaths never entered the witness box.

That experience is the reason we write.  The disquiet it left is not about any single coroner’s exercise of discretion.  It is about a structure that, taken to its logical end, can leave the most important questions about how people died beyond the reach of the only public, judicial inquiry many families will ever get.

A closed shop investigating itself

The RAIB is, in form, an independent statutory body.  Under section 7 of the Railways and Transport Safety Act 2003 it must investigate serious railway accidents and must try to determine what caused them but, critically, it “shall not consider or determine blame or liability.” Its purpose is the improvement of safety, not the apportionment of fault. That is a perfectly sensible design principle. No-blame investigation encourages candour; witnesses speak more freely if they are not building a case against themselves.

But the design has a less comfortable feature that is rarely examined in the courtroom. The expertise that makes the RAIB authoritative is drawn overwhelmingly from within the industry it investigates. Its inspectors are, in many cases, former railway engineers, operators and managers, people who have spent careers inside the very organisations and the very culture under scrutiny. That is precisely what equips them to read the marks on a sleeper or reconstruct a sequence of events from damaged infrastructure. It is also precisely what should make us cautious about treating their conclusions as beyond challenge. An investigation conducted by insiders, however technically excellent, carries an inherent risk of shared assumptions, professional sympathies, and blind spots about systemic and managerial failings that an outsider might press harder. There is nothing improper in any of this. It is simply the nature of a specialist body that recruits from a single industry. The “closed shop” is a feature, not a scandal, but features have consequences, and one consequence is that the body marking the homework was trained in the same school.

Our experience is that the RAIB defends its reports fiercely at inquests and can be notably cloth-eared to criticism of its methods or the scope of what it chose to examine. That institutional confidence is understandable. It is also exactly the dynamic that the law, as currently interpreted, tends to reward.

The coroner’s hands, largely tied

The courts have built a high wall around accident-branch conclusions. The governing principle, drawn from R. (Secretary of State for Transport) v HM Senior Coroner for Norfolk [2016] EWHC 2279 (Admin), through HM Senior Coroner for West Sussex v Chief Constable of Sussex [2022] EWHC 215 (QB), and consolidated in R. (Mid and West Wales Fire and Rescue Service) v HM Assistant Coroner for Carmarthenshire and Pembrokeshire [2023] EWHC 1669 (Admin), is straightforward to state and formidable in effect. Where one of the statutory accident investigation branches has investigated, a coroner should treat its findings as the evidence as to the cause of the accident, and should not re-investigate the same ground, unless there is “credible evidence” that the investigation was “incomplete, flawed or deficient.”

The threshold is, in the courts’ own words, a “high hurdle.” It is expressly “all or nothing”: either the investigation can be relied upon in full, or it cannot be relied upon at all, and there is to be no cherry-picking between conclusions. Minor criticisms do not suffice. Nor does the wisdom of hindsight, the observation that conclusions could have been better expressed, or even the identification of a discrete error, because an inspector “does not have to be and cannot be infallible,” and a sound investigation can still produce an incorrect conclusion. The rationale is the public interest in avoiding duplication and respect for specialist expertise. A coroner, the West Sussex court said, should be “very slow” to find that an expert investigation was incomplete, flawed or deficient.

We do not quarrel with the principle that resources should not be squandered on duplicative inquiries. But the practical effect, as Sandilands illustrates, is that the gateway to a fuller inquest is guarded by a test that is exceptionally difficult to satisfy before the very evidence that might prove the investigation deficient has been heard. There is a circularity here that ought to trouble us: the families could not show the investigation was deficient without exploring the systemic failings, and they were not permitted to explore the systemic failings because they had not first shown the investigation was deficient.

Challenge it anyway

None of this means the door is bolted. It means counsel must approach it with strategy and persistence rather than resignation.

First, the “high hurdle” is a hurdle, not a prohibition. The authorities expressly contemplate that an error may, on proper analysis, reveal that an investigation proceeded on a false premise and was therefore deficient. The task at the pre-inquest review is to identify, with precision and credible material, the specific respect in which the branch’s investigation was incomplete; a line of inquiry it never pursued, a near-miss it failed to examine, an assumption it never tested. Generalised dissatisfaction will fail. A focused, evidenced challenge to a defined gap is a different matter.

Second, the scope of an inquest and the reliability of an accident-branch report are distinct questions. A coroner retains a duty under sections 5 and 10 of the Coroners and Justice Act 2009 to ascertain how the deceased came by their death, and in Article 2 cases, the broader circumstances. The RAIB’s no-blame remit may simply not have covered matters such as regulatory failure, managerial decisions, the engagement of Article 2 that fall squarely within the coronial function. Those are arguments about jurisdiction and scope, not about second-guessing the cause of derailment, and they sit outside the Norfolk line.

 The civil courts are a different country

Here lies the point families most need to hear. The deference shown to accident-branch reports in the coroner’s court does not travel into civil litigation. In Rogers v Hoyle [2014] EWCA Civ 257, the Court of Appeal confirmed that an AAIB report is admissible as evidence in a negligence claim; admissible, examinable, and open to challenge in the ordinary way, with the court free to weigh it against other evidence and reach its own conclusions on fault. The civil court is not bound to treat the report as conclusive, is not engaged in the “all or nothing” exercise, and, crucially, is concerned with liability, the very thing the branches are forbidden to determine. The report becomes one piece of evidence among many, to be cross-examined, contradicted and, where appropriate, rejected, rather than a finding the tribunal must accept whole or not at all.

That contrast is precisely why the inquest, for all its constraints, remains worthy arena to test your case. The inquest is the first and often the only forum in which the relevant material is brought into the open under judicial supervision, on oath, and at public expense. Even where a coroner feels bound by the Norfolk line to treat the branch’s conclusions as the cause of the accident, the inquest process still surfaces the raw material from which a civil claim is built. The supplementary factual evidence an inspector may be asked to give, the documents that can be prised from the branch by application to the High Court, the disclosure of underlying physical evidence, raw data, measurements and expert technical reports that the RAIB will release even when it will not act as an expert witness, all of this is gold dust to the civil practitioner. Witnesses are identified, accounts are pinned down in a setting where they cannot easily be revised later, the operator’s documents are tested in public, and the systemic picture begins to take shape, however imperfectly the coronial conclusion reflects it.

In other words, the value of a proactively conducted inquest is not measured solely by the verdict it produces. A constrained conclusion can sit alongside an evidential record that proves decisive once the matter moves to a forum where the report enjoys no special deference. Counsel who treat the inquest as a dress rehearsal for the civil claim, using the pre-inquest review to define the deficiency arguments, using questioning of the inspector to extract every admissible fact, using the disclosure machinery to its limit, do their clients a service that long outlasts the inquest itself.

The lesson of Sandilands is not despair. It is diligence. Counsel and solicitors should treat the inquest as the opening, not the close, of the accountability process: press the scope arguments hard, build the evidenced case on any genuine deficiency, extract every admissible document, and carry that material into the civil arena where the deference falls away.