What happened to the sub-postmasters
Between 1999 and 2015, the Post Office prosecuted more than 900 sub-postmasters for theft, fraud and false accounting. The evidence against them was, in most cases, shortfalls in branch accounts reported by Horizon, a nationwide accounting IT system supplied by Fujitsu. Sub-postmasters who could not explain the shortfalls were prosecuted. Many were convicted. Some went to prison. Many were bankrupted. At least four are known to have died by suicide.
Horizon was faulty. The system produced phantom shortfalls. The Post Office, as both employer and prosecutor, knew from an early stage that the system had problems but continued to bring prosecutions on its output for more than a decade after internal evidence first raised concerns.
Exonerations began in earnest only after sustained legal action by Alan Bates and the Justice for Subpostmasters Alliance from 2009 onwards. A High Court judgment in 2019 found Horizon had been unreliable. The Court of Appeal quashed the first batch of convictions in 2021. The 2024 ITV dramatisation Mr Bates vs The Post Office drove mass public recognition and a 2024 statutory exoneration process.
The structural pattern
The Horizon case is studied now as a template mass miscarriage of justice. Its structural features are these:
- Institutional-evidence asymmetry. The prosecuting institution held all the evidence that could discredit its own case. It had no incentive to produce that evidence. Defendants had neither the technical expertise nor the resources to obtain it.
- Individualised framing. Each prosecution was treated as an individual case of dishonesty. The pattern — hundreds of people, same employer, same IT system — only became visible when the cases were aggregated.
- Expert deference. Courts accepted Horizon data as reliable on expert evidence from the system’s own operators. Independent forensic IT challenge was difficult to commission and difficult to get into court.
- Loss of public credibility gap. The Post Office was a trusted public institution. Courts, juries and the public gave the benefit of the doubt to the institution’s account over individual defendants’ denials.
- Long tail. Exoneration took twenty-five years. Most defendants were long-since convicted, fined, bankrupted, or dead before the public record shifted.
Why this is relevant to Letby
The structural features of the Letby case are, with specific differences, recognisable from the Horizon template:
- Institutional-evidence asymmetry. The Trust held the Datix records, the infrastructure-incident logs, the outbreak notifications, the staffing rotas, the medical records and the internal-review documents. The defence had limited resources to engage with this material at the speed the trial required.
- Individualised framing. Each alleged incident was treated at trial as an individualised act by one nurse. The pattern of systems failure on the unit — outbreak, staffing, infrastructure, gestation mismatch — was not the prosecution narrative.
- Expert deference. The jury accepted the Crown’s causation expert’s methodology as a sound basis for diagnosis. Independent specialist challenge (Panel, Joint Insulin Report, statistical critiques) has emerged largely after verdict.
- Public-credibility gap. The NHS and the courts are trusted institutions. A verdict delivered by a Crown Court jury on the prosecution narrative of a nurse harming babies carries significant institutional credibility; independent expert challenge faces a much higher bar.
- Long tail risk. The Horizon case shows that a generation may pass between when the evidence becomes available and when the public record catches up. Treating the Letby convictions as “settled” on the 2023 verdicts is the same reasoning that kept Horizon convictions in place from 2009 onwards.
Where the parallel breaks down
The Horizon analogy is not a claim that the Letby case is identical. It is a claim that the features of a mass miscarriage of justice are recognisable in kind. Specific differences include:
- Horizon was a mass case: hundreds of defendants. Letby is one defendant, charged with many counts. The aggregation effect is different.
- Horizon turned on a single technical artefact — an IT system’s reliability. Letby turns on multiple strands: medical causation, statistics, documentary interpretation, institutional response.
- Horizon prosecutions were brought by the institution itself as private prosecutor. Letby was prosecuted by the Crown through the Crown Prosecution Service, with an independent police investigation as the referring body.
None of these differences breaks the structural pattern; they only qualify it. The pattern remains: institutional-evidence asymmetry, individualised framing, expert deference, public-credibility gap, long tail of correction.
What the Horizon case tells us about what is needed here
The Horizon case was corrected because of sustained independent technical expertise (forensic IT experts), organised defendants (the Alliance), sustained public-interest journalism (Nick Wallis’s long-form reporting), senior political attention (David Davis among others), and ultimately a statutory exoneration mechanism driven by the public response to the ITV dramatisation.
The Letby case in 2026 has the first three of those five components in motion: the Shoo Lee Panel and the thirty-one-plus expert reports are the independent technical expertise; the CCRC application is the organised defence; Private Eye, Rachel Aviv, the Guardian investigation and this site are the sustained public-interest journalism. The fourth — senior political attention — is building via Sir David Davis MP and the Bar Council letter. The fifth — a statutory or appellate correction mechanism — is what the CCRC application is asking the Commission to set in motion.