Prepared for Legal Professionals, Journalists & the Department for Education
A structured forensic analysis of the evidential and procedural errors made by the Teaching Regulation Agency — and the grounds on which the determination should be challenged.
Case Summary
Jonathan Ullmer MBE was found guilty of historical misconduct by the Teaching Regulation Agency on the balance of probability in November 2019 — a finding resulting in a lifetime prohibition from teaching with no right of review. He strongly refutes the decision. No criminal charge has ever been brought. The police took no action whatsoever.
The case rested on the testimony of a single complainant: Tom Mitchelson, a professional undercover journalist and author, whose allegations emerged during therapy approximately 25 years after the alleged events. The one additional witness called by the TRA — the complainant's former girlfriend — corroborated nothing beyond the basic fact that time had been spent together.
Core position: Every serious allegation — including all sexual conduct findings upon which the lifetime ban principally rests — was wholly uncorroborated. The panel misapplied the burden of proof, failed entirely to scrutinise the therapeutic origins of the allegations, relied on a covertly edited recording, and subsequently lost the official hearing recording — making meaningful appeal structurally impossible. Each of these failures, individually and collectively, renders the determination unsafe.
The TRA operates on the balance of probability — a civil standard, not a criminal one. No conviction. No arrest. No police action of any kind.
Every serious finding rests on the uncorroborated account of one man — a professional whose career is built on undercover investigation and narrative construction.
A lifetime ban with no right of review — the most severe sanction available — imposed on a decorated headteacher of 30 years based on a single therapy-recovered account.
The Regulatory Context
Understanding the TRA's legal status is essential to understanding how the finding against Jonathan was reached, and why we believe it does not withstand scrutiny.
The TRA is an executive agency of the Department for Education — not a court. It does not apply criminal standards of proof. It does not have the procedural safeguards of a court. Cases are decided by panels of three people, meaning a bare majority of two can impose a lifetime ban.
The civil standard of proof means the panel must find it "more likely than not" that events occurred. Crucially, however, case law (Re B [2008] UKHL 35) requires that the more serious the allegation, the more cogent the evidence needed — a principle the panel did not visibly apply.
The NAHT has formally designated the TRA a "failing regulator." A Guardian whistleblower in 2024 alleged that prohibitions were celebrated internally as successes, and that staff were encouraged to "push back" when evidence was insufficient.
Jonathan was prohibited for life with no provision for review — the most severe outcome available. The proportionality of this sanction, in the circumstances of this case, is a specific ground for legal challenge.
Section One
Each of the following represents a distinct failure in how the TRA conducted this case. Taken together, they form a compelling case that the determination is unsafe and should not be permitted to stand.
The TRA applies a civil standard — the balance of probability — not the criminal standard of beyond reasonable doubt. However, even under the civil standard, serious and inherently improbable allegations require cogent evidence proportionate to their gravity. The panel offered no explanation of how it assessed the credibility of allegations made 25+ years after the alleged events by a single witness with no corroboration.
The House of Lords confirmed that while the civil standard is binary, the more serious the allegation, the more cogent the evidence required. No evidence of this calibration appears in the TRA's written determination.
The hearing was conducted in public and officially recorded — confirmed in the panel's own written determination. That recording has been lost by the TRA. Without it, Jonathan's legal team cannot challenge factual errors in the panel's written account, demonstrate inconsistencies in witness testimony, or prove that his denials were more clearly stated than the record reflects.
This alone may constitute a breach of Article 6 ECHR — the right to a fair hearing. The TRA's own administrative failure should not be permitted to operate as a bar to appeal. It should instead constitute grounds for a fresh hearing before a differently constituted panel.
Critically, the lost recording would have confirmed Tom Mitchelson's oral acknowledgement — attested by Jonathan Ullmer, his wife, and his counsel Andrew Faux — that his allegations first emerged during therapy. The TRA's institutional failure has buried this evidence.
Attested by witnesses to a public hearing: Tom Mitchelson acknowledged in oral evidence that his allegations first emerged during therapy. This was heard by Jonathan Ullmer, his wife, and his counsel Andrew Faux — all present throughout. The official recording confirming this was lost by the TRA. The written determination contains no analysis of this fact whatsoever.
The panel made no inquiry into the therapeutic methods used, did not request therapy notes, and did not call the therapist as a witness. The therapist subsequently became a personal friend of the complainant — a potential breach of BACP professional guidelines that the panel never considered.
No allegations for 25+ years → Therapy commenced → Allegations emerged in therapy (admitted in oral evidence) → Therapist became personal friend → Book planned requiring covert recording → TRA hearing → Book published commercially by HarperCollins.
A conversation was covertly recorded by the complainant in 2018 — nearly 25 years after the events alleged — without Jonathan's knowledge. The recording was edited: its beginning and end were removed before presentation. The panel acknowledged the recording did not prove the allegations, yet repeatedly treated it as supportive while dismissing Jonathan's attempts to correct and clarify. Jonathan maintains his denials were in the excised sections, which were never examined.
Where evidence is ambiguous, a fact-finding tribunal must not systematically favour one interpretation without explanation. Acknowledging the recording was inconclusive while treating it as supportive is logically inconsistent and procedurally unfair.
Jonathan's unblemished 30-year career, MBE, Outstanding Ofsted ratings, and extensive character evidence from colleagues, students and community figures were given little to no weight. The legal direction on good character is clear: it is directly relevant to both propensity and credibility, and panels must demonstrate they have considered it. The written determination contains no meaningful engagement with Jonathan's substantial character evidence.
The panel chair acknowledged during proceedings that she had "no idea what goes on in schools." This admission fundamentally undermines the panel's capacity to evaluate teacher-pupil interactions in a secondary school drama department in the early 1990s. A professional conduct panel must be capable of evaluating professional conduct in context. This one demonstrably was not.
The panel acknowledged that professional standards "were not explicit in the 1990s" — yet proceeded to find misconduct on that basis, without commissioning any historical expert evidence to establish what standards actually prevailed at the time. Behaviours including offering lifts after rehearsals and pastoral mentoring of students were common, accepted practice in 1990s arts education. Judging them by 2019 standards without expert evidence is both analytically unsound and legally untenable.
The panel made no finding on the conduct of Haileybury School governors, who negotiated directly with the complainant, controlled when Jonathan would be interviewed, and allowed the complainant to insert information into Jonathan's severance agreement. The school subsequently provided information to the panel about Jonathan's departure that omitted his serious illness and the full circumstances of his resignation. This went entirely unaddressed.
Section Two
The TRA called one additional witness — Child X, the complainant's former girlfriend during his sixth form years. A careful reading of the written determination reveals that her evidence corroborates nothing beyond the basic fact that time was spent together. Every serious finding rests entirely on the complainant's account alone.
Discussion of masturbation including named girlfriends. Rests solely on the complainant's account. Child X provided no evidence.
Sexual activity in the complainant's presence. Wholly uncorroborated. The panel itself acknowledged it "considered that Pupil A might have made it up."
Sexual activity with the complainant. The most serious finding. Rests entirely on the complainant's account. Child X provided no evidence whatsoever.
Sexual motivation finding. Based entirely on inference from the above uncorroborated allegations. No independent evidence of any kind.
Every sexual allegation upon which this lifetime prohibition principally rests is wholly and entirely uncorroborated. Not one of these findings has a single piece of independent supporting evidence.
Telling complainant they would "become better friends" once he turned 16. No independent witness. Complainant's account only.
Inappropriate personal and sexual comments, including discussing sexual experiences. Wholly uncorroborated. No witness to any of these alleged conversations.
The panel relied in part on a covertly recorded conversation from 2018 — 25 years after events alleged — which was edited before presentation. The panel acknowledged it did not prove the allegations yet treated it as supportive. Jonathan maintains his denials were in the excised sections.
The Kent trip and the first occasion of sharing a bed. No independent witness. Rests entirely on the complainant's account.
Sharing a bed during school years. Child X did not witness this. The panel relied on the complainant's account and elements of the edited covert recording.
Play fighting, including in fields at night near Paglesham. Wholly uncorroborated. Child X provided no evidence on this finding.
Child X confirmed phone calls to Jonathan's home and one Christmas meal. This corroborates only that time was spent together — nothing inappropriate, sexual, or coercive.
The corroboration picture in plain terms: The sole independent TRA witness confirmed the complainant spent time at Jonathan's home and attended one meal together. She confirmed nothing sexual, inappropriate or coercive. Every serious finding — and every finding upon which the lifetime ban rests — depends entirely and solely on the complainant's word alone. No physical evidence. No independent witness. No contemporaneous record. Allegations recovered in therapy 25 years after alleged events, uncorroborated on every finding that matters.
The more serious the allegation, the more cogent the evidence required. A lifetime prohibition based on wholly uncorroborated sexual misconduct allegations — from a professional whose career is built on narrative construction, whose memories emerged in therapy, and whose account was never independently verified — does not meet that standard.
Section Three
A matter of public record, attested by witnesses: Tom Mitchelson acknowledged in oral evidence at the public TRA hearing in November 2019 that his allegations first emerged during therapy. This was heard and can be attested to by Jonathan Ullmer, his wife, and his legal representative Andrew Faux — all present throughout. The official recording confirming this was subsequently lost by the TRA. The panel's written determination contains no analysis of this fact.
There is widespread agreement in the psychological community that newly recovered memories of past trauma can be accurate, inaccurate, or a mixture of both — and that reports based on recovered memories alone are "not reliable enough to be the sole basis for legal decisions" (BPS Working Party on Recovered Memories, 2000; Brewin & Andrews, 2017).
Research by Professor Elizabeth Loftus has demonstrated that false memories of events that never occurred can be created through suggestion, therapy and social influence. The British False Memory Society has documented hundreds of cases in which adults recovered memories of alleged abuse during therapy — memories later shown to be false.
In a study of 496 BFMS cases, 84% of daughters who accused fathers of abuse had undergone therapy before making the accusation. The caseload peaked in 1994 — the same period as the alleged events in Jonathan's case.
Eleanor Laws QC's guidance on recovered memory in legal proceedings notes that where a witness has been subject to suggestive therapy, interview and transcription should be disclosed to the defence. None of this was applied. The therapist was not called. Therapy notes were not requested. The BACP implications of the therapist-complainant personal friendship were never addressed.
Section Four — Publication Concern
The TRA panel specifically found the following allegations not proven. We believe Tom Mitchelson's memoir Don't Ask Me About My Dad (HarperCollins, 2022) presents some or all of these as established fact. If confirmed on review of the published text, this raises a serious legal concern entirely independent of the TRA findings themselves.
A claimant in defamation need not prove a statement is false — the burden lies on the defendant to establish a defence of truth, honest opinion, or public interest. Where a tribunal has specifically declined to find an allegation proven, a publisher asserting it as fact faces a materially more difficult task in establishing a truth defence. The TRA's own written determination becomes the evidential foundation for that challenge. Jonathan reserves all rights in this respect.
Section Five
We identify the following as the strongest grounds on which the determination should be challenged before the courts or the Department for Education.
The permanent loss of the hearing recording denies Jonathan his right to effective appeal and may constitute a breach of Article 6 ECHR. A determination without a reviewable record is procedurally unsafe and should be grounds for a fresh hearing.
The determination fails to explain how the panel assessed the credibility of uncorroborated historic allegations, or how 30 years of good character evidence was weighed against a single therapy-recovered account.
The complainant's oral acknowledgement that allegations emerged in therapy — attested by three witnesses — received no analysis. No expert evidence sought. Therapist not called. Therapy notes not examined. A fundamental procedural failure.
A lifetime ban based solely on uncorroborated, therapy-recovered testimony — where the police took no action and the complainant voluntarily maintained contact for years — may meet the Wednesbury unreasonableness threshold.
A covertly edited recording was admitted without inquiry into what was removed or independent authentication — below the minimum standard of procedural fairness.
Judging 1990s professional conduct by 2019 standards — without historical expert evidence — is legally and analytically unsound. The panel's own admission that standards "were not explicit" at the time undermines the validity of the finding.
Section Six
Article 6 of the European Convention on Human Rights guarantees the right to a fair and public hearing. While the TRA is not a criminal court, it exercises a function that determines civil rights — specifically the right to practise a profession. The European Court of Human Rights has confirmed that Article 6 applies to professional disciplinary proceedings that determine civil rights.
The loss of the official hearing recording is not a minor administrative inconvenience — it directly impairs Jonathan's ability to mount an effective appeal. Without the primary record, it is structurally impossible to demonstrate that the panel misdirected itself, misheard evidence, or made factual errors in its summary of oral testimony.
Case law is clear that institutions cannot rely on their own administrative failures to frustrate appellate rights. The TRA's loss of the recording should not operate as a bar to appeal — it should operate as grounds for a fresh hearing before a differently constituted panel with appropriate expertise.
A lifetime prohibition with no right of review — the most severe sanction available — was imposed on the basis of one uncorroborated witness, no criminal conviction, and allegations recovered in therapy 25 years after the events alleged. The proportionality of this outcome, in these circumstances, is highly questionable and directly susceptible to judicial review challenge.
Section Seven
Jonathan's case does not exist in isolation. The TRA has attracted sustained, serious criticism from professional bodies, unions, the media and its own former employees. The following systemic failures are directly relevant to the fairness of his hearing.
A former TRA employee alleged the agency had "forgotten it was investigating real people," that staff celebrated prohibitions as successes, and were encouraged to "push back" when evidence was insufficient. A teacher in a coma was mocked with vegetable nicknames by senior managers in an open-plan office.
The National Association of Head Teachers formally notified the TRA it considers it a failing regulator. NASUWT General Secretary Patrick Roach stated members had become "suicidal" and "self-harming" as a result of investigations, and that it was "easy to wreck someone's career with impunity."
TRA referrals rose by more than 60% in a single year to nearly 1,700 in 2023–24. Two teachers waited more than eight years for hearings. One headteacher's case cost £400,000 in TRA expenditure alone. A headteacher admitted breaching a teacher's human rights — and then prohibited them anyway.
Multiple teaching unions submitted a claim for judicial review against the TRA at the High Court, alleging non-white teachers are overrepresented in misconduct referrals and that the TRA refused to collect equalities data — continuing a pattern of operating without accountability.
Section Eight
The following steps are recommended for those acting on Jonathan's behalf, in order of priority.
On grounds of lost records (Article 6 ECHR), irrationality of the finding, inadequacy of reasons, failure to scrutinise recovered memory testimony, and retroactive application of professional standards without historical expert evidence.
Given the loss of the official recording, a fresh hearing before a differently constituted panel — with appropriate expertise in 1990s educational practice and psychological evidence standards — should be sought as the primary remedy.
A forensic psychologist experienced in false memory and the reliability of long-delayed testimony should be retained to provide an independent assessment of the complainant's evidence and the therapeutic context in which it was formed.
An education historian with expertise in 1990s teaching practice should be engaged to provide written evidence on safeguarding norms, teacher-student relationships, and professional boundaries as they existed at the time of the alleged events.
A formal complaint to the BACP regarding the therapist's post-therapy personal relationship with the complainant, and its potential influence on the formation of the allegations, should be considered.
A formal complaint to the Department for Education and/or the Parliamentary and Health Service Ombudsman regarding the TRA's institutional failure should be submitted to establish accountability and support any fresh hearing application.