AB v Chief Constable of British Transport Police

The respondent, AB, is 31. He has Autistic Spectrum Disorder (“ASD”) of the Asperger’s type. He suffers from high levels of anxiety as well as communication and processing difficulties. As a coping mechanism, he stims, rubbing fabric between his fingers.

In 2011 a woman complained to police that AB, who had been sitting next to her on a train, touched her inappropriately. In 2014, another woman made a similar complaint. On each occasion the matter was investigated by the appellant’s officers. AB was not charged with any offence. The police retain information about the complaints. AB claims that the retention of the information is unlawful because:

In a careful and detailed reserved judgment following a 5-day trial, and supplementary written submissions, His Honour Judge Sephton KC (“the judge”) found as a fact that the details of the incidents recorded in police pocketbooks accurately reflected what the police had been told. But he found that they were not an accurate account of what had, in fact, happened. He found that the police crime reports in relation to the incidents were therefore inaccurate. He also found that AB had not posed any real risk to the complainants, and that his disability diminishes the risk that he poses generally. Having regard to the policing benefit in retaining the records, and the impact on AB of doing so, he concluded that the retention of the records amounts to a breach of article 8 ECHR. He made an order that the police delete the records. He awarded damages of £15,000 in respect of loss of earnings, a further award of £15,000 for distress, and a further £6,000 for aggravated damages.

The appellant appeals against that order. She says that the judge’s whole approach to the case was wrong. In particular, she says the judge was wrong to make findings of fact about the underlying incidents, wrong to conclude that the records were inaccurate, and wrong in his approach to proportionality. She says that the judge erred in his assessment of damages. If, contrary to the appellant’s case, the judge was entitled to make factual findings about the incidents then she does not challenge the findings that the judge made.

One of the issues raised by the appeal concerns the approach that a judge should take when deciding if police crime reports are accurate within the meaning of data protection legislation. Must the reports accurately describe what in fact occurred in the underlying incident; or should they instead accurately reflect the information provided to the police about the incident? Beyond that, the issues are whether the judge was wrong:

The factual background is set out in detail in the judge’s judgment. Because there is no challenge to the judge’s factual findings, it is sufficient to give a shorter summary.

On 6 December 2011, a young female trainee solicitor was travelling by train. AB was sitting next to her. He touched her with his hand on her thigh. She informed the train guard, who informed the police. A police officer spoke to the complainant the same day and made a note of what she said in her pocketbook. She recorded that the complainant said that AB “put his hands between her legs and tried to push up her dress.”

An Occurrence Summary Report (“OSR”) was completed on the appellant’s computer system. This contains the following summary, including the “modus operandi” or “MO” (I have corrected obvious typing errors in the extracts that follow, and inserted some punctuation):

“BRC Person Reporting: — BRC Offence: sexual assault on female over 13 years. CAFI BRC MO: female adult victim boarded at Bristol. Sat next to male suspect. He had his jacket on his lap which brushed against her left leg. Jacket moves up. His hand goes between her legs under her dress. She asks him to move. She then got up and found guard to inform him. She then sat in foyer area while guard reported it. Police waiting for train in Newport. Victim was then able to identify male suspect to officer.”

The complainant made a witness statement on 7 December 2011. In the statement the complainant did not say that AB had put his hands between her legs or that he had tried to push up her dress.

On 16 December 2011, AB agreed to accept a caution for an offence of sexual assault. Subsequently, when this was challenged, the police accepted that AB was a vulnerable adult and that an appropriate person should have been present. They agreed to expunge the caution. They made an entry on the file that they would take no further action because there was insufficient evidence.

In June 2016, the OSR for the 2011 incident was updated by adding an entry to the “remarks” field of the record. This meant that the entry appeared immediately beneath the OSR MO when viewed on the computer screen. The entry records:

“Following a review of all available evidence it is appropriate to update the MO in this record. The suspect sat next to the victim aboard a train. After a short while the victim noticed her dress hem moving up her thigh. The suspect had his left arm across his abdomen which disappeared inside the right hand side of his jacket. His jacket obscured where his hand actually was. Victim felt the top portion of his fingers against her thigh and could feel his fingers rubbing in a fashion to pull up her dress. Victim believed this was a deliberate act because he was touching her upper thigh and pulling up her dress. She believed it was sexual.”

A further addition to the record was made on 29 March 2018. This states:

“The accuracy of the above entry has been disputed by [AB], who specifically denies that he placed his hand between the victim’s legs or that the touching was sexual. [AB] contends that he was “stimming”, that is fiddling with the material of the victim’s clothing in order to deal with anxiety.

The above entry was made based on an initial account taken by PC Rouse from the victim (noted in her pocket notebook), but was not a verbatim account of the victim’s allegations. Due to the passage of time since the incident, PC Rouse has no specific, independent recollection of the victim’s first account.

It should be noted that the victim provided a statement on 7 December 2011. She did not mention that [AB] had his hand between her legs or under her dress. She did state that her dress hem was moving up and [AB] touched her thigh and that she believed the assault to have been sexual.

A train driver to whom the victim first reported the incident stated that the victim complained of being touched on her side underneath her coat but that the touching was not intimate. The incident did not result in a prosecution.”

On 9 April 2014, a young woman was travelling by train. AB sat next to her. He touched her leg. She reported the matter to the appellant’s officer. The officer recorded the complainant’s account in her pocketbook. In that account, the complainant said that AB “had placed his hand [over my clothes] on my vagina area hidden under his jacket.” The complainant signed the officer’s pocketbook note of her account.

An OSR entry was created. This states:

“BRC Person Reporting: 4292 BRC Offence: sexual assault on a female aged 13 or over no penetration BRC MO: Offender sat on seat next to victim and during train journey put his hand on her knee and started to stroke her knee. Victim looked at him making him remove his hand. Minute later the offender touched the victim’s vagina over her trousers shielding his hand from view using his coat on his lap.”

The complainant was subsequently interviewed in accordance with the “achieving best evidence” procedures. She said that AB touched her “just on my leg... across my leg but just on my leg. It didn’t go between my legs.” She was asked, “did he actually touch the fly area of your jeans?” She replied, “I don’t remember to be honest...”

AB was arrested. He was interviewed on 15 April and 8 May 2014. He said that he had ASD and provided medical evidence. He did not answer the questions that were put to him in interview. On 12 August 2014, the Crown Prosecution Service decided not to prosecute, having received counsel’s advice that although there was sufficient material to prosecute, it was unlikely that a jury would convict in the light of evidence about AB’s autism.

In June 2016, an entry was added to the OSR remarks field:

“Victim sat on train. Male suspect sat down next to her in aisle with rucksack on his lap. Victim saw suspect has his right hand on knee. He moved his hand onto her knee. She felt touching on her knee - a tickling sensation. Victim looked at suspect who moved his hand back onto his knee. A short time later victim felt something on her knee again. Suspect had hand furthest from victim (left hand) across him under his jacket. Suspect pressing hand moving back and forth on victim’s knee. Victim told suspect to stop. Suspect removed hand when victim told him to stop.”

On 28 March 2018, a further entry was made to the OSR:

“The accuracy of the above entry has been disputed by AB, who specifically denies that he touched the victim’s vagina over her trousers or that any touching was sexual. AB contends that he was “stimming”, that is fiddling with the material of the victim’s clothing in order to deal with anxiety.

The victim signed a verbatim record of her account in an officer’s pocket notebook on 7 April 2014 stating that, during the encounter with AB, she “felt something touching [her] vagina over [her] clothes.” However, this initial allegation was not borne out in the ABE interview on 15 April 2014. There was no suggestion from the victim during the interview that AB touched her vagina, only her upper thigh (but not her crotch).

The incident did not result in a prosecution, following CPS advice.”

AB’s parents complained about the references in the police records to AB having committed sexual assaults. In May 2016, the Deputy Chief Constable concluded that the recorded “MO” in respect of the 2011 incident was inaccurate “in that there is no suggestion by the victim (unless when she initially spoke with [the officer]) that [AB] placed his hands between her legs”. He recommended that the MO should be updated. He also said that the force should accept that the MO recorded in respect of the 2014 incident was inaccurate. He recommended that it too should be updated. The judge inferred that the 2016 updates (paragraphs 11 and 17 above) resulted from the Deputy Chief Constable’s recommendations.

In October 2016, the Disclosure and Barring Service (“DBS”) asked the appellant’s officers to disclose relevant information in respect of AB’s proposal to work as a volunteer with other young people with ASD. The appellant’s officers decided that the records of the 2011 and 2014 incidents were not relevant, because AB had ASD which had caused him to “stim” and that he had adopted coping strategies to minimise the risk that he would touch women on trains again.

In December 2017, Bristol City Council was contacted with safeguarding concerns about AB, specifically that he was suffering ongoing emotional trauma because the appellant was maintaining false allegations about him. The Council undertook an enquiry pursuant to section 42 of the Care Act 2014. It drew the appellant’s attention to the effect that the retention of the records was having on AB. It requested, in terms, a triggered review (see paragraph 29 below) of whether the records should continue to be held. No such review was undertaken. The evidence adduced on behalf of the appellant was that the appellant did not undertake triggered reviews because it did not have sufficient resources.

After AB and his parents had raised concerns about the police records, an agreement was reached that they would be reviewed. On 28 May 2020 Claire Rogers, a data quality supervisor employed by the appellant, undertook the review. She concluded:

“The overriding consideration has to be the risk that [AB] may pose to the public… and whether that risk is outweighed by the harm to [him]. [AB] has shown a clear pattern of behaviour that raises concerns re inappropriate touching when in close proximity to females on the train. This behaviour and the possible risks attached to it illustrate a clear policing purpose to retain, both to protect the travelling public and to ensure [AB] is given proper support when communicating with the police. I have considered the harm that could be experienced by the individual in retaining the information. [AB] argues that retention of the records has had a considerable effect on his mental health and wellbeing, and has provided medical records to support this. However, the allegation is a serious one of Sexual Assault and deleting all trace of this could result in a possible pattern of behaviour by [AB] being missed if he were to come to notice again. Unless the nominal comes to police notice again, the retention of this data has limited impact upon him. I am satisfied that it is accordingly proportionate to retain the data so as to safeguard the travelling public and [AB].”

Mrs Pugh, the appellant’s force records manager, agreed with Ms Rogers’ assessment. Mrs Pugh considered that it was lawful and proportionate to retain the records “in the public interest for the prevention and detection of crime as well as retaining the records to allow BTP officers to safeguard the welfare of [AB] if he was to come to police notice again.” She pointed out that both incidents had a similar MO and demonstrated a pattern of behaviour. Ms Rogers, and Mrs Pugh, both considered that the records should be deleted from the police national computer (as opposed to the appellant’s local computer system, and the police national database where it would continue to be available). Assistant Chief Constable Doyle confirmed the decision of Ms Rogers and Mrs Pugh.

AB’s pleaded case is that he has dyslexia and Asperger’s Syndrome and that, as a coping mechanism, “he ‘stims’ or rubs material between his fingers and plays mind puzzles by pressing fabric.” He said that these coping mechanisms were misinterpreted and that this resulted in the 2011 and 2014 complaints. He complained that the records retained by the police inaccurately record that AB put his hands between the legs, and under the dress, of the 2011 complainant. He also implicitly complained that the records of the 2014 incident were inaccurate insofar as they suggested that AB had placed his hand over the complainant’s jeans in the area of her vagina. He said that there had been a breach of the Data Protection Act 1998 because (amongst other matters) the police data had not been kept accurate and up to date, in breach of the fourth data protection principle. He also said that the retention of the data amounted to a breach of his right to respect for private life under article 8 ECHR. He sought aggravated damages. One of the grounds on which he sought aggravated damages was that the appellant’s officers had not responded to the concerns that had been raised by the Bristol safeguarding enquiry (see paragraph 21 above).

The appellant denied the allegations of unlawful retention and processing of personal data and claimed that all reasonable steps had been taken to ensure the accuracy of the data. The appellant’s case was that the officers’ pocketbook entries recorded the initial accounts that the complainants provided and were “expressed as such.” It was said that it was relevant and necessary to retain the initial account that was given by a complainant, even where a different account was subsequently provided, because “[t]he existence of any difference in accounts can be of significant evidentiary and investigative value” and because “[i]t is not the function of a police force to ‘re-write’ a complainant’s initial account.”

The Data Protection Act 1998 was in force until 25 May 2018. The Data Protection Act 2018 was in force thereafter. Part 3 of the 2018 Act deals with the processing of data for the purpose of law enforcement (including the prevention of crime). Unless a relevant exemption applies, the appellant is required to comply with the data protection principles: section 4(4) of the 1998 Act, section 34(3) of the 2018 Act. The fourth data protection principle (“DPP4”) requires that personal data are accurate and, where necessary, up to date: schedule 1 paragraph 4 of the 1998 Act, section 38(1) of the 2018 Act. Data are inaccurate if they are incorrect or misleading as to any matter of fact: section 70(2) of the 1998 Act, section 205(1) of the 2018 Act.

Where the 1998 Act applies, the data protection principles must be interpreted in accordance with part 2 of schedule 1 of the Act: section 4(2). Paragraph 7 of part 2 of schedule 2 of the Act states:

“The fourth principle is not to be regarded as being contravened by reason of any inaccuracy in personal data which accurately record information obtained by the data controller from the data subject or a third party in a case where—

(a)

having regard to the purpose or purposes for which the data were obtained and further processed, the data controller has taken reasonable steps to ensure the accuracy of the data, and

(b)

if the data subject has notified the data controller of the data subject's view that the data are inaccurate, the data indicate that fact.”

Where part 3 of the 2018 Act applies, personal data based on facts must, so far as possible, be distinguished from personal data based on personal assessments: section 38(2). Where relevant and possible, a clear distinction must be made between personal data relating to different categories of data subject (such as those convicted of an offence on the one hand, and those suspected of an offence on the other): section 38(3).

The appellant applies, as its policy framework for the creation and retention of police records, published “authorised professional practice” (previously known as guidance on the management of police information) that is promulgated by the College of Policing. This states that crime reports should record the “modus operandi”, and that this field “should be as complete as possible because it often forms the basis for identifying suspects.” It requires that police records are subject to “scheduled” and “triggered” reviews to ensure that continued retention is justified. Scheduled reviews of the OSRs in this case should take place every 10 years. Triggered reviews must take place where concerns arise (including when disclosing information in accordance with a subject access request, or a DBS request) about the quality of information that is contained within a record.

In order to decide whether personal data are accurate, so as to comply with DPP4, it is necessary first to decide what the personal data mean: NT1 v Google LLC [2018] EWHC 799 (QB); [2019] QB 344 per Warby J at [80] – [83]. When assessing meaning, it is necessary to interpret the words in context: NT1 at [83]. Once the meaning has been established, it is then necessary to establish the facts so as to decide whether that meaning is “incorrect or misleading as to any matter of fact.”

It is unlawful for a public authority to act in a way that is incompatible with a Convention right: section 6(1) of the Human Rights Act 1998. The right to respect for private and family life under article 8 ECHR is a Convention right: section 1(1)(a) of the 1998 Act. Article 8 ECHR is a qualified right: an interference with the right can be justified if it is in accordance with the law and is necessary for, and proportionate to, the need to prevent crime: article 8(2). The “in accordance with the law” criterion requires both that there is a legal framework which sufficiently regulates the interference, and that any interference is compatible with the legal framework: R (Catt) v Association of Chief Police Officers [2015] UKSC 9; [2015] AC 1065 per Lord Sumption JSC at [11]. The authorised professional practice is part of the legal framework: Catt at [11] – [17]. The systematic collection and retention by the police of information about an individual is an interference with the right under article 8: Catt at [6]. Retention of such data is therefore unlawful, as being a breach of section 6(1) of the 1998 Act, unless it is in accordance with the law (including the authorised professional practice) and is proportionate to the risk to the public and thus the need to prevent crime.

There was a wealth of evidence before the judge about AB and the impact on him of ASD. That included evidence from AB, and from each of his parents, and medical evidence. That evidence explains how AB uses stimming as a means of coping with anxiety. This might include stroking his own clothing and feeling other clothing within his touch (even if it is worn by others; he might not appreciate that it is the clothing of another person). The judge accepted that evidence. He drew particular attention to the evidence of Dr Carpenter, a consultant psychiatrist with an interest in ASD. Dr Carpenter considers that AB’s use of stimming provides an explanation for what happened in 2011 and 2014, and that the alternative, that this was a “deliberate quiet sexual provocation” was “almost completely incompatible with the rest of his statement and not typical of someone with his diagnosis.”

The judge explained what findings of fact he considered were necessary, and why:

“It is necessary for me to make findings about… what happened in relation to both incidents in order to determine whether the descriptions complained of are indeed inaccurate (relevant when dealing with the data protection legislation) and to consider the risk posed by [AB] (a necessary enquiry when considering proportionality in relation to Article 8).”

In relation to the 2011 incident, the judge found, on the evidence, applying the civil standard of proof, that AB did not put his hand between the complainant’s legs or up her dress, and that the prospects of proving that he did so to the criminal standard were negligible. In relation to the 2014 incident, the judge found that AB did not put his hand “over or near the victim’s vagina”.

The judge further found that AB suffers from ASD and anxiety, that he “stims” in order to relieve anxiety, and that the 2011 and 2014 incidents involved “stimming” (ie relieving anxiety provoked by ASD) as opposed to any sexual motivation. He stressed that in reaching “this crucial conclusion” he recognised that from the perception of the two complainants they had been subjected “to an unwanted sexually motivated attack”, but this was a mistaken perception.

Having made these findings of fact, the judge separately considered the accuracy of the pocketbook entries and witness statements on the one hand, and the OSRs on the other hand.

In respect of the police pocketbook entries and the witness statements, the judge said that there was no sufficient evidence to persuade him that the information that the appellant’s officers were given was not accurately recorded. In other words, the judge was satisfied that the officers accurately recorded what the complainants said.

In respect of the OSRs, the Judge said the purpose of the “Summary” field was to provide a concise account of the content of the record. The purpose of the “Modus Operandi” field was to describe how the offence was committed:

“It seems to me that the policing purpose of this field is fulfilled if it contains the most accurate account possible of what the suspect actually did or is alleged to have done, so that any characteristic features of the suspect’s behaviour can be used as an aid to identify the suspect or tie him or her into another crime in which a strikingly similar method was used. More relevantly to this action, to the ordinary reader of the record, the “Modus Operandi” field constitutes a description of what the data subject actually did.”

The judge said that the summary fields of the 2011 and 2014 incident are incomplete because they do not mention that AB had ASD and that he stims, and that the reason for AB’s conduct is his ASD, and his conduct was therefore “wholly innocent”. The summary field and the MO for the 2011 incident wrongly record that AB put his hands between the victim’s legs and up her dress. The summary field and the MO for the 2014 incident wrongly record that AB “touched the victim[’s vagina] over her trousers.” The inaccuracies in the MO for each incident were corrected by the 2016 updates, but the records remain incomplete because they do not indicate that AB has ASD and that it is his ASD which is the impetus for his stimming.

The judge acknowledged that the evidence indicated that the MO field remains static and cannot be changed, so as to maintain the integrity of the record for the purpose of crime recording. However, the judge considered that this offended against important rules and principles. It meant that the record was inaccurate and was not kept up to date, contrary to the legislative requirements. Further, an ability to change the OSR would not compromise the integrity of the information held by the police because the underlying evidence would remain available.

The evidence of AB and his parents, and the medical evidence (particularly the report from Dr Carpenter), indicated that AB suffered significant distress and was in a “highly stressed state.” He was unable to relax, or to do things that he would normally enjoy. He was not sleeping well. He had lost his appetite and libido. He had developed a moderate marked depression. He ruminated about matters, but this had not developed into a frank paranoia. He had lost his job as a chef. AB had consulted clinical psychologists to help him with anxiety, and he had undergone counselling.

The judge acknowledged that anxiety was a hallmark of AB’s underlying condition, and there were signs of this before 2011. So, the retention of the records was not wholly responsible for AB’s ongoing psychological problems. Nevertheless, the judge found that a significant stress for AB was the fact that the police held records upon him which he felt were untrue and unfair. AB’s ASD amplified and concentrated his obsession about the records, and it was this obsession which caused him considerable distress. It was a material component in his decision to give up his job.

It was common ground that AB’s article 8 rights were engaged. It followed from the judge’s finding of a breach of the data protection legislation that the retention of data was not in accordance with the law. It followed that there was a breach of Article 8 ECHR.

In case the judge was wrong to find that the OSRs were inaccurate, he went on to consider “whether the retention of the data is proportionate to the policing purposes said to justify retention, having regard to the extent of the interference with [AB]’s Article 8 rights.”

It was common ground between the parties that the retention of police records pursues a legitimate aim (the prevention of crime), that it is a suitable means to achieve that aim, and that the prevention of crime cannot be achieved by a less restrictive alternative. The fact sensitive question that arose was whether a fair balance had been struck between the interference with AB’s right to respect for private life and the aim of preventing crime. That question arose in the context of the retention by the police of records of unsubstantiated allegations that are strongly denied. Assessing proportionality in this context is a “particularly sensitive and difficult exercise”: R (A) v Chief Constable of Kent [2013] EWCA Civ 1706 per Beatson LJ at [1].

The judge recognised that the purpose of retention is to protect the public. The records relate to alleged offences contrary to section 3 of the Sexual Offences Act 2003 (which are normally kept by the police until the alleged offender reaches the age of 100, subject to 10-year reviews). He gave weight to the assessments made by the Secretary of State and the College of Policing as to the length of time that such records should normally be retained. He took account of the evidence that the complainants each found their experience “disturbing and frightening”, and that the two incidents were broadly similar. He considered that a pattern of behaviour had arguably been established. He also considered that there was a risk of recurrence. Retention of the records would therefore assist the police to identify AB in the event of recurrence (even if “only to eliminate him from their enquiries”). This would fulfil a policing purpose even though AB’s conduct did not, on the judge’s findings, amount to a criminal offence.

As against the factors in favour of retention, the judge recognised that the data are sensitive and intimate and that they carry “a stigma of guilt.” He found that the retention of the data had a very significant effect upon AB. He has suffered distress (although it was difficult to separate the distress occasioned by the inaccuracies from that occasioned by the retention). His medical condition had caused him to become obsessional about the issues, and this had amplified his distress. This distress would be likely to continue even if the records were corrected, and even if the processing of them was highly restricted.

The judge took account of the review that had been undertaken (see paragraphs 22 - 23 above). However, he considered that review was flawed because it did not take into account the effect on AB of retaining the records and because those conducting the review had not appreciated the explanation for AB’s conduct.

After taking all these matters into account, the judge concluded that the retention of the data was not proportionate to the policing purposes said to justify retention, having regard to the extent of the interference with AB’s right to privacy. He considered that the balance fell “heavily in favour of [AB’s] Article 8 rights.” He emphasised that this was a highly fact sensitive assessment based on the particular circumstances of the case, and especially the nature of AB’s disability which both diminished the risk of serious harm he posed and increased the distress he experienced as a result of the existence of the records. The judge concluded that from “at least 2016” it was not proportionate to retain the records.

The judge made declarations that the OSRs are inaccurate and that their retention constitutes a breach of AB’s rights under article 8 ECHR.

The judge ordered the appellant to delete all of AB’s personal data that is held in the appellant’s operational systems.

The judge found that the retention of the police records had caused AB significant stress, and that his ASD “amplified and concentrated his obsession about the records.” This was a material component in AB’s decision to give up a job as a chef. He assessed the financial losses in the sum of £15,000. He made a further award of £15,000 in respect of the distress that AB had suffered. He also made an award of aggravated damages in the sum of £6,000.

In assessing the award for aggravated damages, the judge found that the authorised professional practice required triggered reviews of AB’s data. Such reviews had been required following correspondence in 2014, the updates of the OSRs in 2016 and 2018, the DBS check, and the Bristol enquiry.

None of the required reviews had taken place. The judge took into account that the appellant’s officers had not engaged with the enquiry undertaken by the Bristol Safeguarding team, even though that team had asked for a triggered review to be carried out. The evidence adduced by the appellant was that it would be “impossible” to comply with the obligation to conduct triggered reviews, and that it was not something that the appellant could “adhere to”, because significant resources would be required.

Anne Studd KC, for the appellant, argues that the making of factual findings by the judge was “a fundamental error of approach and results in the judgment being unsustainable.” She says that it was unnecessary to make factual findings, and wrong in law to do so, particularly as there was limited direct evidence as to the events in 2011 and 2014. She submits that if the judge’s approach is correct then it follows that in every case concerning the accuracy of data the court will be required to make findings as to the truth of what happened. She points out that in R (YZ) v South Wales Police [2021] EWHC 1060; [2022] EWCA Civ 683 (a claim for judicial review of a decision not to delete a police record) the court did not engage in a fact-finding exercise in respect of the accuracy of the police records.

Ms Studd says that the judge relied on his conclusion that the OSRs are inaccurate for his conclusion that the retention of the data is unlawful, and his factual findings were fundamental to his assessment of proportionality. That is because, in assessing the proportionality of retention of the police records, he had regard to the risk posed by AB. It follows that the judge’s assessment of proportionality “cannot stand.”

The appellant separately challenges the award of damages. She contends that the judge did not identify his starting point, did not grapple with the competing factors in play in relation to AB’s distress, or what deduction should be made for those other factors. In respect of the award of aggravated damages, Ms Studd submits that the judge was wrong to make a finding that the appellant’s officers had not taken part in the enquiry undertaken by the Bristol safeguarding team, because this was not an issue that had been raised.

Louis Browne KC, for AB, submits that the judge was right to make findings of fact as to the accuracy of the police records in order to determine the complaint that there was a breach of DPP4. He had regard to all relevant evidence. He clearly understood the role of the OSRs. He reached unimpeachable findings that the OSRs are inaccurate.

Mr Browne says that the judge properly addressed himself as to the applicable legal principles and the issues he was required to consider. He critically examined the review that had been conducted by Ms Rogers and Mrs Pugh. The review failed to pay any regard to AB’s autism and the medical evidence that had been disclosed on his behalf. After identifying flaws in their approach, he was right to give the review little weight. The judge made findings of fact that were open to him. He carefully analysed the factors that were relevant to proportionality and his approach is unimpeachable. He was right to conclude that the retention of the records is a disproportionate interference with AB’s article 8 rights.

As to damages, Mr Browne points out that leading counsel on both sides had made submissions to the judge by reference to the Judicial College guidelines, that the award that the judge made for distress (£15,000) was within the bracket for which the appellant had submitted (£5,500 - £17,900), and that the appellant’s failure to engage with the Bristol enquiry had been raised both in the pleadings and the evidence.

This was a private law claim for damages. AB’s case (so far as is now relevant) was that the police records are inaccurate, and their retention is a disproportionate interference with his article 8 rights. Witness statements were exchanged. Witnesses were called and cross-examined. The issues for the judge to resolve included whether the police records are accurate and whether their retention is a disproportionate interference with AB’s right to respect for private life. It was necessary for the judge to make findings of fact to resolve the issues that were before him. The judge needed to make findings of fact to decide whether the police records are accurate. And the judge needed to establish the basic facts in order to assess the question of proportionality.

The question of the accuracy of the records was a factual issue for the judge to determine one way or the other. YZ was a claim for judicial review where there was no live evidence and no challenge to the accuracy of the records. The absence of a fact-finding component to the court’s decision in that case does not indicate that the judge in this case was prohibited from making a factual assessment as to the accuracy of the records.

So far as the accuracy of the records is concerned, the real issue of principle between the parties concerns not so much whether fact finding was permissible, but what fact finding was necessary to decide if the police records are accurate. That raises the antecedent question as to the meaning of the records (see paragraphs 66 – 80 below).

So far as proportionality is concerned, a finding, one way or the other, on the civil standard of proof, as to whether a crime was committed is not determinative of the question of whether retention of a crime report is justified. The judge did not suggest otherwise. But in order to decide whether retention of a crime report can be justified it is necessary to analyse the underlying facts. The judge needed to assess the impact of retention on AB’s right to respect for private life, and the extent to which the retention of the records is necessary for the protection of the public. These issues were disputed by the parties and were the subject of oral evidence. In order to resolve those issues, and after finding that the review was flawed, the judge was entitled to make findings in relation to the underlying incidents in 2011 and 2014. That was relevant to the risk to the public posed by AB, and hence to the proportionality balance.

I therefore reject the appellant’s submission, in its broadest form, that the judge was wrong to make findings of fact and that his whole approach to the case was fundamentally flawed.

The parties appear to have been at cross-purposes throughout this litigation as to the accuracy of the OSRs. AB maintains (and the judge accepted) he did not put his hand under the dress of the 2011 complainant, or over the vagina area of the 2014 complainant. Accordingly, he says (and the judge found) the OSRs are inaccurate. This conclusion is based on the premiss that the OSRs mean what they say: the incidents occurred in the manner set out in the OSRs, and the MOs reflect what actually happened.

Conversely, the appellant says that (as the judge found) the 2011 complainant said in her initial report to the appellant’s officer that AB put his hand under her dress, and the 2014 complainant said in her initial report that he put his hand over her vagina area. Accordingly, she says, contrary to the judge’s findings, the records are accurate. This conclusion is based on the premiss that the OSRs mean only that the police were given the information that is contained in the OSRs (not that that underlying accounts given by the complainants are necessarily accurate).

The underlying difference between the parties is that they take a different view about what the OSRs mean. AB’s case proceeds on the assumption that the OSRs mean that he has committed offences in the manner described. The appellant’s case proceeds on the assumption that the OSRs are merely a record of information that has been received by the appellant’s officers.

This fundamental difference in how the cases were put was not confronted in the statements of case, or in the written materials that were before the judge at trial. The appellant initially contended that the judge had not been asked to make findings of fact as to whether the OSRs were accurate. Ms Studd fairly retracted that contention in the light of AB’s trial skeleton argument which did ask the judge to do exactly that – that is, to determine if the OSRs are accurate. However, the issue was framed simply in those terms. The question of what the OSRs meant does not seem to have been squarely confronted in the written materials that the parties put before the judge. But this was the necessary first step on the route to deciding whether the OSRs are accurate (see paragraph 30 above).

The nature, context, and purpose of the OSRs are critical when deciding what they mean. That is because the same or similar words may mean different things depending on the context in which they are recorded:

The context of the data in the present case is the appellant’s OSR database. The appellant’s policy is that all reports of incidents, whether from victims or others, and whether crime related or not, must result in the registration of an incident report. The evidence of the appellant’s data protection and freedom of information manager, which was accepted by the judge, was that an OSR provides a brief summary of police involvement in an incident. It is updated during an investigation and provides an audit trail of the actions taken by the police. The Head of Information Management for the appellant provided evidence, which again was accepted by the judge, that an original allegation of a crime is recorded and retained, even if it is subsequently found to be inaccurate. The reason is that it is necessary to retain the record of the allegation to identify the rationale for any police action that is taken, and to provide an audit trial, and also to help identify any inconsistencies in the accounts that are provided.

The OSRs do not purport to record convictions. Nor do they purport to convey findings of guilt, or assessments of likely guilt, made by the police or anyone else. They contain a narrative of information that has been received by the police. That narrative is entered at the time, or soon after, the information has been received. It is entered into the database before any analysis is made of the accuracy of the underlying information. The information is not assessed or graded in the way that sometimes occurs when the police receive intelligence (for example, to indicate an assessment of the reliability of the source who provided the information, and whether the information was within the source’s direct knowledge).

The judge said that to the “ordinary reader” of an OSR, the “Modus Operandi” constitutes a description of what the data subject actually did. I agree that a member of the public, with no knowledge of the purpose of an OSR, or how it comes to be created, might well understand the MO to identify what the data subject actually did. But there was no evidence that the OSRs were published or otherwise communicated to members of the public. If they are kept in accordance with the appellant’s policy, and guidance from the College of Policing, then they are only available to those with a proper policing purpose to access them. The “ordinary reader” of an OSR can be taken to be a person with an understanding of how OSRs come to be created and their underlying purpose. The reader knows that they are not a database of convictions or findings, that they simply record the information that has been provided to the police and the actions that the police have taken, and that the underlying allegation may or may not be true.

The judge said that “the DBS check reached a conclusion which was very different from the impression given by the OSRs in this case.” In other words, the impression that that judge had gleaned was that the OSRs purported to record the facts of the underlying incidents, whereas those conducting the DBS checks had treated them only as a record of the accounts that had been given to the police. This serves to reinforce the importance of context and recognition that the records are restricted to a limited readership that appreciate what OSRs are intended to convey.

For all these reasons, the OSRs convey information that is provided to the police, and the responsive steps that are taken by the police. The MO is thus the complainant’s account of what happened. It does not purport to convey that the incident occurred in the way that the complainant alleged.

It therefore follows that the accuracy of the data contained in the OSRs depends on whether they accurately reflect what the police were told and what the police did. It does not depend on what actually happened in the underlying incidents.

For the reasons given in paragraphs 66 – 76 above, the data in the OSRs purport to convey the information that was provided to the police.

There is no challenge to the judge’s factual findings. Those factual findings include that the police pocketbooks accurately record what each complainant told the response officer (see paragraph 37 above). The disputed parts of the OSRs faithfully record that which is contained in the police pocketbooks. It follows that the OSRs are accurate. That is so, even though the Deputy Chief Constable raised some concerns about their accuracy (see paragraph 19 above), and, anyway, those concerns were addressed in the 2016 updates (see paragraphs 11 and 17 above).

The OSRs did not make any reference to AB’s disability or his explanation for his conduct. That is because these matters did not form part of the crime reports that had been made to the police. AB had not even been diagnosed with ASD at the time of the 2011 incident. The OSRs, as records of the complaints that had been made, were accurate and up to date. The 2018 additions include AB’s explanation that he suffers from anxiety, that he “stims” to alleviate that anxiety, and that this explains what happened. They do not record the medical evidence as to his condition, but that does not render the records inaccurate or misleading.

After dealing with the accuracy of the OSRs, the judge turned to whether retention of the police records complies with article 8 ECHR. It was (and is) common ground that the retention of the records amounts to an interference with AB’s article 8 rights. The appellant therefore needed to show that the interference could be justified as being (1) in accordance with the law, and (2) proportionate to the aim of preventing crime.

The judge separately considered these two separate questions. His conclusion on the first issue followed from his finding that the OSRs were inaccurate. That meant that their retention was not in accordance with the law because it involved a continuing breach of DPP4. The internal logic of the reasoning is valid, but the conclusion is flawed because it relies on the finding that the OSRs are inaccurate.

AB does not seek to uphold the conclusion on any alternative ground. It is not therefore necessary to decide whether the appellant’s admitted failure to conduct reviews of police records as required by the authorised professional practice (which is an essential component of the legal framework) means that the retention of the records is not in accordance with the law, and is thus incompatible with article 8 ECHR, and is therefore unlawful as being a breach of section 6(1) of the 1998 Act. I note, however, that in R (C) v Commissioner of Police of the Metropolis [2012] EWHC 1681 (Admin); [2012] 1 WLR 3007 Richards LJ drew attention to the need for the police to secure compliance with (what is now) the authorised professional practice so as to be able to justify any article 8 interference (see at [42] – [46]). Ten years on from the decision in C, there is an admitted systemic failure by the appellant to comply with the requirement to carry out triggered reviews. The asserted lack of resource is no answer: a lack of resource does not justify a failure to comply with an unqualified legal obligation.

After deciding that the retention of the data was not in accordance with the law, the judge then considered whether the retention amounts to a disproportionate interference with the appellant’s article 8 right to privacy. He introduced that section of the judgment with the words “In case I am wrong about the OSRs… I have carefully considered whether the retention of the data is proportionate to the policing purposes said to justify retention, having regard to the extent of the interference with [AB]’s Article 8 rights.” This makes it clear that in considering the issue of proportionality, the judge left out of account his finding that the data are inaccurate. Nothing in his analysis of the article 8 issue suggests that his assessment of proportionality was influenced by his finding that the data are inaccurate. Accordingly, I do not accept Ms Studd’s submission that the judge’s factual finding that the OSRs are inaccurate is fundamental to his decision that it is disproportionate to retain the OSRs.

Ms Studd is right that the judge’s view as to the (lack of) risk posed by the appellant was a factor that he took into account when considering proportionality. That is separate from the question of whether the data are accurate. The judge was right to take account of the question of risk. Broadly, and everything else being equal, there is a stronger need for the police to retain data in respect of those who pose a substantial risk of harm to the public than those who pose no real risk of harm. Much of the intricate structure of the authorised professional practice reflects the need to prioritise data collection and retention in respect of those who pose a higher level of risk. The question of risk is relevant when assessing proportionality.

The judge was not bound to accept the appellant’s assessment of risk (but he did give weight to the views of the appellant’s staff). He was entitled to reach his own view. There is no challenge to the finding that he (effectively) made that the appellant poses no real risk to the public. That was a relevant factor when assessing proportionality.

I therefore reject the submission that the judge was wrong to make findings of fact for this purpose. As Mr Browne submitted in his arguments before the judge, the assessment of proportionality is “an exquisitely fact sensitive exercise.”

The appellant’s challenge to the proportionality assessment is limited to the approach taken by the judge and, in particular, the extent to which he relied on his findings of fact. I reject that challenge for the reasons I have given.

The appellant has not separately challenged the judge’s ultimate conclusion on proportonality. The scope to challenge such a conclusion is limited. Questions of proportionality involve an evaluative assessment. In some cases that assessment is black and white, such that there is only one tenable conclusion. Others fall in a grey area where reasonable judges may reach different conclusions. An appellate court may only intervene if the judge’s conclusion is wrong: Civil Procedure Rules, Part 52 rule 21(3)(a), In re B (a child) (care proceedings: threshold criteria) [2013] UKSC 33; [2013] 1 WLR 1911 per Lord Wilson JSC at [36] – [37], Lord Neuberger JSC at [83] – [90], and Lord Clarke JSC at [136] – [139].

Here, the challenge is based on the route that the judge took to his conclusion, not the conclusion itself. I have rejected the appeal that is advanced. Nor do I consider that the judge’s conclusion itself is wrong. It is important to stress the observation that the judge himself made that his conclusion was highly fact specific. In many cases, it will be proportionate to retain crime reports of offences for long periods of time, particularly where the offences raise public protection issues. That is because, generally, the interference with the subject’s article 8 rights is likely to be modest, whereas there is likely to be a compelling public interest in retention. An example is the decision in R (CL) v Chief Constable of Greater Manchester Police [2018] EWHC 3333 (Admin) per Hickinbottom LJ at [112]. It is only because of the very particular facts of the present case, both in terms of the exceptional impact on AB, and the lack of any real risk to the public, that the judge concluded that it was not proportionate to retain the records.

An appeal against an award of damages can only succeed in narrow circumstances. It is not sufficient that the appellate court would award a lesser sum if sitting at first instance. Rather, it must be shown that the judge acted on a wrong principle of law, or misunderstood the facts, or awarded a sum that is wholly erroneous: Santos v Eaton Square Garage Ltd [2007] EWCA Civ 225 per Maurice Kay LJ at [2].

I have concluded, in respectful disagreement with the judge, that the OSRs are accurate. But that does not impinge on the assessment of damages. The damages reflect the loss and damage sustained by AB as a result of the appellant’s unlawful retention of the records. That loss and damage is not reduced because I have taken a different view about the accuracy of the records.

The judge had assistance from both parties as to the principles that applied. There was no significant dispute between the parties as to those principles. The judge referred to the Judicial College guidelines on the assessment of general damages in personal injury cases, and, in particular, the guidelines in respect of psychiatric and psychological injury. He took account of the fact that other factors had caused distress to AB, and he clearly appreciated that the damages awarded should be limited to the distress that was due to the retention of the records. Thus, he assumed that “the psychological distress caused by the retention of the records” (emphasis added) would abate at the conclusion of the case. There are different possible approaches to the assessment of damages. One approach is to consider the appropriate award for the totality of the distress suffered, and then to discount that to reflect the other factors in play (where it is possible to separate out the damage and apportion it to different causes: BAE Systems (Operations) Ltd v Konczak [2017] EWCA Civ 1188 [2018] ICR 1 per Underhill LJ at [56] and [71] - [72]). Alternatively, where it is possible to do so, it is legitimate simply to assess the award for the distress that has been occasioned by the retention of the records. That was the approach the judge took. It followed that there was no reason for the judge to identify a starting point or a discounting factor.

The judge’s award was well within the bracket identified by the appellant. It was also well within the updated middle “Vento” bracket (£9,900 - £29,600) for injury to feelings: Vento v Chief Constable of West Yorkshire [2002] EWCA Civ 1871; [2003] IRLR 102, as updated by the Presidents of the Employment Tribunals for England and Wales, and Scotland, in Presidential Guidance dated 28 March 2022. The combined award for distress and for aggravated damages is also within this bracket. I do not consider that there is any basis for concluding that the award was “wholly erroneous.”

There is no challenge to the judge’s decision to award aggravated damages. Ms Studd accepts that an award for just satisfaction under section 8 of the 1998 Act can accommodate the concept of aggravated damages: Anufrijeva v Southwark London BC [2003] EWCA Civ 1406; [2004] QB 1124 per Lord Woolf CJ at [68], Alseran v Ministry of Defence [2017] EWHC 3289 (QB); [2019] QB 1251 per Leggatt J at [915]. The issue is whether the judge was entitled to take account of the appellant’s failure sufficiently to engage with the Bristol safeguarding enquiry. I agree with Mr Browne’s submission that this matter had been sufficiently raised, and the appellant had a sufficient opportunity to address it. AB was required to set out his grounds for claiming aggravated damages in his Particulars of Claim: Civil Procedure Rules, Part 16, rule 4(1)(c). He did so. He specifically relied on the Bristol enquiry and the appellant’s failure to respond. In his witness statement (which he adopted at trial) he referred to the Bristol enquiry and exhibited the enquiry report. The statements of each of his parents (which they each adopted at trial) also made reference to the enquiry, and his father’s statement expressly referred to the appellant’s failure to co-operate with the Bristol enquiry. The issue was squarely raised.

The police records in this case are intended to reflect the information that was provided to the police, rather than the underlying facts as to what happened. On this issue I have reached a different conclusion from the judge, with the result that I have concluded that the OSRs are accurate. To this narrow extent, the appeal succeeds.

It has not been shown that the judge was wrong to conclude that the retention of the records is a disproportionate interference with AB’s right to respect for private life. I therefore dismiss the appeal against the judge’s declaration that the retention of the records is unlawful. I also dismiss the appeal against the judge’s assessment of damages.