Does the Scottish Police Authority understand the data it is scrutinising
Introduction
On 26 September the Scottish Police Authority questioned the Chief Constable about Police Scotland’s policy on recording the sex of offenders. This blog argues that the discussion of published police data among senior members of the SPA gave the misleading impression that Police Scotland’s statistics on offenders are more credible than those in England and Wales. We argue that the exchange, which did not come across as off the cuff, exposed a surprising degree of confusion about the relevant statistics within the body tasked with holding Police Scotland to account.
Background
The controversy over Police Scotland’s recording policy dates back to 2019. For the most part it has focused on those accused of rape, and a policy that allowed men to be recorded as women. It is only within the last few weeks that the Authority has publicly demonstrated an interest in the issue, prompted by intensive political and media criticism.
Scottish Police Authority Board Meeting
In a discussion at the SPA Board Meeting on 26 September, Alistair Hay (Chair of the Policing Performance Committee) and the SPA Board Chair Martyn Evans compared the status of recorded crime statistics in England and Wales and Scotland. Their respective comments are shown below.
Alistair Hay (SPA Policing Performance Committee Chair)
“I really do appreciate your statement today and I think that provided a lot of clarity to help us going forward. Just before I come to my point, obviously the statistics within Police Scotland are accredited, which I understand is not the case in all forces in England and Wales, by the Office for National Statistics.
And in terms of the development of policy and practice going forward, the authoritative nature of those statistics and that data is something that’s important and we really need to strive to maintain that. So its not the substantive point today but it’s an important issue that we recognise.” [29:58]Martyn Evans (SPA Chair)
… “My final point is this. It’s a point that Alistair made. In Scotland, police recorded statistics are official accredited statistics. That is a very important matter for the Authority.
In England, they are not accredited. They’re not good enough to be accredited and haven’t had accreditation since 2014. I think we should be very proud that our official statistics, police and crime statistics are official and accredited. And as Alistair says, we will want to do everything in your report to maintain that status, which is by following the statistical guidance….
I make that point, I stress that point about official statistics because I believe if this was the other way round, if we were the only police service in England and Wales and Scotland that didn’t have that status we would be under tremendous pressure. We do have that status. And that’s down to the quality of data gathering and data recording by police.” [37:05]
What are accredited Official Statistics
Official statistics are produced by Crown bodies and other recognised organisations, on behalf of the UK government or devolved administrations. More information on what it means for statistics to be described as official can be found here.
Official statistics producers in Scotland include the Scottish Government, Scottish Fire and Rescue Service, and Public Health Scotland. Police Scotland is not on this list. Subject to developing robust enough internal processes, the service could join it. For example, the Police Service of Northern Ireland (PSNI) is a producer of official statistics. However, the service currently relies on the Scottish Government to convert raw policing data into official statistics, which are included in information published on the criminal justice system in Scotland more generally.
Accredited official statistics are a subset of official statistics that have been independently reviewed by the Office for Statistics Regulation (OSR) as complying with the standards of trustworthiness, quality and value in the Code of Practice for Statistics. OSR accreditation can be cancelled or suspended in cases of serious non-compliance.
The Office for National Statistics (ONS), incorrectly cited by Mr Hay, is a producer of official statistics, not a regulator.
What statistics are being discussed?
In the exchange above, the SPA Chair refers variously to “police recorded statistics”, and “police and crime statistics”, but does not name a particular publication.
Criminal Proceedings in Scotland
Statistics on convicted offender demographics, including gender/sex, can be found in ‘Criminal Proceedings in Scotland’. This Scottish Government publication shows data on ‘offences dealt with by courts, sentencing outcomes and characteristics of convicted offenders’.1 These statistics are accredited.
If the Chair had intended to refer to Criminal Proceedings data, his drawn-out comparison with statistics in England and Wales was misplaced. The nearest equivalent publication in England and Wales is Criminal Justice Statistics Quarterly. This is published by the Ministry of Justice and is also accredited.
Of relevance to the discussion on recording sex, data in Criminal Proceedings in Scotland is ‘generally based on how a person presents and is recorded when a person’s details are entered into the [Criminal History System]’. At the time of writing, the statistics show two rapes as having been committed by females, published in 2006/7 and 2018/19 respectively.
In 2020 we queried the latter case with the Scottish Government. We were surprised at the response, which is shown below.
I’ve investigated your query, and it appears that this case may be misclassified. I’ve found nothing in the media that would fit the bill for a female being convicted of rape during 2018-19.
Scottish Government official, email correspondence 25 June 2020
We think it should be clear to all bodies involved in the production and oversight of criminal justice statistics that reliance on ad hoc data interrogation of unpublished records or scanning media report is not a robust or sustainable approach to data collection.
Recorded Crime in Scotland
The reference to loss of accreditation in 2014 and cross-border comparison suggests Mr Hay and Mr Evans were thinking about recorded crime statistics. The annual publication Recorded Crime in Scotland is published by the Scottish Government, based on data provided by Police Scotland.2
Although accredited, this is not relevant to discussion of recording the sex of offenders, as it contains no information on this.3 This is made clear by the Scottish Crime Recording Board (SCRB). In relation to our petition on recording rape accurately, it told the Citizens and Public Petitions Committee Committee:
… [police recorded crime] measure crimes (not people) and… have never included the recording of demographic details on suspects or perpetrators.….
Scottish Crime Recording Board, 16 May 2023
It is also made clear in the Scottish Government User Guide to Recorded Crime.
Currently the Scottish Government collects aggregated numbers of crimes and offences from Police Scotland, meaning that no information about victims, perpetrators or the individual crimes or offences and incidents are available.
Scottish Government, User Guide to Recorded Crime (2021: paragraph 26.1, emphasis added)
Recorded crime in Scotland statistics lost their accreditation in 2014, but regained it in 2016 (see HMICS; paras. 10-11). As discussed by Mr Hay and Mr Evans, the equivalent figures for England [sic] and Wales also lost accredited status in 2014 and have not yet regained it.4 But they are irrelevant to any discussion of how individual offenders’ sex is recorded.
Companion recorded crime bulletins
In addition to the annual Recorded Crime in Scotland bulletin, Scottish Government publish several companion bulletins. These include Homicide, Domestic abuse recorded by the police in Scotland, Drug Seizures and Offender Characteristics, and Recorded Crimes and Offences Involving Firearms,
Of these, only Homicide in Scotland has accredited official statistics status. Homicide statistics in England and Wales are also accredited official statistics. In Scotland, data is collected on an individual case basis and includes demographics (Scottish Government, 2021: 107). It is worth noting that the number of women accused of homicide is very small, which makes the data much more vulnerable to error. Added to this, the incidence of female homicide is much more volatile than male homicide.5 Introducing uncertainty into how people are being categorised makes the year-on-year volatility in these figures even harder to make sense of.
Police Scotland management information data
Police Scotland publish some information management data.6 This is relevant to discussing how sex is recorded; but not to any discussion of accredited statistics. As noted above, Police Scotland is not an official statistics producer.
Conclusion
Stating that accreditation was “very important” to the Authority, the SPA Chair hammered home a supposedly favourable comparison between statistics in Scotland and those in England (and Wales).
Yet for the most relevant publication, the closest statistics in England and Wales are also accredited. The statistic for which the cross-border comparison on accreditation is correct is irrelevant to the discussion of recording the sex of offenders.
At the SPA Board meeting it was agreed that progress on Police Scotland’s review would report back to the Policing Performance Committee, chaired by Mr Hay. For the purposes of undertaking its scrutiny role here, the SPA needs to get a much better handle on the nature of the data and statistics that it is discussing. It might also wish to seek clarity on why, since its formation in 2013, Police Scotland has not chosen to become a producer of official statistics in its own right.
Notes
- This Scottish Government publication covers cases dealt with by courts, sentencing outcomes, out of court disposals, including police disposals, and the characteristics of convicted offenders. It explains, ‘Statistical information on the Scottish Government Criminal Proceedings database is derived from data held on the Criminal History System (CHS), a central database used for the electronic recording of information on persons accused and/or convicted of committing a criminal act. The CHS is maintained by Police Scotland and they are responsible for managing its operation and own the majority of the data’ (see ‘Data and Methodology’, page 55). For data on females convicted by main crime or offence see Table 6c.
- The Scottish Government is responsible for the validation of police recorded crime data and the production and publication of police recorded crime statistics. Police Scotland is responsible for providing police recorded crime data to the Scottish Government and also carries out internal audits.
- Recorded Crime in Scotland shows some data on police recorded crimes under the Domestic Abuse (Scotland) Act 2018 by female and male victims, and on rape and attempted rape and sexual assault by victim gender. They do not provide data on the sex of the suspect. They are a count of crimes, not people.
- Police recorded crime statistics for England and Wales are published quarterly by the Office for National Statistics (ONS) and are a measure of the number of crimes recorded by the police in England and Wales. The Home Office collates the data from the 43 territorial police forces in England and Wales and the British Transport Police and supplies it to the ONS.
- See Figures 1 and 2 below.
- This includes crime data and individual-level stop and search data. Police Scotland also share data with the Scottish Police Authority, for example, in Quarterly Policing Performance reports to the Policing Performance Committee, chaired by Mr Hay.
Figure 1. Homicide by female accused, 2003/4 to 2021/22
Figure 2. Homicide by male accused, 2003/4 to 2021/22
