Government-funded research report sets benchmarks to test feasibility and accuracy of matching education data with GP records to mimic proposed LA databases
ADR Wales [Administrative Data Research Wales] is an organisation with a broad remit across different policy areas. As their Co-Director and Chief Statistician Stephanie Howarth states in the organisation’s introductory video, “Wales as a nation is well-placed to make use of data-linking to help us work across organisations and subject boundaries…”
ADR’s work is further described in this way:
“ADR Wales brings together world-renowned data science experts, leading academics and specialist teams within Welsh Government to produce evidence that shapes future policy decisions in Wales. The partnership is ideally placed to maximise the utility of anonymous and secure data to shape public service delivery, which will ultimately improve the lives of people in Wales.” [Emphasis added]
On 18 September, ADR announced the publication of a report which has relevance for home educating families in Wales. The brief synopsis reads as follows:
“ADR Wales has published a report examining how to estimate the number of children not enrolled in state schools across Wales. Using linked data from the Secure Anonymised Information Linkage (SAIL) Databank, the study supports potential regulations to create a database for tracking children missing from state education.
Under the Education Act 1996, parents are responsible for ensuring their children receive appropriate full-time education, either through school attendance or alternative methods, while local authorities must identify those not receiving it. The Welsh Government is proposing that local authorities develop a database of linked education and GP registration data to monitor this.
The report tested this approach by linking anonymised education and GP records, estimating the number of children missing from state education as of April 2021. Children were grouped into those still in education, those no longer recorded, and those unaccounted for. The report also examines why some children appear in GP data but not in education records, including the effects of attending independent schools or schools in England.
The findings were compared with Office for National Statistics analysis using 2021 Census education data for accuracy.”
The full article adds more context:
“To assist local authorities to fulfil their duty, the Welsh Government is proposing regulations to require local authorities to establish a database of children missing education and issue statutory guidance using existing law. It is envisaged that the database would be produced by linking education data to GP registration data to identify those children missing from state education.
The report carried out by the ADR Wales team, replicates the proposed data linkage method using de-identified education and GP records held within the SAIL Databank to test the feasibility and accuracy of the proposed approach.”
Space precludes in-depth analysis of the report, but the Summary and Introduction provide a helpful review of the report’s purpose, its findings and their limitations. Clearly it was never going to be possible to establish watertight figures when using anonymised data and with so many variables. This is highlighted in the extract below for instance, by the frequent use of words such as ‘may,’ ‘likely,’ and ‘suggests.’ But that was not the intention of this exercise.
“The approach found that approximately 6.4% of children (27,000) in the dataset of GP registrations could not be found in the Pupil Level Annual School Census (PLASC) or Educated Other Than At School (EOTAS) data on 20 April 2021. This is likely to represent an upper band estimate of children missing from state education. Reasons for children not found in PLASC or EOTAS data may include the following:
• Children educated in independent schools (approx. 8,000)
• Electively home-educated children
• Children educated in England
• Data linkage issues, possibly due to discrepancies in recording of name, address or date of birth
• Over-coverage in GP registration data (e.g. due to children moving away from Wales but not de-registering with their GP)
• Other reasons
The approach suggests there could be a wide variation of children missing from the education data between local authorities, potentially between around 3.0% and 15.3%. These differences may be partly explained by the presence of independent schools and proximity to schools in England of some local authorities. The approach also suggests a difference between school ages (possibly between about 4.9% to 9% increasing with age). These are estimates only given the limitations of the approach.”
If there can be no certainty from this research, why then has the WG spent taxpayers’ money commissioning it?
An earlier paragraph in the full article explains:
“The report, which uses linked administrative data held within the SAIL databank, has tested the feasibility of estimating the numbers of children not in state education within local authorities in Wales ahead of a possible regulatory change in Wales.”
The WG has stated that it will be evaluating the pilot schemes whilst they are in progress. This research was commissioned in order to provide them with a measuring rod for that evaluation. If the outcome of the data-harvesting exercise results in figures significantly different from those identified by this research, and particularly if the numbers identified are much lower, then as the Cabinet Secretary stated in her response to written questions from Rhys ab Owen MS:
“The evaluation report will help inform whether separate measures, including a duty to notify local authorities if a child is not receiving education at a maintained school in the area, may be necessary.”
To put it simply, if the numbers of children not registered at a state-maintained school found by the pilot schemes are significantly lower than those identified by this research, then the plans for databases across Wales will be torn up, and the government will have to go back to ‘Go’ and look at primary legislation to make it mandatory for parents to enter their children’s names on LA registers, in a similar way to what was proposed in England through the 2022 Schools Bill.
It is not clear why the WG has sought to avoid bringing forward primary legislation for mandatory registers, but it was a long-standing bone of contention between them and Sally Holland, the previous Children’s Commissioner for Wales, who insisted that this would be necessary throughout her term of office. It seems that the databases have been seen as an off-the-shelf solution and therefore perhaps a simpler option than a measure which, despite numerous Westminster Bills over the last 15 years, has yet to be achieved in England!
Putting this research alongside Neagle’s answer to ab Owen makes clear that the proposed CME databases will stand or fall on the numbers of children who are ‘discovered’ through them. Of course, if they fail to match expectations, HE families and their allies should find it much easier to argue against mandatory registration – if the WG found sufficient time in Senedd to put a Bill forward. That is a battle which has to be faced more immediately in England, when the promised Children’s Wellbeing Bill is tabled.
In the meantime, the existence of a Secure Anonymised Information Linkage Databank (SAIL) is a reminder of how governments are increasingly using big data to oversee the lives of their citizens. It is now the go-to tool when developing things like policy planning and service delivery, not to mention tracking and tracing children.
When we consider the Education Secretary’s recently reiterated intention to populate databases with personalised data sourced from health boards, it is even more worrying. The responses to the recent consultation from home educators should have informed her that the vast majority of home educating families have zero confidence that these will serve any positive purpose in “ultimately improv[ing] the lives of people in Wales.”
