2011 Census data to attainment data

You asked

Around two million attainment records from an extract of the National Pupil Database, which includes KS4 and KS5 qualifications, have been matched to the Census data, producing a significant sample size for analysis. https://www.adruk.org/news-publications/news-blogs/research-initiative-harnesses-linked-government-data-to-improve-childrens-services-110/

• Please provide the dates and available location information from which the relevant attainment records mentioned, come from.

GDPR requites that "You should write clear privacy notices for children so that they are able to understand what will happen to their personal data, and what rights they have." [Ref ICO https://ico.org.uk/for-organisations/guide-to-data-protection/key-data-protection-themes/children/]

• Please provide the privacy notice and communications that explains the linkage between the two datasets, and the purposes to which the data will be put, by whom, that were given to:

a) children and adults in households that completed the 2011 census and

b) to children and parents during collection of the national pupil database.

• Please provide any DPIA or other data processing or policy risk assessment carried out about this project, prior to the linkage.

We said

Thank you for your enquiry about the linkage of 2011 Census data to attainment data. This project represents joint working between Office for National Statistics, Department for Education, Administrative Data Research UK (ADR UK) and the Children's Commissioner for England to address the evidence gap for vulnerable children as identified in the Commissioner's 2017 review of childhood vulnerability in England.

The linkage addresses this gap by enabling us to research how a child's characteristics, and the characteristics of those in their household, could influence attainment. No other data source has this level of insight on children who are vulnerable or of concern by virtue of their identity or nationality, children caring for others, children with a disability, ill-health or developmental difficulties and children from workless or low-income families.

To combine the datasets in to a useful and anonymous resource for research we have linked 2011 Census records to attainment data from a bespoke extract of the National Pupil Database. The extract is named the All Education Dataset for England and was initially designed to enable research in to whether administrative data could support or improve the 2021 Census. Attainment records for the 2010/2011 academic year up to the 2015/2016 are included for schools in England and the data can be disaggregated to small area geographies: 2001 Census Output Areas, 2011 Census Output Areas and Local Authority.

The privacy information provided to the respondents of the 2011 Census did not include the specifics of this project as it hadn't been envisaged at that time however respondents were informed that data could be used for statistical research in the 2011 Census white paper: 'Helping to shape tomorrow. The 2011 Census of Population and Housing in England and Wales'. It noted that 2011 Census data may be used: 'as a base for creating linked samples for subsequent follow-up on topics that cannot be included in the Census itself.' We've made respondents aware of this work by publicising it through workshops with researchers in the field of child welfare and education including BBC's Children in Need, NSPCC and Children's Commissioner for England and through publishing news items on the Children and Young People page of the ADR UK website. Additional information on how ONS uses personal data, and keeps it safe and secure, including our privacy notice, is available on the ONS website. ONS does not hold a copy of the privacy information provided to children or parents during collection of the National Pupils Database as that would be a matter for the Department for Education (DfE). They can be contacted through the DfE contact portal.

The privacy of respondents is important to us and the data used for this project was linked in a secure room with access limited to ONS employees who have passed an enhanced disclosure control check. Directly identifiable data was hashed prior to linkage meaning that names, addresses and dates of birth were not visible in an understandable form to the linkers. Upon transfer to the secure research service this hashed identifiable data was removed, rendering the data anonymous to the end researcher.

Children are a vulnerable population and their data needs to be treated sensitively and with caution. We drafted the Data for Children data protection impact assessment (DPIA) to consider the risks involved with this project and to ensure we had appropriate controls in place to protect the data. This assessment applies to the 'proof of concept' phase of the project, whereby a dataset was created solely for access by ONS researchers for an approved project. Details of this project were presented to the National Statistics Data Ethics Committee and can be viewed in the NS DEC Minute, agendas and papers - Feb 2019. A new, updated DPIA was commissioned prior to this FOI request and will be completed by November when the next phase of linkage is due to begin.