Uses and linking of adminstrative data

Session organiser: Dr. Emma White, Administrative Data Research Centre – England, University of Southampton

1.30pm Monday 7 September

Reconstructing women’s fertility histories through retrospective questions: Is sample survey data reliable?
Francesca Rinesi, Marina Attili, and Claudia Iaccarino; Italian National Institute of Statistics (ISTAT)

During the last decades increasing importance has been given to the study of reproductive histories based on event history approach. Data on reproductive life-course is collected either through longitudinal surveys or - more frequently - through surveys which include retrospective questions. Given the increasing supply of survey data collected retrospectively and their wide use in scientific research, it is becoming crucial to assess their reliability. This paper precisely aims to evaluate the reliability and accuracy of women reproductive histories collected by the Italian National Institute of Statistics with the Follow-up Sample Survey on Births, by comparing self-reported information on childbearing histories with evidence that comes both from the Sample Survey on Births (wave 1) and from Population Registers (list of all live-births from women resident in Italy). In order to do so, deterministic record linkage techniques have been applied. In Italy no unique national identity number can be used as matching key. We therefore use a combination of variables present in all data sources we considered such as complete name, surname, exact date and place of birth. The main strength of this paper is that the comparison of fertility histories from three different data sources (but referring to the same women) takes place at individual level. This comparison makes it possible not only to evaluate the Follow-up Sample Survey’s data quality, but also to correct its records in case of clear inconsistency.


Barriers to exploring the educational outcomes of children looked-after in England using administrative data
Louise Mc Grath-Lone, Lorraine Dearden, Bilal Nasim, Katie Harron & Ruth Gilbert; Administrative Data Research Centre England, University of Southampton

Background: Government statistics indicate that looked after children (LAC) in England have poorer educational outcomes than their peers. However, these analyses are restricted to LAC currently in care who have been continuously looked after at least six months rather than those who have ever been in care. We explored the feasibility of creating cohorts of children ever looked after and followed up for school attainment using Department for Education data, namely Children Looked After (CLA) return and National Pupil Database (NPD).
Methods: Unique Pupil Number (UPN), assigned on enrollment to a state-maintained school or nursery, is the common identifier in CLA and NPD used by DfE to link children’s social care and educational records. Using CLA data, the coverage of UPN among children born in 1998 who entered care was explored.
Results: Of LAC born in 1998 (n=4,668) 31% had no UPN recorded as they exited care before school entry (i.e. age 5). UPN was also missing for 12% of LAC in care during school-going years. Overall, just 67% of this cohort had UPN recorded. Name is not recorded in CLA and available identifiers were not sufficiently specific for probabilistic matching of children without UPN recorded.
Conclusion: Potential effects of early childhood experiences of care on educational outcomes cannot be explored using linked CLA-NPD data as coverage of UPN is incomplete. Going forward the collection of names in CLA should be considered to link children’s social care and educational records.


To be followed by a panel discussion.