Overview of administrative data on children’s social care in England

Providing an overview of the administrative data on children's social care that exists, how the data is currently used, and suggesting ways it could be used for future research

Focus areas


Report documents

Full report
(PDF, 3MB)

Providing an overview of the administrative data on children’s social care that exists, how the data is currently used, and suggesting ways it could be used for future research


This short report gives an overview of administrative data available on children’s social care in England, to encourage researchers and local authorities to make more use of these datasets.


The report introduces the national datasets available in England and gives just a few examples of how they have been used to date and how they can be exploited more in future.

The focus on the report is national datasets – that is, where each local authority in theory collects data in the same way. There are also national administrative data in the family justice arena, e.g. from Cafcass, but the report does not cover these.

How we went about it

This is not a systematic review of the use of these datasets but selected illustrative examples are presented from the results of database searches using some key terms, in addition to studies already known to the authors.
Key findings

There are two main types of administrative data set – aggregate and individual-level.

Aggregate data sets are where someone has already calculated results from data on individuals and made those results available for others to use. The most obvious example is data at the level of a local authority. The Department for Education provide a Local Authority Interactive Tool which allows LAs to compare themselves with others on a range of measures – e.g. demographic characteristics of the population, numbers of children in need and looked after, the social care workforce, education, youth offending and spending. These datasets have been used by researchers to study change over time and what drives differences between LAs. For local authorities themselves they provide a very useful benchmark against other areas with similar populations.

Individual-level data sets are ‘raw’ (but anonymised) data on individual children. The national data sets are those that local authorities are obliged to compile and submit to Government each year and they are the basis for calculating aggregate data and statistics for the whole country. It is possible to request permission to access the national data sets.

The two individual-level data sets available on children’s social care are the Children in Need Census and the Children Looked After return. These have been used to date for research on (e.g.) the costs of child welfare; social inequalities; patterns of intervention over time; and the likelihood of re-referral, amongst other topics.


Here are some of the arguments for using administrative data in social care research, audit and evaluation:

  • The samples are much more representative and much larger than can be achieved through any means of gathering data
  • The data are not subject to reporting bias such as the stigma of reporting some kinds of contact with children’s services.
  • There can be data on the same people over many months and years, so you can look at change over time.
  • Using administrative data rather than surveys or interviews means that people do not have to disclose sensitive issues to researchers. These datasets can often be easily anonymised, reducing ethical difficulties with their re-use.
  • There is potential for children’s social care data to be linked to other data sets – e.g. education or health – making it much easier to assess certain outcomes that would be possible from any other type of study

What next

Local authorities have a lot to gain from doing more analysis of their own individual-level data, for example for monitoring trends in services and conducting strategic needs assessments.

Data linkage is where efforts are now especially needed. It is perfectly possible to link different administrative data sets on an anonymised basis, so that individuals are not identifiable. This is already routine in research on health care. A few research teams are already doing this with social care data, but this approach needs to become more mainstream in future, to make it possible to assess some of the outcomes of social care without the need for expensive and intrusive further data collection.

Read the full report