Skip to main content
Guidance

Ministry of Justice: Data First

Data First is a pioneering data-linking, research and academic engagement programme led by the Ministry of Justice and funded by ADR UK.

Data First unlocks the potential of the wealth of data created by the Ministry of Justice (MOJ) by making linked administrative datasets from across the justice system available for research. The programme is led by MOJ and funded by Administrative Data Research UK (ADR UK), an investment by the Economic and Social Research Council (ESRC).

Data from the courts, prison and probation services in England and Wales have been linked to enable new and innovative analysis of user journeys, interactions, and outcomes across the justice system. The programme is also enhancing the linking of justice data with other government departments, including education data from the Department for Education’s (DfE) National Pupil Database (NPD).

Data First enables researchers across government and academia to access these datasets in an ethical and responsible way via secure platforms in the ONS Secure Research Service and SAIL Databank. 

By working in partnership with academic experts to facilitate and promote research in line with evidence priorities set out in the MOJ Areas of Research Interest (ARI) Data First is generating new insights to inform the development of government policy and drive real progress in improving justice outcomes.

General programme information

The Data First user guide provides further information about the programme, including the processes for accessing the data for research. The privacy and data protection statement provides information about how we use and share data.

Data First datasets

Data First cross-justice system datasets

Data First has shared seven datasets derived from administrative sources across the justice system in England and Wales. These include data from magistrates’ courts, the Crown Court, prisoner custodial journeys, probation services, offender assessments, and both the family and civil courts.

To support analysis, a cross-justice system linking dataset is available. This dataset enables person-level linkage across the seven datasets. The cross-justice system linking dataset also includes a case-level linking table for connecting the magistrates’ courts and Crown Court data. For family and civil courts specifically, while person-level linkage is available for some records, limited personal identifiers make it harder to link individuals appearing in these datasets.

Data First Ministry of Justice-Department for Education (MoJ-DfE) data share

Separately, criminal history data from the Police National Computer (PNC) has been linked to education and social care data in England, sourced from the DfE’s National Pupil Database (NPD) as part of the MoJ-DfE data share. The MoJ-DfE metadata (ODS, 1.54 MB) provides information on the datasets and variables available within the data share.

Data catalogues

Data catalogues are available for all Data First cross-justice system datasets, providing information on the variables contained within each.

Synthetic datasets

Data First has released eight low-fidelity synthetic datasets, these can be accessed via the UK Data Service  (UKDS). These datasets replicate the structure of the Data First cross-justice system datasets, without containing any real information about the individuals or cases in our secure datasets. They have been designed to help researchers become familiar with the data structure, prepare feasible research proposals, and develop analytical plans and code for deployment once research projects and secure access are approved.

To access the synthetic datasets, researchers must register for a UKDS account  and agree with the terms and conditions stated in the UKDS End User Licence Agreement.

Available synthetic datasets include:

Limitations of data linking

Users should note that the accuracy of data linking is determined by the availability of personal identifying information in source data. All available identifiers have been used during the matching process, but the availability of demographic information varies by dataset. The Data First criminal justice datasets (magistrates’ courts, Crown Court, prisoner custodial journey, probation and offender assessments) contain numerous, well-populated identifiers. The family and civil court source datasets however contain fewer identifiers than other datasets. As a result, person-level linkage rates are lower for family and civil court data, which may affect the feasibility of research designs that rely on linked records. Further details can be found within the user guide and relevant data catalogues, and researchers are encouraged to contact us to discuss project ideas at their earliest convenience so we can advise on viability.

Applying for data access  

Data First datasets can be accessed through the ONS Secure Research Service (SRS) or SAIL Databank (except for the MOJ-DfE data share, which is only available through the ONS SRS).

Requests to access data through the ONS SRS require completion of the Secure Access to Data Form, which can be accessed here along with additional supporting documents.

Guidance for completing the application form can be found in the Data Sharing Guidance, and the list of datasets and access routes can be found here. Further information on the process overall is included within the Data First user guide above.

To access data within the SAIL Databank please apply though SAIL.

A register of external research projects which have been approved to use MOJ data is available to view here.

Analytical outputs

Statistical and social research publications using Data First data have been delivered by MOJ analysts or in collaboration with other government departments. Outputs have also been produced by ADR UK-funded Research Fellows.  These publications can be found below:

Through Data First, MOJ has developed a free and open-source software library to enable data linkage at scale. This software has been used to link some of the largest datasets held by MOJ as part of Data First.

Splink is a freely available, open-source Python package that is:

  • faster and more accurate than other free tools
  • able to link large datasets, of tens of millions of records or more
  • developed with advice from academic experts in data linkage
  • able to produce a wide range of interactive data visualisations that help to build effective models, explain linkage predictions, diagnose problems and quality assure models
  • compatible with multiple databases and big data processing engines, meaning it can run on a wider range of computer systems

You can find out more on the Splink website, where you can download and start using Splink. You can also ask us a question or raise an issue on the public GitHub repository. Splink are happy to hear from researchers interested in using the software for their work.

Awards and Recognition

Contact

Contact the Data First team at datafirst@justice.gov.uk if you would like further information or have any queries.

Sign up for emails or print this page