Employment Data Lab: Reflections on the first two years

Abstract

Objectives
Shortly after rollout, the Employment Data Lab was showcased at the 2023 ADR conference. We have now gathered two years of practical experience of linking charity and third-party data with Government datasets to conduct evaluations. We will share our reflections from the first ten projects: both good and bad.

Method
The Employment Data Lab was developed to evaluate local employment programmes run by charities and local authorities. The Data Lab’s unique position, with access to a plethora of administrative datasets, allows impact evaluations of local programmes efficiently and securely. Data lab analysts use the statistical technique of propensity score matching to create a realistic, counterfactual comparison group. They then compare the treatment group of participants with the constructed comparison group to accurately evaluate employment programmes.

Results
We will present an overview of ten completed evaluations, reflecting on using linked administrative datasets in the Employment Data Lab. The Data Lab began with just Department for Work and Pensions (DWP) and Her Majesty’s Revenue and Customs (HMRC) administrative datasets. Overtime the service has been extended to include other datasets involving education and housing and homelessness data. This expansion has allowed the Data Lab to evaluate a wider set of programmes. Moving beyond employment, to support initiatives providing skills/sports training and helping people at risk of homelessness. In time, the service will expand further by adding Ministry of Justice data to help a wider set of charities.

Conclusion
The Data Lab has helped a wide variety of external organisations to harness the power of linked administrative data. Since inception, this free service has been a tremendous success. Furthermore, increased Government emphasis on more devolved programmes and localism mean the Data Lab will go from strength to strength.