Work Programme evaluation - effectiveness of providers

[ employment-and-welfare  randomised-trial  modelling-and-descriptives  ]

Under the Work Programme, employment services in Britain are contracted out to a mix of for-profit (FP) and not-for-profit (NFP) providers who are paid according to the results they achieve. Providers are free to choose whichever strategy they feel is likely to be most effective - the so-called ‘black box’ approach - and are paid according to the results they achieve. As part of the programme of research evaluating the Work Programme, we carried out two projects.

The first project explores the extent to which the characteristics and behaviour of Work Programme providers explain their effectiveness. A particular focus is on differences between FP and NFP providers. The research uses econometric techniques to model variations in effectiveness across providers. We construct measures to capture variations in service delivery across providers and relate these to estimates of provider effectiveness. The research exploits the fact that at least two prime providers operate within each of the 18 “Contract Package Areas” and that individuals entering the Work Programme are assigned to a prime on a random basis. This means that, within each CPA, we can derive experimental estimates of the relative effectiveness of prime providers. We explore the extent to which relative effectiveness is associated with type of service delivery, prime contractor characteristics, individual characteristics and area characteristics.

The second project estimates the impact of the Work Programme on young Jobseeker’s Allowance (JSA) claimants. It does this by exploiting a difference between 24- and 25-year-old unemployed people in the rules governing their participation. For 24-year-olds, participation was compulsory after 9 months on JSA. For 25-year-olds, it was compulsory after 12 months. We 24- and 25-year-olds to be similar with regard to both observed and unobserved characteristics and therefore to be likely to experience similar outcomes, if subject to similar influences. Since they differ in terms of WP participation, this provides a means of estimating the short-term impacts of the Work Programme. This is done within the framework of a duration model in order to take account of the fact that the majority of young people beginning a JSA claim exit relatively quickly and so do not enter the Work Programme.


Dorsett, R. and Lucchino, P. (2016) Work Programme: factors associated with differences in the relative effectiveness of prime providers DWP ad hoc research report no. 26


Department for Work and Pensions