The Ministry of Housing, Communities and Local Government have carried out ground breaking work over several years to put in place a robust, rigorous and innovative impact evaluation of the expanded Troubled Families Programme: a targeted intervention programme for families with multiple issues including crime, anti-social behaviour, truancy, unemployment, mental health problems and domestic abuse.
 The evaluation was set up in 2015 and runs until the end of the Programme in 2020. The research design allows us to observe the impact of the Programme at a national level on a wide range of outcomes it aims to affect, spanning the remit of several government departments. It also enables us to control for family and individual characteristics to ensure that a 'comparison group' can be used to derive a robust estimate of the impact of the Programme on those who have taken part.
 A key aspect of the evaluation is the use of matched nationally held administrative datasets, including the National Pupil Database (which provides educational attainment and attendance as well as data for Children in Need), the Police National Computer (which provides data on offending behaviour) and The Work and Pensions Longitudinal Study (which provides data on employment and benefits). With the help of our trusted third party contractor (ONS), this data is collected and matched to individual level information of families eligible for the programme provided by local authorities (for both those on the programme and in the comparison group).
 There are many advantages of using administrative data to measure outcomes: we have access to a wealth of historical data on a large number of individuals and families which has allowed us to control for a range of covariates and to develop Propensity Score Matching models to compare the outcomes of those on the programme and the comparison group. The data collected from local authorities also means we can analyse the data at an individual and family level, this is extremely rare with datasets of this nature.
 The impact and wider evaluation findings have provided valuable evidence to support the development of the policy over the programme’s lifetime. The impact results have provided an assessment of what effect the programme is having on child and adult outcomes for those on the programme compared to those who have not yet started. In particular, it allows us to consider the impacts on service use, and attributing costs and benefits to these changes in use of services allows us to weigh up the overall public value of the programme. Other elements of the evaluation, such as the area level case studies and family survey have provided some context for our findings.
 The presentation will discuss the evaluation design, the impact results for child social care, crime, employment and benefits, along with findings from the cost-benefit analysis of the programme.