UK Troubled Families Programme — National Evaluation
UK Department for Communities and Local Government / Ipsos MORI · England, United Kingdom · 2017
Summary
The UK Troubled Families Programme is one of the most important null results in British social policy. The government had claimed that 99% of families in the programme were being 'turned around' — a claim based on administrative data that tracked meeting predefined milestones rather than comparing outcomes to a counterfactual. The independent evaluation, which used a matched comparison group, found no statistically significant effect on any outcome. The suppression of the results — leaked before official publication, with full release resisted for months — makes this a case study in institutional resistance to negative findings. The evaluation found that 'turning families around' in administrative data was often a measurement artefact, not a real change. The programme cost approximately £1.3 billion.
Research question
"Does an intensive key worker programme for high-need families reduce crime, truancy, unemployment, and reliance on social services?"
Methodology
Intervention
Local authorities assigned dedicated key workers to 'troubled families' (families with multiple overlapping problems: unemployment, school exclusion, crime involvement, health issues); key workers coordinated services and provided intensive support over 2 years
Assignment
Matched comparison group with difference-in-differences (no randomization — largest family programme evaluation in UK history)
Sample size
Approximately 400,000 families in treatment; matched 50,000 comparison families
Primary outcome
Unemployment; school exclusion; crime involvement; benefit receipt
Effect estimate
No statistically significant effect on any primary outcome; point estimates near zero for employment, crime, school attendance, and benefit receipt
Decision
Programme continued despite null results due to political commitment; a successor programme (Stronger Families) was launched in 2019; evaluation findings largely suppressed until Freedom of Information requests in 2016
Result
Null
No statistically significant effect on any primary outcome; point estimates near zero for employment, crime, school attendance, and benefit receipt
Evidence strength
Moderate
Quasi-experimental design; causal interpretation requires care.
Replication status
Open for replication
Institution
UK Department for Communities and Local Government / Ipsos MORI
Location
England, United Kingdom
Year
2017
Policy area
Benefits Enrollment
Mechanism
Community engagement