Send email for updates
Lloyd C., Wollny I., White C. et al.
[UK] Department for Education, 2011.
Family interventions were at the heart of the UK government’s ambition to ‘turn round’ the lives of 120,000 troubled families in England. In respect of drink and drug problems, substantial remission was seen, but the featured study could not show whether this was due to the interventions, and a report on a successor programme found no significant impacts.
Summary This report from the National Centre for Social Research evaluates family interventions set up to work with highly troubled and challenging families to tackle anti-social behaviour, youth crime, inter-generational disadvantage and worklessness. Though commissioned by the pre-2010 Labour government, its findings reflect work undertaken under the Conservative-Liberal coalition administration since May 2010 and the impact to March 2011 of its commitment to ‘turn round’ the lives of the estimated 120,000 most troubled families across England by March 2015.
The 159 local authorities which contributed data to the study made 12,850 referrals to a family intervention service during the study period, of which a fifth were in relation to adult substance use with or without other problems. Most common reasons for referral were anti-social behaviour (58%) and poor parenting (43%).
The family interventions typically lasted around a year and involved at first an average nine hours a week contact time between family and worker tailing down to six to seven hours. The core work was undertaken by key workers assigned to work intensively with about six families at any one time, building a close and trusting relationship. Their approach was assertive and persistent yet supportive, working with all members of a family so as to address the inter-connectedness of their problems. They managed the family’s problems, coordinated delivery of services, and used a combination of support, rewards and the possibility of sanctions to motivate families to change their behaviour. Persistence and assertiveness was critical to keeping families engaged and ensure they followed their agreed contract or support plan. Family intervention activities included: anger management; one-to-one and group based parenting sessions; educational support and advocacy; and organising activities for family members such as sports and arts-based activities for children and family outings/activities.
Well over 8000 families engaged with the family interventions during the study period. Typically they were highly disadvantaged. Two thirds were lone parent families and in three-quarters no adult was in employment. Of these families, 3675 had left their interventions by the time the study ended, enabling an assessment of the outcomes of the process based on official records completed by family workers. Based on their own continued contact with the family or information from other agencies, the same workers were asked to complete another set of assessments of the 775 of these families who had finished their intervention at least nine and up to 14 months previously. Of these families, 470 were assessed. They tended to be the ones most likely to have achieved successful outcomes.
Based on official statistics, by the time the intervention had ended the 32% of families experiencing drug problems had shrunk to 20% and the 29% with drink problems to 15%. These falls compared well with less steep falls in the proportions of families experiencing mental or physical health problems, but were in the same range as remissions in other behavioural problems such as poor parenting, relationship or family breakdown, and domestic violence.
No specific data was presented in the report about whether the remissions in alcohol and drug problems were sustained nine to 14 months later. Instead these were subsumed under the health domain. Within this domain, 61% of families who had problems before intervention, but had overcome these when they left, were still judged to no longer be experiencing these problems at follow-up. This was the lowest proportion among the domains, topped by education at 89%.
Generally it was not possible from these figures to say how much improvement was due to the family interventions and how much would have happened anyway, because there was no comparison group of families not offered family interventions. However, in respect of families targeted for anti-social behaviour, information on a comparison group of 93 families who would have qualified for family intervention was obtained from 11 local authorities which did not in fact offer these interventions. Of these, 56 could be followed up nine months later, indicating how such families would fare in areas without family intervention projects. In respect of crime and anti-social behaviour, 33% of comparison families were judged to have improved, compared to 63% of matched families who had been offered family intervention in other areas. The results suggested that compared to usual practices, family intervention led to another 30% of families improving. In respect of education and employment and family functioning, there was similar (but because numbers were smaller, less convincing) evidence of added value from family interventions. However, in respect of the health domain which included drug and alcohol problems, the number of comparison families experiencing these problems at the start (just 20) was too small for any conclusions to be drawn.
This report builds on the compelling evidence endorsing the role and value of family interventions. The outcomes reported at the point of exit have remained consistently high since the projects were first set up despite the increasing number of families being worked with. In the current economic climate it is very encouraging that family interventions appear to be achieving a similarly impressive set of results in a shorter time duration (from 13 months in 2010 to 11 months in 2011). However, as there is a link between the length of intervention and success we will need to wait to assess the impact of a shorter duration of intervention in the longer term.
The findings from the families compared with those in areas without family interventions provide the first indication that the positive outcomes achieved by families can be attributed to a family intervention. There is also further encouraging evidence that the outcomes are sustained nine to 14 months after leaving an intervention.
commentary The family interventions evaluated by the featured report became a core part of the Troubled Families Programme launched in April 2012 to take forward the UK goverment’s commitment to turn around the lives of the estimated 120,000 “troubled” families in England. Published five years later in October 2016, a final government-commissioned evaluation report on the programme integrated evidence collected between January 2013 to September 2015 from national administrative datasets, a large-scale face-to-face survey of families which compared families going through the programme with a matched comparison group, longitudinal case study research with 20 local authorities, and ‘snapshot’ telephone interviews with a further 50 local authorities. The report also drew on self-reports from 143 local authorities at three points in time during the evaluation, and a survey of local authorities conducted during the early stages of the programme to map the broad characteristics of local Troubled Families programmes.
Its crucial advance on the featured report was the construction of comparison sets of families not in the programme against whom its impacts could be benchmarked, helping to clarify whether any improvements really were due to the programme. Across the range of outcome measures that the programme aspired to improve including employment, benefit receipt, school attendance, safeguarding and child welfare, analyses of administrative data did not reveal consistent evidence of any significant or systematic impact. Observed outcomes for Troubled Families Programme families were very similar to the outcomes for a matched comparison group of families. The vast majority of impact estimates were statistically insignificant, with a very small number of positive or negative results.
Relative to comparison families, the separate analyses using survey data also found no significant or systemic impact on outcomes related to employment, job-seeking, school attendance, or anti-social behaviour. The measures included:
• GP and emergency department visits in the previous three months;
• self-reported general health of the respondents and young people interviewed, plus proxy data on partners;
• life satisfaction of the respondents and young people interviewed;
• use of non-prescription drugs in previous three months by the respondents and young people interviewed;
• alcohol use in the previous three months by the respondents and young people interviewed, plus proxy data on partners;
• assessments of the well-being and depression levels of the main carer respondent;
• how well parents got on within the family and physical, verbal, emotional and sexual domestic violence.
There was no statistically significant evidence of any impacts of the programme across all of these health and well-being measures.
The final report also had what it considered an important message for commissioners considering payment-by-results schemes, used by the programme to incentivise local authorities to achieve desired outcomes. Evaluators emphasised the need for these schemes “to be structured in such a way that they incentivise those responsible for delivery to improve outcomes relative to ... a counterfactual, rather than simply claiming ‘results-based’ payments for outcomes which would have been likely to occur in any event”.
Last revised 20 October 2016. First uploaded 08 January 2013
Comment/query to editor
Give us your feedback on the site (one-minute survey)
Open Effectiveness Bank home page
Add your name to the mailing list to be alerted to new studies and other site updates
STUDY 2008 Final report on the evaluation of 'Option 2'