Predictors and Effects of Social Service Contact with Teenagers in England
This paper makes use of data from the Longitudinal Survey of Young People in England (LSYPE) to investigate family circumstances of young people who have had statutory social services contact between the ages of 14 and 16 in relation to their behaviour. The research is funded by the Nuffield Foundation. Two broad research questions are addressed: what social factors predict social service contact among young people in England, including structural, familial, individual factors and neighbourhoods characteristics?; and what effect does social service contact have on a young person’s life outcomes? The first research question establishes associative relationships to identify who gets social service contact in England. The second seeks to identify the causal effect of having social service contact on a number of different outcomes including education, aspiration, mental health and locus of control.
Methods
We make use of two methods to answer these questions, multivariate logistic regression analysis and a ‘treatment effects’ analytical technique. The treatment effects technique obtains the doubly-robust inverse-probability-weighted regression-adjusted results (IPWRA) to estimate the effect of social service contact. This method makes it possible to estimate the causal effects of a treatment or intervention on an outcome using non-experimental data. The social world does not allow for random assignment of routine statutory social service contact, therefore we must consider methods which take into account the non-randomization in the allocation of the intervention.
Results
Multivariate logistic regression analysis found increased odds of social service contact for girls, lower social class families, children living with step-parents or foster carers and children with mixed-race backgrounds or special educational needs. There were also increased odds where parents did not get on with their children or argued frequently, did not socialise with their children or did not attend parents’ evenings in school. The police being in touch about the young person also increased the odds of social service contact, as did the child smoking or playing truant.
The examination of the causal effect of social service contact yielded some interesting results. There was no significant difference between those who received social service contact and those who did not for mental health and aspiration to apply to university. There was evidence that those who received social service contact had statistically significant lower odds of achieving five GCSE exams at grade A*-C (the prerequisite to pursue post-compulsory education in England) as well as lower odds of being confident in university acceptance if they applied. There were mixed results for locus of control.
Conclusions & Implications
This paper presents findings which are important for both research and policy in the social work community. Firstly, it makes use of an advanced statistical method (IPWRA) to identify the effect of social service contact. This method can be applied to other research questions in this field. Secondly, the findings help us to understand the ways in which young people respond to social service contact and focus our attention on the role of social services in supporting young people’s education and wellbeing.
496 words