Outcomes of Social Service Contact with Teenagers in England

It is notoriously difficult to demonstrate whether social service contact achieves its intended outcomes, especially where there are no experimental conditions or follow-up. Large longitudinal cohort studies could be powerful... [ view full abstract ]

It is notoriously difficult to demonstrate whether social service contact achieves its intended outcomes, especially where there are no experimental conditions or follow-up. Large longitudinal cohort studies could be powerful data sources, allowing large scale comparison between those using and those not using social services over time. However these data are much under-used in social work research, especially in the UK.

This paper reports on one strand of a larger study of four such datasets in the UK. It draws on the Longitudinal Survey of Young People in England to investigate the effects on young people of social service contact due to problem behaviour in their teenage years.

The analysis uses inverse-probability-weighted regression adjustment, a technique rarely used in social science, to study causation. By adjusting both for the predictors of social work contact and the outcomes of these predictors, it allows us to compare young people who have social services contact with others in similar in similar circumstances who do not, thus isolating as far as possible the effects of social work contact. The outcomes of social work contact considered are educational achievement and aspiration, mental health and locus of control.

Our findings suggest there is no significant difference between teenagers who receive social service contact and those who do not in terms of their mental health outcomes, aspiration to apply to university, or belief that they will struggle in life even if they do well in school. However the evidence suggests that those who receive social service contact have significantly lower odds of standard educational attainment and of being confident in being accepted by university if they apply. Social service contact also yields significantly higher odds of reporting external locus of control. Possible explanations and implications of these findings, along with the limitations of the data, are discussed.