Abstract
A recurring theme in evaluations of Swedish residential youth care is that treatment is often unplanned. In this paper, I show that planned treatment is strongly positively associated with treatment outcomes. In the short term, teenagers with planned treatment are less likely to experience a treatment breakdown or be reassigned to other forms of residential care after completed treatment. In the long term, teenagers with planned treatment are less likely to engage in criminal behaviour or be hospitalized for mental health problems. The results are robust to controlling for a rich set of potentially confounding factors: Even though observable pre-treatment teenager characteristics explain about one quarter of the variation in criminal behavior 5–10 years after treatment, they have almost no predictive power for whether treatment is planned or unplanned.