How deficient is out-of-home care?

Some measurement problems in observational studies of children in care

Abstract

For decades, out-of-home care for children has been criticized based on discouraging results from empirical studies. Registry studies consistently show that children in out-of-home care fare worse than their counterparts from the normal population, and these studies have played an important role in highlighting the adverse situation of this group. This article thoroughly discusses the ques-tion of whether the negative outcomes among children in care can be attributed to the place-ment as such. Research on out-of-home care is based on observational studies and it is widely known that it is difficult to make causal inferences from such a design. This particularly holds true regarding out-of-home care where the placed children constitute a highly selected group. In the absence of randomized controlled trials, we can basically never know whether the poorer outcomes among placed children are due to the placement or to other factors that we lack data on. A central argument is that not even the most advanced statistical techniques can solve the chal-lenge of identifying the causal effect of the placement in observational studies. However, some of the challenges may be overcome through design, where a stronger methodological approach is to include comparison groups that are more similar to the children being placed in out-of-home care. A few methodologically stronger studies adopting such an approach give appreciably fewer indications that the placement has a negative impact on children.