This thesis describes and analyses in three substudies how home-based measures for children are expanding and why an open care idea are established as part of the Swedish child welfare. The first substudy describes the national increment of what today can be considered as a treatment policy – non-institutional care in child welfare. The results are analysed with Kingdon's agenda setting theory and shows that the factors of what Kingdon describes as problems-, politics- and policyflow influenced the national increment. Significant mechanisms have primarily been, the profession, the society's concerns for socially disadvantaged children, negative experiences of institutionalisation and a political position to deinstitutionalise all care. The second substudy focuses on how knowledge and research contributes to ideological, professional, and organizational change in child welfare by analysis of Research & Developments reports and articles from the professional journal Socionomen. Based on the neo-institutional concepts of organisational fields and isoformism the study's results shows that a consequence of adaptation strategies and "rationalized myths" are a number of similarities in how home-based measures are organized and which measures that are used. The third substudy, a case study, analyses how we can understand a local development of ideas in social services for children with concepts from neo institutional theory. Interviews and municipal documents are analysed and shows that the local translation of an open care idea is characterized by discretion, personal preferences and action and affects the choice of method and organization. The study also shows a political mistrust of institutional care while there is at the same time political confidence in the individual social worker's ability to find solutions that allow non-institutional measures to be an alternative to institutional care.

An overall conclusion of the thesis is that there is now a treatment policy in social services for the children, which involves extensive efforts at home and it has led to that more children receive support from social services. One result of non-institutional care increment is that it led to increased costs for municipalities for intervention for children and young people when out-of-home care has not decreased. Another general conclusion is that translation and adaptation of ideas to local contexts is something that generally fits street-level bureaucrats need for discretion.