In contemporary welfare, collaboration between professions, agencies and different municipal departments is increasingly important. Collaboration is seen as a way to address problems and to help people by working across organisational and professional boundaries. Collaboration is thus seen as different from traditional welfare institutions, where decision-making, organisations and roles are well- established and familiar.

However, new ways of organising and managing welfare raises questions about government and political-administrative relations. This dissertation addresses these questions by studying how welfare collaboration is governed. Based on an in-depth case study of collaboration on children and youth, the work of professionals, managers and politicians is observed and analysed as governing practices.
In the study, it is shown how different forms of knowledge are central in the governing practices. Beyond formal institutions and instruments of government, knowledge is put into practices which influence courses of action. Expert knowledge provided by academics; evidence within social work and management; and local knowledge held by welfare professionals – these are different forms of knowledge which important decisions and actions are based upon.

The dissertation shows how collaboration lacks the formal institutional framework that is often associated with the welfare state. But it is also argued that welfare services have always consisted of knowledge-based practices, and that collaboration therefore is not that different from how welfare has been carried out historically.

In conclusion, it is argued that the role of knowledge should be taken into account to a greater extent than is usually done in studies of welfare collaboration. The dissertation contributes to the study of welfare, its organisation and government, and it provides a theoretical contribution to research on knowledge, politics and public administration.