The creation of the welfare state in Sweden as an idea (in the 1930s/ 1940s) and then as practice since the 1950s has undoubtedly been a huge achievement. It is remarkable that a country with a relatively small population and only recent industrialisation could create one of the most comprehensive welfare systems in the world. However, Sweden's welfare system is not, and never was, paradise (Pringle, 1998). In this chapter, I will draw on a recent study of children's welfare in Sweden (Pringle, 2002) to suggest that the Swedish system, and the society that contextualises it, is permeated by discriminatory power relations associated with ageism, racism and sexism; and that these can also be seen as operating within the policies and practices of child protection.

Such a view runs counter to the prominent and positively oriented social policy analysis of Esping-Andersen (1990), who argues that welfare systems that approximate to the ‘social democratic’ welfare regime type (like the Nordic welfare systems) are the most progressive. However, Esping-Andersen's assessment was largely restricted to a class analysis of welfare systems, and was also limited by its focus on income transfers. More recently, some feminist and profeminist commentators have sought to broaden the ambit of welfare under scrutiny to include, for instance, aspects of social care such as day care for children or services for elders (Anttonen and Sipila, 1996). Yet even that broader approach has restricted itself largely to (albeit important) issues of employment and labour in the workplace and the home (Sainsbury, 1999). In this chapter, I want to take the debate further by looking at the performance of the Swedish welfare system in terms of the way it responds to violence against children. Such a broadening of the critique also further problematises the idea of the Nordic welfare regime being, in any simple sense, ‘welfare progressive’ (Chapter One in this volume).

The Swedish study

I pursue these themes through analysis of part of a recently completed major qualitative study of age, ethnicity and gender in the Swedish child welfare system (Pringle, 2002). The data consist of written verbatim transcripts from semi-structured and audio-recorded interviews with 37 respondents.