Oslo introduced a combination of school choice, per capita funding, balanced management and accountability in their public schools. Recent studies point out that this has increased segregation. In this study, teachers have been interviewed about their experiences. Bernstein's "classification" and "framing tools" have been used to analyse the consequences for schools and relations between schools and parents/students. 'Marginalised' and 'privileged' schools find themselves in negative and positive spirals when it comes to popularity. These spirals are classed, raced and, (in upper secondary school), also gendered. Since attracting the 'right' students and avoiding getting the 'wrong' ones is essential for both school categories, school choice creates a "mutual interest" between the school and privileged parents/students in fortifying the latter´s voice. Three findings are especially interesting: 1. Cream skimming occurs in undersubscribed schools in a strictly public-school context. 2. School choice affects internal priorities in marginalised schools so that segregation at the class level increases, thus the educational context may be more segregated than what is indicated by school level information. 3. School choice increases segregation in the local communities, as two schools near each other may have very different student compositions. Segregation is thus not only explained by segregated housing.