The provision of equitable access to education regardless of gender, social class and geographical location has traditionally been a central idea in Nordic education policy and an imperative tool for promoting social justice through education (Holm & Lundahl, 2018). The educational reforms carried out in Sweden in the 1990s brought about deregulation, privatization and marketization (Lundahl, 2016). Politically, solutions to social differentiation in education shifted from strong state governance and ideas of collective uniformity towards beliefs focused on choice, market superiority and private provision education (Dahlstedt, 2011). These reforms consequently instituted a quasi-market setting and vastly changed the organization of the Swedish educational system to one in which access to education is provided through choice. Neoliberal restructuring through privatisation not only entails changes in structural organization, but it also affects human interaction, social practices and individual self-perception (Beach, 2010). These developments of Swedish marketization was (and is) part of a global encompassing movement where public sectors organizing welfare services are subjugated to neoliberal restructuring, commodification and privatization (Beach, 2010; 2018). A critical examination of the Swedish type of reform-related marketization will contribute valuable knowledge to a debate on effectiveness, choice and social justice in educational sectors both in a European and in a global context.

The marketization process in Sweden has materialized spatially in a de-bordering of municipalities as administrative territories and in a new and expansive urbanized mobilizing of students between these, which has restructured the upper secondary school market. Transitioning into this quasi-market system, Swedish students are now navigating a completely transformed educational landscape – one in which new choices, mobilities and spatial trajectories have been made possible. The choices that are available to an individual in a school market depend on multiple factors, including grades, place of residence and the provision of educational services in the local area or within commuting distance. The educational supply in the Swedish upper secondary school market is geographically differentiated; market segmentation is apparent between regions, municipalities and cities, and it includes contexts such as rural and urban (Fjellman et al., 2018; Skolverket, 2011; 2013). These segments are also socially differentiated and increasing school segregation has been attributed mainly to school choice and residential segregation (Yang & Gustafsson, 2016; Andersson et al., 2012; Trumberg, 2011). Post-reform, students now commute further to undertake education and their movement is largely enabled by the school choice mechanism. Within this pattern, the metropolitan school markets are an especially interesting case. Whilst experiencing a massive geographical expansion and encapsulating a large surplus of educational supply, these markets also have heavy inter and intra mobility. Given all of this, it is important to ask who is being mobilized?

Choice-enabled mobility in a marketized sphere is economically, politically and socially constructed (Urry, 2007; Cresswell, 2006) and spatial relations do not operate independent of social processes (Duncan, 1989). School choice decisions and strategies are strongly related to class, ethnicity and family background (Jonsson & Beach, 2015; Kosunen & Carrasco, 2016; Ball, Macrae & Maguire, 2013). Social groups are also positioned differently in mobility flows (Massey 1991; 1993) and analysing who is entrenched in these is crucial for understanding choice implications in equal access to education in a quasi-market setting. Examining who is mobilized over time further contributes important knowledge towards understanding the ongoing restructuring of the Swedish quasi-market and the spatial outcomes of marketization. The aim of the study is to explore the mechanism of educational choices in relation to students’ social background and commuter flows within three Swedish metropolitan upper secondary school markets (Stockholm, Gothenburg and Malmö) between 1998 and 2011 through a socio- spatial theoretical framework.

Register data from the Gothenburg Educational Longitudinal Database (GOLD) were used in the analysis. GOLD contains background characteristics as well as geographic codes for residential municipality and school municipality for all individuals born between 1972 and 1995 in Sweden. Four subsets of first year students residing in and/ or attending an upper secondary education in one of the metropolitan school markets (Stockholm, Gothenburg and Malmö) from cohorts 1998, 2003, 2008 and 2011 were included. The current analysis was carried out in two stages: (1) propensity score analysis and (2) multiple linear regression analysis were applied to an analysis of register data (N=142 816). Using the conditional probability obtained from the propensity score analysis as a dependent variable, a hypothesis that pscore could be predicted by individual characteristics was explored through multiple regressions sorted by public and private providers for each year and school market. Commuter and upper-secondary school markets identified in a previous study (Fjellman et al., 2018) were used as data. To estimate and compare between-group differences in outcomes (e.g., commuters and non-commuters) and to attribute such differences to group belongingness by utilizing observational data, the inherent selection bias needed to be addressed (Caliendo & Koening, 2008). Propensity score analysis was used to address such bias by mimicking some of the characteristics of a randomized controlled trial (Guo & Fraser, 2010). The purpose of using propensity score analysis was to estimate the probability of commuting to undertake education outside one’s residential municipality, controlling for students’ upper secondary program choices. Thus, the research question in focus relates to whether a student’s background characteristics affect the probability of them commuting for upper secondary education, given students’ upper secondary program choices.

A hypothesis was posed: that students’ background characteristics predict the conditional probability of commuting to undertake upper secondary education, given their educational pathways. The results from the analysis support the hypothesis. However, the support is dependent on both temporal, spatial and regional qualities. Student background characteristics can predict to varying degrees the conditional probability of commuting within Swedish upper secondary metropolitan markets. The explanatory success of the predictors included in the models is reliant on educational ownership, as the likelihood of commuting for upper secondary education is interrelated with who provides the educational pathways the students are choosing. Temporal aspects are interesting in relation to what choice is on offer at that time in the educational market and by whom. For example, the different effects of the compulsory grade predictors suggest either high ability students are not commuting to the same extent for private providers or that low ability students are seeking specifically non-local private upper secondary education. How the upper secondary school market space is organized is therefore influenced by choice of educational provider. The effects of compulsory grades being mediated through choice of educational provider essentially enables lower-achieving students with the same educational pathways to enter a corresponding program as their peers in public schools if they are willing to commute to a private school for these observed years. As the markets were growing, the market space for educational opportunities and opportunities to move developed differently for students choosing private providers, even if they chose the same upper secondary programs. If at any point private providers served as an alternative to lower achieving students interested in programs they could not gain access to at a public school, this opportunity was not equally attainable by all student commuters, as student groups were not mobilized evenly across municipal borders.

Andersson, E., Malmberg, B., Östh, J. (2012) Travel-to-school distances in Sweden 2000- 2006: changing school geography with equality implications. Journal of Transport Geography, 23, 35-43. Ball, S., Macrae, S., & Maguire, M. (2013). Choice, pathways and transitions post-16: New youth, new economies in the global city. Routledge. Beach, D. (2018). Structural Injustices in Swedish Education. Palgrave Macmillan, Cham. Beach, D. (2010). Neoliberal Restructuring in Education and Health Professions in Europe Questions of Global Class and Gender. Current Sociology, 58(4), 551-569. Dahlstedt, M. (2011). I val (o) frihetens spår: Segregation, differentiering och två decennier av skolreformer. Pedagogisk forskning i Sverige, 12(1), 20-38. Duncan, S. S. (1989). Uneven development and the difference that space makes. Geoforum, 20(2), 131-139. Caliendo, M., & Kopeinig, S. (2008). Some practical guidance for the implementation of propensity score matching. Journal of economic surveys, 22(1), 31-72. Cresswell, T. (2006). On the move: Mobility in the modern western world. Taylor & Francis. Fjellman, A. M., Yang Hansen, K., & Beach, D. (2018). School choice and implications for equity: the new political geography of the Swedish upper secondary school market. Educational Review, 1-22. Guo, S., & Fraser, M. W. (2010). Propensity Score Analysis - Statistical Methods and Applications. Thousand Oaks CA: Sage Publications Inc. Holm, A. S., & Lundahl, L. (2018). Stimulerande tävlan på gymnasiemarknaden? Dahlstedt & Fejes (Eds.) Skolan, framtiden och marknaden. Stockholm: Studentlitteratur. Kosunen, S., & Carrasco, A. (2016). Parental preferences in school choice: comparing reputational hierarchies of schools in Chile and Finland. Compare: a journal of comparative and international education, 46(2), 172-193. Lundahl, L. (2016). Equality, inclusion and marketization of Nordic education: Introductory notes. Research in comparative and international education, 11(1), 3-12. Massey, D. (1991) A global sense of place. Marxism Today (38) 24-29 Massey, D. (1993). Power-geometry and a progressive sense of place. Bird, J., Curtis, B., Putnam, T., & Tickner, L. (eds.) Mapping the Futures: Local Cultures, Global Change. Trumberg, A. (2011). Den delade skolan- Segregationsprocesser i det svenska skolsystemet. (Doctoral thesis, Örebro studies in Human Geography, 6). Kållerd: Intellecta Infolog. Urry, J. (2007). Mobilities. London: Polity. Yang Hansen, K., & Gustafsson, J. E. (2016). Causes of educational segregation in Sweden–school choice or residential segregation. Educational Research and Evaluation, 22(1-2), 23-44.