The Swedish educational system has gone through a series of reforms during the past three decades, which transformed the previously highly centralized school system into a decentralized and deregulated one with characteristics as privatization, marketization, and autonomy (e.g., Björklund, Clark, Edin, Fredriksson & Krueger, 2005; Lundahl, 2002; SOU, 2014:5). In the late 1980s, the decentralization reform gave each municipality full responsibility to organize and run schools. Introduction of free school choice along with a nationwide voucher system allows independent schools to be run with public funding in a quasi-market system. Meanwhile, Swedish society and economy has undergone a restructuring, and particularly so during periods of economic crises in the early 1990s and the late 2000s. The level of ambition of the welfare system has been lowered, and society has become more polarized with increasing economic differences between households and segregated residential areas. While the goal that educational quality should be uniform over Swedish schools is still maintained, trends towards increased socioeconomic and ethnic segregation of students have been observed in recent studies (Gustafsson & Yang Hansen, 2017). It may be assumed that these trends are at least partially due to the increased frequency of school choice (Levin, 1998; McEwan, 2000; Yang Hansen & Gustafsson, 2016). There also are tendencies towards increased differences in level of achievement between schools (Yang Hansen, Gustafsson & Rosén, 2014), which may be related to differences between municipalities in the amount of support to education (Gustafsson & Yang Hansen, 2011; Gustafsson et al., 2016). There are also indications that the reforms have caused an increasing diversity between schools and municipalities with respect to educational resources and teaching methods (e.g., Holmlund, et al., 2014, Yang Hansen & Gustafsson, 2016; Skolverket, 2009). The increasing segregation of students across schools may also be expected to lead to an increased sorting of teachers over different schools (Hansson & Gustafsson, 2016; Holmlund, et al., 2014). Given the nature of the reforms in the Swedish school system, and the development of the Swedish society, it is possible that these changes at different levels, including municipalities and schools, have brought about the increased relationship between family background and student achievement. Even though the assumption sounds reasonable, there are, however, very few studies that have examine the direct causality between these parties. The aim of the proposed study is to explain the trend of educational segregation in Swedish compulsory schools with respect to educational achievement, socioeconomic status and ethnicity, by teacher and school related resources at both school and municipality levels.

The Gothenburg Educational Longitudinal Database (GOLD) contains ample amount of educational and demographic information on all individuals being born year 1972 and forward. From 1998, a new goal-related grading system is introduced in Sweden, and school marks from the new system are available in GOLD, and these school marks of the 9th graders will be concentrated upon in this study as indicators of educational achievement. School- and municipality-level SES, the proportion immigrant students, and teacher education and competences will also be included. The individual level data become longitudinal at collective levels. This makes it possible to study and explain the development of segregation over an extended period of time of 1998-2014.Two-level Growth Curve modeling technique (Muthén & Muthén, 1998-2017, Hox & Stoel, 2005; Hedeker, 2004) will be applied to each of these variables simultaneously at school and municipality levels. The estimates of intercept and slope parameters in these growth models capture the changes in these aspects across schools and municipalities, and will be linked together to examine the causality among them.

Preliminary results showed that within-time variation in school grade heterogeneity increase over time across schools and municipalities. School average socioeconomic composition as time-varying covariate increasingly related to within-time variation in school grade heterogeneity at the municipal level, and opposite is true for the proportion students with migration background. Proportion of certified teachers at school becomes significantly related to the within-time differences in school grade. The analysis is still on-going. In the next step, we will test the growth of each of these covariates over time, and linking the growth factors of these covariates together so try to establish causality among them.

Björklund, A., Clark, M., Edin, P-E., Fredriksson, P., & Krueger, A. B. (2005). The Market Comes to Education in Sweden. An Evaluation of Sweden’s Surprising School Reforms. New York: Russel Sage Foundation. Hedeker, D. (2004). An introduction to growth modelling. In D. Kaplan (Ed.), Quantitative Methodology for the Social Sciences. Thousand Oaks CA: Sage Publications. Hox, J., & Stoel, R. D. (2005). Multilevel and SEM approaches to growth curve modeling. Wiley StatsRef: Statistics Reference Online. Hoxby, C. M. (2000). Peer Effects in the Classroom: Learning from Gender and Race Variation. .Unpublished manuscript, Cambridge, MA: National Bureau of Economic Research. Gustafsson, J.-E. (2006). Barns Utbildningssituation [Children's Educational Situation]. Stockholm: Rädda Barnen [Save the Children Sweden]. Gustafsson, J.-E., Nielsen, T., & Yang Hansen, K. (2016). School characteristics moderating the relation between student socio-economic status and mathematics achievement in grade 8. Evidence from 50 countries in TIMSS 2011. Studies in Educational Evaluation. http://dx.doi.org/10.1016/j.stueduc.2016.09.004 Gustafsson, J.-E., & Yang Hansen, K., (2017). Changes in the impact of family education on student educational achievement in Sweden 1998 – 2014. Submitted to Scandinavian Journal of Educational Research. http://dx.doi.org/10.1080/00313831.2017.1306799 Levin, H. M. (1998). Educational Vouchers: Effectiveness, Choice and Costs. Journal of Policy Analysis and Management, 17(3), 373-392. McEwan, P. J. (2000). The Potential Impact of Large-Scale Voucher Programs. Review of Educational Research, 70(2), 103-149. Skolverket. (2006). Vad händer i likvärdigheten i svensk skola? En kvantitativ analys av variation och likvärdighet över tid. [What Happened with Equality in Swedish Schools? A Quantitative Study on Variation and Equality over time]. Stockholm: Skolverket. SOU (2014:5). Staten får inte abdikera – om kommunaliseringen av den svenska skolan Betänkande av Utredningen om skolans kommunalisering, Statens offentliga utredningar [The state must not abdicate -- on the municipalisation of the Swedish school system. Report on the inquiry into the municipalisation of the school system]. SOU, 2014:5, Stockholm. Thrupp, M., Lauder, H., & Robinson, T. (2002). School Composition and Peer Effects. International Journal of Educational Research, 37(5), 483-504. Yang Hansen, K. & Gustafsson, J.-E. (2016). Determinants of Country Differences in Effects of Parental Education on Children’s Academic Achievement. Large Scale Assessments in Education. 4(11), DOI: 10.1186/s40536-016-0027-1 Yang Hansen, K., Gustafsson, J.-E. & Rosén, M. (2014) "School Performance Differences and Policy Variations in Finland, Norway and Sweden", Northern Lights on TIMSS and PIRLS 2011. Pp. 25-47. Retrieved from https://www.udir.no/Upload/Forskning/2014/Nlights%20TIMSS%20and%20PIRLS.pdf