The question of declines of socio-economic inequalities in education is an important issue to both sociologists and policy-makers. There are two schools of thought – one maintaining that declines in socio-economic inequalities have occurred, and the opposing school maintaining that such inequalities have remained stable. The no-decline argument is exemplified by Shavit and Blossfeld. They conclude that, apart from Sweden and The Netherlands, there has been no decline in socio-economic inequalities in education in industrialised countries. Shavit and Blossfeld's conclusion rests on a particular methodological approach, odds ratios across educational transitions. In contrast, regression analyses of educational attainment tend to show declines, for example in Australia, Italy, The Netherlands, the United States, as do sibling studies for Germany, The Netherlands and the United States. The contradictory conclusions confuse policy-makers and anyone interested in changes in the reproduction of socio-economic inequality. The debate in this journal began with Hellevik criticizing the methods used in the Shavit and Blossfeld volume and researchers at Nuffield College, Oxford. He argues that more appropriate measures of socio-economic inequality in education, such as the Gini coefficient, show substantial declines in inequality. In response, Kivinen et al (2002) and Marshall and Swift (1999) defend the use of the odds ratio. This research note contributes to this debate by showing how odds ratios can be misleading for evaluating changes in socioeconomic inequalities in education.
Available at: http://works.bepress.com/gary_marks/43/