Final Activity Report Summary - YOUTH INEQUALITIES (Factors influencing the educational inequality of young people: A European and comparative perspective)
The raison d'être of this project was to understand not only why some young people have lower educational attainments but also to understand why educational inequalities are greater in some European countries than in others and to explore the policy implications.
In understanding cross-national differences in educational inequality, a major objective of the first phase was to investigate whether background factors operate in the same way, and with the same force, in different countries. We found that there is a pattern for some countries with high levels of class inequality in educational attainment also to have high levels of ethnic or gender inequalities. In some countries there are compensating mechanisms operating such as low levels of inequality with respect to parental education. From a policy point of view we found mixed results in that the inequalities are general ones that permeate all the background factors but further analysis highlighted particular inequalities such as gender, class or ethnicity.
Our second research objective was to see whether aggregate properties of the society (or of particular regions) such as their level of income inequality, their cultural values, or their specific educational policies and funding arrangements can explain the cross-national variation in educational inequality. Here we drew on the ESS and ISSP data. The ESS was a priority project by the European Strategy Forum on Research Infrastructures (ESFRI). These data were crucial to the multilevel approach that we took. In addition to the individual respondents' educational attainments and measures of social background, measured at level 1, explanatory variables such as aggregate measures of income inequality (derived from the ECHP) and socio-political culture (derived from the ESS and ISSP) were included at level 2. Because of the well known 'small N' problem, we did divide countries into regions, greatly increasing the number of observations but providing us with some contradictory results.
Underpinning this project was its perception of being part of the next generation of research on from Shavit and Blossfeld (1993). With the emergence of systematic cross-national data sets we originally focused on the PISA data. However other projects also used PISA. In order to avoid replication we introduced additional cross-national data to complement the primary research objectives. However our research was not undertaken in isolation of other human scientists using PISA, TIMMS and PIRLS. Synergies were increasingly developed as the project progressed with other scientists met at conferences, workshops, training programs etc. These led to beneficial transfers of knowledge as evidenced in our range of deliverables.
Additional data sets included some longitudinal such as the ECHP, EU-SILC and LIS. They proved a useful tool in enhancing our understanding of contributory factors to educational inequalities in a way not foreseen when this project application was originally written. Contributory factors included poverty, health, other family effects, religion and linguistic divisions amongst others. Longitudinal analysis allows progress to be made in establishing causation. The team members developed synergies with those working on longitudinal data providing a value-added benefit to the project.
From a methodological view the inter-disciplinary content of the project was a primary objective. We adapted approaches made by economists and statisticians to new but fundamentally similar issues in empirical sociology. Firstly, we drew on techniques used to measure aspects of income distribution and applied them to educational inequalities. Secondly, from the literature on earnings discrimination we decomposed differences into components explained by differing characteristics and differing responses, and applied these to the decomposition of educational inequalities.
In understanding cross-national differences in educational inequality, a major objective of the first phase was to investigate whether background factors operate in the same way, and with the same force, in different countries. We found that there is a pattern for some countries with high levels of class inequality in educational attainment also to have high levels of ethnic or gender inequalities. In some countries there are compensating mechanisms operating such as low levels of inequality with respect to parental education. From a policy point of view we found mixed results in that the inequalities are general ones that permeate all the background factors but further analysis highlighted particular inequalities such as gender, class or ethnicity.
Our second research objective was to see whether aggregate properties of the society (or of particular regions) such as their level of income inequality, their cultural values, or their specific educational policies and funding arrangements can explain the cross-national variation in educational inequality. Here we drew on the ESS and ISSP data. The ESS was a priority project by the European Strategy Forum on Research Infrastructures (ESFRI). These data were crucial to the multilevel approach that we took. In addition to the individual respondents' educational attainments and measures of social background, measured at level 1, explanatory variables such as aggregate measures of income inequality (derived from the ECHP) and socio-political culture (derived from the ESS and ISSP) were included at level 2. Because of the well known 'small N' problem, we did divide countries into regions, greatly increasing the number of observations but providing us with some contradictory results.
Underpinning this project was its perception of being part of the next generation of research on from Shavit and Blossfeld (1993). With the emergence of systematic cross-national data sets we originally focused on the PISA data. However other projects also used PISA. In order to avoid replication we introduced additional cross-national data to complement the primary research objectives. However our research was not undertaken in isolation of other human scientists using PISA, TIMMS and PIRLS. Synergies were increasingly developed as the project progressed with other scientists met at conferences, workshops, training programs etc. These led to beneficial transfers of knowledge as evidenced in our range of deliverables.
Additional data sets included some longitudinal such as the ECHP, EU-SILC and LIS. They proved a useful tool in enhancing our understanding of contributory factors to educational inequalities in a way not foreseen when this project application was originally written. Contributory factors included poverty, health, other family effects, religion and linguistic divisions amongst others. Longitudinal analysis allows progress to be made in establishing causation. The team members developed synergies with those working on longitudinal data providing a value-added benefit to the project.
From a methodological view the inter-disciplinary content of the project was a primary objective. We adapted approaches made by economists and statisticians to new but fundamentally similar issues in empirical sociology. Firstly, we drew on techniques used to measure aspects of income distribution and applied them to educational inequalities. Secondly, from the literature on earnings discrimination we decomposed differences into components explained by differing characteristics and differing responses, and applied these to the decomposition of educational inequalities.