Regional inequalities in PISA: the case of Italy and Spain

This technical brief analyses the regional distribution of skills in Italy and Spain. Educational attainment rates have frequently been used as an indicator of regional educational development in EU Member States (MS). These rates indicate significant regional disparities in education within countries. However, recent evidence shows that the quality of education, as measured by the level of specific skills, is more important than the number of years one spends in school, in particular when considering the relationship between the cognitive (and non-cognitive) skills and economic growth. International large scale assessments (ILSA) of student performance measure these cognitive skills in key areas. OECD’s Program for International Student Assessment (PISA) provides a very useful and important source of information of students' performance in key cognitive skills. When analysing PISA data, researchers and commentators often focus on cross-country comparisons. However, vast within-country differences exist, also in terms of educational attainment and PISA test scores. A focus on country averages alone would hence provide only a partial view of the status of education within countries. However, the possibility of exploring within-country differences with PISA data is limited to only a few countries. In this report we focus on regional inequalities in cognitive skills (as measured by PISA test scores) in Italy and Spain, using regional PISA data from the most recent 2015 wave, and we analyse the factors that are associated with these inequalities. In order to insure full comparability between the two countries we define regions at the level of NUTS1 (macro-region), following Eurostat’s official NUTS 2013 classification. We investigate regional inequalities by using descriptive statistics, by running a range of OLS regression models that allow us to analyse the associations between PISA 2015 science scores and the explanatory variables within regions and finally by using the Blinder-Oaxaca decomposition method to specify the factors that are related to within-country differences. The results show that there are significant regional differences in PISA scores within both MS. There are several factors that are associated with regional differences within Italy and Spain. The factors most consistently positively associated with regional science achievement are teacher-directed teaching and epistemological beliefs, while grade repetition and truancy are significantly negatively related with achievement. Still, there is also a range of other relevant factors varying between and within both MS. The Blinder-Oaxaca decomposition also shows that variables such as the socio-economic background, the students’ expected occupation, learning outside school time, truancy, immigrant status and grade repetition matter for within-country differences. Our results suggest that policy makers should focus on finding solutions to limit truancy and rethink grade repetition to leverage scores in lower performing regions. Moreover, our results with regard to epistemological beliefs and teaching practices challenge thinking about how science should be taught in schools in Italy and Spain. The specific results for each region may allow policy makers to consider more in detail how a region stands in comparison to the rest of the country, and the specific factors that need to be addressed to improve the within-country inequality related to educational achievement.