My research topic relates to the effects of school-level inequality on learning outcomes (such as Reading) in PISA 2015.

While there is extensive and long-standing macro-level and micro-level research recognizing the relevance of socioeconomic factors (e.g. wealth) as important predictors of children’s cognitive learning attainment (Coleman, 1966; Del Bello, Patacchini and Zenou, 2015), less attention has been given to the relevance of wealth inequality to different outcomes.

Inequality is customarily explained as an “unfair difference between groups of people in society, when some have more wealth, status or opportunities than others” (Hornby, Wehmeier and Ashby, 2000, sec.Inequality). Even scholars of conflicting sociological schools agree that inequality presents two concurrent characteristics, namely unfairness and wealth/income difference. 

In terms of operationalization, it appears to be a shortfall on assumed inequality measures such as Gini, Hoover or Theil indexes because they do not take account of the above-mentioned dual configuration of inequality, assuming every wealth/income difference among people as being unfair. At the same time, those indexes require income/consumption positive integer data to compute indexes.

Considering that unfairness and wealth differences are two dimensions that compose unfair inequality, I suggest the need for accounting both elements in the measure, which will be addressed in the section regarding the estimation models. 

My primary research question is as follows: does school-level wealth inequality have effects on learning outcomes in PISA 2015?

This empirical study comprises two parts: first, I estimate a country inequality measure, addressing both theoretical and methodological aspects, and performing correlations between aggregated mean score of the inequality index and Gini Index per country. The estimation of inequality mimics a coefficient of variation approach and was developed by the research (proof of inequality axioms are provided). Finally, I perform different random slopes and intercept linear regression models (also known as linear mixed-effects model) in order to test the effects of the interaction of wealth and inequality on reading scores, adding diverse control variables to the analysis. Assumptions and diagnostics tests are performed to ensure quality of models.

Relaying on the PISA’s wealth index, I have computed and inequality measure at school-level, and modelling the effect of the interaction of inequality and wealth on Reading scores across countries, I have found negative predictive parameters on reading outcomes, which affect homogeneously at least half of the countries of PISA. Another interesting finding is the fact that even when the parameter of wealth presents a higher positive magnitude than the negative inequality parameter when all variables are controlled, the interaction between those variables indicates a negative prediction. That suggests inequality could be understood as a contaminant of the predictor wealth on reading scores. However, the analysis of causality and unravelling this moderation are still interesting lines of further investigation.

I only present an excerpt of my reference due to length limitations. Abul Naga, R.H. and Yalcin, T., 2008. Inequality measurement for ordered response health data. Journal of Health Economics, 27(6), pp.1614–1625. Allison, R.A. and Foster, J.E., 2004. Measuring health inequality using qualitative data. Journal of Health Economics, 23(3), pp.505–524. Almås, I., Cappelen, A.W., Lind, J.T., Sørensen, E. and Tungodden, B., 2011. Measuring unfair (in)equality. Journal of Public Economics, [online] 95(7–8), pp.488–499. Balen, J., McManus, D.P., Li, Y.S., Zhao, Z.Y., Yuan, L.P., Utzinger, J., Williams, G.M., Li, Y., Ren, M.Y., Liu, Z.C., Zhou, J. and Raso, G., 2010. Comparison of two approaches for measuring household wealth via an asset-based index in rural and peri-urban settings of Hunan province, China. Emerging Themes in Epidemiology, 7. Banerjee, A.K., 2010. A multidimensional Gini index. Mathematical Social Sciences, [online] 60(2), pp.87–93. Del Bello, C.L., Patacchini, E. and Zenou, Y., 2015. Neighborhood Effects in Education. IZA Discussion Papers. Birdsall, B.N. and Londoño, J.L., 1997. Asset Inequality Matters: An Assessment of the World Bank ’ s Approach to Poverty Reduction. The American Economic Review, 87(2). Booysen, F., van der Berg, S., Burger, R., Maltitz, M. von and Rand, G. du, 2008. Using an Asset Index to Assess Trends in Poverty in Seven Sub-Saharan African Countries. World Development, 36(6), pp.1113–1130. Bound, J., Brown, C. and Mathiowetz, N., 2001. Chapter 59 Measurement error in survey data. [online] Handbook of Econometrics, Elsevier Masson SAS. Available at: . Breunig, R. and Majeed, O., 2016. Inequality or poverty: which is bad for growth? Brown, A. and Croudace, T.J., 2014. Scoring and Estimating Score Precision Using Multidimensional IRT Models. In: S.P. Reise and D.A. Revicki, eds., Handbook of Item Response Theory Modeling: Applications to Typical Performance Assessment. Brown, A. and Croudace, T.J., 2015. Scoring and estimating score precision using IRT. Handbook of Item Response Theory Modeling: Applications to Typical Performance Assessment (a volume in the Multivariate Applications Series), (March), pp.307–333. Cappellari, L. and Jenkins, S.P., 2007. Summarizing multiple deprivation indicators. Chalmers, R.P., 2012. mirt: A multidimensional item response theory package for the R environment. Journal of Statistical Software, [online] 48(6), pp.1–29. Available at: .