I wonder if a combination of "MAIHDA" and intersectionality theory can be used to help policy makers uncover heterogeneous policy effects.

Speaker(s):

Andrew Bell, University of Sheffield

Abstract:

Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (MAIHDA) is an innovative approach to analysing inequalities in society. In the MAIHDA approach, individuals are divided into intersectional strata, based on the combination of sociodemographic identity characteristics (for instance: gender, age, ethnicity, socio-economic status). The method has a number of good statistical properties that allow for reliable calculation of inequalities, in diverse outcomes. We are interested in using a MAIHDA framework to consider inequalities in the effects of policies. That is not only “Are these groups different?”, but “Is the effect of this policy different in different groups?” To do so requires more advanced versions of MAIHDA, utilising random slopes models, but also presents challenges relating to data formats, statistical power, challenges with non-continuous outcomes, different policy implementation, and theoretical understanding in the face of statistical complexity. We hope to work through these challenges here, and through our ESRC-funded grant.