Talking Intersectional Regression - A Theoretical Foundation of Multiple Inequalities in Regression Models and what Regression Models Contribute to Intersectionality Studies

Talking intersectional regression A theoretical foundation of multiple inequalities in regression models and What regression models contribute to intersectionality studies While some discard quantitative approached to intersectionality altogether given the former’s additive and simplifying nature, more recent work argues for a plurality of intersectionality approaches that complement more than contest each other. However, the tension remains between the more theoretical and complex understanding of intersectionality on the one hand and the rather reductive application of intersectionality in regression-based quantitative research via interaction effects. This paper’s starting point is that this tension is partly due to several misunderstandings of the concepts/practices in the other field (‘intersectionality studies’, ‘regression-based social sciences’). Accordingly, I first set out to discuss the theoretical conditions under which interaction terms in regression-based analysis can be considered a valid form of certain types of intersectional (e.g. intercategoricalism). Second, I will explore how the seminal examples provides in Crenshaw’s 1991 article actually inform different ways in which multiple inequalities can be conceptualized, more particularly in both additive and intersection terms. Working from these examples I will consider how this translates to including multiple inequalities in regression-based models. Third, I will combine the two elements above to present a slightly provocative broader pallet of options to conduct intersectional quantitative analysis, while at the same time formulating far more stricter criterions on when a quantitative analysis actually lives up to the standard of intersectionality. Empirics on the context-dependent, gendered and class-based impact of parenthood and employment on political participation will illustrate the arguments. Overall, the paper does not set out to find definite answers, but to further and refine the arguments about intersectionality and quantitative analyses, hopefully speaking to both an audience approaching this debate from intersectionality studies and an audience mainly conducting quantitative social science.