Though feminist economics encourages methodological plurality, quantitative methods and econometrics have overtaken the discipline in recent years. Many feminist economists have demonstrated reasons to be concerned about the increasing foothold of quantitative methods, and others have provided thoughtful criticisms of specific quantitative measurements. However, few have made distinctions between mainstream econometrics and feminist econometrics, and a succinct set of resources for those trying to do feminist quantitative research is difficult to find. Drawing upon insights from feminist economics, queer methods, and intersectional approaches, this article sets forth practical guidelines for feminists using econometric methods. Namely, it considers issues of data cleaning, replicability, survey weighting, comparison groups, non-binary measures of gender, intersectionality, causality claims, identification problems, atheoretical index composition, and measuring “difference.” It raises questions for contemporary feminist economists to consider as we grapple with the methodological identity of our field. HIGHLIGHTS Gender diagnostic estimation can be used to endogenously measure gender. Separate model approaches and CARTs are useful in intersectional work. Econometric models can be structured so they are “studying up.” Oaxaca–Blinder and PCA methods raise some feminist concerns. Cohen’s D and the Index of Similarity can be used to test gender differences.
Read full abstract