BTECH research seminar

Title: The Exclusion of Women from Economics Research: Trends, Determinants, and Implications (with Melanie Wasserman from UCLA Anderson School)

Abstract: Who researchers study affects the generalizability of insights and the frontier of new knowledge in terms of questions asked and innovations considered. Further, it also has the potential to implicitly value (or devalue) certain groups. This paper examines the trends, determinants, and implications of gender-based sample restrictions in empirical labor economics papers published over four decades. First, we find a three-fold decline in the proportion of labor studies that restrict to an all-male sample with a commensurate increase in the proportion using a combined male-female sample ("mixed gender"; increased from 60% to over 80% in that time). While the majority of papers (70% overall) do not include a justification for the sample restriction, the proportion that do has increased over time. Second, regression analysis shows that gender diversity of authorship teams matters for sample restrictions, yet the homophilic tendency cuts both ways: all-male authorship teams are 15% more likely to use all-male samples and all female-authorship teams are 50% more likely to use all-female samples as compared to mixed-gender authorship teams. Third, we show that there are important potential implications of these sample restrictions in that: a) findings can vary substantially based on sample composition, and b) there is a asymmetry in the restriction transparency by sample (over 80% of papers with female-only samples disclosure in title or abstract, while just 50% of papers using male-only samples).