Strategic aggregation: A more equitable approach to creating two-category variables

Aggregating data into two-category variables can facilitate comparisons when variables align neatly with such divisions. However, applying this approach to complex social constructs can present challenges in representing differences between groups without compromising anonymity or participating in their erasure. This commentary highlights the distinction between sex and gender by critiquing two-variable approaches used by the Government of Canada, including the 2021 Canadian Census, which erase intersex people and collapse non-binary identities into binary variables. While the inclusion in the census of gender-diverse people marks a step in the right direction, current approaches may contribute to historical injustices. To mitigate the potential harm these practices may cause, we explore a more equitable approach to two-category data aggregation: one which draws on strategic essentialism to understand the differences between historically dominant and equity-denied sexes and genders without obscuring or erasing their experiences. We also acknowledge that two-category binary variables may be inherently inequitable and stress the importance of transparency around classificatory decisions and consultation with affected communities.