Continuous Gender Identity and Economics

Key takeaways

(file:///C:\Users\scott\Zotero\storage\CH2KDGYY\Brenøe%20et%20al.%20-%202022%20-%20Continuous%20Gender%20Identity%20and%20Economics.pdf)

Bibliography: Brenøe, A.A., Heursen, L., Ranehill, E., Weber, R.A., 2022. Continuous Gender Identity and Economics. AEA Papers and Proceedings 112, 573–577. https://doi.org/10.1257/pandp.20221083

Authors:: Anne Ardila Brenøe, Lea Heursen, Eva Ranehill, Roberto A. Weber

Collections:: Gender Scale

First-page: 3

Abstract

Economic research on gender largely focuses on biological sex, the binary classification as either a “man” or “woman.” We investigate the value of incorporating a measure of continuous gender identity (CGI) into economics by exploring whether it explains variation in economic preferences and behavior beyond the explanatory power of binary sex. First, we validate a novel single-item CGI measure in a survey study, showing that it correlates with measures used in gender research. Second, we use our single-item CGI measure in an incentivized laboratory experiment to assess CGI's power in explaining previously documented gender gaps in four important economic preferences.

Citations

content: "@brenoeContinuousGenderIdentity2022" -file:@brenoeContinuousGenderIdentity2022

Reading notes

Imported on 2025-04-27 17:39

⭐ Important

& On the other hand, substantial variation within biologically “male” and “female” samples in both behavioral tendencies like risk preferences (Nelson, 2015) and in economic outcomes (Goldin, 2016) suggests that a richer classification may capture important dimensions of individual heterogeneity useful for positive economics and for the design of more carefully targeted policy interventions. (p. 3)
& Each item of the BSRI is a characteristic, coded as either feminine (“love children”), masculine (“defend my own beliefs”) or neutral (“conscientious”). Hence, this scale measures masculinity and femininity as independent dimensions (p. 4)
& A drawback of the BSRI is that the masculinity-femininity classifications are derived from somewhat dated gender stereotypes and may therefore measure the extent to which one conforms to these stereotypes and expectations rather than one’s own sense of gender. (p. 4)
& We score the BSRI following the test manual and get results similar to Bem (1981, p. 71). 37.0% (13.3%) of the women (men) in our data classify as feminine, 14.8% (37.9%) as masculine and 27.8% (24.4%) as androgynous. (p. 5)
& We conclude that (i) our single-item measure captures a substantial share of the variation in gender identity measured by other scales and (ii) there is substantial variation in continuous gender identity. We next provide preliminary evidence on the relationship between gender identity and economic preferences. (p. 6)
& findings in the early stage of a larger data collection, they suggest added explanatory power from incorporating selfreported measures of continuous gender identity, mainly in the domain of risk. (p. 8)

⛔ Weaknesses and caveats

! Single Item Continuous Gender Identity (CGI): Our own scale measures first-order perceptions of gender identity (“Where would you put yourself on this scale?”) by eliciting self-placement on a 7-point scale, ranging from “very masculine” to “very feminine.” The main difference with the Magliozzi scale is measurement of masculinity and femininity in a single dimension. (p. 5)
! In the two-dimensional scale by Magliozzi et al. (2016), and in our unidimensional scale, we find high correlations of almost 0.9 between responses to first- and third- order questions. (p. 5)
! We also use principal component analysis to extract a measure of the underlying latent continuous gender identity from the seven existing scales (online Appendix Table A1). (p. 5)