26th May at 13.30 in Building D6/Room Bracco,
Maria Marino (Universitat de Barcelona) will give a seminar on:
Gender bias in redistributive choices: an experimental study in the United States
Abstract: Building on Cappelen et al. (2025), who document that impartial spectators redistribute less to male than to female losers in an unspecified-assignment setting, we run an incentivized online experiment in which spectators decide how to redistribute earnings between a winner and a loser. We vary two dimensions of the spectator's information set: the gender of the loser (male or female) and the type of assignment the workers performed (unspecified, a male-stereotyped task — visual-spatial rotation — or a female-stereotyped task — facial emotion recognition). We replicate the bias against male losers in the unspecified-assignment condition. Once assignment information is introduced, the structure of the bias shifts in ways that aggregate statistics obscure. In the male-stereotyped assignment, the average gap between male and female losers vanishes, but male and female spectators move in opposite directions: conditional on giving, male spectators favor male losers while female
spectators continue to favor female losers, generating offsetting effects that cancel out at the aggregate level. In the female-stereotyped assignment, spectators of both genders remain less willing to give anything to a male loser, even though the amount conditional on giving is statistically indistinguishable across loser genders. We explore a range of potential mechanisms, including beliefs about why the loser underperformed, empathy toward the loser, and broader gender-related attitudes. Beliefs and attitudes correlate with redistribution choices but do not seem, on their own, to account for the full pattern.
spectators continue to favor female losers, generating offsetting effects that cancel out at the aggregate level. In the female-stereotyped assignment, spectators of both genders remain less willing to give anything to a male loser, even though the amount conditional on giving is statistically indistinguishable across loser genders. We explore a range of potential mechanisms, including beliefs about why the loser underperformed, empathy toward the loser, and broader gender-related attitudes. Beliefs and attitudes correlate with redistribution choices but do not seem, on their own, to account for the full pattern.