Gender bias in redistributive choices: an experimental study in United States
How to cite
Marino, Maria. 2024. "Gender bias in redistributive choices: an experimental study in United States." AEA RCT Registry. AEARCTR-0014043. https://doi.org/10.1257/rct.14043-2.1
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
This study investigates the role of stereotypes about gender abilities in explaining the existence of gender bias in redistributive choices, as recently found in \citet{cappelen2023experimental}.
Using a worker/stakeholders design, we assess how gender bias in redistributive choices varies when spectators are provided with information about the type of assignments undertaken by workers. To this aim, we compare redistributive choices when spectators are informed that workers have conducted an unspecified assignment, a male-typed (visual-spatial rotation) assignment, or a female-typed (facial emotion recognition) assignment. Our findings will contribute to understanding gender inequality by exploring individuals' fairness preferences related to stereotypes about gender abilities. Moreover, we aim to provide insights for policy interventions to address gender disparities.
Experimental design
We designed an experiment with two stages and two types of participants: workers and spectators. In the first stage, workers of either gender complete an assignment. In the second stage, workers are paired, and only the most productive worker (the winner) is assigned $6, while the less productive worker (the loser) is assigned $0. Spectators are then randomly paired with one of these worker pairs and must decide whether to redistribute the $6 from the winner to the loser. Spectators are aware of the workers' genders and who is more productive, but not the details of the assignment they completed.
Our study builds on prior work by introducing new treatments to explore how spectators' decisions are influenced by information about the type of assignment performed. We test a 2x3 design where spectators are randomly assigned to one of six conditions that vary in the information provided about the assignment and the gender of the loser. The treatments include two conditions with no assignment information (one where the loser is male and another where the loser is female), and four conditions where spectators receive information about whether the assignment is male or female-typed.
Design details
Cited by
No citation data available for this registration yet.