Who Values Diversity
This paper describes the results from a randomized controlled field experiment designed to measure the extent to which individuals prefer or dislike diversity. Our novel task asks the participants in two pools – a tech conferences and online – to evaluate their user experience with an image search engine tool. For the same set of four search words (boss, professor, clerk, and nurse), subjects randomized into the less diverse image treatment view stereotypical images that one would associate with the profession, while those in the more diverse image treatment view a set of images that counter the corresponding gender and racial stereotypes. We observe that diverse images result in significantly higher ratings across all our participants, finding no evidence of in-group bias in this context. However, women are disproportionately more dissatisfied with the inaccurate gender representation among the less diverse images for the stereotypically high- status words (boss and professor) than men are. For the low-status words (clerk and nurse), where the stereotypes are reversed and the lack of diversity appears to be less salient, we find weaker treatment effects and no heterogeneity in the satisfaction ratings by gender or race.