Algorithmic Crowding Out: The Reproduction of Gender Division in the Exposure of Algorithm‐Driven Job Advertisements

ABSTRACT This study explores how algorithms curate women's exposure to job ads and reproduce gender divisions through mechanisms more subtle than overt stereotyping. Applying quantitative methods, we analysed over 700 algorithm‐driven job ads on a Chinese advertising platform and compared exposure rates between genders. Results reveal that women receive significantly less exposure to job ads than men, with a 32.4% lower exposure rate. This disparity is not attributable to the gender distribution of the platform's user base but is rooted in algorithmic decision‐making processes. The findings indicate that women encounter disproportionately more shopping and entertainment ads alongside fewer job ads in their information feeds. This pattern suggests that sectors with higher commercial value for women audiences receive greater algorithmic visibility, thereby reducing women's exposure to employment‐related information. Additionally, the study finds that job ads for white‐collar positions and high‐paying jobs exhibit lower exposure rates for women, whereas ads for blue‐collar and low‐paying occupations have higher exposure rates among women. To enrich the current understanding of the reproduction of gender stereotypes and segregation in the digital age, we introduce the concept of “algorithmic crowding out,” which encompasses socioeconomic inequalities encoded in algorithms, the systematic exclusion of women from certain job opportunities, and the reproduction of gender orders in algorithm‐driven forms. This paper contributes to the classic dichotomy of production and reproduction in discussing gender division by demonstrating the power of algorithms in connecting the production sector of job opportunities and the reproduction sector of gender stereotypes.