Screening Women Out? Pay Transparency in Job Search
Women sort into low-paying firms for otherwise similar jobs. Whether this reflects preferences or frictions is largely unknown. Using 29 million job applications from Pakistan's largest job search platform, and a discrete-choice experiment, I document that large, high-paying firms are more likely to omit salaries from job ads, and less likely to offer flexibility – which women value more than men. When pay is disclosed, men and women respond similarly. Without disclosure, men search randomly while women sort negatively on pay. A search model rationalizes these facts, showing that non-disclosure transforms small gender differences in amenity preferences into large gender gaps in sorting. To test whether transparency closes the gaps, I field an experiment randomizing mandatory vs. optional pay disclosure in 20,000 jobs across 8,900 firms. Large-firm pay and amenities remain unchanged by transparency. Yet, women's applications to these firms increase 95%, and men's by 59%, reversing the gender gap in directed search. Large firms also overestimate the costs of transparency; treatment raises their pay disclosure rate by 30% post-experiment. I conclude that women do not prefer flexibility to pay. Rather, they turn to flexibility when its price is unknown.