Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We study the removal of information from a market, such as a job-applicant screening tool. We characterize how removal harms groups with relative advantage in that information: typically those for whom the banned information is most precise relative to alternative signals. We illustrate this using recent bans on employers’ use of credit report data. Bans decrease job-finding rates for Black job-seekers by 3 percentage points and increase involuntary separations for Black new hires by 4 percentage points, primarily because other screening tools, such as interviews, have around 60% higher standard deviation of signal noise for Black relative to white job-seekers.