Some people claim there is no systemic discrimination in the US. Our new paper https://ssrn.com/abstract=3628729 finds that a recent policy change reveals how much institutions have been perpetuating pay inequities. 1/

And we find that these new policies--salary history bans (SHBs)--make some real progress toward greater equality for African-Americans and women. 2/

Becker (1957) argued that markets can overcome discrimination. If employer A pays some workers less because of a “taste for discrimination,” those workers can change to employer B who doesn’t discriminate. 3/

BUT if employer B knows our worker’s pay level, he can offer just a bit more than past pay and the worker will accept. Result: the worker is STILL paid less than she is worth. Discrimination is perpetuated. 4/

In order to take advantage of past pay inequities, employers must bargain over wages, not post them in advertisements. Many employers choose not to post salary information for many jobs. 5/

Beginning in 2017, some states passed laws prohibiting employers from asking for salary histories. 6/

Salary history bans appear to have had an immediate and dramatic effect on employer posting behavior. No longer able take advantage of salary histories in bargaining, employers advertise salaries much more often. 7/

To test this, we compare job advertisements in states with bans, to counties in neighboring states that are in the same labor markets. 8/

We find a significant increase in job posting in difference-in differences analyses. 9/

And we find significant increases in average pay for workers who changed jobs under an SHB compared to similar workers in uncovered states. 10/

Moreover, these increases were driven largely by improvements for African-Americans (13-16% on average ) and women (8%) who changed jobs compared to job-changers in uncovered states. 11/

Salary histories appear to aid institutional discrimination. Even if employers are not biased, they perpetuate past inequities. Salary history bans appear to reduce inequality for job changers and we find no evidence of adverse effects on job turnover or job changing rates. 12/12