Perpetuating Wage Inequality: Evidence from Salary History Bans
By James Bessen, Erich Denk, and Chen Meng
A recent paper “Perpetuating Wage Inequality: Evidence from Salary History Bans” published in the Journal of Economic Inequality by scholars from TPRI (James Bessen, Erich Denk, and Chen Meng) has influenced policy. Along with other papers, it prompted the White House to issue Executive Orders banning the use of salary histories in Federal hiring and in hiring by Federal contractors and it was cited in final hiring rules (the Office of Personnel Management and Department of Defense).
Economists have long argued that underpaid workers can sometimes escape discrimination by switching jobs. However, if employers have access to job applicants’ previous pay histories and use this information for a bargaining advantage, this mechanism might not work. In this way, the salary information may help perpetuate the pre-existing inequities. On the other hand, aware of this possibility and the persistence of the gender pay gaps in the US, women’s advocates have pushed for the passage of Salary History Ban (SHB) legislation that prevents employers from asking for salary histories. Since 2016, more than a dozen states and cities have enacted SHB laws and nearly a quarter of the private-sector workers are covered by the law. Correlated with the passage of SHB laws, the share of online job advertisements that list salary information has increased by 20 percentage points in just one year. This paper is the first to examine the causal and comprehensive impacts on employers’ behavior in wage posting, and the actual for job changers especially women and minorities. The results have important policy implications in addressing the pay inequality issue in the US.
The paper has the following research objectives: 1) Develop a theoretical model predicting the outcomes of the removal of salary history. 2) Use an empirical model to test for the change in firms’ wage-posting behavior. 3) Use an empirical model to look at the actual pay of job-changing workers in private sector under the SHB. 4) Examine specifically how the previously disadvantaged groups such as women and non-whites are causally affected, and the size of the impact. 5) Test for any adverse effect or additional effect such as the change in the composition of job changers and turnover rates, utilizing alternative methods and dataset to make sure the results are robust and valid.
The theoretical model is based on Hall and Krueger’s 2010 wage formation model but it is fitted to the context of SHBs. It uses worker’s reservation wage, productivity, and the elasticity of labor supply to construct employer’s profit-maximizing scheme. The employer will choose to bargain or post wage based on the process that gives the higher profit. The theoretical model predicts that employer lose wage bargaining surplus when an SHB is present. The empirical model tests model predictions using difference-in-difference around a control and treatment group. Because the data is rich enough, the paper looks at bordering counties within the same labor market, one is treated under SHB, the other is not. The method is applied to two big datasets. The first is the Burning Glass Technologies (BGT) online job posting data that has 200 million observations and covers over 90% of the US online job data. This proprietary data is used for getting results on firms’ wage posting behavior and salary offered. The second dataset is the Current Population Survey (CPS) — the nationally representative household survey conducted by the Bureau of Labor Statistics, to estimate the actual wage paid and wage gap. Job changing individuals can be identified in the dataset. Additionally, other empirical methods such as Synthetic Control Method and Coarsened Exact Matching are used to make sure the results hold true under a different method.
The paper first finds a significant rise in the probability that employers list salary information in the job ads after SHBs go into effect. Second, after an SHB, job-changing workers in the private sector earn 3.9% more than comparable job-changers not under SHBs relative to incumbent workers on average. Third, there are even larger increases in the pay of job-changing women (6.2%) and non-whites (5.8%). For these previously disadvantaged groups, the pay increases following an SHB represent a sizeable portion (40%-67%) of the residual wage gap measured for job-changing employees, i.e., SHB eliminates a big portion of the wage gap. This suggests that most of this gap is related to bargaining differences rather than productivity differences between workers. The SHBs have little effect on the composition of job changers, don’t result in changes in job turnover. And multi-state employers are not switching hiring to non-SHB states.
Under the SHB, firms advertise salaries more, they pay higher wages for job changers, and the increase is greater for groups subject to historical discrimination. SHBs reveal evidence that previously, employers have exploited the past salary information and this has been an important factor perpetuating wage inequality, especially for women and non-whites.