The American Economic Association's registry for randomized controlled trials
Please fill out this
short user survey
of only 3 questions in order to help us improve the site. We appreciate your feedback!
The Effect of Salary History Bans
Last registered on July 19, 2018
View Trial History
The Effect of Salary History Bans
Initial registration date
July 11, 2018
July 19, 2018 6:50 PM EDT
United States of America
Contact Primary Investigator
Other Primary Investigator(s)
Additional Trial Information
How does banning firms from seeking applicants’ wage histories affect employment and salary offers? We implement a field experiment in which recruiters evaluate job applications that randomly include or exclude salary information questions and disclosures. The experiment mimics laws passed in Massachusetts, California, New York City, and Chicago banning these questions. Applicants also vary by gender and whether their previous salaries are high or low in the distribution of their previous firms salaries. We test several hypotheses about how salary disclosure might impact wage offers and gender inequality.
Agan, Amanda, Bo Cowgill and Laura Gee. 2018. "The Effect of Salary History Bans." AEA RCT Registry. July 19.
Agan, Amanda et al. 2018. "The Effect of Salary History Bans." AEA RCT Registry. July 19.
Sponsors & Partners
There will be three main treatments, which involve the salary history question the recruiters see on the job applications:
- Salary history question is mandatory
- Salary history question is optional
- Salary history question is not there
* When salary history is not there, in some treatments applicants will still disclose salary history "unprompted" in "additional information" section
Intervention Start Date
Intervention End Date
Primary Outcomes (end points)
- The salary the individual applicant is offered by the recruiter
- Whether the applicant is suggested for an interview
- The variation in salaries the recruiter offers to all applicants they see
Primary Outcomes (explanation)
Secondary Outcomes (end points)
- Maximum offer we should accept from applicants
- Salary distribution for candidate suggested by recruiter
Secondary Outcomes (explanation)
Recruiters will be randomly given a set of 8 applications. This set of applications will fall into one of 11 unique potential treatments, some treatments will be run twice for a total of 14 potential treatments. Each set of applications x treatment will be evaluated by two recruiters. The 11 unique treatments involve whether the salary history is disclosed or not, as well as the mix of salaries that are disclosed conditional on disclosure. Recruiters will then respond to several questions about whether the applicant should be interviewed and the take-it-or-leave-it salary they should be offered, as well as some additional questions about salary offers, via an online survey.
Experimental Design Details
The 11 treatments - with some repeated for a total of 14 - are: # Treatment 1: All Disclosed. All high salaries. # Treatment 2: All Disclosed. All Low salaries. # Treatment 3: All Disclosed. Assign "high within firm" one man/woman each from a high wage firm a low wage firm. The rest are "low within firm." # Treatment 4: All Disclosed - Flip 3 # Treatment 5: Same as 4. # Treatment 6: Prompted, Half Disclosed "uniform". Take Treatment 3. Disclose max male and female and min male and female. # Treatment 7: Prompted, Half Disclosed. "more high." Take Treatment 3. Randomly disclose the 3 on the high end of the distribution 1 on the low end # Treatment 8: Prompted, Half Disclosed. "more low." Take Treatment 3. Whoever didn't disclose in treatment 7 now discloses. Whoever did disclose in treatment 7 now does not. # Treatment 9: Unprompted, Half Disclosed. Same as 6, but unprompted. # Treatment 10: Unprompted, Half Disclosed. Same as 7, but unprompted. # Treatment 11: Unprompted, Half Disclosed. Same as 8, but unprompted. # Treatments 12-14 Banned (unprompted, undisclosed)
Randomization done by computer
Was the treatment clustered?
Sample size: planned number of clusters
Approximately 112 recruiters
Sample size: planned number of observations
112 recruiters x 8 applications = 896 applications reviewed
Sample size (or number of clusters) by treatment arms
8 recruiters (64 applications reviewed) in each of 14 treatments - where some treatments are duplicates.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Given that there is little, if any, research on how salary disclosures or salary disclosure bans affect salary offers, we have little basis for guessing the baseline or variance of the effects we hope to estimate, making any estimation of MDE difficult and with too many potential degrees of freedom.
Supporting Documents and Materials
INSTITUTIONAL REVIEW BOARDS (IRBs)
Columbia Human Research Protection Office
IRB Approval Date
IRB Approval Number
Post Trial Information
Is the intervention completed?
Is data collection complete?
Is public data available?
Reports, Papers & Other Materials
REPORTS & OTHER MATERIALS