The short answer
A selection rate is the share of applicants or candidates in a group who are hired, promoted or otherwise selected. The federal Uniform Guidelines define it as "the proportion of applicants or candidates who are hired, promoted, or otherwise selected," and New York City's Local Law 144 rules apply it to each demographic category a tool moves forward. It matters because every adverse impact check and every bias audit starts with it.
Selection rate in hiring explained
The formula: selection rate = number selected from a group / number of applicants from that group. New York City's rules give a simple example: if 100 Hispanic women apply and 40 are selected for an interview after an AEDT is used, their selection rate is 40/100, or 40%.
"Selected" means whatever the step decides. For a resume screen, it is moving to interview. For a first-round interview, it is advancing to the panel. For the whole process, it is getting the offer. A rate is only meaningful when you say which step it describes.
Once you have rates for each group, you divide each by the highest one to get an impact ratio, and compare it with the four-fifths rule.
Example
An illustrative funnel for one group of applicants to an analyst role:
To check adverse impact, you would build the same table for every group and compare rates step by step and overall. The EEOC says to look at the overall process first, then at any step where the total shows a gap.
Common mistakes
- The wrong denominator. The EEOC describes an applicant as someone who has shown interest in being considered, with the exact meaning depending on your process. A person who voluntarily withdraws at any stage is no longer counted as an applicant for that calculation.
- Mixing steps. Comparing one group's interview pass rate with another group's offer rate gives a meaningless ratio.
- Dropping unknowns silently. NYC audits must state how many people were assessed but left out because their category was unknown.
- Reading tiny groups too closely. With 10 people, one decision moves the rate by 10 points.
To see where these rates show up in AI hiring tools, read our guide to AI bias audits.
Related terms
- Impact ratio: one group's selection rate divided by the highest group's.
- Four-fifths rule: the 80% rule of thumb applied to those ratios.
- Adverse impact: a substantially lower selection rate that disadvantages a group.
- Bias audit: an independent review that reports selection rates by category.
- Time to hire: another funnel metric, measuring speed instead of who advances.
Questions
How do you calculate a selection rate?
Divide the number of people from a group who were selected at a step by the number of applicants from that group at the same step. For example, 40 selected out of 100 applicants is a 40% selection rate.
Is a selection rate the same as a pass-through rate?
In practice, yes for a single step: both are the share of people who move forward. The term selection rate is the one used in the federal Uniform Guidelines and New York City's bias audit rules, usually broken out by demographic group.
Should candidates who withdraw be counted?
No. The EEOC's Q&A on the Uniform Guidelines says a person who voluntarily withdraws at any stage is no longer an applicant or candidate for computing adverse impact, though records should be kept for everyone who applied.
Want first-round results that are consistent enough to track by group? Book a 20-minute demo of a live adaptive AI interview, or try one yourself.
This is not legal advice.
