In this article
Before you start
This page belongs to the behavioral interview questions learning path. If you are answering as a candidate rather than as a hiring professional, how to answer interview questions about diversity covers that side.
Why interviewers ask this
A weak answer to this question can end a human resources (HR), recruiting or people-manager interview, because it shows either no plan or a plan that creates legal risk. Candidates report the question in interviews for recruiting, HR and hiring-manager roles; no public source ties it to a named company.
Interviewers use it to separate three kinds of candidate. The first lists values and offers no method. The second proposes preferences or targets that the law treats as risky. The third describes a process: where candidates come from, how they are judged and how the results are checked.
The first two fail on the same point: neither says what changes in the hiring pipeline. A strong answer names concrete practices, says how each one is measured and shows that you know what the law allows. For the neighboring questions in the same interview, see HR interview questions and answers.
Diversity hiring strategy from first principles
- Selection rate
- A selection rate is the share of a group's candidates who pass one stage of hiring, such as the number of offers divided by the number of interviews, so that groups can be compared stage by stage.
- Rooney Rule
- The Rooney Rule is a hiring rule that requires an employer to include at least one candidate from an underrepresented group on the shortlist or interview slate, best known from its use for head coach vacancies in the National Football League (NFL).
In plain words
Think of hiring as a series of gates. Each gate lets some people through and stops others. A diversity hiring strategy is a plan to make sure every gate is open to qualified people from every background, and is judged the same way for all of them. The analogy breaks in one place. Some people never reach the first gate, because they never heard about the job. So the plan works at two points: who arrives and how they are judged. It does not change the standard. It makes sure the standard is applied evenly.
Step 1: Widen the pool without preferring anyone
Widen the pool by recruiting from more places, not by preferring a group. In practice that means posting roles where more people look, writing job posts around the skills the work needs, and asking employees for referrals outside their own circles.
In a study of 829 midsize and large American firms over three decades, Dobbin and Kalev (2016) found that five years after a firm starts a college recruitment program that targets women, the share of women in its management rises by about 10% on average. The authors add that only about 15% of firms have special college recruitment programs for women and minorities.
These are outreach measures. They change who applies, not who is chosen. One law-firm summary of the Equal Employment Opportunity Commission (EEOC) guidance says that as long as initiatives do not exclude individuals or groups, they can still be lawful. The first gate now sees more qualified people.
Step 2: Standardize evaluation with the same questions and written anchors
Standardize evaluation so that every candidate is asked the same questions and scored on the same written anchors. That is a structured interview, and it leaves less room for gut feeling to decide.
Sackett and colleagues (2022) re-analyzed decades of selection research and reported that structured interviews emerged as the top-ranked selection procedure. They also paired each method's validity with its mean Black–White subgroup differences, so the validity–diversity trade-off can be read from the data.
Consistency matters as much as the tool. Dobbin and Kalev report that managers applied written tests selectively and ignored the results, so testing amplified bias rather than quashed it. Everyone gets the test, and the result is used. Structure reduces bias; it does not remove it, which is why the next step exists.
Step 3: Measure pass rates by group at every funnel stage
Measure the selection rate of each group at each stage of the hiring funnel, where the law and company policy allow the data to be collected. Count applicants, interviews and offers, then divide.
A common screening check is the four-fifths rule. Under the Uniform Guidelines on Employee Selection Procedures, 29 Code of Federal Regulations (CFR) 1607.4(D) (read October 2026), a group's selection rate below four-fifths (80%) of the highest group's rate is generally regarded as evidence of adverse impact. The same section cautions that small samples can mislead. The rule is a flag to investigate, not a verdict.
The stage where a gap opens tells you which fix to use. A gap at application points to sourcing or job-post wording. A gap at interview points to screening. A gap between interview and offer points to how interviewers score and decide. Without stage-by-stage numbers, every fix is a guess. The worked example below shows the arithmetic with illustrative counts.
Step 4: Review the results and stay inside employment law
Review the numbers on a schedule, assign an owner and keep every action inside employment law. Dobbin and Kalev put the psychology plainly: "When people know they might have to explain their decisions, they are less likely to act on bias." (Dobbin and Kalev, 2016)
They describe task forces as one of the programs that worked: firms that put in diversity task forces saw increases of 9% to 30% in the representation of white women and of each minority group in management over five years. They also describe Deloitte, where female partners rose from 5% to 14% in eight years after the firm made its task force's findings transparent.
An owner, a review date and a published result turn a hiring plan into something the firm answers for. The legal limits come next, because a plan that ignores them is not a plan an interviewer will trust.
Step-by-step description
- Recruit from more places, without preferring a group.
- Same questions, same written anchors for everyone.
- Track selection rates by group at each stage.
- Review on a schedule and stay inside the law.
- Widen the pool by recruiting from more places, not by preferring a group.
- Standardize evaluation so that every candidate is asked the same questions and scored on the same written anchors.
- Measure the selection rate of each group at each stage of the hiring funnel.
- Review the numbers on a schedule and keep every action inside employment law.
What the law allows in the United States
In the United States, Title VII of the Civil Rights Act governs this. The EEOC says an employment action can be unlawful when it is motivated "in whole or in part" by a protected characteristic, as a law-firm summary of its guidance reports. Another firm's summary states: "Explicit quotas or 'balancing' the workforce by race or gender are unlawful."
On June 29, 2026 the EEOC voted to rescind its voluntary affirmative action guidelines. A law-firm summary reports that the rescission leaves Title VII itself and Supreme Court precedent unchanged. The summaries of the EEOC guidance read that to mean outreach and inclusion work that excludes no one remains lawful, while quotas, set-asides and preferences are the risky part.
Rules differ outside the United States, and this page is not legal advice. In an interview, say that you would check the plan with counsel.
What the evidence says about common programs
Programs that engage people, add contact or create accountability did better than training or selectively applied tests in Dobbin and Kalev's five-year data. Mentoring programs and task forces work because they put managers in contact with women and minorities and make someone answer for the results, which training and selectively applied tests do not.
| Program | Five-year effect on representation in management |
|---|---|
| Targeted college recruitment | About 10% rise in the share of women |
| Mentoring programs | 9% to 24% rise for several groups of women and men |
| Diversity task forces | 9% to 30% rise for white women and each minority group |
| Diversity training | Positive effects rarely last beyond a day or two; it can spark backlash |
| Written manager tests, applied selectively | 4% to 10% decrease |
The authors say the effective programs engage people in working for diversity, increase their contact with women and minorities, and tap into people's desire to look good to others. These are observational data from firms, so they show association over five years, not a controlled test.
Worked example 1: how much better is a structured interview?
Validity here means the correlation between a selection score and later job performance. The revised estimates of Sackett and colleagues (2022) are 0.42 for structured interviews and 0.19 for unstructured interviews. The older estimates of Schmidt and Hunter (1998) are 0.51 and 0.38.
The ratio of 0.42 to 0.19 is about 2.2, and that ratio is the usual comparison, because utility analysis treats validity as linear. Squaring each gives a rough share of performance differences explained, a second view: about 0.18 against about 0.04, roughly 4.9 times as much.1 The direction holds in both sets of estimates: structured interviews score above unstructured ones, and Sackett and colleagues still rank structured interviews first.
Text description of the diagram
- Revised estimates are 0.42 for structured and 0.19 for unstructured interviews.
- Older estimates are 0.51 for structured and 0.38 for unstructured interviews.
Worked example 2: finding the leak in a hiring funnel
Take an illustrative funnel of 1,000 applicants, 100 interviews and 10 offers, split between two groups. All counts are made up for the exercise.1
| Stage | Group A | Group B | Ratio, B to A |
|---|---|---|---|
| Applicants | 600 | 400 | n/a |
| Interviews | 60 | 40 | n/a |
| Offers | 8 | 2 | n/a |
| Applicant to interview | 10.0% | 10.0% | 1.000 |
| Interview to offer | 13.3% | 5.0% | 0.375 |
| Applicant to offer | 1.33% | 0.50% | 0.375 |
An illustrative ratio of 0.375 is far below the 0.8 line of the four-fifths rule. With only 10 offers, though, the gap could be chance, so watch it over more hires before drawing a conclusion. More outreach would not close this gap, because both groups reach interviews at the same rate. The lever is the last stage: a scorecard with written anchors, a check on how interviewers decide, and a review of the offer decision. If the gap had opened at the first stage, sourcing or the wording of the job post would be the place to look.
Text description of the diagram
- In the illustrative funnel the rate ratio is 1.0 from application to interview.
- The illustrative rate ratio falls to 0.375 from interview to offer.
Knowing it is half the job. Saying it is the other half.
Interview Clozo is an AI interview copilot for real interview rounds: it hears the question and puts a strong answer on your screen, and you say it in your own words. See the features or how it works.
Start Free Trial →The Staff-level answer
How would you build a diversity hiring strategy?
I would treat it as a pipeline problem, not a preference problem. First, widen sourcing: more schools, communities and referral channels, and job posts written around the skills the work needs, so more qualified people apply. Second, standardize evaluation: the same questions in the same order, scored on written anchors, because structured interviews are the best-ranked method in the research. Third, track pass rates by group at each stage, so I can see where people drop out and fix that stage. I would not use quotas, set-asides or preferences, because they create legal risk and undermine trust in every hire. I would review the numbers quarterly and report what changed.
What makes it Staff-level: it names a measure for every move, ties each fix to a funnel stage, and states what it refuses to do and why. The spoken structure is the pipeline framing, then the four moves, then the limit. For more on delivery, see how to answer interview questions.
The rubric
| Dimension | Weak answer | Strong answer | Staff answer |
|---|---|---|---|
| Correctness | Values only, or a hiring quota | Names lawful practices such as outreach and structured interviews | Separates outreach from selection and states what the law allows |
| Depth | Lists buzzwords | Explains why each practice helps | Ties each practice to a funnel stage and its evidence |
| Trade-offs | Says diversity and quality conflict | Says the bar stays the same for everyone | Names limits: structure reduces bias, and a diverse slate does not ensure a diverse hire |
| Communication | Vague or defensive | Four clear parts | Opens with the answer, then one measure and one limit |
| Follow-up readiness | No metrics | Names pass rates by stage | Explains how to read a gap and which fix follows |
Three variants of this question
"How would you improve diversity in hiring at our company?"
Start with the funnel: ask to see pass rates by stage, find where groups diverge, and choose the fix for that stage. Then propose outreach and a structured scorecard, and say you would not set quotas.
"What do you think of the Rooney Rule?"
The Rooney Rule requires at least one underrepresented candidate on the shortlist, and an online experiment found that decision-makers under it selected more minority candidates beyond the required one without considerably lowering the utility of those selected. A theoretical model shows that equal representation at the interview stage does not translate into equal hiring outcomes,. A slate widens the pool; it does not ensure a hire.
"Tell me about a time you made hiring more inclusive."
"Tell me about a time you made hiring more inclusive" is the behavioral version of the question; see common interview questions for the general pattern. An illustrative composite: a recruiter notices that interview notes are free-form, introduces a short scorecard with written anchors, applies it to every candidate, and then compares pass rates before and after. The answer ends with what was measured and what changed.
Common mistakes
- "I would make sure we hire a set number from each group." (wrong) That describes a quota, which summaries of the EEOC guidance call unlawful. (right) "I would widen the pool, standardize evaluation and measure each stage."
- "Unconscious bias training fixes hiring." (wrong) Dobbin and Kalev report that its positive effects rarely last beyond a day or two and that it can spark backlash. (right) "Training alone is not enough; I would change the process and add accountability."
- "Structured interviews remove bias." (wrong) They reduce it. (right) "Structure reduces bias, so I still check pass rates by stage."
- "I would lower the bar to widen the pool." (wrong) The aim is a larger pool judged by the same standard. (right) "The bar stays the same and is applied the same way to everyone."
Follow-up questions
How do you answer diversity hiring questions as a candidate?
Describe what you did or would do in the same four moves, and give one measured result. The full candidate-side approach is in how to answer interview questions about diversity.
Is a diversity hiring strategy legal in the United States?
Outreach, pipeline widening and consistent evaluation are generally treated as permissible, while quotas, set-asides and preferences are risky or unlawful, according to the legal summaries cited above. Check specifics with counsel.
Which metrics show whether a diversity hiring strategy works?
Selection rates by group at each funnel stage, compared with the highest group's rate, and the same figures over time. Add interview-score distributions by interviewer to catch uneven scoring.
Does diversity training work?
Reported evidence says its positive effects rarely last beyond a day or two and it can backfire, so it should not be the centerpiece of an answer.
What should you ask an employer about diversity hiring?
Ask how interviews are structured and how pass rates are reviewed. Other good prompts are in questions to ask in an interview.
- Answer with four moves: widen the pool, standardize evaluation, measure each funnel stage and stay lawful.
- Outreach changes who applies; structure changes how people are judged; neither needs a quota.
- Structured interviews rank first in the 2022 re-analysis and also score above unstructured ones in the older estimates.
- Pass rates by stage show where a gap opens and which fix to use.
- Say what you would not do, and why.
Sources
- Revisiting meta-analytic estimates of validity in personnel selection (Sackett, Zhang, Berry, Lievens) — Journal of Applied Psychology 107(11), 2040-2068, 1 Nov 2022
- The validity and utility of selection methods in personnel psychology (Schmidt and Hunter) — Psychological Bulletin 124(2), 262-274, 1 Sep 1998
- Uniform Guidelines on Employee Selection Procedures, 29 CFR 1607.4 (four-fifths rule) — U.S. Code of Federal Regulations (eCFR), 25 Aug 1978
- Why Diversity Programs Fail (Dobbin and Kalev) — Harvard Business Review, 1 Jul 2016
- The Effect of the Rooney Rule on Implicit Bias in the Long Term (Celis et al.) — ACM FAccT 2021, 1 Mar 2021
- Why the Rooney Rule Fumbles: Limitations of Interview-stage Diversity Interventions in Labor Markets — ACM EC 2025, 1 Jul 2025
- EEOC rescinds affirmative action guidance — Legal 500, 21 Aug 2026
- New EEOC guidance advises on acceptable workplace DEI practices — Harris Beach Murtha, 1 Mar 2025
- Insights: EEOC DEI discrimination guidance — Jackson Walker, 1 Mar 2025
