Take an identical resume. Same experience, same skills, same achievements. Change only the name.

Put "John Smith" at the top: 50 callbacks.

Put "Jamal Jackson" at the top: 37 callbacks.

Put "Aisha Patel" at the top: 39 callbacks.

Put "Wei Chen" at the top: 41 callbacks.

This isn't hypothetical. Harvard researchers did this exact study in 2003 and have replicated it multiple times since. The name-based discrimination is consistent, measurable, and costly.

The Research is Clear

Marianne Bertrand and Sendil Mullainathan studied 1,300+ job applications with identical resumes, only varying the names to signal race/ethnicity:

"White-sounding" names received 50% more callbacks than Black-sounding names
Effect held across job types — both skilled and unskilled roles
Effect held across industries — even organizations with diversity statements in their job postings showed the same bias
Extra experience didn't help — Black applicants needed 8+ years of experience to get the same callbacks as white applicants with 6 years

The cost: massive talent pool being filtered out before anyone ever meets them.

For companies: you're rejecting qualified people based on name alone.

The Halo Effect Makes It Worse

Once a resume passes the name test, other biases kick in. This is the halo effect: one positive trait colors how you interpret everything else.

Someone from MIT's name at the top? Their accomplishments get interpreted favorably. A typo becomes "attention to detail issue but otherwise strong." A gap year becomes "taking time to think strategically."

Same typo from a state school? It's "careless." Same gap year? It's "unreliable."

One study found that identical CVs were rated 25% higher when attributed to a man versus a woman in STEM fields. The reviewers thought the man's work was more "impressive" and "creative" — from the exact same project description.

The Affinity Bias Problem

We naturally like people like us. This is affinity bias. In resume screening, it means:

Reviewers rate candidates from their own demographics 15-20% higher on "potential"
Candidates with similar educational backgrounds get interpreted as "stronger fit"
Unexplained resume gaps get excused for similar-background candidates but penalized for different-background candidates
Career path diversity gets seen as "scattered" instead of "creative"

The practical result: your resume screening system is filtering for "similar to the people already here," not "good at the job."

Where the Bias Hides

Resume screening doesn't feel biased when you're doing it. You're being objective — you're reading resumes, not meeting people. But the biases are built into the evaluation:

School prestige bias

MIT engineering degree gets higher weight than Georgia Tech or University of Washington
(Same program, different names)

Company prestige bias

"Google engineer" seems stronger than "smaller startup engineer"
(Might be equally skilled, smaller company just gave them less brand value)

Title consistency bias

Same role under three different company titles gets evaluated differently
"Senior Software Engineer" = strong. "Staff Engineer" = unknown prestige?

Experience framing bias

"Led team of 5" sounds better than "Worked with cross-functional team"
(Might be the same project, different framing)

Resume formatting bias

Simple, clean resume reads as "professional"
Creative resume reads as "less serious" (bias against younger, design-focused candidates)

None of these are about job capability. They're about resume presentation and background prestige.

The Numbers

Resume-screening bias has real costs:

Lost talent:

If 25% of your diverse applicant pool gets filtered out at resume stage due to name-based bias, you're literally rejecting qualified people before they're ever evaluated
For a company hiring 50 people per year with 20% diverse applicant rate, that's 5+ diverse candidates rejected before any human assessment

Reduced quality:

You're not getting the best people. You're getting the best people who happen to have names that pass your unconscious filter
The person from the "unknown" state school might have been stronger, but you'll never know

Cost of bias:

If 5 strong diverse candidates per year are rejected at resume stage, and each would have been in your hire, that's $250k-$500k in lost productivity and innovation diversity

How to Remove Resume Name Bias

The fix is simpler than the problem:

1. Blind Resume Review

Remove names (or use first initials and last names only)
Remove school names (replace with "State University" or "Regional Tech School")
Remove company names (replace with "Fortune 500 Technology Company")
Hide graduation dates

2. Create Evaluation Rubric

Before reviewing any resumes: define what you actually need
"5+ years of experience in X" (not "experience at prestigious company")
"Track record of leading projects" (not "leadership at well-known company")
Score on rubric, don't use intuition

3. Standardize Scoring

All resumes scored on same 5-point scale
Written criteria for each score level
No "potential" or "seems strong" — only documented qualifications

4. Multiple Reviewers

Have 2-3 people score each resume independently
Average the scores
Individual bias gets diluted

5. Verify Blind Process Worked

After hiring, look back: did you end up with "typical" hires or diverse hires?
If your resume screening went blind and you still hired 100% from prestigious schools/companies, something else is filtering

What Changes

Companies that implemented blind resume review reported:

20-30% increase in diverse applicants advancing past resume screening
No decrease in quality (measured by first-year performance)
Actually better quality in some cases (had to evaluate actual skills instead of brand prestige)

The halo effect disappears when you can't see the halo.

The Bottom Line

Resume bias isn't about bad people. It's about a bad system. You're making the resume do too much work — it's deciding both capability AND background prestige, and you're confusing the two.

Remove the system's access to the names, schools, and company brands that trigger unconscious bias. Force the evaluation to be about actual qualifications.

You'll hire better people. And you'll actually see them if they have different-sounding names.

Research references: Bertrand, M., & Mullainathan, S. (2003). Are Emily and Brendan More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination. American Economic Review. Additional replication studies: Quillian et al. (2019) Meta-analysis of field experiments.

How Resume Bias Costs You Top Talent (And You Don't Even Know It)