6/ What does this mean?

Unintended consequence of referral hiring: the gender composition of referrers will be reflected in the gender composition of the candidates they refer.

Potential solutions:
✅ Encourage opposite-gender referrals
✅ Turn to the underrepresented gender for referrals.

5/ Participants who refer an opposite-gender candidate…

… are more likely to have opposite-gender friends, be master’s students, and be socially popular.

… tend to choose referrals from outside their closest network or class.

… are more likely to refer high-achieving students.

4/ Does job type matter?

Participants were randomized across two jobs. We find suggestive evidence that the degree of gender homophily in job referrals is slightly stronger when the job is more consistent with stereotypes associated with the participant’s own gender. (p=0.07, diff=8 ppt)

3/ What drives this?

Gendered friendship networks are important! On average, 79% of participants' closest friends were of the same gender. Since half of referrals were close friends, this explains much of the gendered referral patterns.

2/ Key findings:
🔹 73% of participants referred someone of their own gender.
🔹 71% of female participants referred a woman.
🔹 75% of male participants referred a man.
🔹 This is 45% higher than random referral behavior would suggest.

1/ Study design:
📍 453 university students at a Swedish business school.
📍 Asked to refer a peer for a real job.
📍 Strong financial incentive (~$580 finder’s fee).
📍 Job ads randomized between two roles: one more stereotypically male, one more stereotypically female.