Are You My Mentor? A Field Experiment on Gender, Ethnicity, and Political Self-Starters

Politicians are not significantly biased towards male students in mentorship; in fact, politicians may respond to women more positively. However, students’ ethnicity may negatively affect the quality of the responses.

• Politicians were no more likely to favor men than women in responding to requests for mentorship: 25% of male senders received a reply, versus 27% of female senders.
• Male and female senders did not receive significantly different responses in terms of content. Male and female senders received similarly meaningful responses (11% vs. 13%), were equally likely to be praised (5% vs. 6%), and were similarly likely to get substantive advice (7% vs. 8%). Sender gender also did not impact likelihood of an official discouraging an individual from seeking office (1% vs. 1.5%).
• Female students were slightly more likely to be encouraged than male counterparts in terms of offers for help and length of responses.
  • 3% of male senders versus 5% of female senders received offers of help (marginally significant p = 0.09).
  • Female senders received slightly longer responses than male senders on average (145 characters in length vs. 170 characters in length, p = 0.04).
• Male legislators responded to female students slightly more frequently than to male students (27% vs. 24%). There were no differences in response rates to male and female student by female legislators, partisan affiliations (i.e. Democrat vs. Republican) or level of office (i.e. state vs. local).
• Ethnicity may have impacted response quality. Male senders with Latinx last names were significantly less likely to receive a quality response than non-Latinx senders.
  • There was no difference in response rates between Latinx and non-Latinx senders (24% vs. 26%), but there was a significant difference in receiving praise (3% vs. 6%, p = .03), a meaningful response (7% vs. 13%, p = .06), substantive advice (4% vs. 8%, p = .07), or an offer for help (2% v. 4%, p = .08)
  • These differences are primarily driven by the male Latino senders, as female Latina and non-Latina senders do not differ on any of the response measures.
  • Male Latino senders received fewer responses than male non-Latino senders (21% vs. 26%, p = 0.03), received less praise (2% vs. 5%, p = 0.02), and received shorter messages on average (94 characters vs. 152 characters, p < 0.01).

Women who are self-starters and initiate their own political careers are just as likely as men to benefit from small acts of mentorship within politics, which is an encouraging sign in closing a representation gap in the United States government.

The researchers constructed a database of 8,189 elected and appointed officials at city, county, and state levels of U.S. government using information supplied by the New Organizing Institute. Officials received an email from a student that indicated the student was working on a class project about politicians’ career paths, and that they were interested in the field and seeking advice. Officials were block randomized by gender and state. Officials were then randomly assigned within blocks to receive a standardized email from a fictional male or female student. The only indicator for student gender was a name associated with the generated email addresses, which were selected as “typical” names based on Census and Social Security data.

From the 8,189 emails, there were 2,127 responses from public officials (26% response rate). Responses from elected officials were counted, and their length was recorded by log word count and character count. Outcome variables also included whether a reply was meaningful and if it included praise, an offer of help, a warning against running, and/or substantive advice.