Can Female Doctors Cure the Gender STEMM Gap? Evidence from Randomly Assigned General Practitioners

Girls exposed to female doctors in childhood are significantly more likely to pursue traditionally male-dominated education programs in high school and college, such as STEMM. This effect is particularly strong in high-ability girls with parents who have low levels of education, suggesting that exposure to female doctors can contribute to reducing inequality and improving intergenerational mobility.

• Girls assigned to female primary care doctors in childhood are 4 percentage points more likely to choose a STEMM program in high school than girls with non-female primary care doctors.
  • Girls who were exposed to a female GP had a 0.047 unit increase in STEMM GPA.
  • Having female primary care doctors can close the gender gap in high school and college STEMM choice each by almost 20%. Girls assigned to female primary care doctors in childhood are 2 percentage points more likely to choose a STEMM program in college than girls with non-female primary care doctors.
• Girls assigned to female primary care doctors were also more likely to perform better than girls with non-female primary care doctors in both STEMM and non-STEMM fields—although the impact was less for non-STEMM fields.
• Having a female doctor in childhood did not significantly impact girls with mothers with a college degree or more. All statistically significant effects were found to be driven mainly by girls whose mothers had less than a college education.
• Being assigned to male primary care doctors in childhood had no impact on boys’ eventual choice in educational fields nor on their academic performance.

Researchers used Norwegian government’s administrative records of Norwegian residents born between 1988 and 1996 (and graduated from high school between 2006 and 2014), to investigate the impact of gender of the primary care doctor on girls’ long-term educational choices. The government data provided unique individual identifiers through which different sets of data from the universal healthcare system and the education system could be linked to individuals over time. The healthcare registry allowed researchers to track changes in a child’s government-assigned general practitioner (GP) due to causes outside of the patient’s control (such as the patient’s original GP was retiring) between 2002 and 2011. They then analyzed government records to see whether young girls who were randomly assigned a female GP would go on to make different life choices as high school and college students regarding STEMM education.

To check the robustness of their findings, researchers investigated whether the girls with female GPs were experiencing different educational outcomes due to factors other than the “role model effect,” such as due to health (because having a same-gender GP impacted their health) or due to the mother (in Norway, typically the mother would be assigned the same GP as their child). They found no evidence suggesting that either of these alternative effects would explain the different educational decisions of girls with same-gender GPs. GP assignment provides an opportunity to investigate intermediary role model involvement, compared to studies that examine the effectiveness of teachers or one-off mentorship programs. Patients would typically see their GP 30 times before they made their choice about their high school degree track (STEMM or non-STEMM). This is less frequent exposure than to a teacher they may see daily but more frequent exposure than a role model they may meet for a short period of time through a mentorship program.