Women are significantly underrepresented in the finance profession. In the United States, for example, women make up almost 60% of college graduates but less than one-fifth of investment professionals and one-tenth of mutual fund managers. This disproportion is reflected in the CFA Institute membership demographic makeup: As of June 2016, women made up 18.4% of the CFA Institute global membership. By encouraging women to consider careers in finance, the profession can help diversify and expand its talent pool. But to accomplish this task, the profession must try to understand why a gender imbalance currently exists.

Some potential reasons might include discrimination, a lack of the requisite education or training in finance, and a need for greater time flexibility than the industry is typically seen to offer. Another might be parental influence. Other studies have shown that role models have a considerable effect on educational attainment and career choice. This study explores the extent to which women have been influenced to consider a career in finance by role models—particularly female role models.

To explore the hypothesis that role models affect the career choices of men and women differently, the authors look at the parents of CFA Institute members who work in an occupation related to science, technology, engineering, and mathematics (STEM)—including finance and economics. They choose these occupations because past studies have shown that women are typically underrepresented in STEM fields, and investment management—like other STEM fields—is math intensive.

The authors use data from an electronic survey sent by CFA Institute to its global membership in June 2016 (about 135,000 people). Respondents were asked whether their father and/or mother had worked in a STEM occupation before they were 14 years old and whether they have brothers and/or sisters who currently work in a STEM field. The authors use univariate analysis to examine the responses from 5,020 members who answered all four questions—a fifth of whom were women.  They then estimate a linear probability model to understand the marginal effects of STEM parents and siblings on the probability that a CFA Institute member is a woman.

The authors also hypothesize that any difference in the influence STEM parents might have on girls and boys could be related to the math gender gap. It is possible that girls benefit more from having STEM parents involved in their math education than boys do and are better prepared to pursue a career in finance. To test this, the authors estimate a regression on data from the Programme for International Student Assessment (PISA)—a global survey that measures math performance of 15-year-old students—collected in six waves from 2000 to 2015.

The authors’ analysis of the CFA Institute survey data strongly suggests that girls and boys are influenced differently by their parents’ career choices. Among members of CFA Institute, women are more likely to have had a STEM parent—in particular, a STEM mother—than men. Having a STEM mother increases the baseline rate at which women become members by 48% more than for men; a STEM father increases the baseline rate for women by 29% more than for men.  These effects, the authors show, are both large and statistically significant. Women and men CFA Institute members are both more likely to have a STEM family member than a member of the general population, but not to the same extent.

A similar pattern emerges when assessing STEM siblings of CFA Institute members; women’s sisters and brothers have a higher probability of working in a STEM field than men’s. Again, this shows that gender modulates outcomes—though with siblings, it is more difficult to predict who (if anyone) influenced the career choice of whom.

After running a regression of the data from PISA, the authors find that while having a STEM parent is associated with higher student math scores, girls consistently score lower than boys. Measuring the effect STEM parents have on girls versus boys does not produce any convincing evidence that either a STEM mother or father reduces the math gap.

By finding a strong correlation between early role models and career choice in women, the authors raise intriguing possibilities for how the gender imbalance in the finance profession might be redressed. The authors note that they were unable to find a connecting mechanism—whether, for example, women might be inspired by a parent to pursue a male-dominated career or whether STEM parents affect the career choices of their daughters indirectly by encouraging academic achievement. But regardless of the precise mechanism, the study makes a compelling case that providing role models for girls at formative stages of their lives might make them more willing to consider a career in finance.