Women and the Choice to Study Economics

Abstract

Underrepresentation of women in economics is documented in many studies. Investigation of its sources at the undergraduate level is examined through students’ decisions to persist in economics, either beyond an introductory course or in their major choices. The authors add to the literature by analyzing students’ decisions to take their first introductory economics course, an intermediate theory course, and ultimately major in economics, using the Multiple-Institution Database for Investigating Engineering Longitudinal Development. Results indicate that a smaller percentage of women take economics at all levels—introductory courses, theory courses, and majoring in economics. Even after controlling for aptitude, demographic characteristics, prior interest, course performance, environment, and course timing, persistent gender differences in the likelihood of partaking in economic education beyond the introductory course decision endure.

Keywords:

• economics education
• economics major
• undergraduate economics

JEL code:

• A2

Notes

1. Some students enroll in introductory micro, introductory macro, an introductory concepts course, or some combination of all three courses simultaneously. We do not differentiate which or how many introductory courses the student has taken in the first semester that they enroll in an introductory economics course.

2. Dynan and Rouse (1997) found considerably higher proportions of students majoring in economics than national averages, and suggested that this difference lay in the fact that their data were drawn exclusively from a liberal arts institution with limited major selections. In a similar line of reasoning, the presence of schools of business and engineering at the institutions in the MIDFIELD database likely decreases the number of majors in our sample.

3. The cumulative GPA is measured through the semester of enrollment in the student's first introductory economics course, but does not include their grade from the introductory course. This cumulative GPA is used to calculate the student's relative grade measure.

4. Because the first introductory course in economics can be introductory microeconomics, introductory macroeconomics, an introductory concepts course, or some combination of all three courses, we define a student's grade in the introductory course as the average of the introductory courses they took in the first semester they took any introductory course.

5. Because relative grade is a composite of course grade and the student's overall GPA, the impact of changes in grades or GPA on persistence will differ across students with different initial grades or GPAs. Thus, we describe the impact of changes in GPA using a student with an average course grade. Similarly, the impact of changes in cumulative GPA is demonstrated for a student with an average course grade and GPA.

6. The sample average cumulative GPA (at the point of completing the first introductory course) is 2.571 for male and 2.909 for female students, as reported in table 2. The standard deviation for male and female students is 0.665 and 0.615, respectively. Calculations for percentages are based on coefficients in model 3.

7. This partial letter grade improvement translates into a 0.3 increase in course grade. While the direct impact of this on the probability of taking intermediate theory is to increase it by 3.0 percent (based on model [3]), the relative grade influence reduces this probability by 1.0 percent.

8. This partial letter grade improvement translates into a 0.3 increase in course grade. While the direct impact of this on the probability of majoring is to decrease it by 8.8 percent (based on model [3]), the relative grade influence increases the probability by 6.7 percent.

9. The sample average cumulative GPA (at the point of completing the first theory course) is 2.848 for male and 2.958 for female students. The standard deviation of cumulative GPA for male and female students is 0.627 and 0.603, respectively. Calculations for percentages are based on coefficients in model (3).