Gail Greenfield, principal at Mercer LLC, presented the results of her study of career outcomes of women who received bachelor’s degrees in engineering.1 Her analysis was based on data from the Department of Education’s Baccalaureate and Beyond (B&B) Longitudinal Study, which gathers information on the postgraduation experiences—education, employment, and family experiences—of people who have graduated with a bachelor’s degree. The B&B study also collects information on degree recipients’ undergraduate experiences (e.g., major field of study, how they paid for college) and demographic information such as gender, ethnicity, and age. It follows multiple cohorts of students over time, drawing its initial cohorts from the National Postsecondary Student Aid Study (NPSAS).
The B&B study follows three cohorts of a representative sample of graduating seniors in all majors over time:
- Cohort 1, about 11,000 students who graduated with a bachelor’s degree during the 1992–1993 academic year. These students were surveyed in their last year of college and then one year, four years, and ten years after graduation (in 1994, 1997, and 2003, respectively).
- Cohort 2, a similarly sized group of students who graduated with a bachelor’s degree during the 1999–2000 academic year. For this cohort only one follow-up survey was conducted, one year after graduation (2001).
- Cohort 3, a larger sample (nearly19,000 students) who graduated with a bachelor’s degree during the 2007–2008 academic year. A follow-up survey was conducted one year after graduation (2009), and another is in progress.
Greenfield focused her presentation on her analysis of Cohort 1, the group for which 10 years of data are available, and used the other two cohorts primarily for comparisons. She used the B&B data to address three questions:
- What are the career outcomes of women who receive bachelor’s degrees in engineering?
- How do these outcomes compare to men who receive bachelor’s degrees in engineering and to women who receive bachelor’s degrees in other “career-oriented” majors (e.g., business and management, education, and nursing)?
- What factors help explain these observed career outcomes?
LIMITATIONS OF THE B&B DATA
The primary limitation of the B&B data are the small number of engineers represented in each survey, noted Greenfield. Of the approximately 8,000 people (out of 11,000) in Cohort 1 who responded to all four surveys, fewer than 7 percent (520 people) graduated with a degree in engineering, and only about 80 of those were women (Table 2-1). In addition, the online tool that provides access to the data does not permit the pooling of cohorts and limits the ways in which variables can be used and created. Last, the survey does not distinguish managers and supervisors who oversee engineering-related work as opposed to work of any other type. This limitation is most problematic for Cohort 1 as one might expect a sizable number of these graduates to have progressed to manager or supervisor by ten years after graduation.
Greenfield examined two key career outcomes: (1) the percentage of graduates in the labor force (“labor force
1The slides of Greenfield’s presentation are in Appendix F.
TABLE 2-1 Number of engineering graduates responding to each survey, by cohort and survey year
|Engineering graduates responding to the survey|
*Includes the base-year survey in 1993.
SOURCE: National Center for Education Statistics
participation”)2 and (2) the percentage of employed graduates working in the field of their major (“retention”), using the following two types of analysis:
- Descriptive comparisons: To the extent that the B&B surveys are representative, descriptive comparisons depict the population of individuals who received undergraduate degrees in engineering in the graduation year under consideration (e.g., 1992–1993). Such statistics are useful for understanding gender similarities and differences in career outcomes for those with bachelor’s degrees in engineering.
- Multiple regression: Regression analysis can yield insights into the reasons for observed differences in career outcomes. It is used to examine the relationship between an outcome of interest (e.g., labor force participation) and a variable hypothesized to be related to this outcome (e.g., gender), holding constant other factors that may also be related to the outcome (e.g., age, marital status, undergraduate debt).
DESCRIPTIVE COMPARISONS: LABOR FORCE PARTICIPATION
Greenfield found that for engineering bachelor’s degree recipients in Cohort 1, the labor force participation rate is quite high for both men and women (Figure 2-1), although women’s participation is consistently lower than men’s in the three survey years. This difference in labor force participation is not statistically significant until 10 years after graduation, when the men’s participation rate is more than 10 percentage points higher.3
Greenfield’s analysis revealed that labor force participation for women with undergraduate engineering degrees is quite similar to that of women in other career-oriented majors. One year after graduation, the participation for women engineering graduates in Cohort 1 was 88.6 percent—similar to, though a little lower than, the rate for other career-oriented majors (specifically, education, business and management, and health). Four years later, that rate increased to 92 percent, again almost indistinguishable from the other fields. But ten years after graduation, labor
FIGURE 2-1 Labor force participation rate by gender for engineering bachelor’s degree recipients, Cohort 1 of the National Center for Education Statistics Baccalaureate and Beyond Longitudinal Study. * Difference between men’s and women’s rates is statistically significant at the 5 percent level. SOURCE: National Center for Education Statistics.
2Labor force participation rate is the percentage of working-age persons in an economy who are (1) employed or (2) unemployed but looking for a job.
3Greenfield also compared labor force participation across all three cohorts of women engineering graduates one year after their graduation. For Cohort 1, labor force participation was 88.6 percent (1994); for Cohort 2, 93 percent (2001); and for Cohort 3, 87.7 percent (2009). The pattern—lower for Cohort 1, higher for Cohort 2, lower for Cohort 3—was the same for men in those years.
FIGURE 2-2 Percentage of employed engineering bachelor’s degree recipients in an engineering/architecture occupation, Cohort 1 of the National Center for Education Statistics Baccalaureate and Beyond Longitudinal Study. ^ Difference between numbers for women and men is statistically significant at the 10 percent level. * Difference between numbers for women and men is statistically significant at the 5 percent level. The decline in the percent retained from four to ten years after graduation is statistically significant at the 5 percent level for both men and women. SOURCE: National Center for Education Statistics.
force participation had declined for women in all four fields, most notably for those in education, where participation had dropped to 78.7 percent; for those with degrees in business and management, engineering, and health-related fields, the percentages participating in the labor force were in the high 80s.
DESCRIPTIVE COMPARISONS: RETENTION IN ENGINEERING
Greenfield analyzed the B&B data to determine the rates at which labor force participants with bachelor’s degrees in engineering remain in engineering or leave for other types of jobs, and whether women are staying in engineering at the same rates as men.4 She compared the retention rates of employed men and women who were engineering bachelor’s degree recipients in Cohort 1 at various points after graduation (Figure 2-2).
Female engineers have higher retention rates in engineering careers than male engineers in engineering careers one year, four years, and ten years after graduation
Surprisingly, Greenfield said, retention rates in engineering careers 1, 4, and 10 years after graduation were actually higher for women engineering graduates than for men, although by the 10th year the rate had dropped substantially for both sexes. This does not take into consideration the substantial migration of male engineers into engineering management career paths, where female engineers are less well represented.
When comparing the retention rates of women engineering graduates one year after graduation to those of women graduates in other career-oriented majors, the former remain in their field at a rate (67.8 percent) similar to that of their peers in education (69.7 percent) and health (75.3 percent), and significantly higher than for those in business and management (52.2 percent). Four years after graduation, the pattern is similar. Ten years after graduation, however, women’s retention in engineering has dropped to 52.5 percent, lower (although not significantly) than that of the women graduates in business and management (58.6 percent), and significantly lower than that of health professions (72.2 percent) and education (72.5 percent).
Next, Greenfield looked at retention rates 1 year after graduation for women across all three cohorts and found a significant drop between Cohort 1 and Cohort 3 (Figure 2-3). Specifically, retention in engineering was 36.4 percent for Cohort 3 in 2009, as compared to 67.8 percent for Cohort 1 in 1994.
Ten years after graduation, women’s retention in engineering careers is significantly lower than women with degrees in other career-oriented fields such as health, business, and education.
The pattern was not the same for the men, whose retention rates across cohorts held steady between 50 percent and 60 percent (Figure 2-4).
Why did those who left engineering for another field decide to leave? When the B&B researchers asked the engineering graduates in Cohort 3 who were working outside of engineering their primary reason for doing so, the most frequently given answer was the same for both men and women: A job was not available in their field (Table 2-2). However, while that was the primary reason for nearly half the women, it was the primary reason for only one-third of the men. The second most frequently given reason for men (32 percent) was pay/promotion opportunities; in other words, there were more lucrative opportunities
4In the B&B study, engineers and architects are grouped in a single occupation category.
FIGURE 2-3 Percentage of employed women with a bachelor’s degree working in the field of their major (engineering) one year after graduation, by cohort of the National Center for Education Statistics Baccalaureate and Beyond Longitudinal Study (B&B). ^ Difference between this and prior cohort is statistically significant at the 10 percent level. SOURCE: National Center for Education Statistics.
FIGURE 2-4 Percentage of employed men with a bachelor’s degree working in the field of their major one year after graduation, by cohort of the National Center for Education Statistics Baccalaureate and Beyond Longitudinal Study (B&B). SOURCE: National Center for Education Statistics.
elsewhere. In contrast, only 11 percent of women cited this as their primary reason. For women the second most frequent reason was a change in career interests; although only about 16 percent of the women respondents cited this reason, this was still almost four times the percentage of men who selected this answer.
MULTIPLE REGRESSION RESULTS
To interpret the B&B data and shed light on the reasons for observed differences in career outcomes, Greenfield conducted logistic regression analyses. Specifically, she modeled the following two outcomes for Cohort 1:
- Likelihood that graduates are participating in the labor force (“participation”)
- Likelihood that graduates are working in the field of their major (“retention”)
The 15 predictors considered included gender, ethnicity, and undergraduate GPA, among other factors (see list below). The regression models revealed that changing one predictor (when all others are held constant) changes the odds of a graduate either participating or being retained in the labor force.
The participation regression model was used to determine the factors or characteristics related to whether women graduates with a degree in either engineering or another “career-oriented” major are participating in the labor force. The model revealed that women engineering graduates in Cohort 1 are generally no more or less likely to be participating in the labor force than their peers in other career-oriented majors. The model also showed that both groups of women graduates:
- are less likely to be in the labor force if they have children;
- are less likely to be in the labor force one and ten
TABLE 2-2 Primary Reasons for Working Outside Degree Field, Engineering Bachelor’s Degree Recipients in Cohort 3
|Job in field not available (47.6%)||Job in field not available (33.8%)|
|Change in career interests (16.6%)||Pay / promotion opportunities (32%)|
|Job location (12.4%)||Other factors (25.1%)|
|Pay / promotion opportunities (11%)||Change in career interests (4.6%)|
|Working conditions (8.5%)||Job location (4.0%)|
|Other factors (4.0%)||Working conditions (0.5%)|
|Family-related reasons (0%)||Family-related reasons (0%)|
years after graduation if their spouse’s income is relatively high;
- are more likely to be in the labor force one and four years after graduation if they incurred debt in college;
- are less likely to be in the labor force one year after graduation if they are enrolled in school, but are more likely ten years after graduation; and
- are more likely to be in the labor force ten years after graduation if they have a graduate degree.
The focus of the retention regression model was on factors or characteristics related to engineering graduates’ retention in the field of engineering. According to the results, women with undergraduate engineering degrees are not retained at lower rates than men with undergraduate engineering degrees; in fact, they are more likely than men to be working in engineering four years after graduation. The regression model also found that both men and women were more likely to be working in engineering one year and ten years after graduation if their undergraduate GPA was 3.5 or above.
In terms of the findings’ implications, Greenfield clarified that participation in the labor force is not a concern: women are participating in the labor force at high rates for all three cohorts. It is retention in engineering that appears to be a challenge for both men and women: ten years after graduation, only about 50 percent of men and women in Cohort 1 were working in engineering jobs. In Cohort 3 both men and women cited a lack of jobs in their field as their primary reason for working outside engineering. Moreover, the retention rate appears to be worsening for women, since the one-year retention rate fell more than 30 percentage points between Cohort 1 in 1994 and Cohort 3 in 2009 (Figure 2-3).
Erin Cadwalader from the Association for Women in Science (AWIS) commented on Greenfield’s finding that a higher number of women than men said there were no jobs available: could this be because there are more women in certain subfields of engineering? Greenfield explained that the data did not capture that information, but acknowledged the possibility that more women might have a degree in an engineering subfield that is not in demand in the marketplace. Given what she had heard anecdotally, she said she was shocked to see that almost half the women said there was not a job available; she assumed that what they meant was “there was not a job that fit my skills or requirements.”
Thomas Perry from the American Society of Mechanical Engineers (ASME) pointed out that, in looking at the difference in the reasons men and women gave for working outside their degree field, the variations between the men’s and women’s answers tend to have a magnitude between 0 and 3 to 4 times; the number who cited “working conditions,” in contrast, has a variation of 16 times. Greenfield replied that she had tried to find the survey to see what was meant by “working conditions” but was unable to. For men, she noted, the reasons given for working outside the field were more dispersed; for example, the number who cited “pay and promotion opportunities” is not much lower than those who cited “jobs not available.” For women, on the other hand, the reasons were concentrated in “jobs not available” and “changes in career interests.”
Andy Richman from Carnegie Mellon University asked whether Greenfield had looked at the possible influence of the economy on retention rates—for example, when Cohort 1 graduated in 1992 unemployment rates were quite high. Greenfield responded that she had looked at recessions (there was not a recession in 1992) but not unemployment rates. Richman explained that some research shows that for those who graduate during a recession, their career earnings never recover. Follow-up data for Cohort 3, who graduated
|Fifteen Predictors Considered in the Regression Model|
|Gender||undergraduate debt||Citizenship Status|
|Age at graduation||Undergraduate institution (Carnegie Classification)||Spousal employment|
|Ethnicity||Parental education||Spousal income|
|Undergraduate GPA||Marital status||Graduate degree|
|undergraduate degree specialty/subfield||Parental status||Employment status|
Note: Not all predictors were included in the final set of models.
during the latest recession, may reveal its impacts on their careers over time.
Jennifer Hunt, chief economist at the Department of Labor, suggested comparing the rates at which men and women leave engineering with the rates for other fields. Her research found that women were exiting at higher rates than men in all fields, but maybe that has flipped and is now the opposite, she said. It could be that women are leaving engineering less and also leaving other fields less as well—that women are now staying longer with careers in general. Pursuing this question is a good idea, said Greenfield, noting that she has the data to do so.
Echoing Richman’s question about the economy’s effects on retention, Patricia Taboada-Serrano from the Rochester Institute of Technology asked whether Greenfield could speculate about the reason for the dramatic drop in women’s retention rates between Cohort 1 in 1994 (67.8 percent) and Cohort 3 in 2009 (36.4 percent). Was it mostly a response to job market conditions, or were there other reasons behind it? Greenfield did not have this information and hoped some of those at the workshop would do the research to find an answer.
Joanne Cohoon of the University of Virginia wondered whether the decline in retention could be related to the increasing representation of women with engineering degrees. Women have been successfully recruited into the line of study, she explained, but their educational experience may have been such that they concluded, “I’m going to finish this degree, but I’m getting out of the field.” This information will have to be captured in future surveys.