Abstract

1. Introduction

Despite the increasing number of women pursuing higher education and entering the labor market, the fields of science, technology, engineering, and mathematics (STEM) remain male-dominated. The nuclear industry is a particularly notable example. According to the Nuclear Energy Agency (NEA) within the Organisation for Economic Co-operation and Development (OECD), women constitute only 24.9% of the nuclear workforce in its 17 member countries [1].¹ To improve female representation, the International Atomic Energy Agency (IAEA) has initiated several programs, including the Women in Nuclear Security Initiative (WINSI) and the Marie Sklodowska-Curie Fellowship Programme, while the NEA collects data on women in the nuclear sector globally to develop policies and practices that improve gender balance in the field [2,3]. On the research front, scholars have tried to identify the factors that could discourage women from choosing nuclear engineering. A commonly studied factor is gender differences in the perception towards nuclear energy. Studies have shown that women tend to have a greater opposition to nuclear energy compared to men, largely due to concerns about radiation exposure and associated risks during pregnancy, as well as a lack of knowledge and familiarity with the field relative to men [[4], [5], [6], [7]]. These differences may deter women from pursuing a degree in nuclear engineering and a career in nuclear-related fields.

This study aims to contribute to the existing research and policy efforts by examining female representation among faculty in nuclear engineering. Our focus pertains to the findings from research in economics and education that show having a role model has an important effect for historically under-represented populations, including women, to break stereotypes and pursue the corresponding career path. Specifically relating to faculty composition, various studies report evidence suggesting that limited opportunities to meet women in the same field of work might dissuade women from pursuing their field of interest (see Ref. [[8], [9], [10], [11], [12], [13]]). In fact, this line of studies supports a series of gender-conscious policies and practices promoting female scholarship and gender equality in STEM fields, such as affirmative action in faculty recruitment and promotion, research grant awards and funding programs, and mentoring and training services at the national, local, and institutional levels ([14]; [23]; [[15], [16], [17]]).

As a result, we assess the extent to which nuclear engineering is successful in attracting women to its workforce by comparing its female faculty share with that of other fields across all majors. To do so, we collect college-major level information among four-year Korean colleges from 2000 to 2022. The information includes faculty size, the percentage of female faculty members, as well as the number of degree recipients and gender composition in undergraduate, master's, and doctoral programs for each college-major in a college.

During the sample period, the nuclear engineering major had the third lowest percentage of full-time faculty members who were women at 2.55%, after mechanical engineering (1.63%) and electrical and electronic engineering (1.93%). The share of female faculty members in nuclear engineering increased by 5.08 percentage points from 2000 to 2022, which was the fifth smallest increase among all majors. The majors with smaller increases were physical science (1.99%pts), electrical and electronic engineering (3.40%pts), mechanical engineering (3.43%pts), and arts and athletics (4.24%pts).

However, once we control for various accounting factors using multivariate linear regression models, the nuclear engineering major is comparable to electrical and electronic engineering, mechanical engineering, and physical science in terms of both the overall and time-trend female faculty share. Moreover, contrary to the raw statistics, we find that the nuclear engineering major has an annual increase in female faculty share similar to food and nutrition science and nursing and health sciences, which have high shares of female faculty.

Our findings imply two important policy implications for nuclear engineering. First, assessing whether women are under-represented in a specific field requires careful examinations of compounding factors rather than relying on simple statistics. As shown by our results, nuclear engineering appears to underperform in recruiting female faculty based on simple statistics, but it is comparable to other fields once we control for other factors that could affect female faculty share. Second, to increase female representation in faculty, a field should engage in attracting women at the early stage of their career. We find that a one percentage point increase in female share among bachelor's degree recipients is associated with a 0.127 percentage point increase in female share of full-time faculty members. For nuclear engineering, an increase in the proportion of female bachelor's degree recipients from 20.12% (the average in 2022) to 38.20%, which would align with chemical engineering—the field with the highest proportion of female bachelor's degree recipients among engineering fields—could lead to a rise in the percentage of female faculty from 5.08% (the average in 2022) to 7.38%, representing a 45.28% increase.

The rest of this study is organized as follows: Section Ⅱ describes the status of female representation in the nuclear field in South Korea. Section Ⅲ provides the details about the data and empirical framework. Section Ⅳ presents the findings, and Section Ⅴ concludes.

2. Data and summary statistics

3. Statistical analysis

3.2. Results

The main findings are presented in columns (1) and (2) of Table 3. We use mechanical engineering as a baseline, and all reported estimates for other college majors are relative to mechanical engineering. The estimated value of 

${\alpha }_m$

for nuclear engineering is −1.637, and for chemical engineering, it is −2.186. These results indicate that, compared to mechanical engineering, the female faculty share is 1.637 percentage points lower in nuclear engineering and 2.186 percentage points lower in chemical engineering, even after controlling for other factors that could impact the female faculty share. The difference between mechanical engineering and nuclear engineering (i.e., −1.637) is not statistically significant at the 10% level, while the difference between mechanical engineering and chemical engineering (i.e., −2.186) is statistically significant at the 1% level. This indicates that, given the size of the faculty, undergraduate and graduate programs, and female shares among students, nuclear engineering departments have, on average, the same female faculty share as mechanical engineering departments. However, all else being equal, chemical engineering departments tend to have a smaller share of female faculty than mechanical engineering departments, and this difference is statistically significant at the 1% level.

 

Table 3

	With all degree recipients				With bachelor's degree recipients only
	${\hat{\alpha }}_m$		${\hat{\beta }}_m$		${\hat{\alpha }}_m$		${\hat{\beta }}_m$
	(1)		(2)		(3)		(4)
Field
(Ref. = Mechanical eng.)	–	–	−0.022	(0.030)	–	–	−0.013	(0.029)
- Nuclear eng.	−1.637	(3.228)	−0.122	(0.163)	−1.501	(3.270)	−0.086	(0.165)
- Chemical eng.	−2.186***	(0.832)	0.068	(0.052)	−2.121***	(0.799)	0.099*	(0.052)
- Civil eng.	−2.936***	(0.733)	0.170***	(0.034)	−2.554***	(0.699)	0.188***	(0.034)
- Computer sci.	−1.141	(0.820)	0.206***	(0.040)	−0.699	(0.783)	0.207***	(0.039)
- Electrical and electronic eng.	−0.688	(0.829)	0.037	(0.046)	−0.493	(0.765)	0.023	(0.042)
- Industrial eng.	−4.106***	(0.940)	0.241***	(0.061)	−3.926***	(0.917)	0.277***	(0.063)
- Unclassified, other eng.	−2.437	(2.006)	0.251**	(0.119)	−1.864	(1.940)	0.256**	(0.117)
- Agricultural sci.	−1.800*	(1.007)	0.188***	(0.054)	−1.537	(0.988)	0.218***	(0.054)
- Biology	1.507	(1.067)	0.249***	(0.065)	2.120**	(1.039)	0.284***	(0.065)
- Chemistry	−3.149**	(1.350)	0.148**	(0.073)	−2.043	(1.299)	0.165**	(0.072)
- Food and nutrition sci.	29.754***	(3.366)	0.123	(0.170)	31.533***	(3.485)	0.121	(0.175)
- Math and statistics	−0.927	(1.298)	0.306***	(0.073)	−0.234	(1.275)	0.318***	(0.072)
- Physical sci.	1.116	(1.224)	0.068	(0.068)	1.357	(1.188)	0.095	(0.068)
- Unclassified, other sci.	3.135	(2.298)	0.475**	(0.219)	3.721	(2.310)	0.445**	(0.221)
- Nursing and health sci.	50.047***	(3.572)	−0.084	(0.155)	51.652***	(3.748)	−0.076	(0.159)
- Medicine and pharmacy	4.872***	(0.988)	0.583***	(0.035)	4.736***	(0.970)	0.675***	(0.034)
- Business and economics	−5.535***	(0.806)	0.661***	(0.039)	−4.885***	(0.767)	0.702***	(0.038)
- Other social sci.	8.180***	(1.408)	0.501***	(0.072)	8.825***	(1.384)	0.540***	(0.071)
- Education	10.961***	(1.408)	0.971***	(0.066)	11.458***	(1.407)	1.007***	(0.067)
- Humanities	6.983***	(1.000)	0.757***	(0.040)	7.981***	(0.957)	0.762***	(0.039)
- Arts and athletics	20.927***	(1.333)	0.178***	(0.064)	21.763***	(1.322)	0.179***	(0.064)
$R^2$	0.443				0.438
${\bar{y}}_{m,c,t}$	20.28
N	164,526

 

Notes: The table displays coefficient estimates of weighted regression, with standard errors clustered at the major by institution level in parentheses. Other controls include the number of faculty members, the number and female share of bachelor’, master's, and doctoral degree recipients within a major, a full set of dummies for institution, and a full set of year dummies. The asterisks *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.

 

As for major-specific time trends, we find that the estimated 

${\beta }_m$

is −0.022 for mechanical engineering and −0.122 for nuclear engineering, and both estimates are not statistically significant from zero at a 10% significance level. This suggests that, conditional on the size of the faculty, undergraduate and graduate programs, and their female shares among students, the share of female faculty in mechanical engineering departments decreases by 2.2 percentage points per year on average, but this trend is not statistically significant. Similarly, holding other factors constant, each year, the female faculty share in nuclear engineering departments decreases by an additional 12.2 percentage points compared to mechanical engineering departments (i.e., a total decrease of 14.4 percentage points per year, obtained by adding −0.122 to −0.022), but this difference is not statistically significant at a 10% significance level. That is, nuclear engineering departments share the same time trend with mechanical engineering departments.

 

We have categorized majors into six groups based on the signs and statistical significance of our estimated values 

${\alpha }_m$

and 

${\beta }_m$

(i.e., 

${\hat{\alpha }}_m$

and 

${\hat{\beta }}_m$

) at the 10% level. These categories include: (1) both 

${\hat{\alpha }}_m$

and 

${\hat{\beta }}_m$

are statistically indifferent from zero, (2) both 

${\hat{\alpha }}_m$

and 

${\hat{\beta }}_m$

are positive, (3) 

${\hat{\alpha }}_m$

is negative and 

${\hat{\beta }}_m$

is positive, (4) 

${\hat{\alpha }}_m$

is positive and 

${\hat{\beta }}_m$

is not statistically different from zero, (5) 

${\hat{\alpha }}_m$

is negative and 

${\hat{\beta }}_m$

is not statistically different from zero, and (6) 

${\hat{\alpha }}_m$

is not statistically different from zero and 

${\hat{\beta }}_m$

is positive. In Fig. 1, the x-axis and y-axis represent the signs and magnitudes of 

${\hat{\alpha }}_m$

and 

${\hat{\beta }}_m$

, respectively.

 

1. Download : Download high-res image (901KB)
2. Download : Download full-size image

Fig. 1

At the coordinate (0,0) on the graph, we place four college majors that belong to group (1), including mechanical engineering (the baseline), nuclear engineering, electrical and electronic engineering, and physical science. These college majors exhibit similar performance to mechanical engineering in terms of both the overall and time-trend female faculty shares. The college majors belonging to group (2) can be regarded as those that surpass nuclear engineering majors in terms of the overall and time-trend female faculty shares, as they have significantly positive 

${\hat{\alpha }}_m$

and 

${\hat{\beta }}_m$

.

 

In the upper left quadrant of the graph, we can find the college majors belonging to group (3), which have negative 

${\hat{\alpha }}_m$

and positive 

${\hat{\beta }}_m$

at a 10% significance level. These majors include chemistry, civil engineering, agricultural science, industrial engineering, and business and economics. They perform better than nuclear engineering majors in terms of the time trend, that is, how fast the share of female faculty increases over time, but not in the overall level when considering the share of female students among all students.

 

The remaining three groups (groups (4) to (6)) are comparable to nuclear engineering in terms of either 

${\hat{\alpha }}_m$

or 

${\hat{\beta }}_m$

, but not both. For instance, in terms of the time trend, nuclear engineering shows a similar annual increase in the female faculty share as that of food and nutrition science and nursing and health sciences, which have high shares of female faculty in our sample (45.11%, and 63.15%, respectively).

 

In summary, in terms of securing women in faculty positions, nuclear engineering majors perform strictly worse than five college majors in group (2)– namely, arts and athletics, education, humanities, medicine and pharmacy, and other social sciences. However, nuclear engineering performs no worse than the rest of the college majors considering various compounding factors that could affect female faculty share.

Finally, Table 4 presents the estimates for the female share and number of bachelor's, master's, and doctoral degree recipients as well as full-time faculty members. Column (1) demonstrates that a 1 percentage point increase in the female share among bachelor's, master's, and doctoral degree recipients corresponds to an increase in the female faculty share of approximately 0.127, 0.052, and 0.021 percentage points, respectively. From 2000 to 2022, the female shares among students increased by 3.72 percentage points for undergraduates, 22.49 percentage points for master's degree recipients, and 18.30 percentage points for PhD holders, respectively (see Table 2). Multiplying these numbers with the corresponding estimates and adding them up results in a 2.03 percentage point increase. That means that 2.03 percentage points out of the total increase in the female faculty share (i.e., 5.08 percentage points) are attributable to the increased share of female students among all students at all levels of degree programs, which accounts for 39.89 percent. Finally, we find a negative association between the size of the student body and female faculty share, but a positive association between the number of total faculty members and female faculty share.

Table 4

	With all degree recipients		With bachelor's degree recipients only
	(1)		(2)
Female share of degree recipients
- bachelor's	0.127***	(0.007)	0.140***	(0.007)
- master's	0.052***	(0.005)	–	–
- doctoral	0.021***	(0.007)	–	–
Number of degree recipients
- bachelor's	−0.017***	(0.003)	−0.018***	(0.003)
- master's	−0.010	(0.009)	–	–
- doctoral	−0.085***	(0.020)	–	–
Number of full-time faculty members	0.008***	(0.003)	0.001	(0.003)
$R^2$	0.443		0.438
${\bar{y}}_{m,c,t}$	20.28
N	164,526

 

Notes: The table displays coefficient estimates of weighted regression, with standard errors clustered at the major by institution level in parentheses. Other controls include a full set of dummies for institution, and a full set of year dummies, and a full set of institution-specific linear time trends. The asterisks *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively.

 

As a robustness check, we run the regression model again while excluding the variables for female share and number of master's and doctoral degree recipients. This is because, in certain fields, pursuing a graduate degree abroad is considered crucial for in-depth learning and a smoother transition to work post-graduation. Consequently, the female share and the number of master's and doctoral degree recipients from domestic colleges may not accurately reflect the number of bachelor's degree recipients who pursue graduate studies abroad. To address this limitation, we used an alternative approach to assess the extent to which our results are sensitive to this data constraint. We estimated the female faculty share based solely on information from undergraduates, excluding the number of master's and doctoral degree recipients. While this alternative specification may diminish the predictive power of the estimation results by omitting relevant explanatory variables, it helps avoid potential biases arising from the systematic exclusion of the number of degree holders from abroad. This alternative approach allows us to compare the results with our baseline findings and analyze the impact of limited information on degree recipients from abroad. The results, presented in columns (3) and (4) of Table 3 and column (2) of Table 4, show no significant changes in the outcomes. All coefficients exhibit the same sign and similar magnitudes. The only difference is found in the significance of estimates for agricultural science, chemistry, biology, and chemical engineering, which might be significantly influenced by the number of master's and doctoral degree recipients from abroad. These findings from the alternative approach provide evidence supporting the validity of our results.

4. Conclusion

In this study, we construct a novel dataset that includes panel information for every college major at all Korean universities from 2000 to 2022. The objective is to gauge the success of nuclear engineering in attracting female faculty in comparison to other majors. The study adopts a quantitative approach, and unlike a basic statistical analysis that places the nuclear engineering major third from the bottom in terms of the share of female faculty, our findings demonstrate that it performs comparably to other fields when controlling for the major-specific number and gender composition of degree recipients and college fixed effects.

Our study emphasizes the need for thorough and rigorous quantitative analysis when measuring gender gaps and developing policies to address them. In this regard, our findings raise some concerns regarding a recent gender-related policy implemented in Korea. In January 2020, the [19] Official Act was amended to set a minimum yearly target for the female faculty share in all public colleges, starting from 19.8% in 2023 and reaching 25.0% by 2030. Given that the share of female faculty is only 5.08% in 2022, nuclear engineering majors should actively recruit female faculty to comply with this new policy. However, no study has provided clear statistical evidence suggesting that public colleges are less female faculty-friendly than private colleges in their hiring practices. Conversely, the policy goal seems to be based on the observation that the overall female faculty share in public colleges (17.70% in 2020) is lower than that of private colleges (27.06% in 2020). This gap could be attributed to gender-biased hiring practices in public colleges, but it could also be influenced by other observable differences, as public colleges in Korea tend to have larger engineering schools than private colleges. In the context of nuclear engineering, the female faculty share at public colleges (3.13%) is 2.68 percentage points lower than that at private colleges (5.81%). However, this difference does not necessarily imply that public colleges underperform compared to private colleges under a similar interpretation of our main findings. Without identifying the underlying factors hindering women's status, such a policy may not achieve its ultimate goal. Instead, our findings highlight the importance of attracting female students at all degree levels. This suggests that policymakers might be better off promoting female representation in STEM fields, including nuclear engineering, early on, rather than focusing on achieving a specific share of women among faculty.