Global sex segregation in higher education: trends and explanations

• In Event: Gender equity in Higher education

Proposal

While the world has witnessed a significant increase in women’s college enrollment rates over the past decades, concerns have been raised about whether greater educational expansion led to greater incorporation of women into STEM-related fields that are traditionally dominated by men. It is important because the sex segregation by field of study directly contribute to the occupational sex segregation and gender earning gap. For example, in the US, recent research indicates that 36% of the occupational sex segregation was explained by uneven gender distribution across fields of study (Zhang and Weeden 2022).
The evolutionary narrative predicts that the sex segregation in the higher education will eventually decline as the gender egalitarian values become institutionalized with economic development (Inglehart and Norris 2003). Increasing women participations in higher education among most countries partially supported the evolutionary account. Yet, in recent years, accumulated evidence has shown that women continue to be underrepresented in STEM fields with economic development (Barone 2011; Charles and Bradley 2002, 2009). In addition, the sex segregation by fields of study is more pronounced among advanced industrialized countries and more gender egalitarian countries, a phenomenon famously known as the “gender equality paradox (Stoet and Geary 2018)” or “paradox of affluence (Liu 2019)”.

To explain the paradox, Charles and Bradley (2009) purposed the theory of “indulging gendered selves”. This theory suggests that the traditional gender-role ideology have declined in developed countries, but the gender-essentialism beliefs persist. The strong emphasis on self-expression and self-realization in affluent Western countries reinforce the gender-essentialism ideas by encouraging women to purse feminine fields and averting them from masculine subjects such as natural sciences and mathematics. In contrast, women in less developed countries have strong instrumental motivation to choose lucrative and scientific fields.

Charles and Bradley (2009) used international data from 44 countries in the year 1999 to test these hypotheses. As supporting evidence, they found that economic development, measured as per capital GDP, is positively associated with women representation in humanities, social sciences, and health, but negatively associated with women’s representation in engineering, math, and natural science for developing countries. On the other hand, the girls’ affinity for math is positively associated with women’s representation in engineering, math, and natural science but negatively associated with humanities, social sciences, and health fields among advanced industrial countries.

However, the validity of “indulging gendered selves” theory cannot be fully evaluated on the basis of cross-sectional data. More research could be done to directly examine the trend of sex segregation by fields of study across countries given the increasing data availability. From a historical perspective, will continued educational expansion increase the magnitude of sex segregation and how does it vary by national contexts? Did the advance industrial countries always have higher levels of sex segregation in higher education than developing countries (persistent sex segregation/gender inequality)? Will the high level of sex segregation in high education reduce in recent years in advanced industrial countries given the policy efforts in promoting women’s STEM participation (modernization/institutionalization theory)? Will the low level of sex segregation in developing countries converge to the level of affluent countries with economic development (indulging gendered self)? Is the “gender equality paradox” an ecological resulting from cross-national comparison at a given point in time?

This study aims to expand and update the previous research on cross-national differences in sex segregation in higher education using data from both historical archives and recent governmental reports. From 1965 to 1999, UNESCO has collected and published gender-specific statistics on graduates by field of education across the world in their statistical yearbooks. Since 2000, the Global Education Digest Report (GEDR) has continued to collect these data for all countries with available data until the year of 2012. While the GEDR stop publishing statistics on graduates by field after 2012, the OECD database collected those data for OECD member countries up to 2019. The OECD database are accessible online and ready to use. I have finished converting GED reports from 2000 to 2012 into dataset.

In the first step, I will use the index of association (A) to measure the sex segregation by fields of study for each country. The advantage of A over index of dissimilarity (D) is that A accounts for the marginal effects. The size of A represents the magnitude of sex segregation by field of study, with number of 1 indicates zero sex segregation. The larger the number, the stronger the segregation. The formula of the Index of Association (A) can be written as:
𝐴 = exp{(1/J)×∑{In(𝑊𝑗/𝑀𝑗)−[(1/𝐽)×∑In(𝑊𝑗/𝑀𝑗)]}2 }1/2
where Wj and Mj denote the number of women and men graduates in field of J. The intensity of sex segregation within each field can be represented by the Aj:
𝐴j = In(𝑊𝑗/𝑀𝑗)−[(1/𝐽)×∑In(𝑊𝑗/𝑀𝑗)]
Different from the aggregate level index A, Aj can have negative values. Within each field, the negative number indicates women are less representative in each field. The positive number indicates women are overrepresented in this specific field. The size of number reflects intensity of segregation. The number of zero indicates gender equity in distribution.

In the second step, I will apply a multivariate log-linear model to examine macrolevel and structural correlates of sex segregation by field of study. Following the previous literature, I will include gender equality index, economic development index, the structure of economic sector, size of tertiary education, female labor force participation, and other indicators into the analytical models. The goal is to investigate which variables explain the most variation in cross-country and cross-time sex segregation by field of study.

This study has important theoretical implications. While modernization/institutionalization theory predicts a declining trend in sex segregation with economic development, the indulging gendered self theory predict an increasing trend. In addition, empirical evidence from the US and western Europe indicates a pattern of persistent sex segregation. With the historical data at hand, these theories can be tested and evaluated. This study also has the potential to assist policymakers in developing gender equality policies for women in higher education.

Author

• Kai Feng