Women earn bachelor’s degrees at higher rates than men in most economically developed countries. Within these countries however, women continue to lag in postsecondary majors and careers in science, technology, engineering, and mathematics (STEM) fields.  This has considerable implications for women’s earnings.  While the relationship between economic development and urbanization and gender equity in educational achievement has been assumed to be positive, surprisingly and intriguingly increasing evidence supports a puzzling finding:  developing countries tend to have narrower gender gaps in mathematical and scientific achievement than those with higher levels of economic development and gender equity, including in the US and Europe. 

Despite the surprising finding, this phenomenon is not well-studied beyond a few select measures on cross-national educational assessments.  No studies prior to this one, to our knowledge, have analyzed variation in postsecondary STEM gender gaps within developing countries. Furthermore, given the rapid increase in higher education participation across the globe, there may be critical albeit insufficiently understood socioeconomic factors associated with women’s differential participation in STEM fields.  Given the rise of university participation in less developed countries and the global movement of women into the labor force, it is important to examine postsecondary sex segregation in distinct social contexts.

This study aims to address these deficits by examining sex segregation in STEM within a rapidly developing, non-Western country:  Cambodia.  While women’s rising postsecondary achievements have been studied extensively, much remains unknown about the mechanisms influencing their share of global increases in STEM higher education. It seems important then to examine gender inequality from a non-Western, lower-income perspective. Using merged data from the Cambodian General Census and the Cambodian Ministry of Education, Youth, and Sport, as well as primary qualitative data collected at a number of universities in Cambodia, we investigate the following questions:

1. How does women’s share of postsecondary enrollment in STEM fields vary between wealthy, urbanized areas and poorer rural areas in a non-Western developing country, Cambodia?
2. How does intra-country variation in socioeconomic development, urbanization, and socio-economic measures of gender equity affect women’s share of postsecondary enrollment in specific STEM and STEM-related fields within and across rural and urbanized provinces in Cambodia?
3. What explains STEM sex segregation within a particular developing country, and what might that tell us about this phenomenon elsewhere?

To address these questions, we procured and merged data from Cambodia’s Ministry of Education, Youth, and Sport (MoEYS) and the General Population Census of Cambodia 2008.  These provide data on Cambodian university enrollment by gender and by major for the 2011–2012 academic year.  We divide the list of Cambodian majors reported in the data into two mutually exclusive categories, STEM and non-STEM, using the U.S. National Center for Education Statistics definitions of STEM, while recognizing that some of these majors may have particular relevance to the burgeoning Cambodian economy as well as other rapidly developing economies.  STEM includes majors in the following clusters: agricultural and environmental sciences, biological sciences, engineering, health (including medical/dental), natural sciences, accounting and information technology. Non-STEM includes majors in the following clusters: arts and humanities, business, civil service/professions, and social sciences. To enhance the inclusion of recently opened universities and STEM programs, we focused our analyses on students enrolled in their first two years.  The final analytic dataset describes the enrollment patterns of 98,751 first and second year Cambodian university students majoring in 683 fields at 83 campuses located in 18 provinces. 

To examine how gender STEM enrollments varied by degree of urbanization and socioeconomic development levels across Cambodia, we used data from the 2008 Cambodian Census aggregated to the province-level.  An urban density scale was developed based on variation in (a) population density (calculated as persons/km²) and (b) urbanicity (percent urban population).   Further, a high resources scale was created indicating variation in mean levels of resource development and infrastructure access among provincial households. Average household size was also considered, as it has been shown to be an important factor influencing a family’s relative investment in the education of a particular child. 

Finally, three variables serve as indicators of variation in gender equity in each province. First, a female educational access indicator, composed of:  (a) the percentage of women aged 25 years or older with postsecondary educational attainment, (b) female literacy rate (as percentage of population ages 15 and above), and (c) the average age at marriage for women. In addition, we control for the percentage of females employed in each province, measured by the proportion of economically active females, a metric commonly used to capture women’s employment rates in developing countries. We also account for the provincial percentage of women enrolled in postsecondary education (PSE) (irrespective of degree field).

Altogether, findings present a complex but consistent pattern. First, women are better represented in postsecondary education overall, and in postsecondary traditional STEM and IT majors in particular, at universities in less urbanized areas. Further, higher levels of socioeconomic development as measured by household access to key infrastructure varies inversely with women’s share of STEM and STEM-related majors, with the exception of health where the relationship is positive. Finally, results identify an intriguing and primarily negative relationship between traditional gender equity measures and women’s share of specific STEM and STEM-related majors.  These surprising findings defy conventional expectations that greater urbanization, resource access and gender equity within a developing country should be uniformly associated with higher educational outcomes for women. 

However, some results point to a more complex picture.  Urban density has mixed effects on women’s share of majors, including a positive effect on women’s share of STEM and IT majors, but a negative relationship with accounting. Access to household and infrastructure resources positively predicts women’s share of health majors, but negatively predicts women’s share of IT majors.  With respect to gender equity, measures of female educational access and higher provincial rates of well-educated and employed women are inversely associated with women’s participation in health and information technology, critical fields for Cambodia’s development goals. Overall, these socioeconomic provincial differences suggest the importance of more in-depth case studies on the relationship between socioeconomic development and sex segregation in higher education in Cambodia, as well as in other developing countries.

Should the search for female scientists shift away from its focus on the most socioeconomically developed provinces and countries? One possibility is that globalization – occurring first in the major cities – might drive women away from STEM, implanting gendered ideologies where they did not previously exist. On the other hand, Cambodia and other less developed countries have rich and distinct cultural histories, including how gender functions as a social category, and thus it may be a premature assumption to conclude differences are driven by encroaching global forces, especially given this is the first study of gendered postsecondary education pathways in this non-Western and industrializing country context. Future empirical work may shed further light on global variation in the gendering of mathematics and science, including how our narrative compares to other, distinct socio-cultural contexts.

The feature image is from the U.S. Bureau of Labor Statistics.