Sex Ratios and Female Educational Attainment in China


A review of The Relationship Between Cohort Sex Ratios and Relative Female Educational Attainment in Post-1949 China, by Christina Jenq

Both the rapid increasing female educational attainment and the abnormally imbalanced sex ratio (male/female) in post-1949 China have been extensively discussed by many scholars. The majority of the relevant literature tries to sort out how the improvement in female education affects sex ratio; however, Christina Jenq’s dissertation takes a different perspective and looks into this question in the other way around. Not only does Jenq highlight the negative correlation between the sex ratio and female educational attainment in the second half of the 20th century China, more importantly, this dissertation attempts to unveil the true mechanisms that explain how an imbalanced sex ratio leads to lower rates of female years of schooling.

This study mainly relates to two large lines of literature. One is more contextual—the dramatic demographic changes in China after 1949, with a particular focus on the imbalanced sex ratio; and the other explains the worldwide increasing trend in female educational outcome. Firstly, the combination of the one-child policy and strong cultural roots for son bias in China caused a great crisis in the national sex ratio, reaching as high as 120 males per 100 females in more recent cohorts. Secondly, the closing gender gap in educational attainment around the globe during the 20th century has been a hotly discussed topic in many social science fields. The existing literature focuses on unveiling socioeconomic and cultural factors to explain such change. For instance, the lifting of entry barriers for women in many labor markets has dramatically increased the education return for women, leading to a huge boost in educational demands of women. On the other hand, the prevalence of public schools open to females extraordinarily increased the education supply for women. From a cultural perspective, after World War II, the general movement for gender equality increased the bargaining power of females in resource allocation within a family, therefore increasing the possibility for women to access education. Similarly, the weakening concept of traditional marriage and higher divorce rates led to a smaller opportunity costs for women seeking education. Thus, there are many varieties of factors that have contributed to the rapid increase of female educational attainment. Departing from the traditional academic sociological and historical literature on this topic, this dissertation examines whether, apart from these socio-economic explanations, a biological variable— imbalanced sex ratio— influences increasing rates of female educational attainment.

The Part 1 of this dissertation lays out foundations for further analysis, and outlines five possible channels that could generate the negative correlation between sex ratio and female educational attainment, which will be the theoretical framework to be tested in the later empirical analysis. The first possible mechanism the author argues is that the expansion of basic infrastructure like healthcare and education supply has had a positive impact on females’ life expectancy. This trend lowers the sex ratio, but drives up females’ educational incentives. Secondly, a framework of ‘brain vs. brawn’ suggests that the comparative advantages of males in physical tasks would motivate females to invest more in education. According to this theory, one would observe increasing female educational attainment in an environment of improving nutritional access. Thirdly, family characteristics, such as the order of birth and gender of her sibling, are also proved to be highly relevant to female educational outcomes. Additionally, the adoption of ultrasound allows a negative selection of females to be born into higher social-economic families, therefore better female educational outcomes. Lastly, an economic theory of ‘premarital investment’ emphasizes that education decisions are also related to the marriage market. Accordingly, the prediction of which could go either way depends on the relationship between sex ratio and female education.

The second section of Part 1 explains in detail the methodological approach this study adopts. The empirical strategy of this study includes carefully chosen proxy variables in a linear specification of female educational attainment and sex ratio which reflects the different possible mechanisms. The author claims that, by adding in the full set of control variables that account for all components of the error term, plus province and birth year fixed effects, this linear specification can be treated as a difference-in-difference identification strategy, which establishes not only raw correlation but causal relationship between sex ratio and female educational attainment (pp.17-18). To explore heterogeneity in the subpopulation, this study separately examines the urban and rural population, as well as the cohort born between 1949 and 1975 and those born after 1975 when the one-child policy was first informally implemented.

Following the analytical framework set up in Part 1, Part 2 of this dissertation provides a description of data, including its coverage and representativeness. Instead of simply introducing the dataset employed by this study, this section has a larger ambition in that it validates the calculation of sex ratios across different waves of China Population Censuses (1982, 1990, 2000, and 2005), and further quantifies the nature of their biases. Comparing  the 2010 China Family Panel Survey and the China Population Censuses, two selection problems of Chinese census data are identified: both males and lower-educated groups are under-sampled in census data. This finding has broad implications for researchers using Chinese Population Census data. Additionally, because of the deep understanding of the different bias nature of each dataset, the author’s choices of data to calculate different indicators make this study more convincing.

Part 3 is the backbone of this dissertation, which summarizes its empirical results and provides some implications of this study. There are rich empirical results shown in this section, among which the main findings are consistent with the first two hypothesized mechanisms that the author proposed in Part One. Firstly, looking at the socialist period from, 1949 to 1975, this study finds that the main mechanism to explain the negative correlation between sex ratio and female educational outcome is through the increasing supply of public health. However, the influence channel changes for the vpost-1975 period when not only the one child policy was implemented, but also when comprehensive economic reforms took place in China. Through the reform period, the results of this dissertation suggest that the negative correlation between sex ratio and female education is largely driven by economic prosperity, particularly for the urban sub-sample. Secondly, as suggested by the “brains vs. brawn” theory, higher birth weight, as a proxy of access to nutrition, is found to be associated with higher relative female years of schooling; this is especially true for rural samples. Unfortunately, the lack of availability of data prevents the author from extending her empirical tests on the other three possible mechanisms listed in the previous section.

This dissertation addresses two individual questions. One would be of interest to economists—what explains the correlation between sex ratio and female educational outcome; and the other will attract attention from demographers—the selection biases of Chinese population census.

Pei Gao
Department of Economic History
London School of Economics and Political Science

Primary Sources
China Family Panel Survey 2010
Chinese Population Census: 1982, 1990, and 2000
Chinese Statistic Yearbooks

Dissertation Information
The University of Chicago. 2014. 94 pp. Primary Advisors: Hoyt Bleakley and Emily Oster.

Image:  Picture of a poster from the Cultural Revolution, taken at the Shanghai Propaganda Poster Art Centre.

Leave a Reply