Educational Assortative Mating in Hong Kong: 1981–2011

Abstract

This article documents trends in educational assortative mating using samples of the 1981 to 2011 Hong Kong population censuses/by-censuses, with a particular focus on specific education groups and the Hong Kong-mainland China cross-border marriages. Results show an overall declining trend in educational assortative mating, which is driven mostly by the great increase of intermarriage between those with associate qualifications and those with lower secondary or less levels of education. Growing number of cross-border marriages is also accompanied by the substantive decline in educational homogamy for women from mainland China in recent years. It implies that immigrants from mainland China are more likely to exchange education for the Hong Kong permanent residency by marrying spouses with less education in Hong Kong in later periods. These findings suggest that a general declining trend in educational assortative mating does not necessarily indicate an increased openness of a society.

Acknowledgements

The author is deeply grateful to Colin Mills from the University of Oxford, Professor Xiaogang Wu from the Hong Kong University of Science & Technology (HKUST), and the anonymous reviewers for their comments and suggestions on this article.

Notes

The definition of cross-border marriages is based on the residence status of wives and husbands at the time of marriage rather than their birthplaces. It is common to see that both spouses were born in mainland China, but one (usually the groom) is usually a Hong Kong permanent resident, while the other is a mainland-Chinese resident at the time of marriage.

The number of cross-border marriages is based on the number of registered cross-border marriages in Hong Kong plus half of the number of issued Certificate of Absence of Marriage Record (CAMR) for each year.

Because the census/by-census data contain no information on marriage timing, I only focus on prevailing marriages in population from 1981 to 2011. It should be noted that there is a great overlap of the population across the census data, and the trends of the prevailing marriages represent all married couples at a given time. As there are only about 2 percent to 4 percent of the people aged 60 or above who have never been married during this period, the problem that only married population are selected is unlikely to be an issue. In 2011, among the population aged 21 to 60, the proportion of those divorced or separated was around 5 percent, while the proportion of those who remarried only accounted for around 3 percent (Yip et al. 2014). Therefore, the bias brought by selective marital dissolution and remarriage is limited.

Data used are the 1 percent random samples of the 1981 and 1986 population censuses or by-censuses, and the 5 percent random samples of the 1991, 1996, 2001, 2006, and 2011 population censuses or by-censuses. Due to the much smaller sample sizes in 1981 and 1986, the quasi-confidence intervals of the estimates scale factors for these two years are much wider than that of the rest.

Quasi-symmetry design matrix (main diagonal blocked):

• Table

Wife’s hypergamy design matrix:.

Table

Parameters for the main diagonal cells are also set to be time-varying in Models 3 and 4.

Couples with wives born in places other than Hong Kong and mainland China are too heterogeneous and occupy less than 7 percent of the total population. Those couples are not included.

Additional information

Notes on contributors

Muzhi Zhou

Muzhi Zhou is a PhD student in the Department of Sociology at the University of Oxford, United Kingdom. She is also affiliated with St. Antony’s College. Her areas of interest included family and marriage, social demography, and quantitative methodology.