Marigolds at an Indian wedding

The skewed sex ratio at birth in India, the result of sex-selective abortion, has led demographers to anticipate that there will be an excess of men in the future who will be unable to marry. Marriage markets, however, are not only structured by demographic factors such as age and sex, but also by factors such as education. An analysis of future marriage, then, should also consider another significant change that is also taking place in these populations: the closing of the gender gap in education, and the expansion of female education at the secondary and university-levels. How might future marriage patterns be affected by these changes in the age, sex and educational structure of India’s population?

In a recent paper in Demography, along with my co-authors Albert Esteve and Joan Garcia-Roman, I approached this question by developing some scenarios for future marriage patterns in India. We did this in two steps: first we characterized existing marriage patterns and pairing norms using micro-level data from two national-level surveys in India; the Indian National Family Health Survey and the India Socio-Economic Survey available on IPUMS-International. From a demographic perspective, contemporary marriage norms in India can be characterized in three ways. Firstly, marriage is nearly universal, with very few people remaining never married by age 50. In India we found that by age 50, 0.6 percent of women were never married and 1.2 percent of men were never married. Secondly, marriage tends to take place generally early in life, especially for women. Thirdly, it is often asymmetric with men often marrying women who are younger and less educated than they are. In this way, India is not uncommon; in many parts of the world, similarly educated men and women marry, but when spouses have different educational levels, men tend to marry women who are less educated than they are.

We then carried out the following counterfactual exercise: we applied contemporary marriage norms to population projections until 2050. These projections were first disaggregated by sex and age, and second, disaggregated by sex, age and educational attainment. In the first case in (the sex and age scenario), our model suggested that proportions never married increased more significantly among men. But, when we disaggregated the population by sex, age and education, our model showed a steeper rise in proportions never married among women from 0.7% in 2010 to nearly 9% in 2050, with the most significant decline in marriage rates experienced by university-educated women.  For men, the percentage never married at the age of 50 also rose from 1.4% in 2010 to 5% by the age of 50, with the least educated likely to have the most trouble finding a wife.

Norms are of course not static, and our model is at best a model that usefully highlights the ‘mismatch’ between prevailing marriage norms and the way the Indian population is changing. To see how greater flexibility in pairing norms might offset a significant decline in marriage prevalence, we also developed another scenario in which we allowed for greater educational and age asymmetry in marriage patterns over time — for example, by allowing more equal partnerships with women marrying men who are equally educated as themselves, as well as pairings where women are more educated than their spouses. This helped alleviate the mismatch, with proportions never marrying not rising as steeply as when norms remain inflexible as in the scenario before.

How and for whom will marriage change in India in the future? A lot will depend on if, and how, the tensions between marriage asymmetry and prevalence that lie ahead are negotiated. If education provides women with greater autonomy, but traditional roles and expectations for women and men in marriage persist, might many more women choose to forego marriage in India like they have in South Korea and Japan? Or will Indian couples and their families be willing to change their conceptions of what makes a suitable match? These are not questions that a model can answer – but hopefully discussions that it can stir.

—

The study described here is: Kashyap, Ridhi, Albert Esteve, and Joan García-Román. “Potential (Mis) match? Marriage Markets Amidst Sociodemographic Change in India, 2005–2050.” Demography (2015): 52(1) 183 – 208.