China’s marriage market

I decided to start out with this:

…in order to illustrate that you could probably write a not too dissimilar post about other countries as well. Also, it’s a nice image. Image credit: Wikipedia. “Description: Sex ratio total population. Pink = Female higher than male, Green = Equal, Blue = Male higher than female.”

This post will only deal with China. Here’s some related stuff about India.

So anyway, I was skimming a few world bank working papers and I found this one (pdf), which I decided to cover in a bit of detail here. It’s called China’s Marriage Market and Upcoming Challenges for Elderly Men and it’s written by Monica Das Gupta, Avraham Ebenstein & Ethan Jennings Sharygin. Some stuff from the paper:

“The Chinese census in 2005 reflected a staggering sex ratio at birth of 119, implying that each year there are roughly 1 million more boys born than girls.3 For cohorts born between 1985 and 2005, we estimate that there are 27 million more men than women4, implying a large number of men will fail to marry. […]

We demonstrate two key facts regarding the Chinese marriage market using historical census microdata from 1990 and 2000. First, economic status is a crucial predictor of marital probability for men in China. We use years of education as the closest proxy for status, and document that while there is almost universal marriage for highly educated men, lower rates of marriage prevail among men of lower education. By contrast, the marriage market for women cleared: women across the educational distribution enjoy nearly universal marriage, and are able to engage in hypergamy, choosing spouses of higher status and income. Second, since many women migrate for the purpose of marriage, it seems very likely that in the coming decades the collapse of marital prospects for men will occur in poor areas of the country with low educational attainment. […]

The results paint a grim picture for China’s ability to care for these men under the current policy structure of social assistance and social insurance programs that are primarily locally funded (Wang 2006, World Bank 2009). We estimate that in the absence of major redistribution of education and employment opportunities across China, the marriage squeeze will be in China’s poorer regions with large minority populations.7 Thus it will not necessarily be the more prosperous eastern regions of China with the most skewed sex ratio at birth that will experience high marriage failure rates among men. Rather, the poorer provinces ─ with more balanced sex ratios at birth ─ will bear a disproportionate share of the social and economic burden of China’s unmarried and childless men.”

How big is the difference in marriage rates between the successful males and the not quite so successful males, I hear you ask? Well, the paper states that: “over 98% of college graduates successfully marry by age 35 whereas the proportion is under 90% for men with less than a primary education.” One way to look at those numbers is that ‘that’s actually not that big of a difference’ – it’s around 9 out of 10 or more in both cases, right? But who are we actually comparing again? – another way to look at that is that males with less than a primary education are more than 5 times as likely to not succesfully marry by age 35. To me, that sounds like a huge difference, and it’s expected to get even worse over time: “over 10 percent of men with less than primary school education aged 30+ in 2030 are projected never to marry, and this figure increases to almost half in 2050”. Of course one might argue that economic growth increases mobility (so that even poor men might be able to move to find females willing to marry them) and ‘historical data are historical data’ which perhaps shouldn’t be given as much weight, given how much Chinese society has changed over the past decades. But rural China is still very poor and it isn’t growing very much compared to the rest – many of the people who have not left already for the urban provinces are people who can’t afford to, and they can’t really afford to save either so there’s not in my mind any compelling reason to think they will be able to afford to move in the future. Incidentally, it’s not really that hard to set up a model where you have decreased mobility over time even though the poor group has a positive net savings rate. Property prices are functions of local economic conditions, and if an area experiences significant income growth whereas another area does not and the people living in the poorer area are neither able to save enough money over time to at least keep up with the income growth of the richer area nor can afford to move there in the short run, the relative property price differential and the costs of moving will go up over time, even though the poor single guy might have a significant positive net savings rate. A very simplified model illustrating this could go along these lines:

Average income of ‘poor area’ residents: 10.
Average income of ‘rich area’ residents: 100.
Poor area income growth rate: 0%.
Rich area income growth rate: 10%

I shall assume that income growth rates and housing price growth rates are identical. In reality, housing prices are probably growing faster than income for the relevant demographic in the rich area and slower than income in the poor area. Let’s say the poor guy saves 20% of his income/year, i.e. 2 mu (‘monetary units’)/period. Say he invests that money in the rich area, earning 10%/year. After 10 years, he’ll have saved ~35 mu. How much will a house in the rich area that used to cost 100 mu cost after 10 years? 259. At the beginning, the poor guy was 98 mu short of being able to buy a house in the rich area – after ten years he’s now more than 200 mu short, even though he had a very high savings rate given his income and even though he earned a quite nice return on investment during that period. The property price differential was 90 mu to begin with, it’s 249 mu after 10 years. Maybe the effect sizes won’t be as large as assumed in the paper, but some of the dynamics described in the paper will probably play out to some degree.

Some more numbers and stuff related to these remarks from the paper:

“Poverty in China is heavily concentrated in the rural areas. Different measures of poverty all paint the same picture: while nearly 30 percent of the rural population was poor in 2005, this applied to only 5 percent or less of the urban population […] The vast majority of the poor in 2003 lived in rural areas, and poverty is most heavily concentrated in the northwestern and southwestern regions […] Both rural and urban incomes have continued to grow, but the rural-urban gap has continued to widen […]

Significant proportions of urban workers are covered by formal social insurance programs: in 2007, around half of workers had pension coverage, 45 percent had Basic Medical Insurance, and 40 percent had unemployment insurance and work injury insurance […] The rural pension system (funded mainly by personal contributions and collective subsidies) covered only about 10-11% of the rural labor force (World Bank 2009: Table 6.65), and coverage of the farm-based elderly population appeared to be particularly limited. Beneficiaries were highly concentrated in a few (mostly wealthy) provinces. […]

Since men who are not as educated, healthy, and able to earn well tend to fail to attract a bride, they are likely to be heavily represented among those who are unable to save adequately for their old age, or labor heavily into their old age. They are the most vulnerable to income and illness shocks, since they cannot smooth fluctuations in household income by pooling earnings from spouses or children. Unmarried individuals are also more likely to be living without family to serve as caregivers (Table 5). For example, in the 2000 census, 65% of those aged 65-80 who had ever-married were co-residing with younger kin, compared with only 20% of those never-married. Moreover, levels of co-residence have dropped sharply in recent decades (Table 5), and this trend can be expected to continue. The men who fail to marry are among the least likely to be able to save for their old age, to work in their old age, and to have access to old age support from family members.”

Last, a few tables (click to view full size):

Wu Bao, Di Bao and Tekun Hu are various social assistance programs: “The Te Kun program provides cash assistance to very poor and incapacitated residents of less-developed areas, at the discretion of the local officials. The Wu Bao program, dating from the 1950s, sought to ensure that no section of the population remained destitute.11 In 2006, the State Council issued regulations that shift financing responsibility for wubao from village reserves to local fiscal budgets (World Bank 2008:79-80). The Di Bao program, also known as the Minimum Living Standard Scheme, provides subsidies and in-kind transfers to those living below a certain poverty line.”

More than 45 % of the total income of Chinese urban residents above the age of 60 comes from pensions; the number for rural residents in the same age group is about one-tenth of that, 4.6 %. Also take note of the family support numbers.


November 26, 2011 - Posted by | Data, Demographics, Economics, Geography, marriage

No comments yet.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: