Heterogeneity in Contemporary Chinese Marriage

Abstract

Is universal and early marriage still true in China after decades of dramatic socioeconomic changes? Based on the 2005 Population Survey data, we find that by age 35 to 39 years, almost all women are married, and less than 5% of men remain single with the singulate mean age at marriage (SMAM) in the country at 25.7 and 23.5 for men and women, respectively. There are notable regional variations in marriage prevalence and timing across China, likely due to economic development, migration, and cultural norms particularly for ethnic minorities. Those who live in the East and in urban areas tend to enter marriage later. Universal and early marriage is particularly true for women with no education and least so for men with no education. College education delays marriage for both men and women, but most of them eventually marry. We discuss the implications of findings for highly educated women and poorly educated men.

Keywords

marriage China education leftover women surplus men

Introduction

Family scholars have been perplexed by contemporary China’s universal and early marriage phenomenon not found in Western industrialized countries or its East Asian and Southeast Asian neighbors (Jones & Gubhaju, 2009). At the same time, there are increasing concerns over the surplus men resulting from the imbalanced sex ratio that arose largely due to the strict one-child family policy in China since the late 1970s (Huang, 2014; Poston & Glover, 2005; Trent & South, 2011; Tuljapurkar, Li, & Feldman, 1995). Since the first decade of this century, a heated debate has also arisen among policy makers, the public, and scholars over the “leftover” women, who are college educated and remain single in their late 20s and 30s (Cai & Tian, 2013; Gaetano, 2010; Qian, 2012; To, 2013). This article addresses the following questions: To what extent is universal and early marriage still true in contemporary China? What are the heterogeneities within the national pattern of marriage? How are education and gender interwoven in shaping marriage formation patterns?

This study has two motivations. First, although recent research has focused on determinants of marriage formation in China (Tian, 2013; Yeung & Hu, 2013; Yu & Xie, 2013), little attention has been paid to the geographic heterogeneity found across the country. We examine the overall marriage formation pattern in China and further differentiate it by region and discuss the relevance of variations in economic, social, and cultural contexts across China in shaping marriage formation behavior. Second, the derogatory labeling of “leftover women” tends to stigmatize the small proportion of college educated women while ignoring the looming marriageability issue facing a large number of excess young men in the marriage market in the coming decades. We investigate the marriage formation patterns for both social groups and examine the interplay between gender and education.

Using the 2005 One-percent Population Survey data, we first calculate the singulate mean age at marriage (SMAM) for the entire country and for the 25 provinces for which data are available (for convenience of wording, we refer to metropolitan Beijing, Shanghai, Tianjin, and Chongqing and ethnic autonomous regions of Xinjiang, Xizang, Ningxia, Inner Mongolia, and Guangxi as provinces even though they are administrative districts at the provincial level). Second, we select provinces from each region to illustrate the heterogeneity in marriage prevalence across the country and how gender and education differentiate marriage prevalence within each province. Finally, we estimate marriage timing both at the national and provincial levels in China to examine factors that shape family formation patterns in China.

Background

Heterogeneity Within China

Much research on social and family changes in China has concentrated on the impact of socioeconomic development in China before and after the reform (Tian, 2013; Yeung & Ji, 2013; Yu & Xie, 2013). These studies neglect the importance of regional heterogeneity in marriage formation within China, for example, variations in contextual factors, such as cultural norms and ethnic difference. Thornton and Fricke (1987) and Thornton (2005) argue that there is no solo developmental trajectory of family changes, and it is critical to study family behaviors in the local socioeconomic and cultural context. Recent research in Asian countries also emphasizes the role of historical culture and local institutional contexts in shaping marriage formation behavior (Ji, 2013; Park, 2013; Yeung & Alipio, 2013).

In term of socioeconomic development and local culture, China is conventionally categorized into three regions: the East, Central, and West. The highly urbanized East has a dense population with a relatively higher education level and a highly developed economy. For example, most of China’s mega cities with a population of more than 10 million and highly developed economy such as Shanghai, Beijing, Tianjin, and Guangzhou are in this region. This region is also the destination of massive migration from the West and Central agricultural provinces. Historically, this region is well integrated into the global economy and is regarded as culturally more open and progressive (Yuan, 2012).

In the Central and West, education level is low, and agriculture plays an important role. The economy is less integrated into the global system, and many local residents migrate to the East, with Sichuan and Henan as two big migrant sending provinces. Compared to the East, the Central and the West are more isolated inland, and tend to stick to tradition (Yuan, 2012).

The West has a more concentrated ethnic minority population, with many residing in the remote inland. Tibet, Guangxi, Ningxia and Xinjiang are four of the five ethnic autonomous provinces with less developed economies, and ethnic groups have unique cultural traditions stemming from their own religious-cultural history, distinct from the mainstream Chinese culture (Yuan, 2012; Zhao, 2007). These ethnic groups are less influenced by the Confucian pro-marriage and pro-fertility tradition. For example, Zhao (2007) points out that among the Hui, the third largest ethnic group in China, and mostly residing in Northwest China, Islam plays a crucial role in the society, although the Hui also adhere closely to Chinese culture. They are different from Muslims in other parts of the world, with the exception of marriage customs where they strictly follow the Muslim tradition (Zhao, 2007). Research in Malaysia and Indonesia finds that Muslims tend to marry early and have higher marriage rates compared to Chinese and other non-Muslims (Jones, 1980, 2001; Tey, 2011). Therefore, Hui’s ethnic and religious tradition of early and near universal marriage may play an important role in their marriage formation.

Likewise, other ethnic groups in Southwest as well as South China also have their unique marriage customs, such as the “delayed transfer marriage” practiced among various non-Han ethnic groups in Qing Dynasty in south China such as the Zhuang in Guangxi, the Miao in Guizhou, and the Li in Hainan (Siu, 1990; Xu, 1983). In this practice, wives usually stay long-term in their natal homes after marriage before eventually moving into the husbands’ homes, contrary to Han Confucian tradition (Siu, 1990; Xu, 1983). With increasing educational attainment and migration and the changing gender roles in China in the past six decades, marriage patterns in regions with high concentration of ethnic minorities may have undergone changes as a result of a complex interplay between traditional cultural norms and socioeconomic development.

Gender and Education in Relation to Marriage Formation

Education and gender have been shown to be two key factors in the literature of marriage formation. In the Western family literature, early research generally relates men’s education to higher marriage rates and women’s education to later and fewer marriages (Becker, 1991). In the past few decades, research has examined the relationship between gender, education, and marriage formation in the social context of changing gender equality and women’s changing economic role (Goldstein & Kenney, 2001; Kalmijn, 2013; McDonald, 2000; Ono, 2003; Oppenheimer, 1988, 1994; Sweeney, 2002).

McDonald (2000) argues that the different institutional context of gender equality is the important background of puzzling family formation behaviors across different countries. Oppenheimer (1988, 1994) notes that after the 1960s in the United States, women’s labor force participation increased dramatically, men’s labor market circumstances deteriorated, especially after 1990, mass education spread rapidly, and gender ideology became more egalitarian. These changing institutional factors have made women’s earning capacity more attractive in the marriage market since the 1970s. A considerable body of research in the West reports the positive effects of women’s education on marriage formation (Goldscheider & Waite, 1986; Goldstein & Kenney, 2001; Oppenheimer, 1994; Sweeney, 2002). Ono (2003) demonstrates that in three countries from three continents on a spectrum of gender role differentiation from segregated to egalitarian—Japan, the United States, and Sweden—the effect of women’s economic role on marriage timing switches from negative to positive. Fukuda (2013) recently reports that in Japan, due to a more egalitarian gender relationship and the greater labor market uncertainties, the association between women’ earning capacity and marriage propensity for the 1960s and the 1970s birth cohort has shifted from negative to positive.

In other Asian societies including China, education is still reported to have a gendered pattern in marriage formation, delaying women’s but accelerating men’s marriage formation (Jones & Gubhaju, 2009; Raymo & Iwasawa, 2005; Tian, 2013). However, it has also been reported that education can delay both men’s and women’s entry into marriage (Yeung & Ji, 2013; Yu & Xie, 2013). At the same time, research on China reports an increasing educational homogamy, and the pattern tends to be pronounced among college graduates (Han, 2010; Qian, 2012). However, studies have found that marriage chances can decline significantly for college educated women who are older, particularly those in their 30s (Cai & Tian, 2013; Cai & Wang, 2011; Yu & Xie, 2013).

Another important body of research examining gender differentials in marriage formation pertains to age and status hypergamy (Glick & Lin, 1986; Presser, 1975). Furthermore, due to the “gender double standard of aging”—as men and women age, women’s marriage market shrinks much faster than men’s—unmarried older men will marry even younger women (England & McClintock, 2009). In view of this, researchers believe that both age and status hypergamy are underlying challenges in the marriage market faced by older college educated women, the so-called “leftover” women in China (Cai & Tian, 2013).

Gender and Education Dynamics in China

Since the 1950s, the Chinese government has promoted gender equality in various social dimensions: choice of spouse, property ownership, right to participate fully in education, and the labor market (Bauer, Feng, Riley, & Zhao, 1992; Maurer-Fazio, Connelly, Chen, & Tang, 2011; Whyte, 1984). After the launch of economic reforms in 1978, and with the deepening of marketization, the Marxist ideology that has nurtured gender equality norms lost ground to the traditional Chinese culture, which endorses gender hierarchy (Cohen & Wang, 2008; Zuo & Bian, 2001).

With the rapid expansion of education over the last half a century in China, the gender gap in education has narrowed dramatically. Chinese women’s college enrollment surpassed that of males in recent years (Yeung, 2013). However, gains in education do not directly translate to gains in the labor market. Since the economic reform and marketization in the 1980s, research has shown that women are heavily concentrated in relatively low-paying sectors or in unpaid labor such as agriculture, and they face various forms of discrimination in the labor market (Cao & Hu, 2007; Cohen & Wang, 2008; Zhang, 2002; Zhang, Hannum, & Wang, 2008). Facing the deteriorating labor market and the “gender double standard of aging” (England & McClintock, 2009), China’s educated women may experience a double disadvantage in the areas of job and mate search. It is hence important to investigate whether they delay or forgo marriage and to what extent they do so. While focusing on educated women, we should not ignore the challenges that less or uneducated men may encounter in the marriage market. It is likely that less educated and uneducated men will be most disadvantaged due to both the heavily skewed sex ratio in the marriage market and the cultural norm of status hypergamy.

Universal, Early Marriage in China

In contrast to Western societies and other East and Southeast Asian countries where there has been a notable decline in marriage formation when women’s educational achievement increased remarkably, marriage remains near universal in China till today (Jones & Gubhaju, 2009; Yeung & Hu, 2013). Based on 2000 and 2005 data, Jones and Gubhaju (2009) show that about 2% of women remain single at ages 30 to 34 years in China, and education only contributes moderately to delayed marriage.

Distinct from many Western societies, given the weak social welfare system in contemporary China and strong family values, Chinese young adults feel the pressure to procreate to continue the family line and provide old age security. In addition, Chinese policy stipulates that couples have to obtain birth license for childbearing and for the child to gain a legal hukou status, on the condition that they are legally married. Marriage thus is the only legitimate social institution for begetting a child in China. Many young Chinese men and women feel tremendous social and family pressure to get married early in order to have a child soon afterwards. In the context of early, near universal marriages in China, we expect college education to delay both men’s and women’s marriage timing, but they will eventually get married, rather than forgo marriage. Another unique context in China is the skewed sex ratio as a result of the one-child policy launched in 1979. According to reports by the National Bureau of Statistics of China (2011), the sex ratio at birth was 116.9 in 2000 and 118.6 in 2005. A substantial number of men may be unable to find mates even though they may desire to get married. We expect that less educated or uneducated men are at a disadvantaged position in the marriage market, considering the fact that educated men are usually regarded as attractive mates. It is important to investigate the interplay between gender and education in relation to marriage formation at the local level.

Data and Method

Sample and Measure

This study employs data from the 2005 One-percent Population Survey of China to calculate the singulate mean age of marriage (SMAM). For our multivariate analysis, we draw data from a random 20% sample of the 2005 One-percent Population Survey. We include individuals aged 30 years and above at the time of interview for the analysis on timing of marriage. We remove 3,766 respondents who were married but had a missing value on the age at marriage. In addition, 561 individuals are removed due to missing values in education and gender. The final analysis sample is 1,483,104 individuals. For the Cox hazard model analysis, we further exclude 2,635 individuals who have missing values on the covariates in the model: hukou status, ethnicity, migration, and family economic resources. Aware of the limitations of the cross-sectional nature of the data and the restricted number of variables in the survey, we are thus cautious in making any causal interpretations.

Timing of marriage is the dependent variable, measured in months. To explain the variation in the timing of first marriage, we include variables at the individual, family, and community levels in the model, with gender and education being the focal interest of this study. Gender is coded as 1 if female and 0 if male. Education has four categories: no education, less than high school, high school, and some college or above education, with no education as the reference category. In the Cox proportional hazard models, we also control for age, hukou status, ethnicity, migration, family economic resources, region, and provincial characteristics. Hukou status is coded as 1 if urban and 0 if rural. Ethnicity is coded as 1 if Han and 0 if ethnic minority. Migration is coded as 1 if the respondent was away from where his/her hukou was registered at the time of the interview and 0 otherwise. The census data do not provide data for family socioeconomic background. We create a crude index of family financial resources taking the sum of three dichotomized indicators: whether there is tap water and toilet in the house and whether the house is made of concrete or mixed versus other materials. This index ranges from 0 to 3. Region is categorized as the East, Central, and West, with the East as the reference group. At the provincial level, we measure the percentage of population with professional occupations to attempt to capture the community effect of the socioeconomic development of the province.

Results

Marriage Prevalence: The Overall and Region-Specific Patterns

On the whole, marriage is relatively early and near universal in China, but there is notable regional heterogeneity. Table 1 shows SMAM by gender at the national and provincial levels. Unfortunately, data for six provinces are not available for the province-specific level calculation. However, these provinces are included in the calculation of SMAM for the entire nation. On average, SMAM is 23.5 and 25.7 years for women and men, respectively. The East provinces tend to have later SMAM than the Central and West provinces. The capital city of China, Beijing, in the East has the highest SMAM, 25.8 and 27.6 years for women and men, respectively. Metropolitan Shanghai and Guangdong province, both in the East, have slightly lower figures than Beijing, 24.9 and 26.5, and 25.6 and 27.5, for men and women, respectively. All the Central provinces for which we have data have similar SMAM, around 23 for women and 25 for men.

Table 1.

Singulate Mean Age at Marriage (SMAM) in China, for the Entire Country and by Province, in 2005.

	SMAM		% Single by age 30-34		% Single by age 35-39
	Female	Male	Female	Male	Female	Male
Entire nation	23.5	25.7	2.1	9.7	0.7	4.8
East
Beijing	25.8	27.6	6.4	13.9	2.5	5.9
Hebei	23.2	24	1.3	5.6	0.4	3.9
Liaoning	23.7	25.7	3.3	9.6	1.1	4.8
Shanghai	24.9	26.5	5.4	12.4	1.9	6.4
Jiangsu	23	24.6	1.1	5.1	0.4	3.3
Zhejiang	23.4	25.9	1.5	8.7	0.5	4.7
Fujian	23.8	26.8	2.0	9.7	0.7	4.4
Shandong	24.1	24.6	1.2	5.8	0.3	3.8
Guangdong	25.6	27.5	5.6	13.6	1.9	5.6
Hainan	24.4	27.4	5.8	19.5	1.7	9.1
Central
Heilongjiang	23	25.2	1.9	8.9	0.7	4.1
Anhui	22.7	24.7	0.9	7.1	0.3	4
Jiangxi	22.6	25.8	1.2	8.8	0.3	3.5
Henan	23.5	24.7	1.2	6.8	0.2	3.8
West
Inner Mongolia	23	25	1.4	7.9	0.5	4.8
Guangxi	24.3	27.5	4.2	18.2	1.1	9.3
Chongqing	22.9	25.7	1.6	11.6	0.4	5.7
Sichuan	22.4	25.1	1.7	11.3	0.6	5.9
Guizhou	22.3	25.7	1.6	11.4	0.4	5.2
Shannxi	23.1	25.7	1.1	9.5	0.3	5
Gansu	22.8	25.5	1.3	9.6	0.4	5.1
Qinghai	22	25.2	3.2	11.6	1.3	5.7
Ningxia	22	24.2	1.3	4.7	0.5	2.2
Xinjiang	22.4	25.6	3.1	11.0	1	3.6

Note. Data for Tianjin, Shanxi, Jilin, Hubei, Hunan, Yunnan, and Tibet are not available for the province-specific calculation. However, data for these provinces are included in the calculation of the SMAM for the entire nation.

Source. 2005 One-percent Population Survey.

In contrast, Ningxia, the Hui ethnic autonomous province in the West, which is not necessarily the economically least developed province, has the lowest SMAM, 22 for women and 24.2 for men. This is consistent with our early discussion of its unique Muslim culture in relation to marriage. We find a surprising pattern in Guangxi, the Zhuang ethnic autonomous province in the West, as it emerges as an outlier with SMAM 24.3 for women and 27.5 for men, higher than in many other economically more advanced provinces. A less pronounced case is Hainan, an Eastern province with relatively high proportion of Li ethnic group but economically much less advanced than other Eastern provinces, which also has relatively high SMAM, 24.4 for women and 27.4 for men. We do not fully understand what factors shaped these marriage patterns. In the case of Guangxi, it is likely that the low educational attainment of Zhuang men and the strong patriarchal norms make it challenging for Zhuang men to find a marriage partner. The traditional “delayed transfer marriage” may also be part of the story (Siu, 1990; Xu, 1983). However, ethnic subculture may not be the only explanation for these patterns. In Guangxi, the Zhuang ethnic group accounts for only about one third of the total population with the rest mostly being Han. Guangxi is also a province with a high out-migration flow, with Guigang, Nanning, Qinzhou, and Liuzhou among the highest sending cities in the country, mostly to Guangdong. Thus, the high age at first marriage observed in Guangxi is likely a result of a complex combination of socioeconomic development, notably increased education and migration, and ethnic cultural norms. Hainan has also experienced rapid socioeconomic transformation in recent years. We know of little research on marriage patterns in Hainan to provide useful insight. Further investigation is needed to understand the family formation behavior in these regions.

The last four columns of Table 1 show that the timing of first marriage varies across regions. Overall, by age 30 to 34 years, only about 2% of women and 10% of men remain single, and almost no women and about 5% of men remain so by age 35 to 39 years. Beijing, Shanghai, Guangdong, and Hainan in the East and Guangxi in the West lead the trend of late entry into marriage, with more than 5% of women and between 12% to 20% of men remaining single at age 30 to 34 years; and at age 35 to 39 years, the proportion remaining single is highest for Beijing women (2.5%) and Guangxi men (9.3%). The regional heterogeneity suggests that the pace to marriage formation is not linear with regard to economic development or urbanization. It is thus important to understand the regional customs and ethnic practices in the local context when studying marriage formation in China.

To further illustrate the variations in trajectory to marriage at different ages and the ultimate marriage rates across China, we parsimoniously select four provinces from the three regions: Beijing in the East, Henan in the Central, and Ningxia and Guangxi in the West. Beijing, the metropolitan area with the highest SMAM in China, has a population of about 20 million with high levels of education and socioeconomic development. Henan, an agricultural province with the SMAM close to the national average, is one of the most populous and largest migrant sending provinces. Guangxi, the Zhuang ethnic autonomous province, with the highest SMAM in the West, is an agricultural province in which the largest minority group in China, the Zhuang, resides. Not least advanced in terms of socioeconomic development and educational level, Ningxia, a Muslim Hui ethnic autonomous province located in the remote inland in Northwestern China, has the lowest SMAM in the country based on the 2005 data.

Figure 1 shows Kaplan–Meier survival curves on the proportion of respondents remaining single by age for the four provinces. Regional heterogeneity is noticeable, with Ningxia showing the fastest pace to marriage entry: by ages 25, 30, and 35 there are 81%, 96%, and 99% of respondents, respectively, already married, with a similar pattern in Henan. In contrast, for respondents from metropolitan Beijing, by age 25, 30, and 35 there are 56%, 90%, and 96%, respectively, already married. Beijing and Henan converge at age 35 years, and there is no noticeable difference beyond that point, although Ningxia experiences a higher marriage rate as seen in Figure 1.

Figure 1.

Proportion of respondents remaining single for selected provinces in 2005.

Yet, for largely rural and ethnic-concentrated Guangxi province, the survival curve crosses over that of Beijing in the late twenties; about 66% of individuals are married by age 25 years, 89% by age 30 years, and 94% by age 35 years. This is remarkable in the near universal and early marriage context in China. Considering its marginal status in terms of socioeconomic development and geographic location in China, it is likely that many in Guangxi may face challenges if they have not yet married by their late twenties. However, we cannot rule out the likelihood that these people may have chosen to delay their marriage due to the rising education and migration in this province noted earlier. The relatively late marriage entry and lower marriage prevalence in Guangxi province could be partially due to a complex interplay between economic development and cultural norms that we do not fully understand. The relatively early and high marriage rates in Ningxia compared to other provinces could be partially attributed to the Muslim religious tradition of the Hui ethnic group.

Marriage Prevalence: Interwoven Stories of Gender and Education

To understand the dynamics of marriage formation in each of the four provinces, we estimate survival curves of proportion remaining single by education and gender for the above four provinces in Figure 2. We use a dichotomous variable of education, those with some college versus those with less than some college education. For Henan and Guangxi provinces, as college graduates are only a small minority, we also dichotomize education as high school and above versus less than high school education. The survival curve patterns are almost identical to those based on the education coding of some college versus less than some college education in Figure 2. Results are thus not presented here but available upon request.

Figure 2.

Proportion of respondents remaining single by gender and education for selected provinces.

The common patterns among the four provinces are that marriage is still close to universal and much earlier than in other East Asian and Western countries; this is more so for women with low or no education and less so for men with low or no education. Eventually college graduates, both men and women, do get married. Although college education does delay marriage formation for men and women in their twenties, they catch up in their late twenties and early thirties. This is shown in the crossover of survival curves of college educated men and women and that of less educated men. A more notable pattern is the poorly educated or uneducated women in Beijing who marry early and have higher marriage rates than others and the poorly educated or uneducated men in Henan and Guangxi who marry later beyond the late twenties and eventually have lower rates of marriage than others. Ningxia witnesses a convergence of all four groups in their 30s.

Marriage Timing: Interwoven Stories of Gender and Education

Table 2 and Figure 3 summarize the results from the Cox proportional hazard models that examine the interaction effects between gender and education on marriage timing for the national sample and for the four selected provinces. We group education into four categories: no education, less than high school, high school, and some college or above education. For the whole country (first column in Table 2), the youngest cohort tends to marry later than other cohorts. Those with an urban hukou have a hazard of marrying 20% lower than those with a rural hukou. Current migrants have a slightly lower hazard of marrying. Those in the West and Central have a hazard of marrying more than 10% higher than those in the East. However, ethnicity and family economic resources do not differentiate individuals in terms of marriage timing. It is likely that regional difference has captured variations in these variables because they tend to have effects in the province-specific models (in columns 2-5).

Table 2.

Hazards of First Marriage for Respondents 30 Years and Older.

Variables	China	Beijing	Henan	Guangxi	Ningxia
Gender (Female = 1)	3.730***	6.964***	3.547***	2.996***	3.764***
Gender (Female = 1)	(−0.155)	(−0.783)	(−0.133)	(−0.151)	(−0.230)
Education (ref = No education)
<High school	1.524***	2.077***	1.778***	1.266***	1.367***
<High school	(−0.050)	(−0.192)	(−0.059)	(−0.057)	(−0.063)
High school	1.525***	1.696***	1.823***	1.327***	1.281***
High school	(−0.054)	(−0.159)	(−0.065)	(−0.064)	(−0.069)
College	1.312***	1.342**	1.486***	1.138**	1.217***
College	(−0.047)	(−0.126)	(−0.056)	(−0.057)	(−0.068)
Interaction between female and education
<High school	0.514***	0.248***	0.436***	0.661***	0.556***
<High school	(−0.024)	(−0.028)	(−0.017)	(−0.034)	(−0.036)
High school	0.362***	0.220***	0.331***	0.460***	0.399***
High school	(−0.016)	(−0.025)	(−0.014)	(−0.026)	(−0.028)
College	0.351***	0.207***	0.347***	0.445***	0.345***
College	(−0.014)	(−0.024)	(−0.016)	(−0.028)	(−0.025)
Age cohort (ref = 30-35)
35-40	1.171***	1.168***	1.225***	1.221***	1.081**
35-40	(−0.015)	(−0.030)	(−0.018)	(−0.024)	(−0.032)
40-45	1.234***	1.342***	1.245***	1.416***	1.256***
40-45	(−0.024)	(−0.034)	(−0.019)	(−0.029)	(−0.042)
45-50	1.016	1.139***	0.947***	1.265***	1.070*
45-50	(−0.025)	(−0.027)	(−0.015)	(−0.025)	(−0.036)
≥50	1.152***	1.166***	1.076***	1.220***	1.358***
≥50	(−0.028)	(−0.026)	(−0.015)	(−0.020)	(−0.043)
Hukou status (Urban = 1)	0.792***	0.739***	0.837***	0.794***	0.767***
Hukou status (Urban = 1)	(−0.017)	(−0.017)	(−0.012)	(−0.014)	(−0.025)
Ethnicity (Han = 1)	1.076	1.079*	1.051	0.923***	0.766***
Ethnicity (Han = 1)	(−0.053)	(−0.038)	(−0.043)	(−0.011)	(−0.020)
Migration (Current migrants = 1)	0.950***	0.998	0.942**	0.937**	0.952+
Migration (Current migrants = 1)	(−0.011)	(−0.015)	(−0.017)	(−0.019)	(−0.028)
Family financial resources	0.994	0.986	0.976***	1.053***	0.929***
Family financial resources	(−0.008)	(−0.010)	(−0.006)	(−0.007)	(−0.012)
Region (ref = East)
Central	1.159**
Central	(−0.056)
West	1.126*
West	(−0.067)
Provincial characteristics
% Professional occupation	0.995
% Professional occupation	(−0.003)
Observations	1,479,513	27,347	51,719	35,900	12,403

Note. The values in parentheses are standard errors.

p < .05.**p < .01. ***p < .001.

Figure 3.

Hazards of first marriage by gender and education, aged 30 years and older in 2005.

In addition, there is a highly significant interaction between education and gender on timing of marriage. Figure 3 presents hazards for men and women of different levels of education. Consistent with our expectation, women with no education have the highest hazard of marrying whereas men with no education have the lowest hazard: The former is almost 4 times as high as that of the latter. College educated women have a hazard that is 70% higher than that of uneducated men. Women of all educational levels have higher hazards than their male counterparts. Yet, the effect on men is more nuanced. Men with no education have lower hazards of marrying than all other men. Compared to uneducated men, the hazard of marrying is 30% higher for college graduates and 50% higher for those with a high school or less than high school degree. Hence, college education delays both men’s and women’s marriage formation, but the effect is much stronger for women. In contrast, having no education massively accelerates women’s marriage entry but postpones men’s marriage timing.

We estimate the same Cox proportional hazard models to examine the interaction between gender and education on timing of marriage for the four selected provinces. Columns 2 to 5 in Table 2 all show significant interactions between gender and education. To better illustrate the effects, hazards of the eight education-by-gender groups are shown in Figure 4.

Figure 4.

Hazards of first marriage by gender and education, aged 30 and older in 2005.

Overall, uneducated men have the lowest hazards of marrying among all the eight social groups, and uneducated women have the highest hazards of marrying. A common pattern across the four provinces is that education is negatively related to marriage timing for women, and women marry earlier than men with equivalent education. There is more heterogeneity in education effects on men’s marriage. Uneducated men have lower hazards of marriage than men with any levels of education; yet, among educated men, college graduates have lower hazards of marriage than their moderately educated counterparts. Consistent with the national pattern (shown in Figure 3), the delaying effect of education on marriage is stronger for women than for men.

In Beijing, women with no education have the hazard of marrying almost 7 times that of men with no education. The massive effect of no education on women’s marriage timing may be due to Beijing having the lowest proportion of uneducated women among all regions in China (calculated from the data of this study). According to the norm of status hypergamy, these women have a much larger pool of potential male partners in the marriage market. Educated men in general have higher hazards of marriage than do men with no education. The disadvantaged status of men with no education in the marriage market still holds for the provinces of Henan, Guangxi, and Ningxia; the disadvantage being most pronounced in Henan, one of the most populous provinces that send millions of migrant workers to the coastal and metropolitan areas. Yet, the education effect on men’s marriage timing in these three provinces is less pronounced than that in Beijing, and variations among college educated men are trivial, particularly for the two ethnic autonomous provinces, Guangxi and Ningxia.

Discussion and Conclusion

Using the 2005 One-percent Population Survey and a random sample of the survey, this study illustrates geographic heterogeneity in marriage formation in China with a focus on four regions. Near universal and early marriage largely remains true in contemporary China. Yet regional variations in marriage prevalence are notable and more prominent at younger ages of the twenties, but they converge fast after age 30. The East metropolitan areas witness a slower pace to marriage entry than in other places. Provinces with more ethnic minorities seem to show more heterogeneity in marriage formation. For example, Ningxia, the Hui ethnic autonomous region, has the earliest SMAM, whereas Guangxi, the Zhuang ethnic autonomous region, and Hainan province with a relatively high concentration of the Li ethnic group have relatively high SMAM. Guangxi has the latest age at marriage for men in the country. This is a puzzling finding that we do not fully understand. We speculate that it could have to do with the ethnic minority group finding it a challenge to negotiate their traditional norms of marriage and gender roles in a rapidly changing society in which education attainment and migration are on the rise, which may make it harder to find a marriage partner.

As noted, variations in marriage formation prevalence and timing do not strictly fit economic development stages. Local cultural norms, ethnic traditions, and religious practices are important factors in understanding regional variations in marriage formation in China. Migration may also play a critical role in shaping family formation patterns in modern China. For future studies, it is important to develop contextually specific hypotheses and to collect related measures to understand better regional heterogeneities in marriage formation. It is also important to investigate how these cultural variations intertwine with different levels of economic development and urbanization in the local context.

Gender and education largely differentiate men’s and women’s marriage formation in the fashion that we had expected. Although higher education tends to delay both men and women’s marriage formation with the effect stronger for women, most college graduates do not forgo marriage. However, men with less or no education not only postpone marriage but also have difficulty in finding mates.

Women with less or no education and college graduated women and men eventually get married after age 30. Men with less or no education have lower marriage rates as they grow older. With respect to marriage timing, education postpones women’s marriage, but college education delays both men and women’s marriage. Men with no education have the lowest hazards of marriage. Educated men marry later than do men with no education; and college educated men marry later than do moderately educated men.

Based on the 2005 data, we show that although college educated women have the latest marriage age among all women almost all of them ultimately get married, even in Beijing. Hence, the so-called “leftover” women issue does not have much empirical basis then. Female college graduates may delay marriage formation for career development as they are faced with increasing discrimination in the job market in recent decades. At the same time, they can also afford to spend more time searching for a compatible mate compared to less educated women, due to their relatively favorable financial situation. However, constrained by the social norm of status hypergamy and “gender double standard of aging,” they may face more challenges in searching for a mate than their male counterparts face. Nonetheless, evidence shows that most eventually marry by age 35.

As higher education continues to expand, it is likely that more Chinese women will further postpone their marriage formation. As late marriage becomes more socially acceptable, it is difficult to predict to what extent Chinese women will eventually forgo marriage in the future, a pattern that can be observed in societies such as South Korea, Singapore, Taiwan, and Hong Kong. Chinese women who choose to remain single voluntarily may continue to be derogated by the society as “leftover” women. It is urgent for the public and policy makers to promote gender egalitarian ideology and respect women’s choice in relation to marriage formation.

A severe challenge falls on men with less or no education. The social norm of status hypergamy and the severely skewed sex ratio have produced a double whammy for these men, particularly those living in remote, rural areas. Studies have estimated that the excess males in China will continue to increase in the next two to three decades, reaching 30 to 50 million, and cause a severe marriage squeeze for males (Chen, 2004; Huang, 2014; Poston, Conde, & DeSalvo, 2011). With the continuing expansion of higher education and a massive excess of young men coming to marriage age, the issue of “leftover” men will become much more severe. Therefore, the context of universal marriage in China will transform into a context of involuntary singlehood for many men. But this will not necessarily result in less marriage for poorly educated women because the pool of “surplus” men with higher levels of education will also be growing. Following the social norm of hypergamy, these women can always marry men of similar or better education. This is thus different from the case in the United States, where some lower educated women cannot find marriageable men due to the deterioration of poorly educated men’s labor market situation. Policy makers should consider measures to raise educational attainment and improve employment opportunities for these disadvantaged men as well as to introduce policies that facilitate marrying foreign brides if these men desire to do so.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Bauer

Feng

Riley

N. E.

Zhao

(1992). Gender inequality in urban China: Education and employment. Modern China, 18, 333-370.

Becker

G. S.

(1991). A treatise on the family. Cambridge, MA: Harvard University Press.

Cai

Tian

F. F.

(2013). Cultural expectations, market mismatch and individual choices: The making of “Shengnu” in China. Paper presented at the Population Association of America Annual Meeting, New Orleans, LA.

Cai

Wang

(2011). (Re)emergence of late marriage in New Shanghai. Paper presented at the Conference on Marriage in Cosmopolitan China, Hong Kong University, Hong Kong.

Cao

(2007). Gender and job mobility in postsocialist China: A longitudinal study of job changes in six coastal cities. Social Forces, 85, 1535-1560.

Chen

(2004). Examining the marriage market of China and European Union. Nanjing, China: Nanjing University Press.

Cohen

P. N.

Wang

(2008). Market and gender pay equity: Have Chinese reforms narrowed the gap? In Davis

Wang

(Eds.), Creating wealth and poverty in postsocialist China (pp. 37-53). Stanford, CA: Stanford University Press.

England

McClintock

E. A.

(2009). The gendered double standard of aging in U.S. marriage markets. Population and Development Review, 35, 797-816.

Fukuda

(2013). The changing role of women’s earnings in marriage formation in Japan. Annals of the American Academy of Political and Social Science, 646, 107-127.

10.

Gaetano

A. M.

(2010). Single women in urban China and the “unmarried crisis.” Lund, Sweden: Centre for East and South-East Asian Studies, Lund University.

11.

Glick

P. C.

Lin

S. L.

(1986). Recent changes in divorce and remarriage. Journal of Marriage and the Family, 48, 737-747.

12.

Goldscheider

F. K.

Waite

L. J.

(1986). Sex differences in the entry into marriage. American Journal of Sociology, 92, 91-109.

13.

Goldstein

J. R.

Kenney

C. T.

(2001). Marriage delayed or marriage forgone? New cohort forecasts of first marriage for U.S. women. American Sociological Review, 66, 506-519.

14.

Han

(2010). Trends in educational assortative marriage in China from 1970 to 2000. Demographic Research, 22, 733-770.

15.

Huang

(2014). Marriage squeeze in China: Past, present and future. Journal of Family Issues.

16.

Y. C.

(2013). Negotiating marriage and schooling: Nepalese women’s transition to adulthood. Annals of the American Academy of Political and Social Science, 646, 194-213.

17.

Jones

G. W.

(1980). Trends in marriage and divorce in Peninsular Malaysia. Population Studies, 34, 279-292.

18.

Jones

G. W.

(2001) Which Indonesian marry youngest, and why? Journal of Southeast Asian Studies, 32, 67-78.

19.

Jones

G. W.

Gubhaju

(2009). Factors influencing changes in mean age at first marriage and proportions never marrying in the low-fertility countries of East and Southeast Asia. Asian Population Studies, 5, 237-265.

20.

Kalmijn

(2013). The educational gradient in marriage: A comparison of 25 European countries. Demography, 50, 1499-1520.

21.

Maurer-Fazio

Connelly

Chen

Tang

(2011). Childcare, eldercare, and labor force participation of married women in urban China, 1982-2000. Journal of Human Resources, 46, 261-294.

22.

McDonald

(2000). Gender equity, social institutions and the future of fertility. Journal of Population Research, 17(1), 1-16.

23.

National Bureau of Statistics of China. (2011). Shi yi wu” jing ji she hui fa zhan cheng jiu xi lie bao gao zhi san: Ren kou zong liang ping wen zeng zhang, jiu ye xing shi bao chi wen ding [Series report for the eleventh five-year socioeconomic development, No. 3: Steady growth in total population, employment patterns remain stable]. Retrieved from http://www.stats.gov.cn/ztjc/ztfx/sywcj/201103/t20110302_71315.html

24.

Ono

(2003). Women’s economic standing, marriage timing, and cross-national contexts of gender. Journal of Marriage and Family, 65, 275-286.

25.

Oppenheimer

V. K.

(1988). A theory of marriage timing. American Journal of Sociology, 94, 563-591.

26.

Oppenheimer

V. K.

(1994). Women’s rising employment and the future of the family in industrial societies. Population and Development Review, 20, 293-342.

27.

Park

(2013). The transition to adulthood among Korean youths transition markers in productive and reproductive spheres. Annals of the American Academy of Political and Social Science, 646, 129-148.

28.

Poston

D. L.

Jr. Conde

DeSalvo

(2011). China’s unbalanced sex ratio at birth, millions of excess bachelors and societal implications. Vulnerable Children and Youth Studies, 6, 314-320.

29.

Poston

D. L.

Jr. Glover

K. S.

(2005). Too many males: Marriage market implications of gender imbalances in China. Genus, 61, 119-140.

30.

Presser

H. B.

(1975). Age differences between spouses: Trends, patterns, and social implications. American Behavioral Scientist, 19, 190-205.

31.

Qian

(2012). Marriage squeeze for highly educated women? Gender differences in assortative marriage in urban China (Unpublished master’s thesis), Ohio State University, Columbus.

32.

Raymo

J. M.

Iwasawa

(2005). Marriage market mismatches in Japan: An alternative view of the relationship between women’s education and marriage. American Sociological Review, 70, 801-822.

33.

Siu

H. F.

(1990). Where were the women? Rethinking marriage resistance and regional culture in South China. Late Imperial China, 11(2), 32-62.

34.

Sweeney

M. M.

(2002). Two decades of family change: The shifting economic foundations of marriage. American Sociological Review, 67, 132-147.

35.

Tey

N. P.

(2011). Understanding marriage and divorce trends in Peninsular Malaysia. In Jones

G. W.

Hull

T. H.

Mohamad

(Eds.), Changing marriage patterns in Southeast Asia: Economic and socio-cultural dimensions (pp. 137-155). New York, NY: Routledge.

36.

Thornton

(2005). Reading history sideways: The fallacy and enduring impact of the developmental paradigm on family life. Chicago, IL: University of Chicago.

37.

Thornton

Fricke

T. E.

(1987). Social change and the family: Comparative perspectives from the West, China, and South Asia. Sociological Forum, 2, 746-779.

38.

Tian

F. F.

(2013). Transition to first marriage in reform-era urban China: The persistent effect of education in a period of rapid social change. Population Research and Policy Review, 32, 529-552.

39.

(2013). Understanding Sheng Nu (“leftover women”): The phenomenon of late marriage among Chinese professional women. Symbolic Interaction, 36(1), 1-20.

40.

Trent

South

S. J.

(2011). Too many men? Sex ratios and women’s partnering behavior in China. Social Forces, 90, 247-268.

41.

Tuljapurkar

Feldman

M. W.

(1995). High sex ratios in China’s future. Science, 267, 874-876.

42.

Whyte

M. K.

(Ed.). (1984). Sexual inequality under socialism: The Chinese case in perspective. Class and social stratification in post-revolution China. New York, NY: Cambridge University Press.

43.

(1983). Qing Bai Lei Chao. Juan 5 [Qing Petty Matters Anthology. Vol. 5]. Taipei, Taiwan: Commercial Press.

44.

Yeung

W. J.

(2013). College expansion policy and social stratification in China. Chinese Sociological Review, 45(4), 54-80.

45.

Yeung

W. J.

Alipio

(2013). Transitioning to adulthood in Asia: Courtship, marriage, and work: An introduction. Annals of the American Academy of Political and Social Science, 646, 6-27.

46.

Yeung

W. J.

(2013). Coming of age in times of change: The transition to adulthood in China. Annals of the American Academy of Political and Social Science, 646, 149-171.

47.

Yeung

W. J.

Y. C.

(2013). Are economic foundations of marriage shifting in China? Evidence in a socialist context. Paper presented at the Population Association of America Annual Meeting, New Orleans, LA.

48.

Xie

(2013). Changes in the determinants of marriage entry in post-reform urban China. Paper presented at the Population Association of America Annual Meeting, New Orleans, LA.

49.

Yuan

(2012). Guan yu zhong guo di yu wen hua de li lun si kao [Theoretical thoughts about China’s regional culture]. Journal of Peking University (Philosophy and Social Sciences), 49(1), 11-20.

50.

Zhang

(2002). Labor market segmentation and income inequality in urban China. Sociological Quarterly, 43(1), 27-44.

51.

Zhang

Hannum

Wang

(2008). Gender-based employment and income differences in urban China: Considering the contributions of marriage and parenthood. Social Forces, 86, 1529-1560.

52.

Zhao

(2007). She hui zhuang xing qi xi bei shao shu ming zu ju ming jia zhi guan de shan bian [Changes in values among northwestern ethnic groups]. Beijing, China: People’s Publishing House.

53.

Zuo

Bian

(2001). Gendered resources, division of housework, and perceived fairness—A case in urban China. Journal of Marriage and Family, 63, 1122-1133.