Demographic Research:Volume 31,Article 45 market insecurity for women,women have strong incentives to marry up in status in categories such as age and education,to achieve high living standards through marriage. Relatedly,the breadwinner role of the husband and the homemaker role of the wife remain firmly in place in Chinese families(Qian and Qian 2014;Zuo 2003;Zuo and Bian 2001).Strikingly,since the 2000s,there has been a growing emphasis on traditional gender roles among Chinese men and women (Attane 2012).Urban women's domestic responsibilities are further reinforced by the unequal role given to mothers to raise the perfect child under the one-child family policy (Evans 2010;Greenhalgh 2010).Indeed,career-oriented women are commonly criticized as "selfish," “nonfeminine,.”and“irresponsible to household needs,”whereas husbands'failure to fulfill the provider role is often the primary source of marital conflict(Zuo and Bian 2001).This suggests that women value economic prospects in a potential mate,and that women with high earning potentials and career aspirations may not find marriage beneficial,due to clashes between career and family.Thus,we hypothesize that educational attainment is positively associated with men's but negatively associated with women's likelihood of marriage in urban China. Meanwhile,attitudes toward gender equality vary substantially by gender itself.In China,women hold more egalitarian work and family values than men (Chang 1999). Highly educated individuals are more likely to hold egalitarian gender attitudes,but the educational effect is stronger for women than for men(Shu 2004).This suggests that highly educated women may not want to compromise their careers when they search for marriageable partners.In contrast,highly educated men with less egalitarian gender attitudes may prefer to marry less-educated women.As a result,the marriage pool may be small among highly educated women due to a shortage of similarly educated men who share similar levels of egalitarian gender-role expectations.This contributes to a "marriage squeeze"for highly educated women in urban China (Jones 2007). In addition,parents often meddle with their children's marriage(Jennings,Axinn, and Ghimire 2012).Unlike Western cultures,China has long-standing,strong, intergenerational family ties.Although arranged marriages are banned in China,parents remain actively involved in their children's choice of mate(Riley 1994;Xu and Whyte 1990).Parents attempt to ensure that their children meet,date,and marry the"right" person.Their influence is twofold.Indirectly,they socialize their children about gender roles and mate choices well before their children are ready for marriage (Riley 1994). Directly,they often disapprove their children's inclinations to form nontraditional marriages,in which the wife is older than the husband or the husband has less schooling than the wife. http://www.demographic-research.org 1341
Demographic Research: Volume 31, Article 45 http://www.demographic-research.org 1341 market insecurity for women, women have strong incentives to marry up in status in categories such as age and education, to achieve high living standards through marriage. Relatedly, the breadwinner role of the husband and the homemaker role of the wife remain firmly in place in Chinese families (Qian and Qian 2014; Zuo 2003; Zuo and Bian 2001). Strikingly, since the 2000s, there has been a growing emphasis on traditional gender roles among Chinese men and women (Attané 2012). Urban women‟s domestic responsibilities are further reinforced by the unequal role given to mothers to raise the perfect child under the one-child family policy (Evans 2010; Greenhalgh 2010). Indeed, career-oriented women are commonly criticized as “selfish,” “nonfeminine,” and “irresponsible to household needs,” whereas husbands‟ failure to fulfill the provider role is often the primary source of marital conflict (Zuo and Bian 2001). This suggests that women value economic prospects in a potential mate, and that women with high earning potentials and career aspirations may not find marriage beneficial, due to clashes between career and family. Thus, we hypothesize that educational attainment is positively associated with men‟s but negatively associated with women‟s likelihood of marriage in urban China. Meanwhile, attitudes toward gender equality vary substantially by gender itself. In China, women hold more egalitarian work and family values than men (Chang 1999). Highly educated individuals are more likely to hold egalitarian gender attitudes, but the educational effect is stronger for women than for men (Shu 2004). This suggests that highly educated women may not want to compromise their careers when they search for marriageable partners. In contrast, highly educated men with less egalitarian gender attitudes may prefer to marry less-educated women. As a result, the marriage pool may be small among highly educated women due to a shortage of similarly educated men who share similar levels of egalitarian gender-role expectations. This contributes to a “marriage squeeze” for highly educated women in urban China (Jones 2007). In addition, parents often meddle with their children‟s marriage (Jennings, Axinn, and Ghimire 2012). Unlike Western cultures, China has long-standing, strong, intergenerational family ties. Although arranged marriages are banned in China, parents remain actively involved in their children‟s choice of mate (Riley 1994; Xu and Whyte 1990). Parents attempt to ensure that their children meet, date, and marry the “right” person. Their influence is twofold. Indirectly, they socialize their children about gender roles and mate choices well before their children are ready for marriage (Riley 1994). Directly, they often disapprove their children‟s inclinations to form nontraditional marriages, in which the wife is older than the husband or the husband has less schooling than the wife
Qian and Oian:The gender divide in urban China:Singlehood and assortative mating by age and education 2.3 The current study In China,increasing educational opportunities have equalized the playing field in college education between men and women(Treiman 2013;Yeung 2013).Meanwhile, education is now strongly tied to occupational prestige;earnings returns to education-- college education in particular--have increased rapidly(Bian and Logan 1996;Zhang and Zhao 2007;Zhao and Zhou 2007).Since education plays an increasingly important role in determining an individual's socioeconomic position,it is not a surprise that educational homogamy in urban China has increased between the 1970s and the late 1990s,as seen in the United States (Han 2010).Nevertheless,traditional gender roles are stronger today than in the recent past,which is predicted to produce different patterns of marriage formation and assortative mating from those in the United States and other developed countries (Blossfeld 1995;Qian 1998;Schwartz and Mare 2005). Based on the discussion above,we formulate the following hypotheses regarding marriage formation and assortative mating in urban China: 1)College-educated women have lower marriage rates than their male counterparts,especially among those who marry at later ages. 2)The patterns of educational and age assortative marriage are gender asymmetrical: 2.a)Among marriages in which two spouses have different levels of educational attainment,husbands tend to have more education than wives; 2.b)Among marriages in which two spouses differ in age,wives tend to be younger than husbands. 3)The pattern of educational assortative mating varies by age at first marriage (Qian 1998):individuals marrying at later ages are less likely to form educational homogamy.This pattern is expected to be more evident among men than among women. 1342 http://www.demographic-research.org
Qian and Qian: The gender divide in urban China: Singlehood and assortative mating by age and education 1342 http://www.demographic-research.org 2.3 The current study In China, increasing educational opportunities have equalized the playing field in college education between men and women (Treiman 2013; Yeung 2013). Meanwhile, education is now strongly tied to occupational prestige; earnings returns to education -- college education in particular -- have increased rapidly (Bian and Logan 1996; Zhang and Zhao 2007; Zhao and Zhou 2007). Since education plays an increasingly important role in determining an individual‟s socioeconomic position, it is not a surprise that educational homogamy in urban China has increased between the 1970s and the late 1990s, as seen in the United States (Han 2010). Nevertheless, traditional gender roles are stronger today than in the recent past, which is predicted to produce different patterns of marriage formation and assortative mating from those in the United States and other developed countries (Blossfeld 1995; Qian 1998; Schwartz and Mare 2005). Based on the discussion above, we formulate the following hypotheses regarding marriage formation and assortative mating in urban China: 1) College-educated women have lower marriage rates than their male counterparts, especially among those who marry at later ages. 2) The patterns of educational and age assortative marriage are gender asymmetrical: 2.a) Among marriages in which two spouses have different levels of educational attainment, husbands tend to have more education than wives; 2.b) Among marriages in which two spouses differ in age, wives tend to be younger than husbands. 3) The pattern of educational assortative mating varies by age at first marriage (Qian 1998): individuals marrying at later ages are less likely to form educational homogamy. This pattern is expected to be more evident among men than among women
Demographic Research:Volume 31,Article 45 3.Data,sample,and measurement 3.1 Data In this study,we pool nationally representative samples from the Chinese General Social Surveys (CGSS)conducted in 2003,2005,2006,and 2008.The CGSS is a repeated cross-sectional survey that started in 2003.Currently,four waves of data and related information are available online through a public data archive-Chinese Social Survey Open Database (www.cssod.org).The 2003 through 2008 waves of the CGSS used a multi-stage,stratified,random sampling method and included nationally representative samples of adults aged 18 and above (see Bian and Li 2012 for more details).The CGSS data are ideal for this study because the surveys collected information on respondents'current marital status and year of first marriage along with other sociodemographic characteristics for respondents and their spouses. 3.2 Sample We restrict the analysis to urban residents because of strong urban and rural differences in education and marriage (Han 2010;Treiman 2012)and,more importantly,most highly educated men and women in China live in urban areas.Our analytic sample includes the couples first married between 2000 and 2008'and their never married counterparts at risk of forming such marriages over the same period.Specifically,we include the couples first married at ages 20 to 49 years between 2000 and 2008 and their never married counterparts in the same age range.We use this age range because the legal minimum age at marriage is 20 for women and 22 for men in China and almost no one marries for the first time after age 49 (Lindgren 2009).This study provides an important update,as no prior research has examined assortative marriage patterns since the 2000s,right after the 1999 college expansion in China. We carry out the analysis in two steps.First,we calculate first marriage rates by education and age among men and women,respectively.To calculate first marriage rates,we follow Raymo and Iwasawa(2005)to reconstruct the population at risk of first marriage.Specifically,we generate person-year data for never married men and We rely only on respondents'marital order to distinguish first marriages from higher-order marriages, because our data include information on whether the respondent is married for the first time,but have no marital order information of the spouse.Fortunately,marriages contracted between spouses of different marriage orders are rare.We performed additional analysis using data from the 2006 CGSS,which is the only survey that has information on the marriage order of the spouse,and showed that only 10 out of 676 first- time-married respondents had a remarried spouse.Thus,a lack of controls over the marriage order of the spouse is unlikely to bias the results. http://www.demographic-research.org 1343
Demographic Research: Volume 31, Article 45 http://www.demographic-research.org 1343 3. Data, sample, and measurement 3.1 Data In this study, we pool nationally representative samples from the Chinese General Social Surveys (CGSS) conducted in 2003, 2005, 2006, and 2008. The CGSS is a repeated cross-sectional survey that started in 2003. Currently, four waves of data and related information are available online through a public data archive – Chinese Social Survey Open Database (www.cssod.org). The 2003 through 2008 waves of the CGSS used a multi-stage, stratified, random sampling method and included nationally representative samples of adults aged 18 and above (see Bian and Li 2012 for more details). The CGSS data are ideal for this study because the surveys collected information on respondents‟ current marital status and year of first marriage along with other sociodemographic characteristics for respondents and their spouses. 3.2 Sample We restrict the analysis to urban residents because of strong urban and rural differences in education and marriage (Han 2010; Treiman 2012) and, more importantly, most highly educated men and women in China live in urban areas. Our analytic sample includes the couples first married between 2000 and 20083 and their never married counterparts at risk of forming such marriages over the same period. Specifically, we include the couples first married at ages 20 to 49 years between 2000 and 2008 and their never married counterparts in the same age range. We use this age range because the legal minimum age at marriage is 20 for women and 22 for men in China and almost no one marries for the first time after age 49 (Lindgren 2009). This study provides an important update, as no prior research has examined assortative marriage patterns since the 2000s, right after the 1999 college expansion in China. We carry out the analysis in two steps. First, we calculate first marriage rates by education and age among men and women, respectively. To calculate first marriage rates, we follow Raymo and Iwasawa (2005) to reconstruct the population at risk of first marriage. Specifically, we generate person-year data for never married men and 3 We rely only on respondents‟ marital order to distinguish first marriages from higher-order marriages, because our data include information on whether the respondent is married for the first time, but have no marital order information of the spouse. Fortunately, marriages contracted between spouses of different marriage orders are rare. We performed additional analysis using data from the 2006 CGSS, which is the only survey that has information on the marriage order of the spouse, and showed that only 10 out of 676 firsttime-married respondents had a remarried spouse. Thus, a lack of controls over the marriage order of the spouse is unlikely to bias the results
Qian and Oian:The gender divide in urban China:Singlehood and assortative mating by age and education women,as each of them was at risk of first marriage at every single age,starting at age 20 until 49 years old or until censored by the surveys.We then generate person-year data among those married after 2000-they remained single after age 20 until the age they married.We merge the data for the married and unmarried individuals for men and women,separately,and create never married person-year data starting at age 20 and ending at age 49,the age censored by the surveys,or the age they married,for the period between 2000 and 2008.We do not distinguish marriages within the period, because the data are sparse and because the substantive results are similar if the period is further classified into specific years.The final number of person-years is 11,311 for women and 14,156 for men. Second,we explore assortative mating patterns by education and age.Dropping six observations with missing data results in a sample of 2,151 couples first married between ages 20 and 49 during the 2000-2008 period. 3.3 Measurement The CGSS asked respondents their educational attainment,and,if married,their spouses'education at the time of the survey.We do not have information on educational attainment at the time of the wedding.However,these two measures should be similar because we focus on newlyweds,and educational upgrading after marriage is rare in China.We classify educational attainment into four groups:less than senior high school,senior high school,vocational college (Da Zhuan),and four-year college or higher.Due to compulsory nine-year education in urban China,all people with junior high school or less are collapsed into one educational group. Individuals are grouped into three age categories:20-24,25-29,and 30-49,which reflects three typical ages of marriage in China:early,normal,and late.Note that in this study,age of a never married person refers to age at the time of surveys,while age of a married person refers to age at first marriage.We do not further disaggregate the age category for 30 and above into consistent five-year age groups because a small proportion of people marry for the first time after age 30 in China (Jones and Gubhaju 2009).Indeed,in our sample,2.5%and 5.5%of person-years are between 35 and 49 years of age for women and for men,respectively,while only 4.5 percent of couples are those having at least one spouse older than 34.Results are substantively the same if we limit the age range to 20-34 years(results available upon request). 1344 http://www.demographic-research.org
Qian and Qian: The gender divide in urban China: Singlehood and assortative mating by age and education 1344 http://www.demographic-research.org women, as each of them was at risk of first marriage at every single age, starting at age 20 until 49 years old or until censored by the surveys. We then generate person-year data among those married after 2000 – they remained single after age 20 until the age they married. We merge the data for the married and unmarried individuals for men and women, separately, and create never married person-year data starting at age 20 and ending at age 49, the age censored by the surveys, or the age they married, for the period between 2000 and 2008. We do not distinguish marriages within the period, because the data are sparse and because the substantive results are similar if the period is further classified into specific years. The final number of person-years is 11,311 for women and 14,156 for men. Second, we explore assortative mating patterns by education and age. Dropping six observations with missing data results in a sample of 2,151 couples first married between ages 20 and 49 during the 2000–2008 period. 3.3 Measurement The CGSS asked respondents their educational attainment, and, if married, their spouses‟ education at the time of the survey. We do not have information on educational attainment at the time of the wedding. However, these two measures should be similar because we focus on newlyweds, and educational upgrading after marriage is rare in China. We classify educational attainment into four groups: less than senior high school, senior high school, vocational college (Da Zhuan), and four-year college or higher. Due to compulsory nine-year education in urban China, all people with junior high school or less are collapsed into one educational group. Individuals are grouped into three age categories: 20–24, 25–29, and 30–49, which reflects three typical ages of marriage in China: early, normal, and late. Note that in this study, age of a never married person refers to age at the time of surveys, while age of a married person refers to age at first marriage. We do not further disaggregate the age category for 30 and above into consistent five-year age groups because a small proportion of people marry for the first time after age 30 in China (Jones and Gubhaju 2009). Indeed, in our sample, 2.5% and 5.5% of person-years are between 35 and 49 years of age for women and for men, respectively, while only 4.5 percent of couples are those having at least one spouse older than 34. Results are substantively the same if we limit the age range to 20–34 years (results available upon request)
Demographic Research:Volume 31,Article 45 4.Methods 4.1 Marriage rate Marriage rate is commonly used to measure the incidence of marriage (e.g.,Preston and Richards 1975;Qian and Preston 1993;Raymo and Iwasawa 2005).The first marriage rate is based on the numerator(number of first marriages occurring within a given time period)and the denominator ("person-years"at the risk of first marriage during that time period)(Preston,Heuveline,and Guillot 2001).It takes the form: Marriage Rate [O,T]=- number of first marriages occuring between time 0 and T person-years of exposure to the risk of occurence of first marriage (1) For each person-year observation in our sample,if the event (i.e.,first marriage) does not occur,that is,the person remains single in that year,the person-year observation then contributes one year to the denominator.If the person marries in that year,this person contributes half a year to the denominator because marriages in a given year are assumed to happen evenly throughout the year,which holds true for large populations.We follow the convention of multiplying the results from equation (1)by 1,000.Thus,marriage rates in this paper indicate incidence of first marriages per 1000 person-years. 4.2 Log-linear models When studying age and education assortative mating,we use log-linear models.Log- linear models control for gender differences in the marginal distributions of age and education so as to study assortative mating patterns net of the effects of population structure (Hout 1983;Kalmijn 2010).Given that people tend to marry within their educational and age groups,we add an education homogamy parameter and an age homogamy parameter to the cells on the main education and age diagonals, respectively.In order to examine gender differences in marriage patterns,we add education and age hypergamy parameters in the models to explore gender asymmetries. For education,we constrain the cells in which the husband is better educated than the wife into one parameter,and for age,we create a parameter in which the husband married at an older age group than the wife.Note that our age assortative mating parameters are age-group specific.Thus,age hypergamous marriages only consist of those in which the husband and the wife belong to different age groups.For our purpose,age homogamous marriages consist of those in which the husband and the http://www.demographic-research.org 1345
Demographic Research: Volume 31, Article 45 http://www.demographic-research.org 1345 4. Methods 4.1 Marriage rate Marriage rate is commonly used to measure the incidence of marriage (e.g., Preston and Richards 1975; Qian and Preston 1993; Raymo and Iwasawa 2005). The first marriage rate is based on the numerator (number of first marriages occurring within a given time period) and the denominator (“person-years” at the risk of first marriage during that time period) (Preston, Heuveline, and Guillot 2001). It takes the form: Marriage Rate [0, T] = (1) For each person-year observation in our sample, if the event (i.e., first marriage) does not occur, that is, the person remains single in that year, the person-year observation then contributes one year to the denominator. If the person marries in that year, this person contributes half a year to the denominator because marriages in a given year are assumed to happen evenly throughout the year, which holds true for large populations. We follow the convention of multiplying the results from equation (1) by 1,000. Thus, marriage rates in this paper indicate incidence of first marriages per 1000 person-years. 4.2 Log-linear models When studying age and education assortative mating, we use log-linear models. Loglinear models control for gender differences in the marginal distributions of age and education so as to study assortative mating patterns net of the effects of population structure (Hout 1983; Kalmijn 2010). Given that people tend to marry within their educational and age groups, we add an education homogamy parameter and an age homogamy parameter to the cells on the main education and age diagonals, respectively. In order to examine gender differences in marriage patterns, we add education and age hypergamy parameters in the models to explore gender asymmetries. For education, we constrain the cells in which the husband is better educated than the wife into one parameter, and for age, we create a parameter in which the husband married at an older age group than the wife. Note that our age assortative mating parameters are age-group specific. Thus, age hypergamous marriages only consist of those in which the husband and the wife belong to different age groups. For our purpose, age homogamous marriages consist of those in which the husband and the