复旦大学：《文化人类学 Cultural Anthropology》阅读资料_China’s One-Child Policy and the Empowerment of Urban Daughters

VANESSA L. FONG Research articles Chinas One-Child Policy and the Empowerment of Urban Daughters ABSTRACT Urban daughters have benefited from the demographic pattern produced by China's one-child policy. In the system of trilineal kinship that has long characterized most of Chinese society, parents had little incentive to invest in their daughters. Singleton daughters, however, enjoy unprecedented parental support because they do not have to compete with brothers for parental invest- ment. Low fertility enabled mothers to get paid work and, thus, gain the ability to demonstrate their filiality by providing their own parents with financial support Because their mothers have already proven that daughters can provide their parents with old age sup- and because singletons have no brothers for their parents to favor, daughters have more power than ever before to defy disad- tageous gender norms while using equivocal ones to their own advantage. [Keywords: gender, family, fertility, demography, China N 1998, WHEN I FIRST started tutoring Ding Na, the scores, checking and rechecking her arithmetic, her eyes daughter of two factory workers in Dalian City, China, wide. "Are you sure you heard correctly? "her mother thought her father's attitude exemplified the asked. Ding Na was sure. She had scored higher than she bias against daughters portrayed in many studies of Chi- had ever scored on a practice exam in high school, and lese family life(Greenhalgh 1985a, 1994b; Harrell 1982; well above the likely cutoff for her top-choice four-year Salaff 1995; Wolf 1968, 1972). Although studious and well college. She shouted with joy as we congratulated her. Her behaved, Ding Na was often criticized by her father, who father beamed at her with tears in his eyes and said, " I was liked to remind her that he had always wanted a son. He wrong to have wanted a son. a daughter like you is worth worried that she might not score high enough to get into ten sons good four-year college, even though she usually ranked in e experiences of girls like Ding Na are quite differ the top 20 percent of her high school class on practice ex- ent from those of daughters who grew up in the patril ams."What will you do if you don't get into a good col- ineal, patrilocal, and patriarchal world described in classic lege? " he lamented. " If you were a boy, you could study studies of gender in Chinese societies(Andors 1983; Croll abroad while supporting yourself as a laborer, but what 1995: Greenhalgh 1985a, 1994b; Jaschok and Miers 1994; can a girl do abroad besides sit around waiting for remit- Stacey 1983; Watson 1986, 1996; Wolf 1968, 1972). The afford? Although her mother praised her devastating effect of gender norms on daughters of th for being more willing to help with chores than most world is evident in the life stories of women born prior to other teenagers, whenever Ding Na had trouble helping the 1950s, and to a lesser extent in those of women born her father carry groceries or move furniture, he snapped, in the 1950s and 1960s. girls born after Chinas one-child Girls are so useless. A boy would have no trouble with this. policy began in 1979, however, have more power to chal On July 26, 1999, when Ding Na's college entrance lenge detrimental gender norms and use helpful ones than between Ding Na and her father in a different light. I sence of brothers for their parents to favor y and the ab- exam scores were released, I began to see the relationship ever before, thanks to the decline of patrili stayed up with Ding Na and her parents as we waited well In this article, I argue that urban daughters born un past our bedtimes for her scores to become available der Chinas one-child policy have benefited from the hrough an automated phone hotline at midnight. After demographic pattern produced by that policy. By compa her call finally went through, she wrote down her subject ing the experiences of daughters born in the 1980s with AMERICAN ANTHROPOLOGIST 104(4) 3-1109. COPYRIGHT 2002. AMERICAN ANTHROPOLOGICAL ASSOCIATION

VANESSA L. FONG Research Articles China’s One-Child Policy and the Empowerment of Urban Daughters ABSTRACT Urban daughters have benefited from the demographic pattern produced by China’s one-child policy. In the system of patrilineal kinship that has long characterized most of Chinese society, parents had little incentive to invest in their daughters. Singleton daughters, however, enjoy unprecedented parental support because they do not have to compete with brothers for parental invest￾ment. Low fertility enabled mothers to get paid work and, thus, gain the ability to demonstrate their filiality by providing their own parents with financial support. Because their mothers have already proven that daughters can provide their parents with old age sup￾port, and because singletons have no brothers for their parents to favor, daughters have more power than ever before to defy disad￾vantageous gender norms while using equivocal ones to their own advantage. [Keywords: gender, family, fertility, demography, China] IN 1998, WHEN I FIRST started tutoring Ding Na, the daughter of two factory workers in Dalian City, China, I thought her father’s attitude exemplified the parental bias against daughters portrayed in many studies of Chi￾nese family life (Greenhalgh 1985a, 1994b; Harrell 1982; Salaff 1995; Wolf 1968, 1972). Although studious and well behaved, Ding Na was often criticized by her father, who liked to remind her that he had always wanted a son. He worried that she might not score high enough to get into a good four-year college, even though she usually ranked in the top 20 percent of her high school class on practice ex￾ams. “What will you do if you don’t get into a good col￾lege?” he lamented. “If you were a boy, you could study abroad while supporting yourself as a laborer, but what can a girl do abroad besides sit around waiting for remit￾tances I can’t afford?” Although her mother praised her for being more willing to help with chores than most other teenagers, whenever Ding Na had trouble helping her father carry groceries or move furniture, he snapped, “Girls are so useless. A boy would have no trouble with this.” On July 26, 1999, when Ding Na’s college entrance exam scores were released, I began to see the relationship between Ding Na and her father in a different light. I stayed up with Ding Na and her parents as we waited well past our bedtimes for her scores to become available through an automated phone hotline at midnight. After her call finally went through, she wrote down her subject scores, checking and rechecking her arithmetic, her eyes wide. “Are you sure you heard correctly?” her mother asked. Ding Na was sure. She had scored higher than she had ever scored on a practice exam in high school, and well above the likely cutoff for her top-choice four-year college. She shouted with joy as we congratulated her. Her father beamed at her with tears in his eyes and said, “I was wrong to have wanted a son. A daughter like you is worth ten sons.” The experiences of girls like Ding Na are quite differ￾ent from those of daughters who grew up in the patril￾ineal, patrilocal, and patriarchal world described in classic studies of gender in Chinese societies (Andors 1983; Croll 1995; Greenhalgh 1985a, 1994b; Jaschok and Miers 1994; Stacey 1983; Watson 1986, 1996; Wolf 1968, 1972). The devastating effect of gender norms on daughters of that world is evident in the life stories of women born prior to the 1950s, and to a lesser extent in those of women born in the 1950s and 1960s. Girls born after China’s one-child policy began in 1979, however, have more power to chal￾lenge detrimental gender norms and use helpful ones than ever before, thanks to the decline of patriliny and the ab￾sence of brothers for their parents to favor.1 In this article, I argue that urban daughters born un￾der China’s one-child policy have benefited from the demographic pattern produced by that policy. By compar￾ing the experiences of daughters born in the 1980s with AMERICAN ANTHROPOLOGIST 104(4):1098–1109. COPYRIGHT © 2002, AMERICAN ANTHROPOLOGICAL ASSOCIATION

Fong China's One-Child Policy 1099 the experiences of their mothers and grandmothers, I well-compensated work rather than using it to bear and show how singleton daughters have unprecedented power rear large numbers of children. Fertility is especially low to deal with gender norms in ways that benefit them. Al- when most women are expected to work at jobs incompat though I argue that low fertility has been a key factor in ible with childrearing. A high rate of female employment the empowerment of urban Chinese daughters, I do not is one of the strongest correlates of low fertility(Burggraf claim that it is the only necessary and sufficient factor. 1997; Essock- Vitale and McGuire 1988: 229, 233; Felmlee Low fertility can only empower daughters in areas where 1993; Gerson 1985; Sander 1990; Weinberg 1976).School- opportunities for employment and education are already ing is also likely to cause women to learn childrearing available to women. In the Chinese countryside, where practices that reduce infant mortality and, thus, reduce such opportunities remain out of reach for many women, the need to have large numbers of children, as Robert A ompulsory low fertility tends to frustrate women more Levine and his coauthors found in a 1983 study of Mexi- than it empowers them In cities like Dalian, however, it is can mothers' education and childcare practices(Levine et clear that daughters would have been less able to take ad- al. 1991 vantage of available opportunities if they had to compete Michelle Zimbalist Rosaldo attributed gender inequal- with brothers for family resources, and if their mothers ity to a universal"opposition between the'domestic'ori had not demonstrated that women can support their par- entation of women and the extra-domestic or public'ties ents in old age that, in most societies, are primarily available to men (Rosaldo 1974: 17-18). The public sphere offers greater THEORETICAL CONSIDERATIONS possibilities for empowerment because, unlike the domes Studies of many developed and developing societies tic sphere, it involves formal roles, rights, and duties; the worldwide have documented a high correlation between authority that comes from lack of intimacy; the opportu- nity to achieve rather than be ascribed status; the power to low fertility and women's empowerment(Abadian 1996, create"culture"; the tendency to be categorized as the Balk 1997; Davis 1986; Dharmalingam and Morgan 1996 Keyfitz 1986; Sathar 1988). Although these studies have norm"rather than the anomaly", and control over the focused on low fertility as a cause and effect of mothers production of goods with greater cultural value(rosaldo empowerment, my findings suggest that more attention 1974: 25-35). This theory was later criticized by Rosaldo herself (1980)as well as by other feminist anthropologist should be paid to how low fertility affects daughters. The for relying on dichotomies that do not exist in all societies effects of China's one-child policy on mothers are equivo-( Collier and Yanagisako 1987; Mac Cormack and Strathern cal. On one hand, it has freed mothers from heavy child bearing and child-rearing burdens; on the other hand, it standing gender systems in societies like China, in which has deprived mothers of the freedom to choose their fam- gender inequality has long been based on distinctions be- by men and a nd enforcement tactics. The policy's effects on urban subordinate domestic sphere associated with women daughters, however, are largely beneficial In such societies, the adoption of a modern economy Low resistance to the one-child policy in cities like tends to increase women's employment rates and parental Dalian can be attributed to the rapid pace with which people bias against daughters tends to decrease when daughters in such cities have internalized the same cultural model of are seen as capable of earning money. This pattern was modernization that has caused fertility decline in many documented in late 1980s Taiwan(Stafford 1995) and in lates with the degree to which it has adopted a modern Rosenweig and Schultz 1982). When d i et al. 1995 societies worldwide. A society's fertility rate usually corre 1970s-1980s India (Kishor 1993: Murt economy in which child mortality is low, most people live modernization, the fertility transition enables and com- in urban environments in which children consume a lot pels women to devote themselves to work and education more than they produce, most mothers as well as fathers rather than motherhood. This is not always beneficial to work at jobs incompatible with childrearing, and exten- the first generation of women to experience the fertility sive education is widespread for both genders and seen as transition, since they tend to have been socialized to de the road to socioeconomic success. All of these factors are sire large numbers of children and may suffer when they kely to be both causes and effects of low fertility cannot realize this desire. It is much more beneficial. how Parents are likely to want few children in a modern ever, for daughters born to low-fertility mothers, since economy, in which children cannot contribute much to these daughters tend to be socialized from childhood to family income even though they cost a lot of time and value the educational and career success that the modern money to raise and educate(Aries 1996: 413; Handwerker economy and the fertility transition enable them to pur 3; Knodel et al. 1984; Oshima 1983). Daughters sue. Among my survey respondents, 32 percent(N without brothers are more likely to be encouraged to pur- 1, 215)of girls indicated that they hoped to remain child- sue advanced education and demanding careers that tend less all their lives. The fertility transition has also enabled to reduce fertility. Highly educated daughters have signifi- urban Chinese daughters to receive heavy parental invest- cant incentives to use their time to pursue prestigious and ment and remain filial all their livesan ideal that has

the experiences of their mothers and grandmothers, I show how singleton daughters have unprecedented power to deal with gender norms in ways that benefit them. Al￾though I argue that low fertility has been a key factor in the empowerment of urban Chinese daughters, I do not claim that it is the only necessary and sufficient factor. Low fertility can only empower daughters in areas where opportunities for employment and education are already available to women. In the Chinese countryside, where such opportunities remain out of reach for many women, compulsory low fertility tends to frustrate women more than it empowers them. In cities like Dalian, however, it is clear that daughters would have been less able to take ad￾vantage of available opportunities if they had to compete with brothers for family resources, and if their mothers had not demonstrated that women can support their par￾ents in old age. THEORETICAL CONSIDERATIONS Studies of many developed and developing societies worldwide have documented a high correlation between low fertility and women’s empowerment (Abadian 1996; Balk 1997; Davis 1986; Dharmalingam and Morgan 1996; Keyfitz 1986; Sathar 1988). Although these studies have focused on low fertility as a cause and effect of mothers’ empowerment, my findings suggest that more attention should be paid to how low fertility affects daughters. The effects of China’s one-child policy on mothers are equivo￾cal. On one hand, it has freed mothers from heavy child￾bearing and child-rearing burdens; on the other hand, it has deprived mothers of the freedom to choose their fam￾ily size and subjected them to intrusive state surveillance and enforcement tactics. The policy’s effects on urban daughters, however, are largely beneficial. Low resistance to the one-child policy in cities like Dalian can be attributed to the rapid pace with which people in such cities have internalized the same cultural model of modernization that has caused fertility decline in many societies worldwide. A society’s fertility rate usually corre￾lates with the degree to which it has adopted a modern economy in which child mortality is low, most people live in urban environments in which children consume a lot more than they produce, most mothers as well as fathers work at jobs incompatible with childrearing, and exten￾sive education is widespread for both genders and seen as the road to socioeconomic success. All of these factors are likely to be both causes and effects of low fertility. Parents are likely to want few children in a modern economy, in which children cannot contribute much to family income even though they cost a lot of time and money to raise and educate (Aries 1996:413; Handwerker 1986:3; Knodel et al. 1984; Oshima 1983). Daughters without brothers are more likely to be encouraged to pur￾sue advanced education and demanding careers that tend to reduce fertility. Highly educated daughters have signifi￾cant incentives to use their time to pursue prestigious and well-compensated work rather than using it to bear and rear large numbers of children. Fertility is especially low when most women are expected to work at jobs incompat￾ible with childrearing. A high rate of female employment is one of the strongest correlates of low fertility (Burggraf 1997; Essock-Vitale and McGuire 1988:229, 233; Felmlee 1993; Gerson 1985; Sander 1990; Weinberg 1976). School￾ing is also likely to cause women to learn childrearing practices that reduce infant mortality and, thus, reduce the need to have large numbers of children, as Robert A. Levine and his coauthors found in a 1983 study of Mexi￾can mothers’ education and childcare practices (LeVine et al. 1991). Michelle Zimbalist Rosaldo attributed gender inequal￾ity to a universal “opposition between the ‘domestic’ ori￾entation of women and the extra-domestic or ‘public’ ties that, in most societies, are primarily available to men” (Rosaldo 1974:17–18). The public sphere offers greater possibilities for empowerment because, unlike the domes￾tic sphere, it involves formal roles, rights, and duties; the authority that comes from lack of intimacy; the opportu￾nity to achieve rather than be ascribed status; the power to create “culture”; the tendency to be categorized as the “norm” rather than the “anomaly”; and control over the production of goods with greater cultural value (Rosaldo 1974:25–35). This theory was later criticized by Rosaldo herself (1980) as well as by other feminist anthropologists for relying on dichotomies that do not exist in all societies (Collier and Yanagisako 1987; MacCormack and Strathern 1980). Still, Rosaldo’s argument is useful for under￾standing gender systems in societies like China, in which gender inequality has long been based on distinctions be￾tween a superior public sphere dominated by men and a subordinate domestic sphere associated with women. In such societies, the adoption of a modern economy tends to increase women’s employment rates and parental bias against daughters tends to decrease when daughters are seen as capable of earning money. This pattern was documented in late 1980s Taiwan (Stafford 1995) and in 1970s–1980s India (Kishor 1993; Murthi et al. 1995; Rosenweig and Schultz 1982). When accompanied by modernization, the fertility transition enables and com￾pels women to devote themselves to work and education rather than motherhood. This is not always beneficial to the first generation of women to experience the fertility transition, since they tend to have been socialized to de￾sire large numbers of children and may suffer when they cannot realize this desire. It is much more beneficial, how￾ever, for daughters born to low-fertility mothers, since these daughters tend to be socialized from childhood to value the educational and career success that the modern economy and the fertility transition enable them to pur￾sue. Among my survey respondents, 32 percent (N = 1,215) of girls indicated that they hoped to remain child￾less all their lives. The fertility transition has also enabled urban Chinese daughters to receive heavy parental invest￾ment and remain filial all their lives—an ideal that has Fong • China’s One-Child Policy 1099

1100 American Anthropologist VoL 104, No 4. December 2002 long been valued by Chinese people of both genders but the experiences of the highest-ranked university graduates was usually only attainable by men who dominate intellectual discourse, or of the rural citi zens who constituted 64 percent( China Population Infor METHODS AND REPRESENTATIONS mation and Research Center 2001)of the Chinese popula Ding Na is one of the students I tutored in English during tion in 2000 two years of fieldwork (1997, 1998-2000) conducted in Dalian,a large coastal city(1999 urban population: THE ONE-CHILD POLICY 1, 977, 214)in Liaoning Province, northeastern China. To The primary aim of China's one-child policy is not to em- learn about the experiences of singletons, I conducted par- power women but, rather, to promote modernization by ticipant observation in a junior high school, a vocational reducing the number of people who must compete for re- high school, a college prep high school, and the homes of sources, both in the family and the nation. while the goal 107 families that invited me to tutor their children in Eng- of emancipating women from the burdens of high fertility lish or provide information about going abroad. I estab- was prominent in campaigns to promote the use of cor lished long-term relationships with 31 of these families traceptive technology during the 1950s and 1960s, gov- and participated in their social lives, leisure time, and eve- ernment propaganda promoting the one-child policy that ryday activities. I also conducted a survey of 2, 273 stu- began in 1979 tended to mention women's empowerment dents at the schools I studied. Only two of the 31 families only as an auxiliary benefit of the policy (White 1994) I befriended had more than one child. Only six percent of Contraceptive technology has enjoyed official approval in my survey respondents(n= 2, 167) had siblings the People's Republic of China since 1954, although it did The schools where I conducted my survey enrolled not become widely available until 1962. Family planning students from a wide variety of socioeconomic backgrounds, was voluntary until 1970, when Premier Zhou Enlai initi- although the most disadvantaged teenagers(such as those ated a population control campaign with paramount who were disabled or lacked urban citizenship)and the leader Mao Zedongs blessing. This campaign encouraged most elite teenagers(who were more likely to attend pi families to have no more than two children but it was un- vate schools, keypoint high schools, or study abroad pro- evenly enforced Strictly enforced fertility limitation be- grams)were underrepresented. Because of the midlevel gan in 1978, when government officials set a population statuses of the schools I studied, my survey results seem target of no more than 1.2 billion people by the year 2000, unlikely to deviate too far from the norms that might be and decided that a nationwide one-child policy was the found by a census or random sample of Dalian teenagers. only way to avoid exceeding this target (Liu Zheng 1981 Dalian's educational system divided high schools into six Peng Xizhe 1991). Despite widespread rural resistance that ranks of prestige. The nonkeypoint college prep high led to a de facto two-child policy in the countryside school I studied belonged in the second most prestigious ( Greenhalgh 1994a), China had close to its target popula category, and the vocational high school I studied be- tion in 2000, when a nationwide census counted a popula- longed in the fifth most prestigious category. The junior tion of 1. 27 billion( Chu 2001). In 1970, when population gh school I studied had the widest range of achievement control policies began, China's total fertility rate was six levels and socioeconomic statuses, since it admitted all births per woman; in 1980, two years after the start of the primary school graduates in its district without consider- one-child policy, Chinas total fertility rate was down to ing their exam scores or ability to pay. Almost all Dalian two births per woman( Coale and Chen 1987: Whyte and teenagers attended primary and junior high school and Gu 1987: 473). Farmers had higher fertility than urban most went on to secondary education as well(Dalian Shi residents even before the one-child policy, and two-child Jiaoyu Zhi Bian Zuan Bangongshi [ Dalian City Education families are the norm in rural areas, where farmers'over Records Compilation Office] 1999: 219-221, 394-426) whelming desire for sons who can serve as labor resources The tutoring and information I provided was only and old-age insurance has made the one-child policy diffi- useful to those who believed they had some chance of get- cult to enforce( Greenhalgh 1990, 1994a; Greenhalgh et ting high school or college degrees, going abroad, or get- al. 1994; White 1987, 2000). In urban areas, however, the ting work that required English skills. I suspect that most vast majority of women who married after 1978 have only urban singletons held this belief, since 94 percent (N one child. Compliance with the policy has remained high 2, 192) of survey respondents indicated that they were tu- in cities like Dalian even during the 1990s, when the costs tored or took private classes at some point in their lives, of violating the policy were reduced by rising incomes and and I seldom heard of urban singletons who thought they the decline of the state sector and its surveillance and en had no possible chance of upward mobility. Still, I cannot forcement mechanisms claim to have known families from all areas of Chinas so- Much of the literature on China's one-child policy has cioeconomic pyramid. Like my survey sample, my ethno- emphasized that compulsory fertility limitation harms wo- raphic sample does not include youth from the narrow, men. American opponents of China's one-child policy have extremely elite top or the wide, impoverished, rural bot- focused on abuses associated with the policy, such as cad- tom of that pyramid. My findings are not representative of res killing babies or physically forcing women to undergo

long been valued by Chinese people of both genders but was usually only attainable by men. METHODS AND REPRESENTATIONS Ding Na is one of the students I tutored in English during two years of fieldwork (1997, 1998–2000) conducted in Dalian, a large coastal city (1999 urban population: 1,977,214) in Liaoning Province, northeastern China. To learn about the experiences of singletons, I conducted par￾ticipant observation in a junior high school, a vocational high school, a college prep high school, and the homes of 107 families that invited me to tutor their children in Eng￾lish or provide information about going abroad. I estab￾lished long-term relationships with 31 of these families and participated in their social lives, leisure time, and eve￾ryday activities. I also conducted a survey of 2,273 stu￾dents at the schools I studied.2 Only two of the 31 families I befriended had more than one child. Only six percent of my survey respondents (N = 2,167) had siblings. The schools where I conducted my survey enrolled students from a wide variety of socioeconomic backgrounds, although the most disadvantaged teenagers (such as those who were disabled or lacked urban citizenship) and the most elite teenagers (who were more likely to attend pri￾vate schools, keypoint high schools, or study abroad pro￾grams) were underrepresented. Because of the midlevel statuses of the schools I studied, my survey results seem unlikely to deviate too far from the norms that might be found by a census or random sample of Dalian teenagers. Dalian’s educational system divided high schools into six ranks of prestige. The nonkeypoint college prep high school I studied belonged in the second most prestigious category, and the vocational high school I studied be￾longed in the fifth most prestigious category. The junior high school I studied had the widest range of achievement levels and socioeconomic statuses, since it admitted all primary school graduates in its district without consider￾ing their exam scores or ability to pay. Almost all Dalian teenagers attended primary and junior high school and most went on to secondary education as well (Dalian Shi Jiaoyu Zhi Bian Zuan Bangongshi [Dalian City Education Records Compilation Office] 1999:219–221, 394–426). The tutoring and information I provided was only useful to those who believed they had some chance of get￾ting high school or college degrees, going abroad, or get￾ting work that required English skills. I suspect that most urban singletons held this belief, since 94 percent (N = 2,192) of survey respondents indicated that they were tu￾tored or took private classes at some point in their lives, and I seldom heard of urban singletons who thought they had no possible chance of upward mobility. Still, I cannot claim to have known families from all areas of China’s so￾cioeconomic pyramid. Like my survey sample, my ethno￾graphic sample does not include youth from the narrow, extremely elite top or the wide, impoverished, rural bot￾tom of that pyramid. My findings are not representative of the experiences of the highest-ranked university graduates who dominate intellectual discourse, or of the rural citi￾zens who constituted 64 percent (China Population Infor￾mation and Research Center 2001) of the Chinese popula￾tion in 2000. THE ONE-CHILD POLICY The primary aim of China’s one-child policy is not to em￾power women but, rather, to promote modernization by reducing the number of people who must compete for re￾sources, both in the family and the nation. While the goal of emancipating women from the burdens of high fertility was prominent in campaigns to promote the use of con￾traceptive technology during the 1950s and 1960s, gov￾ernment propaganda promoting the one-child policy that began in 1979 tended to mention women’s empowerment only as an auxiliary benefit of the policy (White 1994). Contraceptive technology has enjoyed official approval in the People’s Republic of China since 1954, although it did not become widely available until 1962. Family planning was voluntary until 1970, when Premier Zhou Enlai initi￾ated a population control campaign with paramount leader Mao Zedong’s blessing. This campaign encouraged families to have no more than two children, but it was un￾evenly enforced. Strictly enforced fertility limitation be￾gan in 1978, when government officials set a population target of no more than 1.2 billion people by the year 2000, and decided that a nationwide one-child policy was the only way to avoid exceeding this target (Liu Zheng 1981; Peng Xizhe 1991). Despite widespread rural resistance that led to a de facto two-child policy in the countryside (Greenhalgh 1994a), China had close to its target popula￾tion in 2000, when a nationwide census counted a popula￾tion of 1.27 billion (Chu 2001). In 1970, when population control policies began, China’s total fertility rate was six births per woman; in 1980, two years after the start of the one-child policy, China’s total fertility rate was down to two births per woman (Coale and Chen 1987; Whyte and Gu 1987:473). Farmers had higher fertility than urban residents even before the one-child policy, and two-child families are the norm in rural areas, where farmers’ over￾whelming desire for sons who can serve as labor resources and old-age insurance has made the one-child policy diffi￾cult to enforce (Greenhalgh 1990, 1994a; Greenhalgh et al. 1994; White 1987, 2000). In urban areas, however, the vast majority of women who married after 1978 have only one child. Compliance with the policy has remained high in cities like Dalian even during the 1990s, when the costs of violating the policy were reduced by rising incomes and the decline of the state sector and its surveillance and en￾forcement mechanisms. Much of the literature on China’s one-child policy has emphasized that compulsory fertility limitation harms wo￾men. American opponents of China’s one-child policy have focused on abuses associated with the policy, such as cad￾res killing babies or physically forcing women to undergo 1100 American Anthropologist • Vol. 104, No. 4 • December 2002

Fong China's One-Child Policy 1101 sterilizations or abortions (Aird 1990; Mosher 1993) Earlier studies attribute much of the male dominance Scholarly studies of the one-child policy have focused on in Chinese societies to parents' preferential treatment of the suffering of women who long for additional children; sons over daughters( Greenhalgh 1985a, 1994b; Salaff are blamed by husbands and parents-in-law for giving 1995; Wolf 1968, 1972). My students' grandparents told birth to singleton daughters instead of sons; and face sur- me that, in their youth, daughters could not live with reillance, gynecological exams, mandatory contraception, their parents after marriage or provide nursing care or eco- fines, and the loss of benefits or jobs(Anagnost 1988, nomic support for their elderly parents. A sig 1995;Greenhalgh and Li 1995; Kaufman 1993; Wolf stacle to equality between daughters and sons in previous 1985). Even some Chinese feminists have expressed alarm generations was the assumption that daughters would not at the problems the one-child policy has caused for moth be able to support their parents in old age. Because of this ers(Greenhalgh 2001). Demographers have found an in- assumption, parents avoided investing family resources in creasingly skewed Chinese gender ratio, which may result daughters from female infanticide, parents' refusal to register daugh- Because most of my students' grandmothers lacked ters, parents' abandonment or lethal neglect of daughters, the financial resources to support their own parents, they sex selection through selective abortion, or some combi- could not contest the cultural expectation that daughters nation of these factors(Arnold and Liu Zhaoxiang 1986; would be less filial than sons. As early as the 1920s, lead Coale and Banister 1994; Johnson 1996; Li Yongping and g Chinese feminists of both genders have advocated Peng Xizhe 2000; Zeng et al. 1993) paid work as a key to womens emancipation (Lan and Fong 1999). Motivated both by feminist ideals and by a I do not discount the suffering caused by the one- child policy. However, i think a balanced view of the ef- desire to mobilize womens labor for national development, the Communist government began providing women fects of the policy must also take into account the ways in with employment opportunities soon after it took control which the low fertility produced by that policy has em of China in 1949. Yet many of my students' grandmothers powered urban daughters. As scholars working in other told me that they were too busy bearing and rearing chil- Chinese cities have pointed out(Gates 1993; Milwertz dren to take advantage of these opportunities. According 1997), the one-child policy seldom results in extreme acts to the high school and junior high school students I sur- of enforcement or resistance in urban areas, in which de- veyed in 1999, 81 percent(N= 1,998)of their fathers and sire for high fertility is far less intense than in rural areas 2 percent (N= 2,006)of their mothers had at least three While medical techniques for detecting the sex of fetuses siblings who survived infancy. "I got up at dawn, and by have been available (although illegal) in Dalian since the the time I had shopped, cooked, cleaned, and sewed clothes 1980s, the mothers of boys I tutored denied ever having for my five children, the sun would be down, "a grand sed abortion to avoid having daughters and maintained mother told me. Who would have done these things if I that only farmers would do this. Among my survey re- had gotten a job? Grandmothers were far more likely spondents, boys' parents were indeed more likely than than their husbands or children to have remained unem- girls'parents to have lived in rural areas. While I heard ployed all their lives rumors about farmers committing infanticide, physically The maternity leaves and medical problems caused by forcing women to undergo sterilizations or abortions, or frequent childbearing also hindered the careers of those abandoning or lethally neglecting daughters, I never women who did paid work during the 1950s and 1960s. "I heard of such abuses occurring in Dalian. Most Dalian par got to work upstairs in the factory office because I had ents I knew told me that it was acceptable to have just one gone to school, but I couldn't take a position of responsi- child, even if that child was female, and some even told bility because I always had to take time off when I got they were glad they had daughters instead of sons. pregnant, a grandmother told me. "After my fourth child They knew from their own experience that daughters my health was bad all the time, and I had to quit my job could fulfill the filial obligations once reserved for sons Grandmothers were far less likely than their husbands or Unlike their rural counterparts, my female students' par- children to work as cadres, managers, or white-collar workers ents were not desperate to have sons at any cost. at any point in their lives Many scholars writing about women's status in China THE LEGACY OF LOW-FERTILITY MOTHERS (Honig and Hershatter 1988; Stacey 1983; Wolf 1985)and elsewhere (Goldman 1993: Hochschild and Machung My students' mothers were able to begin the transforma- 1989 Molyneux 1985; Randall 1992; Steil 1995: Stockman tion of their societys kinship system from a patrilineal, et al. 1995)have argued that working women are bur- patrilateral, and patrilocal one to a bilineal, bilateral, and dened by having to work both a"first shift"of paid work neolocal one. This was at least partly because of the paid and a"second shift"of housework. Yet the single shift of work their low fertility enabled them to do. Paid work en- housework that a housewife did to take care of numerous abled women to provide their own parents with financial children in the 1950s and 1960s seems at least as exhaust- support in old age and, thus, prove that daughters could ing and time consuming as the combined first and second be as filial as sons hints of an employed mother who only had to take care of

sterilizations or abortions (Aird 1990; Mosher 1993). Scholarly studies of the one-child policy have focused on the suffering of women who long for additional children; are blamed by husbands and parents-in-law for giving birth to singleton daughters instead of sons; and face sur￾veillance, gynecological exams, mandatory contraception, fines, and the loss of benefits or jobs (Anagnost 1988, 1995; Greenhalgh and Li 1995; Kaufman 1993; Wolf 1985). Even some Chinese feminists have expressed alarm at the problems the one-child policy has caused for moth￾ers (Greenhalgh 2001). Demographers have found an in￾creasingly skewed Chinese gender ratio, which may result from female infanticide, parents’ refusal to register daugh￾ters, parents’ abandonment or lethal neglect of daughters, sex selection through selective abortion, or some combi￾nation of these factors (Arnold and Liu Zhaoxiang 1986; Coale and Banister 1994; Johnson 1996; Li Yongping and Peng Xizhe 2000; Zeng et al. 1993). I do not discount the suffering caused by the one￾child policy. However, I think a balanced view of the ef￾fects of the policy must also take into account the ways in which the low fertility produced by that policy has em￾powered urban daughters. As scholars working in other Chinese cities have pointed out (Gates 1993; Milwertz 1997), the one-child policy seldom results in extreme acts of enforcement or resistance in urban areas, in which de￾sire for high fertility is far less intense than in rural areas. While medical techniques for detecting the sex of fetuses have been available (although illegal) in Dalian since the 1980s, the mothers of boys I tutored denied ever having used abortion to avoid having daughters and maintained that only farmers would do this. Among my survey re￾spondents, boys’ parents were indeed more likely than girls’ parents to have lived in rural areas.3 While I heard rumors about farmers committing infanticide, physically forcing women to undergo sterilizations or abortions, or abandoning or lethally neglecting daughters, I never heard of such abuses occurring in Dalian. Most Dalian par￾ents I knew told me that it was acceptable to have just one child, even if that child was female, and some even told me they were glad they had daughters instead of sons. They knew from their own experience that daughters could fulfill the filial obligations once reserved for sons. Unlike their rural counterparts, my female students’ par￾ents were not desperate to have sons at any cost. THE LEGACY OF LOW-FERTILITY MOTHERS My students’ mothers were able to begin the transforma￾tion of their society’s kinship system from a patrilineal, patrilateral, and patrilocal one to a bilineal, bilateral, and neolocal one. This was at least partly because of the paid work their low fertility enabled them to do. Paid work en￾abled women to provide their own parents with financial support in old age and, thus, prove that daughters could be as filial as sons. Earlier studies attribute much of the male dominance in Chinese societies to parents’ preferential treatment of sons over daughters (Greenhalgh 1985a, 1994b; Salaff 1995; Wolf 1968, 1972). My students’ grandparents told me that, in their youth, daughters could not live with their parents after marriage or provide nursing care or eco￾nomic support for their elderly parents. A significant ob￾stacle to equality between daughters and sons in previous generations was the assumption that daughters would not be able to support their parents in old age. Because of this assumption, parents avoided investing family resources in daughters. Because most of my students’ grandmothers lacked the financial resources to support their own parents, they could not contest the cultural expectation that daughters would be less filial than sons. As early as the 1920s, lead￾ing Chinese feminists of both genders have advocated paid work as a key to women’s emancipation (Lan and Fong 1999). Motivated both by feminist ideals and by a desire to mobilize women’s labor for national development, the Communist government began providing women with employment opportunities soon after it took control of China in 1949. Yet many of my students’ grandmothers told me that they were too busy bearing and rearing chil￾dren to take advantage of these opportunities. According to the high school and junior high school students I sur￾veyed in 1999, 81 percent (N = 1,998) of their fathers and 82 percent (N = 2,006) of their mothers had at least three siblings who survived infancy. “I got up at dawn, and by the time I had shopped, cooked, cleaned, and sewed clothes for my five children, the sun would be down,” a grand￾mother told me. “Who would have done these things if I had gotten a job?” Grandmothers were far more likely than their husbands or children to have remained unem￾ployed all their lives.4 The maternity leaves and medical problems caused by frequent childbearing also hindered the careers of those women who did paid work during the 1950s and 1960s. “I got to work upstairs in the factory office because I had gone to school, but I couldn’t take a position of responsi￾bility because I always had to take time off when I got pregnant,” a grandmother told me. “After my fourth child, my health was bad all the time, and I had to quit my job.” Grandmothers were far less likely than their husbands or children to work as cadres, managers, or white-collar workers at any point in their lives.5 Many scholars writing about women’s status in China (Honig and Hershatter 1988; Stacey 1983; Wolf 1985) and elsewhere (Goldman 1993; Hochschild and Machung 1989; Molyneux 1985; Randall 1992; Steil 1995; Stockman et al. 1995) have argued that working women are bur￾dened by having to work both a “first shift” of paid work and a “second shift” of housework. Yet the single shift of housework that a housewife did to take care of numerous children in the 1950s and 1960s seems at least as exhaust￾ing and time consuming as the combined first and second shifts of an employed mother who only had to take care of Fong • China’s One-Child Policy 1101

1102 American Anthropologist VoL 104, No 4. December 2002 one child in the 1980s and 1990s. Both generations of and let wealthier siblings pick up the slack. Because most women worked all day, every day. The main difference is men earned more than most women, these wealthier sib that employed mothers had part of their work valorized lings were more likely to be brothers rather than sisters with monthly wages, which constantly reminded them as Still, my students'mothers had at least proven daughters well as their husbands, parents, and parents-in-law of were capable of providing financial support for their par their power and indispensability ents. This reassured my students' parents that their daugh Because of the Chinese government's policy of assign ters could have the same capability, especially if they were ing apartments that were too small to accommodate joint given the resources to take full advantage of socioeco families, most urban Chinese people have "networked nomic opportunities families"(Davis and Harrell 1993; Unger 1993), in which married children live neolocally but in close proximity to DEALING STRATEGICALLY WITH GENDER NORMS oth sets of parents. Only 17 percent of my survey respon- The strategy of raising a brotherless daughter to fill the dents(N=2,188)indicated that at least one grandparent kinship role usually reserved for sons was occasionally lived in their home. Neolocality allows couples consider practiced even in prerevolutionary China ( ordan 1972 able flexibility in the negotiation of relationships with 91-92: Pasternak 1985; Rofel 1999: 80-94). The approp both sets of parents. In the flexible kinship system en- ateness of such a strateg oo proclaimed by legends like joyed by urban families, paid work gave women the lever that of Mulan, a girl who took her fathers place in the age they needed to maintain ties to their own parents. As a army because he had no son old enough to do so. As a junior high school student's mother told her husband when he complained that she was giving too much money rare and difficult last resort, the strategy of raising a to her parents, Why shouldnt I give them the money daughter as a son"(guniang dang erzi yang) had little influ I' ve earned? You should be grateful that I don't give all my ence on dominant chinese cultural models or the scholars wages to them! who studied them. This strategy gained popularity after Elderly parents who were widowed or disabled usually the one-child policy made it a necessity for half of my stu- moved into an adult childs household. Which child they dents’ families. ended up living with depended less on gender than on in- Parents whose love, hope, and need for old-age sup terpersonal dynamics and on the amount of time and liv. port are all pinned on just one child tend to do whatever is necessary to make that child happy and successful, re- families, elderly parents rotated between all their children, gardless of the child's gender. Daughters and their parents staying a few weeks to a few months in the household of face the extra challenge of winning happiness and success each son or daughter. Regardless of their gender, adult in a society structured by gender norms that have long dis- children tended to contribute as much in care, compan advantaged women. They meet this challenge with a stra- ionship, money, and gifts to their parents as they could af. tegic combination of conformity and resistance ford. Many of my students' mothers provided monetary For academically unsuccessful daughters of poor par- support and nursing care for their own elderly parents ents, gender norms provide a means of upward mobility (often getting their husbands to help), most who per- through marriage and job markets unavailable to their formed annual worship rituals for their husbands' de- male counterparts. Women face a glass ceiling produced ceased parents also did so for their own deceased parents by their extra burden of domestic responsibility, by gender and some inherited money, goods, and housing from their norms that favor men in elite professions, and by in- parents. While 12 percent of my survey respondents(N= equalities between elite husbands and their less elite, hy 2, 187)were living with at least one paternal grandparent pergamous wives. Women also enjoy the protection of a at the time of the survey, 5 percent(N=2, 188)were living glass floor created by the hypergamous marriage system, with at least one maternal grandparent. Because of my stu- by gender norms that favor nonelite women in the educa dents'mothers'success in diverting resources to their own tional system, and by the rapidly expanding market for parents, my students'families accept that daughters can feminine jobs in the service and light industry sectors be as filial as sons This glass floor makes it less likely that women will sink to My students'mothers were not able to completely the bottom of society, into poverty, crime, and unemploy obliterate patrilineal assumptions Because women tended ment. Men have neither the obstacle of the glass ceiling to earn less than men, they also tended to contribute less nor the protection of the glass floor. While elite men are to their parents than their brothers could. This became es- more likely than their female counterparts to rise to the pecially apparent in the 1990s, after the economic reforms top of their society, nonelite men are also more likely than caused layoffs and early retirements that disproportion- their female counterparts to fall to the bottom. ately targeted women. According to survey respondents, 25 percent of their mothers(N= 2, 190)and 12 percent of ple's expectations of how males and females would be- their fathers(N= 2, 190)have been laid off or given early have. They were not interested in debating the extent to retirement. Men and women who lost their jobs tended to which such expectations corresponded with the way peo- reduce the financial support they provided their parents ple actually behaved. Rather, they focused on weighing

one child in the 1980s and 1990s. Both generations of women worked all day, every day. The main difference is that employed mothers had part of their work valorized with monthly wages, which constantly reminded them as well as their husbands, parents, and parents-in-law of their power and indispensability. Because of the Chinese government’s policy of assign￾ing apartments that were too small to accommodate joint families, most urban Chinese people have “networked families” (Davis and Harrell 1993; Unger 1993), in which married children live neolocally but in close proximity to both sets of parents. Only 17 percent of my survey respon￾dents (N = 2,188) indicated that at least one grandparent lived in their home. Neolocality allows couples consider￾able flexibility in the negotiation of relationships with both sets of parents. In the flexible kinship system en￾joyed by urban families, paid work gave women the lever￾age they needed to maintain ties to their own parents. As a junior high school student’s mother told her husband when he complained that she was giving too much money to her parents, “Why shouldn’t I give them the money I’ve earned? You should be grateful that I don’t give all my wages to them!” Elderly parents who were widowed or disabled usually moved into an adult child’s household. Which child they ended up living with depended less on gender than on in￾terpersonal dynamics and on the amount of time and liv￾ing space each child’s household could spare. In many families, elderly parents rotated between all their children, staying a few weeks to a few months in the household of each son or daughter. Regardless of their gender, adult children tended to contribute as much in care, compan￾ionship, money, and gifts to their parents as they could af￾ford. Many of my students’ mothers provided monetary support and nursing care for their own elderly parents (often getting their husbands to help), most who per￾formed annual worship rituals for their husbands’ de￾ceased parents also did so for their own deceased parents, and some inherited money, goods, and housing from their parents. While 12 percent of my survey respondents (N = 2,187) were living with at least one paternal grandparent at the time of the survey, 5 percent (N = 2,188) were living with at least one maternal grandparent. Because of my stu￾dents’ mothers’ success in diverting resources to their own parents, my students’ families accept that daughters can be as filial as sons. My students’ mothers were not able to completely obliterate patrilineal assumptions. Because women tended to earn less than men, they also tended to contribute less to their parents than their brothers could. This became es￾pecially apparent in the 1990s, after the economic reforms caused layoffs and early retirements that disproportion￾ately targeted women. According to survey respondents, 25 percent of their mothers (N = 2,190) and 12 percent of their fathers (N = 2,190) have been laid off or given early retirement. Men and women who lost their jobs tended to reduce the financial support they provided their parents and let wealthier siblings pick up the slack. Because most men earned more than most women, these wealthier sib￾lings were more likely to be brothers rather than sisters. Still, my students’ mothers had at least proven daughters were capable of providing financial support for their par￾ents. This reassured my students’ parents that their daugh￾ters could have the same capability, especially if they were given the resources to take full advantage of socioeco￾nomic opportunities. DEALING STRATEGICALLY WITH GENDER NORMS The strategy of raising a brotherless daughter to fill the kinship role usually reserved for sons was occasionally practiced even in prerevolutionary China (Jordan 1972: 91–92; Pasternak 1985; Rofel 1999:80–94). The appropri￾ateness of such a strategy was proclaimed by legends like that of Mulan, a girl who took her father’s place in the army because he had no son old enough to do so.6 As a rare and difficult last resort, the strategy of “raising a daughter as a son” (guniang dang erzi yang) had little influ￾ence on dominant Chinese cultural models or the scholars who studied them. This strategy gained popularity after the one-child policy made it a necessity for half of my stu￾dents’ families. Parents whose love, hope, and need for old-age sup￾port are all pinned on just one child tend to do whatever is necessary to make that child happy and successful, re￾gardless of the child’s gender. Daughters and their parents face the extra challenge of winning happiness and success in a society structured by gender norms that have long dis￾advantaged women. They meet this challenge with a stra￾tegic combination of conformity and resistance. For academically unsuccessful daughters of poor par￾ents, gender norms provide a means of upward mobility through marriage and job markets unavailable to their male counterparts. Women face a glass ceiling produced by their extra burden of domestic responsibility, by gender norms that favor men in elite professions, and by in￾equalities between elite husbands and their less elite, hy￾pergamous wives. Women also enjoy the protection of a glass floor created by the hypergamous marriage system, by gender norms that favor nonelite women in the educa￾tional system, and by the rapidly expanding market for feminine jobs in the service and light industry sectors. This glass floor makes it less likely that women will sink to the bottom of society, into poverty, crime, and unemploy￾ment. Men have neither the obstacle of the glass ceiling nor the protection of the glass floor. While elite men are more likely than their female counterparts to rise to the top of their society, nonelite men are also more likely than their female counterparts to fall to the bottom. My students and their parents often talked about peo￾ple’s expectations of how males and females would be￾have. They were not interested in debating the extent to which such expectations corresponded with the way peo￾ple actually behaved. Rather, they focused on weighing 1102 American Anthropologist • Vol. 104, No. 4 • December 2002

Fong China's One-Child Policy 1103 the costs and benefits of disregarding invoking, trans- do work that would fund their brothers'education(Green forming, or conforming to particular expectations on par- haigh 1985b, 1994b; Lan and Fong 1999: Wolf 1968, ticular occasions I translate these expectations as "gender 1972). Brotherless daughters, however, are encouraged to norms."Although powerful, these norms are recognized, make full use of their academic talents because they are talked about, and open to challenge. They are thus compa- their parents' only objects of investment, and only hope rable to what Pierre Bourdieu called"orthodoxy, "a system for old-age support of acceptable ways of thinking and speaking the natural In the educational systems of Britain, Canada, the and social world, which rejects heretical remarks as blas- United States, Belgium, Morocco, and Algeria, girls from phemies"(1977: 169). Unlike doxa, which Bourdieu defined stigmatized minority backgrounds have tended to outper as the"self-evident and natural order which goes without form their male counterparts, who are more likely to rebel saying and therefore goes unquestioned"(1977: 166), or- against school discipline(which is identified with their thodoxy is defined in opposition to heterodoxy and, thus, ethnic oppressors)(Gibson 1997). Although they were not unable to conceal its own arbitrariness ethnic minorities, economically disadvantaged teenagers Judith Butler proposed that the task of feminism is"to in Dalian experienced a similar phenomenon. Girls at the repeat and, through a radical proliferation of gender, to schools I studied tended to have higher overall scores than displace the very gender norms that enable the repetition boys. This advantage, however, was balanced out by elite itself"(1990: 148). By parodying norms and performing schools'emphasis on math and science(which boys fa the possibilities they exclude or condemn, Butler argued, vored) over the social sciences and humanities(which people can liberate themselves and others from the con- girls favored). High school entrance exams tested students traints of these norms. Because they have the full support on more science and math subjects than humanities and of their parents, singleton daughters have unprecedented social science subjects, and four-year colleges accepted freedom to engage in this kind of play. At the same time, more science and math majors than humanities and socia however, their freedom is limited by a socioeconomic sys- science majors. These factors constituted a significant bias tem that remains structured by class and gender inequ against girls at the highest levels of academic achieve- ies. While more elite women might have the wherewithal ment, but not at the middle and lower levels where the to seek the total liberation that Butler proposed, my majority of students found themselves. mostly nonelite students and their parents find that they Gender norms structure Dalians job market, but not must choose their battles. Therefore, they do not try to always to women's disadvantage and men's advantage eradicate all gender norms. Rather, they only try to do Rather, they work in favor of younger women and aca away with ones most likely to hurt their own interests, demically unsuccessful women from lower-class families such as those that portray daughters as less filial and les even as they work against older women, elite women, and worthy of parental investment than sons. At the same poor, academically unsuccessful men. time, they conform to other gender norms, such as those Stereotypically feminine traits are seen as ideal for most jobs in light industry and the service sector. Stereo- that portray women as more patient and meticulous than typically masculine traits are seen as ideal for most jobs in men, when they feel that such norms may further their in- terests. They seek happiness and success, not liberation status professions open only to a tiny elite. This means er se. While previous generations have also done this, that elite women are less likely to get elite work than their daughters born after the one-child policy have more famil- male counterparts, but also that nonelite women are more ial support for their strategies than ever before EDUCATION AND WORK parts. Daughters are therefore counseled both to conform to gender norms that can give them an advantage in the Parents of daughters as well as sons believe that success in general job market and to disregard those that might ex education and work will be the key determinant of their clude them from elite professional work. children's(and, thus, their own) future happiness. Like Women are rare in the most prestigious and best-paid sons,daughters are their parents'only hope for the future. professions, partly because they are hindered by their"sec- I have never heard of any Dalian daughter's parents want- ond shift"of domestic work and partly because of many ing her to become a housewife with no paid work. While a employers belief that women do not have enough daring woman has the option of relying on her husband's income, and creativity to do elite work. Focusing on biases against it is not as desirable as having an income of her own. older women and elite women, recent studies have argued Girls who conform to gender norms are more studious that post-Mao economic reforms have intensified dis- and obedient than their male counterparts and, thus, crimination against women( Croll 1995; Honig and Her- more successful in the educational system at all levels be- shatter 1988; Hooper 1998; Kerr et al. 1996; Summerfield sides the very highest. The greater studiousness of girls 1994). I found, however, that the consequences of those was of limited use in previous generations because parent eforms are more complicated for the majority of youths, were reluctant to spend money on daughters' education who are of average or below-average education and family and sometimes even made daughters drop out of school background. The same economic reforms that encourage

the costs and benefits of disregarding, invoking, trans￾forming, or conforming to particular expectations on par￾ticular occasions. I translate these expectations as “gender norms.” Although powerful, these norms are recognized, talked about, and open to challenge. They are thus compa￾rable to what Pierre Bourdieu called “orthodoxy,” a system “of acceptable ways of thinking and speaking the natural and social world, which rejects heretical remarks as blas￾phemies” (1977:169). Unlike doxa, which Bourdieu defined as the “self-evident and natural order which goes without saying and therefore goes unquestioned” (1977:166), or￾thodoxy is defined in opposition to heterodoxy and, thus, unable to conceal its own arbitrariness. Judith Butler proposed that the task of feminism is “to repeat and, through a radical proliferation of gender, to displace the very gender norms that enable the repetition itself” (1990:148). By parodying norms and performing the possibilities they exclude or condemn, Butler argued, people can liberate themselves and others from the con￾straints of these norms. Because they have the full support of their parents, singleton daughters have unprecedented freedom to engage in this kind of play. At the same time, however, their freedom is limited by a socioeconomic sys￾tem that remains structured by class and gender inequali￾ties. While more elite women might have the wherewithal to seek the total liberation that Butler proposed, my mostly nonelite students and their parents find that they must choose their battles. Therefore, they do not try to eradicate all gender norms. Rather, they only try to do away with ones most likely to hurt their own interests, such as those that portray daughters as less filial and less worthy of parental investment than sons. At the same time, they conform to other gender norms, such as those that portray women as more patient and meticulous than men, when they feel that such norms may further their in￾terests. They seek happiness and success, not liberation per se. While previous generations have also done this, daughters born after the one-child policy have more famil￾ial support for their strategies than ever before. EDUCATION AND WORK Parents of daughters as well as sons believe that success in education and work will be the key determinant of their children’s (and, thus, their own) future happiness. Like sons, daughters are their parents’ only hope for the future. I have never heard of any Dalian daughter’s parents want￾ing her to become a housewife with no paid work. While a woman has the option of relying on her husband’s income, it is not as desirable as having an income of her own. Girls who conform to gender norms are more studious and obedient than their male counterparts and, thus, more successful in the educational system at all levels be￾sides the very highest. The greater studiousness of girls was of limited use in previous generations because parents were reluctant to spend money on daughters’ education and sometimes even made daughters drop out of school to do work that would fund their brothers’ education (Green￾halgh 1985b, 1994b; Lan and Fong 1999; Wolf 1968, 1972). Brotherless daughters, however, are encouraged to make full use of their academic talents because they are their parents’ only objects of investment, and only hope for old-age support. In the educational systems of Britain, Canada, the United States, Belgium, Morocco, and Algeria, girls from stigmatized minority backgrounds have tended to outper￾form their male counterparts, who are more likely to rebel against school discipline (which is identified with their ethnic oppressors) (Gibson 1997). Although they were not ethnic minorities, economically disadvantaged teenagers in Dalian experienced a similar phenomenon. Girls at the schools I studied tended to have higher overall scores than boys.7 This advantage, however, was balanced out by elite schools’ emphasis on math and science (which boys fa￾vored) over the social sciences and humanities (which girls favored). High school entrance exams tested students on more science and math subjects than humanities and social science subjects, and four-year colleges accepted more science and math majors than humanities and social science majors. These factors constituted a significant bias against girls at the highest levels of academic achieve￾ment, but not at the middle and lower levels where the majority of students found themselves. Gender norms structure Dalian’s job market, but not always to women’s disadvantage and men’s advantage. Rather, they work in favor of younger women and aca￾demically unsuccessful women from lower-class families even as they work against older women, elite women, and poor, academically unsuccessful men. Stereotypically feminine traits are seen as ideal for most jobs in light industry and the service sector. Stereo￾typically masculine traits are seen as ideal for most jobs in the rapidly shrinking heavy-industry sector and in high￾status professions open only to a tiny elite. This means that elite women are less likely to get elite work than their male counterparts, but also that nonelite women are more likely to avoid unemployment than their male counter￾parts. Daughters are therefore counseled both to conform to gender norms that can give them an advantage in the general job market and to disregard those that might ex￾clude them from elite professional work. Women are rare in the most prestigious and best-paid professions, partly because they are hindered by their “sec￾ond shift” of domestic work and partly because of many employers’ belief that women do not have enough daring and creativity to do elite work. Focusing on biases against older women and elite women, recent studies have argued that post-Mao economic reforms have intensified dis￾crimination against women (Croll 1995; Honig and Her￾shatter 1988; Hooper 1998; Kerr et al. 1996; Summerfield 1994). I found, however, that the consequences of those reforms are more complicated for the majority of youths, who are of average or below-average education and family background. The same economic reforms that encourage Fong • China’s One-Child Policy 1103

1104 American Anthropologist VoL 104, No 4. December 2002 state enterprises to discriminate against middle-aged women women alike, but women who fall short by those stand have also created service and light-industry jobs that favor ards can compensate with pleasant personalities, physical young women. Physical attractiveness and stereotypically attractiveness, and the ability and willingness to do house- feminine positive traits can compensate for a woman's work Men can use these qualities to compensate as well, lack of education and family connections, but the poorly but not nearly to the extent women can educated son of powerless parents is simply out of luck In the marriage market created by the one-child pol- Recognizing the midlevel job market's greater de- icy, women enjoy several advantages. As in the past, mand for female workers, the educational system admits grooms are expected to provide marital housing. The abil- more girls than boys at the high school level. Greater edu- ity to live up to this expectation remains an important de cational opportunities for girls were reflected in the mate. terminant of whether a man can win a bride. Thus, a son rials published by Dalian's Bureau of Education and given and his parents must try to buy, rent, borrow, or inherit to Dalian's graduating junior high school class of 1999. At extra housing by the time the son is ready to marry. A the technical-school level(sixth-rate), there were 1, 346 daughter and her parents, on the other hand, can consider places open to both boys and girls and 4, 492 places re- the ability to provide or contribute to the purchase of served for girls, but only 4, 301 places reserved for boys. At marital housing an extra bonus to enhance the daughter's the vocational-school level(fifth-rate), there were 2, 949 marriagiability and comfort, rather than a requirement places open to both boys and girls and 5, 189 places re- Brotherless daughters and their parents see this as an ad served for girls, but only 3, 849 places reserved for boys. vantage, rather than a sign that daughters are valued less Several all-female private college-preparatory high schools than sons. Singletons of either gender face no competition (third-rate)were established in the Dalian area during the for parental investment or inheritance. They and their 1990s, but no all-male schools. At the second-rate college parents just have to decide what form the wealth transfer prep high school I studied, 52 percent of students(N= will take. Unlike sons' parents, daughters' parents can in 781)were female and 48 percent were male. According to vest all their savings in their daughters'education, rather teachers, students, and education officials, only the small than saving part of it for the purchase of marital housing minority of schools classified as keypoint college prep The need to purchase housing to attract a spouse is thus a schools(first-rate) had more boys than girls. A study con- disadvantage for sons and their parents. This disadvantage ducted in 1990s Shanghai(an eastern coastal city with a became particularly onerous after the housing reforms of olitical economy similar to Dalian's) found that the in- 1997, which allowed work units to sell apartments on the comes of unmarried young women in Shanghai exceeded private market instead of assigning them to workers in ex- those of unmarried young men (Wang Zheng 2000: 75). change for low, subsidized rents. a male vocational high Most of the boys and about a quarter of the girls in my sur- school student told me that he could have gone to a college vey sample indicated that girls had an easier time getting prep high school if his parents, who ran a small shop, had jobs than boys did spent all their savings on extra fees and bribes that would Drawing on evidence from France and Kabylia, Pierre have gotten him in despite his low exam score. He said, Bourdieu argued that, because of the strength of symbolic modes of masculine domination " in work as in educa They gave me a choice. Either they could use their to send me to the college prep high school, or they tion, the progress made by women must not conceal tl use it to buy an apartment for me so that I'll be able corresponding progress made by men, so that, as in a a wife when the time comes. I don't like to study, and I handicap race, the structure of the gaps is maintained" didn't think I could make it to college even if I went to a (2001: 91). I found, however, that the gap between male ollege prep school, so I chose the apartment. and female statuses was much narrower in my students generation than in their parents'and grandparents' gen- while men prefer to marry women of the same status erations. Although they still face a glass ceiling perpetu- Thus, women can gain upward mobility through hyper ated by the symbolic structures of masculine domination gamous marriage, while men are often forced to choose that Bourdieu described, singleton daughters are not hin- between permanent bachelorhood and marriage to some dered by the parental discrimination that disadvantaged one of lower status. Although it produces inequality be- neir mothers and grandmothers. The removal of this dis tween husbands and wives, hypergamous marriage is in advantage has enabled singleton daughters to make th some ways more favorable to women than to men. Unlike best use of their glass floor, and in some cases push the women, men seldom gain upward mobility through mar- limits of their glass ceiling. riage. Because of a low divorce rate, parental and school prohibitions against dating among teenagers, legal and so- cial prohibitions against pregnancy outside wedlock, and a Many parents told me that girls are more fortunate than strict social prohibition against premarital sex that applies boys because girls have more paths to upward mobility. especially to women but also to men, Dalian women are Family background, career success, and educational attain- relatively protected from the feminization of poverty per- ment are important spouse selection criteria for men and vasive in societies with high rates of single motherhood. o

state enterprises to discriminate against middle-aged women have also created service and light-industry jobs that favor young women. Physical attractiveness and stereotypically feminine positive traits can compensate for a woman’s lack of education and family connections, but the poorly educated son of powerless parents is simply out of luck. Recognizing the midlevel job market’s greater de￾mand for female workers, the educational system admits more girls than boys at the high school level. Greater edu￾cational opportunities for girls were reflected in the mate￾rials published by Dalian’s Bureau of Education and given to Dalian’s graduating junior high school class of 1999. At the technical-school level (sixth-rate), there were 1,346 places open to both boys and girls and 4,492 places re￾served for girls, but only 4,301 places reserved for boys. At the vocational-school level (fifth-rate), there were 2,949 places open to both boys and girls and 5,189 places re￾served for girls, but only 3,849 places reserved for boys. Several all-female private college-preparatory high schools (third-rate) were established in the Dalian area during the 1990s, but no all-male schools. At the second-rate college prep high school I studied, 52 percent of students (N = 781) were female and 48 percent were male. According to teachers, students, and education officials, only the small minority of schools classified as keypoint college prep schools (first-rate) had more boys than girls. A study con￾ducted in 1990s Shanghai (an eastern coastal city with a political economy similar to Dalian’s) found that the in￾comes of unmarried young women in Shanghai exceeded those of unmarried young men (Wang Zheng 2000:75). Most of the boys and about a quarter of the girls in my sur￾vey sample indicated that girls had an easier time getting jobs than boys did.8 Drawing on evidence from France and Kabylia, Pierre Bourdieu argued that, because of the strength of symbolic modes of masculine domination, “in work as in educa￾tion, the progress made by women must not conceal the corresponding progress made by men, so that, as in a handicap race, the structure of the gaps is maintained” (2001:91). I found, however, that the gap between male and female statuses was much narrower in my students’ generation than in their parents’ and grandparents’ gen￾erations. Although they still face a glass ceiling perpetu￾ated by the symbolic structures of masculine domination that Bourdieu described, singleton daughters are not hin￾dered by the parental discrimination that disadvantaged their mothers and grandmothers. The removal of this dis￾advantage has enabled singleton daughters to make the best use of their glass floor, and in some cases push the limits of their glass ceiling. MARRIAGE Many parents told me that girls are more fortunate than boys because girls have more paths to upward mobility. Family background, career success, and educational attain￾ment are important spouse selection criteria for men and women alike,9 but women who fall short by those stand￾ards can compensate with pleasant personalities, physical attractiveness, and the ability and willingness to do house￾work. Men can use these qualities to compensate as well, but not nearly to the extent women can. In the marriage market created by the one-child pol￾icy, women enjoy several advantages. As in the past, grooms are expected to provide marital housing. The abil￾ity to live up to this expectation remains an important de￾terminant of whether a man can win a bride. Thus, a son and his parents must try to buy, rent, borrow, or inherit extra housing by the time the son is ready to marry. A daughter and her parents, on the other hand, can consider the ability to provide or contribute to the purchase of marital housing an extra bonus to enhance the daughter’s marriagiability and comfort, rather than a requirement. Brotherless daughters and their parents see this as an ad￾vantage, rather than a sign that daughters are valued less than sons. Singletons of either gender face no competition for parental investment or inheritance. They and their parents just have to decide what form the wealth transfer will take. Unlike sons’ parents, daughters’ parents can in￾vest all their savings in their daughters’ education, rather than saving part of it for the purchase of marital housing. The need to purchase housing to attract a spouse is thus a disadvantage for sons and their parents. This disadvantage became particularly onerous after the housing reforms of 1997, which allowed work units to sell apartments on the private market instead of assigning them to workers in ex￾change for low, subsidized rents. A male vocational high school student told me that he could have gone to a college￾prep high school if his parents, who ran a small shop, had spent all their savings on extra fees and bribes that would have gotten him in despite his low exam score. He said, They gave me a choice. Either they could use their savings to send me to the college prep high school, or they could use it to buy an apartment for me so that I’ll be able to get a wife when the time comes. I don’t like to study, and I didn’t think I could make it to college even if I went to a college prep school, so I chose the apartment. In Dalian, women prefer to marry men of higher status, while men prefer to marry women of the same status. Thus, women can gain upward mobility through hyper￾gamous marriage, while men are often forced to choose between permanent bachelorhood and marriage to some￾one of lower status. Although it produces inequality be￾tween husbands and wives, hypergamous marriage is in some ways more favorable to women than to men. Unlike women, men seldom gain upward mobility through mar￾riage. Because of a low divorce rate, parental and school prohibitions against dating among teenagers, legal and so￾cial prohibitions against pregnancy outside wedlock, and a strict social prohibition against premarital sex that applies especially to women but also to men, Dalian women are relatively protected from the feminization of poverty per￾vasive in societies with high rates of single motherhood.10 1104 American Anthropologist • Vol. 104, No. 4 • December 2002

Fong China's One-Child Policy 1105 Because of womens preference for marrying up, men their fathers did any housework at all. The percentage of have difficulty obtaining brides of similar socioeconomic female respondents who indicated that they wanted to do status as themselves. I often heard boys and their parents more housework than their husbands was a lot smaller complain about how difficult it will be to find brides. This than the percentage of respondents who indicated that is partly because of China's skewed gender ratio. In addi- their mothers did more housework than their fathers tion, young Dalian men fear that there is a shortage of ur- When I asked boys who indicated that they wanted to do ban women willing to marry men of equal status. Women more housework than their wives why they chose that re- already at the top of Dalian's socioeconomic hierarchy as- sponse, some said they would have to do a lot of house pire to marry even higher-status men from wealthier cities work to win and keep wives, since they are not likely to or foreign countries. Dalian men who cannot find local get good jobs or neolocal housing in time for their mar brides can acquire brides from the countryside, where riages. Others who already had girlfriends pointed out that women are eager to gain urban residency through marriage. their girlfriends are unlikely to do much housework after Most urban men, however, consider this unacceptable. marriage. As one college prep high school student said, Matthew Kohrman(1999) found that even a disabled Bei- " my girlfriend is too lazy to even buy her own snacks, so I ng man considered permanent bachelorhood or marriage have to run down to the shop and get them for her during to a disabled urban woman preferable to marriage to a lunch. How can I expect her to do even her fair share of nondisabled rural woman, because he felt that as an urban the housework? an he could never feel "real love" toward a rural woman Boys and girls alike recognize that greater gender some of my male students likewise told me that they would equality in the distribution of housework is expected for prefer lifelong bachelorhood to marriage with"country their generation than for previous generations. As a junior bumpkins high school student replied when her mother, a retired factory worker, asked what she will do after marriage if she CHANGING DOMESTIC ROLES never learns to cook, "My husband will cook! Who says In my students' parents'generation, men are expected to women have to be the ones to cook earn more, have better jobs, and do less housework than heir wives, who are expected to take primary responsibil- CONCLUSION ity for domestic work, usually at the expense of their ca- The benefits enjoyed by singleton daughters result from reers. Survey respondents mothers were far more likely the demographic pattern produced by Chinas one-child than their fathers to do household chores. Still, my stu- policy, and not necessarily from the compulsory nature of lents'fathers were far more likely to help with housework that policy. Global processes of industrialization, modern than my students' grandfathers, most of whom told me ization, and urbanization have led to low fertility in all de that they did no housework at all. In the families of a few veloped countries and many developing countries world students I tutored, husbands did even more work than wide. These processes would probably have caused a their wives. This was particularly likely when the mother fertility transition in cities like Dalian even without a one worked or earned more than the father. The mother of the child policy. Such a transition would have occurred more family I lived with cheerfully did all the housework while slowly, and produced fewer brotherless daughters, than she and her husband both worked at nine-to-five jobs. the transition mandated by the one-child policy. Still, Things changed when she rented a fruit stall, where she even a daughter with one brother is likely to enjoy more sold fruit from 8: 30 a. m. to 7: 30 p. m, seven days a week, resources than a daughter with several brothers while his factory increasingly sent him home with no Singleton daughters deal with gender norms in ways work and no pay and eventually laid him off. Suddenly, that seem likely to further their own interests. People of she was making more money and doing more hours of every generation have tried to use gender norms to attain paid work than he. Although she took pride in being a their own desires(whether they involved socioeconomic good wife and virtuous mother, "she realized that her success or the maintenance of strong ties to ones parents time had become a lot more valuable than her husband's but the efforts of Chinese daughters born prior to the one time and started pressuring him to do more housework. child policy were severely hindered by a patrilineal system He reluctantly agreed, and from then on had dinner wait- that overwhelmingly favored sons at the expense of their ing for her when she got back home at 8 p.m sisters. In contrast, urban singleton daughters enjoy un g 4e Survey respondents expect the division of domestic precedented support for their effort to challenge norms in their own marriages to be more egalitarian than in that work against them while utilizing those that work in heir parents'marriages by their their favor. When daughters are not systematically ex sponses to a question about how much housework they cluded from familial resources, norms that once went nt to do after marriage. The percentage of male re- hand in hand with patriarchy become tools that girls can ondents who indicated that they wanted to do at least use as well as boy half the housework after marriage was somewhat higher Chinas social structure is still characterized by than the percentage of respondents who indicated that inequality, particularly at the upper levels of the academic

Because of women’s preference for marrying up, men have difficulty obtaining brides of similar socioeconomic status as themselves. I often heard boys and their parents complain about how difficult it will be to find brides. This is partly because of China’s skewed gender ratio.11 In addi￾tion, young Dalian men fear that there is a shortage of ur￾ban women willing to marry men of equal status. Women already at the top of Dalian’s socioeconomic hierarchy as￾pire to marry even higher-status men from wealthier cities or foreign countries. Dalian men who cannot find local brides can acquire brides from the countryside, where women are eager to gain urban residency through marriage. Most urban men, however, consider this unacceptable. Matthew Kohrman (1999) found that even a disabled Bei￾jing man considered permanent bachelorhood or marriage to a disabled urban woman preferable to marriage to a nondisabled rural woman, because he felt that as an urban man he could never feel “real love” toward a rural woman. Some of my male students likewise told me that they would prefer lifelong bachelorhood to marriage with “country bumpkins.” CHANGING DOMESTIC ROLES In my students’ parents’ generation, men are expected to earn more, have better jobs, and do less housework than their wives, who are expected to take primary responsibil￾ity for domestic work, usually at the expense of their ca￾reers. Survey respondents’ mothers were far more likely than their fathers to do household chores.12 Still, my stu￾dents’ fathers were far more likely to help with housework than my students’ grandfathers, most of whom told me that they did no housework at all. In the families of a few students I tutored, husbands did even more work than their wives. This was particularly likely when the mother worked or earned more than the father. The mother of the family I lived with cheerfully did all the housework while she and her husband both worked at nine-to-five jobs. Things changed when she rented a fruit stall, where she sold fruit from 8:30 a.m. to 7:30 p.m., seven days a week, while his factory increasingly sent him home with no work and no pay and eventually laid him off. Suddenly, she was making more money and doing more hours of paid work than he. Although she took pride in being a “good wife and virtuous mother,” she realized that her time had become a lot more valuable than her husband’s time and started pressuring him to do more housework. He reluctantly agreed, and from then on had dinner wait￾ing for her when she got back home at 8 p.m. Survey respondents expect the division of domestic work in their own marriages to be more egalitarian than in their parents’ marriages, as demonstrated by their re￾sponses to a question about how much housework they want to do after marriage.13 The percentage of male re￾spondents who indicated that they wanted to do at least half the housework after marriage was somewhat higher than the percentage of respondents who indicated that their fathers did any housework at all. The percentage of female respondents who indicated that they wanted to do more housework than their husbands was a lot smaller than the percentage of respondents who indicated that their mothers did more housework than their fathers. When I asked boys who indicated that they wanted to do more housework than their wives why they chose that re￾sponse, some said they would have to do a lot of house￾work to win and keep wives, since they are not likely to get good jobs or neolocal housing in time for their mar￾riages. Others who already had girlfriends pointed out that their girlfriends are unlikely to do much housework after marriage. As one college prep high school student said, “my girlfriend is too lazy to even buy her own snacks, so I have to run down to the shop and get them for her during lunch. How can I expect her to do even her fair share of the housework?” Boys and girls alike recognize that greater gender equality in the distribution of housework is expected for their generation than for previous generations. As a junior high school student replied when her mother, a retired factory worker, asked what she will do after marriage if she never learns to cook, “My husband will cook! Who says women have to be the ones to cook?” CONCLUSION The benefits enjoyed by singleton daughters result from the demographic pattern produced by China’s one-child policy, and not necessarily from the compulsory nature of that policy. Global processes of industrialization, modern￾ization, and urbanization have led to low fertility in all de￾veloped countries and many developing countries world￾wide. These processes would probably have caused a fertility transition in cities like Dalian even without a one￾child policy. Such a transition would have occurred more slowly, and produced fewer brotherless daughters, than the transition mandated by the one-child policy. Still, even a daughter with one brother is likely to enjoy more resources than a daughter with several brothers. Singleton daughters deal with gender norms in ways that seem likely to further their own interests. People of every generation have tried to use gender norms to attain their own desires (whether they involved socioeconomic success or the maintenance of strong ties to one’s parents), but the efforts of Chinese daughters born prior to the one￾child policy were severely hindered by a patrilineal system that overwhelmingly favored sons at the expense of their sisters. In contrast, urban singleton daughters enjoy un￾precedented support for their effort to challenge norms that work against them while utilizing those that work in their favor. When daughters are not systematically ex￾cluded from familial resources, norms that once went hand in hand with patriarchy become tools that girls can use as well as boys. China’s social structure is still characterized by gender inequality, particularly at the upper levels of the academic Fong • China’s One-Child Policy 1105

1106 American Anthropologist Vol 104, No 4. December 2002 and socioeconomic hierarchies. But brotherless daughters aartiteular Nicole s ons aale. sue ong at that pres have the power to make the best use of their glass floor Nonini,Ichiro Numizaki,and Jesook Shu-Min Huang, Don am also grateful to Re- and push the limits of their glass ceiling, thanks to the pa- gina ahr me tarqui Barak andu sai au andel spang ac De denied. Daughters empowered by the support of parents Graeme Lang, Tianshu Pan, Sonja Plesset, Jinbao Qian, Erica James with no sons to favor are able to defy detrimental norms Razafimbahiny, Tam Tai, James L. Watson, and Rubie Watson, for while strategically using ones that give them advantages am grateful to Fran Mascia-Lees, Susan H. Lees, TaraJ. Pearson, and in the educational system and the job and marriage mar- an anonymous reviewer for the American Anthropologist for their kets. Parents of singletons only complain about their chil dren's gender when they believe their children are con- cle was funded by a Beinecke Brothers Memorial Fellowship, an Andrew w. Mellon Grant for Predissertation Research and a na. forming to disadvantageous gender norms, not when their tional Science Foundation Graduate Research Fellowship children are conforming to advantageous ones. Complaints 1. The degree to which any given gender norm is detrimental or like those of Ding Na's father can thus be seen as discursive helpful depends on the characteristics of the individual: The norm strategies adopted on specific occasions to exhort a beloved associating women with gentleness, for instance, might help a vo child to challenge detrimental gender norms. cational high school graduate win a secretarial iots but prevent a unemployment that plagues her male espite his moment of epiphany on learning his female college graduate from becoming a manager(since managers daughters excellent college entrance exam scores, Ding are not supposed to be gentle Na's father continues to remind her that he always wanted 2. This survey was administered in 1999 to most of the students in grades 8-9 at the jur high school, grades 10-ll at the voca a son whenever he finds fault with her. He continues to tional high school, and grades 10-12 at the college prep high fret about her future, demanding that she succeed in col chool,753 were from the vocational high school, and 782 were lege and worrying that she will not be able to find a good from the college prep high school. The junior high school and col b afterward, especially since she has chosen to major in lege prep high school had balanced gender ratios, while the voca omputer programming, a subject considered difficult for cialized in female-d women. Yet Ding Na takes her fathers commentaries in ted majors such as business and tourism gs down by gender or school only whe stride. "He criticizes me only because he wants to push me dealing with survey responses that vary significantly by gender or to do better, "she told me. Indeed, I noticed that, though school he criticized her in her presence, he also boasted about her 3. On my survey, 38 percent of female respondents (N in her absence and 29 percent of male respondents(N=852)indicated th had at least one parent who had not lived in the countryside While Ding Na was away at college, I had dinner with her parents and paternal uncle, whose singleton daughter dents indicated that both parents had lived in the countryside. was still in high school. Ding Na's father's brother talked 4. Percent of relatives who never did paid work, according to sur- bout his fear that his own daughter would not succeed in vey respondents: paternal grandmother (N= 1, 716)36 percent maternal grandmother(N= 1, 493)34 percent, paternal grandfa- the science major she had chosen, since "science is harder ther(N= 1, 651)0 percent, maternal gra er(N=1,748)0per for girls. "Ding Na's father, however, reassured him by cent, mother(N= 1,995)0 percent, father(N= 1,964)0 percent quoting a famous line from "The Red Detachment of Wo- 5. Percent of relatives who were ever cadres, managers, or white- men"(Hongse Niangzijun), a revolutionary model opera about Communist women soldiers. "In ancient times there 15 percent, paternal grandfather(N=1, 863)42 percent, maternal was Hua Mulan: in modern times there is the red detach father(N= 1, 984)48 percent. ment of Women. "He then raised his glass of beer in a toast 6. The Mulan ballad is said to have originated in the fifth or sixth and added, "in the future it will be up to our daughters. when it was included in the prestigious Song dynasty(960-1279 C E )anthology Yuefu(Luo Genze 1996). The legend of Mulan has VANESSA L. FoNG Postdoctoral Fellow, Population Studies inspired many novels, plays, and poems, and even a Disney ani- Center, University of Michigan at Ann Arbor, 48104 ated movie. Mulan Girls' High School is the name of an expe sive private girls schools established in Dalian during the 1990s 7. Average percentile ranks of singleton respondents by gender NOTES school: girls grades 8-9(N=361), 57th percentile; boys grade students, parents, and tea shared their lives with mty 325), 42nd percentile; girls in college prep hig The earliest version of this presented at the City Univer- chool, grades 10-12(N= 201), 44th percentile. Percentile March 13. 2000. I thank audience members are based comparisons of the January 2000 final exam scores of pho commented on my presentation, especially Graeme Lang all students in each grade level. The best possible percentile rank is read and commented on a later version Later versio 100, and the worst possible is 1 this article won honorable mentions for the 2001 Elsie Clews par. 8. Survey Question: "Who has a harder time getting jobs-male Society and the 2001 Sylvia Forman Graduate Student Paper Prize percent respo=1, 181): 33 percent responded"females, "26 tales, "41 percent responded"no difference warded by the Association for Feminist Anthropology. Another version of this article was presented at a joint conference of the sponded"males, "32 percent responded"no difference. American Ethnological Society, the Canadian Anthropology Soci- 9. Although important in young people's choice of marriage part. y, and the Society for Cultural Anthropology, held at McGill Uni- ners in 1990s urban China, romantic love is still sometimes ou versity, Montreal, Canada, May 3-6, 2001.I weighed by socioeconomic factors

and socioeconomic hierarchies. But brotherless daughters have the power to make the best use of their glass floor and push the limits of their glass ceiling, thanks to the pa￾rental support that their mothers and grandmothers were denied. Daughters empowered by the support of parents with no sons to favor are able to defy detrimental norms while strategically using ones that give them advantages in the educational system and the job and marriage mar￾kets. Parents of singletons only complain about their chil￾dren’s gender when they believe their children are con￾forming to disadvantageous gender norms, not when their children are conforming to advantageous ones. Complaints like those of Ding Na’s father can thus be seen as discursive strategies adopted on specific occasions to exhort a beloved child to challenge detrimental gender norms. Despite his moment of epiphany on learning his daughter’s excellent college entrance exam scores, Ding Na’s father continues to remind her that he always wanted a son whenever he finds fault with her. He continues to fret about her future, demanding that she succeed in col￾lege and worrying that she will not be able to find a good job afterward, especially since she has chosen to major in computer programming, a subject considered difficult for women. Yet Ding Na takes her father’s commentaries in stride. “He criticizes me only because he wants to push me to do better,” she told me. Indeed, I noticed that, though he criticized her in her presence, he also boasted about her in her absence. While Ding Na was away at college, I had dinner with her parents and paternal uncle, whose singleton daughter was still in high school. Ding Na’s father’s brother talked about his fear that his own daughter would not succeed in the science major she had chosen, since “science is harder for girls.” Ding Na’s father, however, reassured him by quoting a famous line from “The Red Detachment of Wo￾men” (Hongse Niangzijun), a revolutionary model opera about Communist women soldiers: “In ancient times there was Hua Mulan; in modern times there is the Red Detach￾ment of Women.” He then raised his glass of beer in a toast and added, “in the future it will be up to our daughters.” VANESSA L. FONG Postdoctoral Fellow, Population Studies Center, University of Michigan at Ann Arbor, 48104 NOTES Acknowledgments. I am most deeply indebted to the Dalian City students, parents, and teachers who shared their lives with me. The earliest version of this article was presented at the City Univer￾sity of Hong Kong on March 13, 2000. I thank audience members who commented on my presentation, especially Graeme Lang, who also read and commented on a later version. Later versions of this article won honorable mentions for the 2001 Elsie Clews Par￾sons Student Paper Prize awarded by the American Ethnological Society and the 2001 Sylvia Forman Graduate Student Paper Prize awarded by the Association for Feminist Anthropology. Another version of this article was presented at a joint conference of the American Ethnological Society, the Canadian Anthropology Soci￾ety, and the Society for Cultural Anthropology, held at McGill Uni￾versity, Montreal, Canada, May 3–6, 2001. I thank audience members and fellow panelists who gave me suggestions at that presentation, particularly Nicole Constable, Suhong Chae, Shu-Min Huang, Don Nonini, Ichiro Numizaki, and Jesook Song. I am also grateful to Re￾gina Abrami, Narquis Barak, Manduhai Buyandel, Shanghan Du Deborah Gewertz, Miriam Goheen, Susan Greenhalgh, Michael Herzfeld, Xiaojiang Hu, William Jankowiak, Arthur Kleinman, Graeme Lang, Tianshu Pan, Sonja Plesset, Jinbao Qian, Erica James Razafimbahiny, Tam Tai, James L. Watson, and Rubie Watson, for the suggestions they gave me about various drafts of this article. I am grateful to Fran Mascia-Lees, Susan H. Lees, Tara J. Pearson, and an anonymous reviewer for the American Anthropologist for their careful readings and detailed suggestions. The research for this arti￾cle was funded by a Beinecke Brothers Memorial Fellowship, an Andrew W. Mellon Grant for Predissertation Research, and a Na￾tional Science Foundation Graduate Research Fellowship. 1. The degree to which any given gender norm is detrimental or helpful depends on the characteristics of the individual: The norm associating women with gentleness, for instance, might help a vo￾cational high school graduate win a secretarial job and avoid the unemployment that plagues her male counterparts but prevent a female college graduate from becoming a manager (since managers are not supposed to be gentle). 2. This survey was administered in 1999 to most of the students in grades 8–9 at the junior high school, grades 10–11 at the voca￾tional high school, and grades 10–12 at the college prep high school. Of the 2,273 respondents, 738 were from the junior high school, 753 were from the vocational high school, and 782 were from the college prep high school. The junior high school and col￾lege prep high school had balanced gender ratios, while the voca￾tional high school was 71 percent (N = 752) female because it spe￾cialized in female-dominated majors such as business and tourism. I break my statistical findings down by gender or school only when dealing with survey responses that vary significantly by gender or school. 3. On my survey, 38 percent of female respondents (N = 1,254) and 29 percent of male respondents (N = 852) indicated that they had at least one parent who had not lived in the countryside, while 62 percent of female respondents and 71 percent of male respon￾dents indicated that both parents had lived in the countryside. 4. Percent of relatives who never did paid work, according to sur￾vey respondents: paternal grandmother (N = 1,716) 36 percent, maternal grandmother (N = 1,493) 34 percent, paternal grandfa￾ther (N = 1,651) 0 percent, maternal grandfather (N = 1,748) 0 per￾cent, mother (N = 1,995) 0 percent, father (N = 1,964) 0 percent. 5. Percent of relatives who were ever cadres, managers, or white￾collar workers, according to survey respondents: paternal grand￾mother (N = 1,871) 14 percent, maternal grandmother (N = 1,875) 15 percent, paternal grandfather (N = 1,863) 42 percent, maternal grandfather (N = 1,871) 45 percent, mother (N = 2,009) 38 percent, father (N = 1,984) 48 percent. 6. The Mulan ballad is said to have originated in the fifth or sixth century C.E., and it became part of the official Chinese canon when it was included in the prestigious Song dynasty (960–1279 C.E.) anthology Yuefu (Luo Genze 1996). The legend of Mulan has inspired many novels, plays, and poems, and even a Disney ani￾mated movie. Mulan Girls’ High School is the name of an expen￾sive private girls’ schools established in Dalian during the 1990s. 7. Average percentile ranks of singleton respondents by gender and school: girls grades 8–9 (N = 361), 57th percentile; boys grades 8–9 (N = 325), 42nd percentile; girls in college prep high school, grades 10–12 (N = 262), 54th percentile; boys in college prep high school, grades 10–12 (N = 201), 44th percentile. Percentile ranks are based on comparisons of the January 2000 final exam scores of all students in each grade level. The best possible percentile rank is 100, and the worst possible is 1. 8. Survey Question: “Who has a harder time getting jobs—males or females?” Girls (N = 1,181): 33 percent responded “females,” 26 percent responded “males,” 41 percent responded “no difference.” Boys (N = 788): 16 percent responded “females,” 53 percent re￾sponded “males,” 32 percent responded “no difference.” 9. Although important in young people’s choice of marriage part￾ners in 1990s urban China, romantic love is still sometimes out￾weighed by socioeconomic factors. 1106 American Anthropologist • Vol. 104, No. 4 • December 2002

Fong China's One-Child Policy 1107 10. Most of my students live in two-parent homes. Among my sur. Coale, Ansley J. and Judith Banister ey respondents, 91 percent (N= 2, 188)indicated that they were 1994 Five Decades of Missing Females in China. Demography living with both their parents). Childbearing outside of wedlock is 31(3):459479 legal, socially scandalous, and almost nonexistent in Dalian Coale, Ansley J, and Chen ShengLi Most unmarried women who get pregnant have abortions, whic 1987 Basic Data on Fertility in the Provinces of China, 1940-82. re readily available and less stigmatizing than the alternatives. In Honolulu: The East-West Population Institute 987, less than one percent of women were estimated to have re- Collier, Jane Fishburne, and Sylvia Junko Yanagisako mained single through age 50( Zeng Yi 2000: 93) 1987 Gender and Kinship: Essays toward a Unified Analysis. Stan- ford: Stanford University Press. 11. China's gender ratio imbalance has increased steadily since Croll, Elisabeth J. 1995 Chan 1990 census, there were 1.083 males for every female born be. nce, and Self-Perception in Twentieth-Century China Hong tween 1980 and 1984( Coale and Banister 1994: 461), the period ong: Hong Kong University Press. when most of my students were born. According to Chinas 1995 Dalian Shi jiaoyu Zhi Bian Zuan Bangongshi census, there were 1.17 males for every female born in 1995(Li Yongping and Peng Xizhe 2000: 71). o tion l slian Jiaoyu Yaolan 1997-1998(a Survey of Dalian Educa- 97-1998). Dalian: Dalian Shi Jiaoyu Zhi Bian Zuan Ban- Percentages of mothers and fathers who did various kinds of gshi (Dalian City Education Records household chores, according rey respondents: 94 percent of Davis, Deborah, and Stevan harrell nothers(N= 2, 198)and 41 percent of fathers(N= 2, 199)cleaned 94 percent of mothers(N= 2, 195)and 42 percent of fathers(N Life In Chinese families in the post.Mao era. D. Davis and s har. rell, eds. Pp. 1-24. Berkeley: University of California Press. cent of fathers(N=2,196) shopped for groceries; and 88 percent of Davis, k ingsley 1986 Wives and Work: The Sex role Revolution and Its Conse mothers(N= 2, 194)and 59 percent of fathers(N= 2, 194)cooked. quences. Population and Development Review 10(3): 397-418 13. Percentages of respondents (girls (N=1, 159), boys [N=839) Dharmalingam, A, and S. Philip Morgan 1996 Womens Work, Autonomy and Birth Control: Evidence and 17 percent of boys want to do more housework than their from Two South Indian Villages. Population Studies 50(2) mouse: 63 percent of girls and 48 percent of boys want to do half the housework; 12 percent of girls and 35 percent of boys want to Essock- Vitale. Susan M. and MichaelT McGuire 1988 What 70 Million Years Hath Wrought: Sexual Histories and productive Success of a Random Sample of American Womer In Human Reproductive Behaviour: A Darwinian Perspective. L. REFERENCES CITED L Betzig, M. B. Mulder, and P. Turke, eds. Pp 221-235. Cam Abadan, Sousan bridge: Cambridge University Press 1996 Women's Autonomy and Its Impact on Fertility. World De Felmlee, Diane h 793-1809 1993 The DynamicInterdependence of womens Employment Aird, John and Fertility. Social Science Research 22(4): 333-360 1990 Slaughter of the Innocents: Coercive Birth Controlin China. Gates, Hill Washington, DC: AEI Press. 1993 Cultural Support for Birth Limitation among Urban Capital Anagnost, Ann Owning Women. In Chinese Families in the Post-Mao Era. D 1988 Family Violence and Magical Violence: The"Woman-as-vic Davis and s Harrell, eds. Pp 251-274. Berkeley: University of tim"in Chinas One-Child Family Policy. Women and Language California press 1(2):16-22 1995 ASurfeit of Bodies: Population and the Rationality of the 1985 Hard Choices: How Women Decide about work, Career, and tate in Post-Mao China In Conceiving the New World Order Motherhood. Berkeley: University of California Press. The Global Politics of Reproduction. F D. Ginsburg and R Rapp, ibson, Margaret ds. Pp 22-41. Berkeley: University of California Press 1997 Complicating the Immigrant/Involuntary Minority Typo- Andor, Phyllis 1983 The Unfinished Liberation of Chinese women, 1949-1980 Goldman, Wendy Z. Bloomington: Indiana University Press 993 Women, the State and Revolution. Cambridge: Cambridge Aries, Philip 1996 Centuries of Childhood. London: pimlico Greenhalgh, Susan Arnold, Fred, and Liu Zhaoxiang 1985a Is Inequality Demographically Induced? The Family Cycle 1986 Sex Preference, Fertility, and Family Planning in China Population and Development Review 12(2): 221-246. 985b Sexual Stratification: The Other Side of"Growth with Eq- 1997 Defying Gender Norms in Rural Bangladesh: A Social Demo- uity. "Population and Development Review 11: 265-314 1990 The Evolution of the One-Child Policy in Shaanxi, 1979-88. graphic Analysis. Population Studies 51(2): 153-172 The China Quarterly 122(une): 191-229 994a Controlling Births and Bodiesin Village China. American 1977 Outline of a Theory of Practice. R Nice, trans. Cambridge: Ethnologist 21(1): 3-30. ambridge University Press. 1994b De-Orientalizing the Chinese Family Firm. American Eth- 001 Masculine Domination. Stanford: Stanford University Press. logist2l(4):746-775 Burggraf, Shirley P. 1997 The Feminine Economy and Economic Man: Reviving the 2001 Fresh Winds in Beijing: Chinese Feminists Speak Out on the One-Child Policy and Womens Lives. Signs 26(3): 847-886 Role of Family in the Post-Industrial Age. Reading, MA: Addison- Greenhalgh, Susan, and Jiali 1995 Engendering Reproductive Policy and Practice in Peasant China: For a Feminist Demography of Reproduction. Signs 1990 Gender Trouble: Feminism and the Subversion of Identity Greenhalgh, Susan, Li Nan, and Zhu Huzhu China Population Information and Research Center 1994 Restraining Population Growth in Three Chinese village 2001 Major Figures of the 2000 Population Census. National Bu- 1988-93. Population and Development Review 20(2): 365-39 reau of Statistics, Peoples Republic of China Handwerker W. Penn Chu, Henry 1986 Modern Demographic Transition: An Analysis of Subsis- 2001 India Joins China as Member of the Billion-Population Club. tence Choices and Reproductive Consequences. American An Los Angeles Times, March 29: 9 thropologist 88(2): 400-417

10. Most of my students live in two-parent homes. Among my sur￾vey respondents, 91 percent (N = 2,188) indicated that they were living with both their parents). Childbearing outside of wedlock is illegal, socially scandalous, and almost nonexistent in Dalian. Most unmarried women who get pregnant have abortions, which are readily available and less stigmatizing than the alternatives. In 1987, less than one percent of women were estimated to have re￾mained single through age 50 (Zeng Yi 2000:93). 11. China’s gender ratio imbalance has increased steadily since the implementation of the one-child policy. According to China’s 1990 census, there were 1.083 males for every female born be￾tween 1980 and 1984 (Coale and Banister 1994:461), the period when most of my students were born. According to China’s 1995 census, there were 1.17 males for every female born in 1995 (Li Yongping and Peng Xizhe 2000:71). 12. Percentages of mothers and fathers who did various kinds of household chores, according to survey respondents: 94 percent of mothers (N = 2,198) and 41 percent of fathers (N = 2,199) cleaned; 94 percent of mothers (N = 2,195) and 42 percent of fathers (N = 2,194) did laundry; 94 percent of mothers (N = 2,196) and 54 per￾cent of fathers (N = 2,196) shopped for groceries; and 88 percent of mothers (N = 2,194) and 59 percent of fathers (N = 2,194) cooked. 13. Percentages of respondents (girls [N = 1,159], boys [N = 839]) who want to do various amounts of housework: 25 percent of girls and 17 percent of boys want to do more housework than their spouse; 63 percent of girls and 48 percent of boys want to do half the housework; 12 percent of girls and 35 percent of boys want to do less housework than their spouse. REFERENCES CITED Abadian, Sousan 1996 Women’s Autonomy and Its Impact on Fertility. WorldDe￾velopment 24(12):1793–1809. Aird, John S. 1990 Slaughter of the Innocents: Coercive Birth Control in China. Washington, DC: AEI Press. Anagnost, Ann 1988 FamilyViolence and Magical Violence: The “Woman-as-Vic￾tim”inChina’s One-ChildFamily Policy. Women and Language 1(2):16–22. 1995 A Surfeit of Bodies: Population and the Rationality of the State in Post-Mao China.In Conceiving the New WorldOrder: The Global Politics of Reproduction. F. D. Ginsburg and R. Rapp, eds. Pp. 22–41. Berkeley: University of California Press. Andors, Phyllis 1983 The Unfinished Liberation of ChineseWomen, 1949–1980. Bloomington: Indiana University Press. Aries, Philippe 1996 Centuries of Childhood. London: Pimlico. Arnold, Fred, and Liu Zhaoxiang 1986 Sex Preference, Fertility, and FamilyPlanning in China. Population and Development Review 12(2):221–246. Balk, Deborah 1997 Defying Gender Norms inRural Bangladesh: A SocialDemo￾graphic Analysis. Population Studies 51(2):153–172. Bourdieu, Pierre 1977 Outline of a Theory of Practice. R. Nice, trans. Cambridge: Cambridge University Press. 2001 Masculine Domination. Stanford: Stanford University Press. Burggraf, Shirley P. 1997 The Feminine Economy and EconomicMan: Reviving the Roleof Familyin the Post-Industrial Age. Reading, MA: Addison￾Wesley. Butler, Judith P. 1990 Gender Trouble : Feminismand the Subversion of Identity. NewYork: Routledge. China Population Information and Research Center 2001 Major Figures of the 2000 Population Census. National Bu￾reau of Statistics, People’s Republic of China. Chu, Henry 2001 IndiaJoins China as Member of the Billion-Population Club. Los AngelesTimes, March 29: 9. Coale, Ansley J., and Judith Banister 1994 Five Decades of Missing Females in China.Demography 31(3):459–479. Coale, Ansley J., and Chen ShengLi 1987 Basic Data on Fertility inthe Provinces of China, 1940–82. Honolulu: The East-WestPopulation Institute. Collier, Jane Fishburne, and Sylvia Junko Yanagisako 1987 Gender and Kinship: Essays toward a Unified Analysis. Stan￾ford: Stanford University Press. Croll, Elisabeth J. 1995 Changing Identities of Chinese Women: Rhetoric, Experi￾ence, and Self-Perception in Twentieth-Century China. Hong Kong: Hong Kong University Press. Dalian Shi Jiaoyu Zhi Bian Zuan Bangongshi 1999 Dalian Jiaoyu Yaolan 1997–1998 (a Survey of DalianEduca￾tion 1997–1998). Dalian: Dalian Shi Jiaoyu Zhi Bian Zuan Ban￾gongshi (DalianCity Education Records Compilation Office). Davis, Deborah, and Stevan Harrell 1993 Introduction: The Impact of Post-Mao Reforms on Family Life.In ChineseFamilies in the Post-Mao Era.D. Davisand S. Har￾rell,eds. Pp. 1–24. Berkeley: University of CaliforniaPress. Davis, Kingsley 1986 Wives and Work: The Sex Role Revolution and Its Conse￾quences. Population and Development Review 10(3):397–418. Dharmalingam, A., and S. Philip Morgan 1996 Women’s Work, Autonomy and Birth Control: Evidence from Two SouthIndian Villages. Population Studies50(2): 187–202. Essock-Vitale, Susan M., and Michael T. McGuire 1988 What70 Million Years HathWrought: SexualHistories and Reproductive Success of a Random Sampleof American Women. InHuman Reproductive Behaviour: A DarwinianPerspective. L. L. Betzig, M. B.Mulder, and P. Turke, eds. Pp. 221–235. Cam￾bridge: Cambridge University Press. Felmlee, Diane H. 1993 The DynamicInterdependence of Women’s Employment and Fertility. Social Science Research 22(4):333–360. Gates, Hill 1993 Cultural Support for Birth Limitation among UrbanCapital￾Owning Women.In Chinese Familiesinthe Post-Mao Era. D. Davis and S. Harrell, eds. Pp. 251–274. Berkeley: University of CaliforniaPress. Gerson, Kathleen 1985 Hard Choices: How Women Decide about Work, Career, and Motherhood. Berkeley: University of California Press. Gibson, Margaret 1997 Complicating the Immigrant/Involuntary Minority Typo￾logy. Anthropology and Education Quarterly28(3):431–454. Goldman, Wendy Z. 1993 Women, the State and Revolution. Cambridge: Cambridge University Press. Greenhalgh, Susan 1985a Is Inequality DemographicallyInduced? The FamilyCycle and the Distribution of Income inTaiwan. American Anthropolo￾gist 87(3):571–594. 1985b Sexual Stratification: The Other Side of “Growth with Eq￾uity.” Population and Development Review 11:265–314. 1990 The Evolution of the One-Child Policy inShaanxi, 1979–88. The China Quarterly 122(June):191–229. 1994a Controlling Births and BodiesinVillage China. American Ethnologist21(1):3–30. 1994b De-Orientalizing the ChineseFamilyFirm. American Eth￾nologist 21(4):746–775. 2001 Fresh Winds in Beijing: Chinese Feminists SpeakOut on the One-ChildPolicy and Women’s Lives. Signs 26(3):847–886. Greenhalgh, Susan, and Jiali Li 1995 Engendering Reproductive Policy and Practice in Peasant China: For a Feminist Demography of Reproduction. Signs 20(3):601–642. Greenhalgh, Susan, Li Nan, and Zhu Chuzhu 1994 Restraining Population Growth inThree Chinese Villages, 1988–93. Population and Development Review20(2):365–396. Handwerker, W. Penn 1986 Modern DemographicTransition: An Analysisof Subsis￾tence Choicesand Reproductive Consequences. American An￾thropologist 88(2):400–417. Fong • China’s One-Child Policy 1107