Preferred Citation: Davis, Deborah, and Stevan Harrell, editors. Chinese Families in the Post-Mao Era. Berkeley:  University of California Press,  c1993 1993. http://ark.cdlib.org/ark:/13030/ft3q2nb257/

Chinese Families in the Post-Mao Era Edited by Deborah Davis and Stevan Harrell UNIVERSITY OF CALIFORNIA PRESS Berkeley · Los Angeles · Oxford © 1993 The Regents of the University of California

Preferred Citation: Davis, Deborah, and Stevan Harrell, editors. Chinese Families in the Post-Mao Era. Berkeley:  University of California Press,  c1993 1993. http://ark.cdlib.org/ark:/13030/ft3q2nb257/

ACKNOWLEDGMENTS

The chapters in this volume were first presented at a conference on "Family Strategies in Post-Mao China," held at Roche Harbor, Washington, June 12-17, 1990. Financial support for the conference and subsequent editorial expenses were provided by a grant from the Joint Committee on Chinese Studies of the American Council of Learned Societies and the Social Science Research Council. In addition to our nine fellow participants who worked so speedily and intensely on paper revisions, the editors would like to acknowledge the intellectual contributions of four other conference discussants: Anthony Carter, Shumin Huang, William Lavely, and Susan Watkins, as well as conference assistant Fu Yunqi. During the several phases of revisions, we also greatly benefited from the careful reading and thoughtful criticisms of two anonymous readers and Dorothy Solinger. For help with photo reproduction, we thank Joey Liao.

One
INTRODUCTION: The Impact of Post-Mao Reforms on Family Life

Deborah Davis and Stevan Harrell

In the decade after the Communist victory in 1949, state orthodoxy created a new institutional and moral environment for Chinese families. The collectivization of the economy and the elimination of most private property destroyed much of the economic motivation that had previously shaped family loyalties, and the frontal attack on ancestor worship and lineage organization struck directly at the cultural and religious core of the extended family. It was not simply the case, however, as many predicted, that communism destroyed the traditional Chinese family. On the contrary, many key policies actually stabilized and strengthened families. For example, large investments in public health and famine relief dramatically reduced mortality; fewer infants died, more children survived to marry, old age became the norm, and people from all social classes had larger and more complex kin networks than had been possible before 1949.^[1] Similarly, the severe restrictions on internal migration not only served the interest of the government in controlling individual autonomy, but they also intensified the flow of intergenerational aid because they tied most adult men (and their sons) to the villages and towns of their birth.^[2] Thus the Communist revolution created contradictions. On the one hand, it undercut the power and authority of patriarchs and destroyed the economic logic of family farms and businesses. On the other hand, it created demographic and material conditions conducive to large, multigenerational households with extensive economic and social to nearby kin. In short, Chinese families between 1950 and 1976

[1] Edward Tu, Jersey Liang, and Shaomin Li, "Mortality Decline and Chinese Family Structure," Journal of Gerontology 44, no. 4 (July 1989): 157-68.

[2] Deborah Davis-Friedmann, Long Lives , 2d ed. (Stanford: Stanford University Press, 1991), chaps. 3-4.

survived and reproduced themselves within a paradoxical environment: the often repressive egalitarianism of communism permitted more Chinese parents and children than ever before to realize core ideals of traditional Chinese familism, while at the same time the revolution eliminated many of the original incentives for wanting to realize those ideals.

With the shift away from collectivization and toward entrepreneurship and privatization after 1978, the Maoist compromise between state policies and family self-interest entered a drastically altered environment. This was especially the case in the countryside, where the Deng reforms dismantled the People's Communes as a core political and economic unit and abruptly subcontracted more than 80 percent of all farmland to individual families on the basis of fifteen-year leases. Commune and village assets were similarly redistributed, rented, or sold to the highest bidders, and by the mid-1980s most rural families were operating in a political economy of family tenancy where the agents of state authority had less control over the labor, land, and loyalty of rural residents than at any time since the Land Reform of the early 1950s.^[3]

In the cities, the Deng reforms were far more modest and incremental. Private entrepreneurship was encouraged rhetorically, but the great majority of urban residents remained wage earners, and most households were units of consumption.^[4] Moreover, the paradoxical relationship between the Maoist state and traditional familism had never been as pronounced in the cities as it had been in the villages. Urban families had very few welfare functions, income was highly individuated, and relations with kin could be very weak. Few households had continued as units of production following the collectivization of artisan workshops and services in the mid-1950s, and generous medical and pension plans for all state employees reduced the

[3] For detailed discussion and varied interpretations of this process, see Philip C. C. Huang, The Peasant Family and Rural Development in the Yangzi Delta (Stanford: Stanford University Press, 1990), 286, 321-22; Victor Nee and Su Sijin, "Institutional Change and Economic Growth in China," Journal of Asian Studies 49, no. 1 (Feb. 1990): 3-25; Sulamith H. Potter and Jack M. Potter, China's Peasants (Cambridge: Cambridge University Press, 1990), 158-79; Louis Putterman, "Entering the Post-Collective Economy in North China," Modern China 15, no. 3 (July 1989): 275-320; Helen Siu, Agents and Victims in South China (New Haven: Yale University Press, 1989), 273-90; Vivienne Shue, The Reach of the State (Stanford: Stanford University Press, 1988), esp. 148-52; Terry Sicular, "Rural Marketing and Exchange," in The Political Economy of Reform in Post-Mao China , ed. Elizabeth Perry and Christine Wong (Cambridge: Harvard University Press, 1985), 83-110.

[4] Deborah Davis, "Urban Job Mobility," in Chinese Society on the Eve of Tiananmen , ed. Deborah Davis and Ezra F. Vogel, 85-108 (Cambridge: Harvard University Press, 1990); Dorothy Solinger, "Capitalist Measures with Chinese Characteristics," Problems of Communism 38 (Jan.-Feb. 1989): 19-33. In December 1984 there were 86.3 million employees in state enterprises and 3.3 million working privately. By December 1989 these totals were 101 million and 6.5 million respectively. Zhongguo Tongji Nianjian (ZGTJNJ ) 1988 , 153, and Renmin Ribao (RMRB ), Feb. 21, 1990.

need for kin-based networks of mutual obligation. For these reasons, the main thrust of the Deng reforms in the cities had virtually no direct impact on the urban household economy. Nevertheless, urban families also entered into a new relationship with the state during the 1980s.

First, there was the challenge of prosperity, and if there were no great disparities of wealth, there was increasing economic differentiation and competition. Between 1978 and 1989 urban wages trebled, and consumer items such as color televisions, refrigerators, and electric fans, which in the 1970s had represented unimaginable luxury, became commonplace.^[5] Equally important was the availability of the products of capitalist culture—Hong Kong and Taiwan soap operas, American movies and magazines, and the dance and video music of European and Japanese teenagers. Thus, while urban reforms did not commodify economic relationships within (and between) families as did decollectivization of agriculture, they did introduce a new range of cultural choices as well as the prosperity to permit individuals to act on preference, not merely respond to necessity.^[6]

At the same time, however, the state continued to play a highly intrusive and coercive role vis-à-vis all Chinese families. Simultaneous with policies to decollectivize agriculture, encourage petty capitalism, and permit importation of the cultural products of the bourgeois West, the post-Mao leaders designed and implemented the most draconian birth control campaign in history. As first outlined in 1979, the birth control regulations prohibited any woman from having more than one child unless her first had died of a noninheritable disease or handicap. In one stroke of the bureaucratic pen, Chinese leaders decided that henceforth only half of all families would have a son to carry on the family name, that the sibling relationship would disappear, and that failure to use contraception would be a punishable offense. In terms of economic and political initiatives the reform leaders moved to disengage the state from control over land, labor, and markets; in terms of family reproduction they demanded that the state control the fertility of

[5] In 1978 urban staff and workers had an average annual wage of 615 yuan; by 1989 it had risen to 1,950 yuan. ZGTJNJ, 1989 (Chinese Statistical Yearbook ), 125; Guowuyuan Gongbao (GWYGB ) (Bulletin of the State Council), March 3, 1990, 85. For gains in consumer items, ZGTJNJ 1989 , 728.

[6] There are indications that the state-dependency of urban families may decline more rapidly in the near future. Nicholas Lardy reports that by 1990, 45 percent of China's industrial output was the product of the nonstate sector. While this includes many products of rural industries as well as urban collective enterprises, the growth rates of the private sector are about eight times those of the state sector, and continuation of current trends would mean that in a few years, a greater percentage of urban families in China will be unambiguously in the private sector and cast further adrift from the state-organized welfare net than is now the case. See Lardy, "Redefining U.S—China Economic Relations," NBR Analysis No. 5 , National Bureau of Asian and Soviet Research, Seattle, June 1991.

over one hundred million married couples. Rarely had the paradox of state power been more acute than between 1979 and 1983.

By the mid-1980s, however, the leadership had retreated from its original extreme position, and the ambitions of the state to dictate family size had moderated.^[7] Rural families that could show that acceptance of the one-child limit would cause economic hardship or whose members were ethnic minorities were permitted to have a second or even a third child. And even in the cities, where pensions and social services eliminated the economic rationale for additional children, a variety of exceptions permitted some families a second birth.^[8] Nevertheless, regardless of these modifications, the one-child policy continued to create tension and contradictions. For example, in designating the rural household as the basic unit of agricultural production, decollectivization increased the value of child labor. Yet the one-child quota demanded that families drastically curb fertility. If the one-child policy really succeeded, then approximately half the women born after 1955 would not have a son in their home to provide care for them in old age. There would have to be either a revolution in rural pension systems or massive construction of old-age homes. But with the collapse of the communes, rural pensions became even less extensive than they had been in the late 1970s, and village supports for the childless elderly precipitously declined in quality and coverage.^[9]

In urban areas the clash between private and public interests created by the introduction of the one-child policy appeared in a different form. Thus even though young urban couples did not find the one-child campaign a threat to their immediate prosperity or to financial security in old age, they did perceive the campaign as an oppressive intervention of state power into an arena of family life that had remained private even during years of high political mobilization such as the Cultural Revolution.

The primary question, and the question that unites the chapters of this book, is this: What have been the initial consequences of these powerful and

[7] Karen Hardee-Cleaveland and Judith Banister, "Fertility Policy and Implementation in China 1983-1986," Population and Development Review 14, no. 2 (June 1988): 276; Steven Mosher, "Birth Control: A View from a Chinese Village," Asian Survey 22, no. 4 (April 1982); Delia Davin, "The Single-Child Family Policy in the Countryside," in China's One-Child Family Policy , ed. Elisabeth Croll, Delia Davin, and Penny Kane (London: Macmillan, 1985) 74; Joan Kaufman, "Family Planning Policy and Practice in China," Population and Development Review 15, no. 4 (Dec. 1989): 707-30.

[8] Birth control policies in minority areas were set locally and generally were looser than in Han areas. In urban areas, couples were permitted a second child when both spouses were only children, when the first child was seriously handicapped by a noninheritable defect, or when in remarriage the woman had had no children of her own. Nevertheless, second children are rare in the cities. Griffith Feeny, "Recent Fertility Dynamics in China," Population and Development Review 15, no. 2 (June 1989): 297-322.

[9] Deborah Davis-Friedmann, Long Lives , 110-11.

contradictory initiatives for Chinese families in the 1980s? In particular, our authors have asked: What have been the consequences for household composition, for marriage arrangements, for fertility decisions, and for provision of care to kin unable to compete in the more individualized and commodified society of the 1980s? In terms of size and composition, has the percentage of joint and extended families declined? Has marriage age fallen and have parents lost control over a child's choice of spouse? Has veneration of ancestors become more elaborate, and have lineages returned as key brokers of local power? Where, overall, are the lines now drawn between public and private responsibilities, between the Deng party-state and Chinese society?

Drawing on William Goode's classic study of families in periods of economic change and growing prosperity, most observers would predict that the reforms of the 1980s would encourage nuclear households more focused on conjugal loyalties.^[10] With greater affluence, higher rates of urban living, and increases in long-distance migration, the Goode model predicts that the influence of corporate kin groups should decline, brideprice and dowry become less prevalent, divorce rates rise, women's rights improve, age of marriage increase, and parents lose control over children's courtship and mate choices.^[11]

But the Goode model, which drew heavily on the experience of Western Europe and North America and could not have used material on China after 1960, never dealt with a state as intrusive and coercive as the Chinese. In Goode's overview, state regulation of family life was a consequence of industrialization and urbanization. Clearly, in the People's Republic of China (PRC) state power and policies have been the creators, not the creations, of a transformed society. Thus, for example, formal, corporate kin groups such as lineage organizations were specifically replaced by Communist Party organizations and a people's militia as early as 1950, and in most parts of China family shrines and temples disappeared even as China remained an essentially rural and agrarian society.^[12] The Marriage Law of 1950 prohibited concubinage and child betrothal and permitted women to sue for divorce. Within the first three years of the law nearly two million divorces were granted,^[13] and concubinage and child marriage became almost unknown by the end of the decade. We could go on at length documenting how state policies in the areas of public health, education, and migration similarly pushed Chinese families toward a "modern family" form at a speed unknown in European or American experience. But here we empha-

[10] William S. Goode, The Family (Englewood Cliffs, N.J.: Prentice-Hall), 183-85.

[11] Ibid.

[12] C. K. Yang, The Chinese Family in the Communist Revolution (Cambridge: MIT Press, 1965), chap. 11.

[13] Beijing Review , Sept. 12, 1983.

size that because the state was so powerful in the past in shaping the resources and choices of family members and in defining the functional division of responsibility between private and public authorities, the retreat of the state during the 1980s may have permitted a revival of cultural preferences and economic forces that the Maoist state held in check but did not eliminate.

The interaction between these cultural preferences and economic forces is partially predicted by Jack Goody's models of the evolution of the family. These models stress the importance of property in shaping the functions a family holds in a particular society. For Goody the contrast is between periods when families derive their livelihood and social status from land or other productive resources they own or control, and periods when property belonged to the collective and status was based on political criteria. According to this model, extended-family ties and use of dowry and brideprice should weaken in the Maoist period but revive with the return of differences in social status within and among local communities after the Deng reform.^[14]

In addition, it is possible that normative expectations rooted in sources independent of rational economic choices may prove decisive in an era when the party-state no longer intrudes as directly into cultural and religious life. Thus we may find by the late 1980s that in areas where government authority has dramatically retreated (most notably in rural areas) the dynamic of change for Chinese families may parallel that documented in John Caldwell's study of the demographic transition in western Africa, where family composition shifted in response to new ideologies rather than exclusively in response to new economic incentives and rationality.^[15] Similarly, John Hajnal's study of marriage age and household type in Europe and Asia suggests that the key explanation of different household arrangements is likely to be culturally specific "rules for family formation" rather than the more global processes of urbanization, industrialization, or commodification.^[16]

At the same time, in those areas where the state has strengthened its hand in the 1980s, we have a situation that is, as Susan Greenhalgh points out in chapter 9, unprecedented in world history and untouched by major theoreticians. The state, by implementing the planned-birth campaign, initially put itself at cross-purposes with just those cultural rules that shaped

[14] See Jack Goody, Production and Reproduction (Cambridge University Press, 1976), and The Oriental, the Ancient, and the Primitive (Cambridge University Press, 1990).

[15] John C. Caldwell, "Toward a Restatement of Demographic Transition," Population and Development Review 2, nos. 3 and 4 (Sept. and Dec. 1976): 321-66.

[16] John Hajnal, "Two Kinds of Pre-industrial Household Formation Systems," Population and Development Review 8, no. 3 (Sept. 1982): 449-94.

the prerevolutionary Chinese family. That it later had to compromise and loosen the restrictions, at least in the rural areas, testifies to the resiliency of those cultural rules in the villages. That the compromise was much less evident in the cities suggests that urban couples held values fundamentally different from those of their rural peers or that the power of the party-state was still so dominant that urban men and women could not act on their personal preferences.

Household Composition and Division

During most of the Maoist era, economic and political pressures fostered "family convergence." That is, in contrast to the two decades prior to 1949 when war, famines, and infectious disease prevented many from realizing the stable multigenerational households they desired, the overall legacy of the Maoist political economy was a remarkably uniform household structure, where differences in composition were increasingly due to the phase of the family life cycle, rather than to education or wealth, as had been true before 1949.

In the 1980s the new freedom to work outside of one's natal village and the introduction of the one-child campaign altered the external parameters shaping household size and composition, and in fact one can observe remarkable change within a short period. For example, in terms of average household size, families became noticeably smaller over the decade of the 1980s, and the number of households grew at a faster rate than the population.^[17] In 1980 rural households averaged 5.5 members and urban households 4.4; by 1988 they had fallen to 4.9 and 3.6, respectively.^[18] These declines, however, are explained almost entirely by declining fertility in response to the one-child campaign.^[19] Thus macrodemographic indicators lend little support to an interpretation in line with Goode's initial predic-

[17] Between 1982 and 1987 the number of people living in families rose by 8.2%, from 971 million to 1,051 million, while the number of family households rose by 12.7%, from 220 million to 248 million. ZGTJNJ 1988 , 103, and Zhongguo Shehui Tongji Ziliao 1987 , 31.

[18] These figures should be treated as best estimates. For the 1980s figures there is no consistent national estimate, and for 1988 one must consider that after 1984 the definition of urban was relaxed to include many village residents in urban suburbs and also residents of small rural towns who had previously been excluded. The estimates used here are from Martin K. Whyte and William Parish, Urban Life in Contemporary China (Chicago: University of Chicago Press, 1984), 155 n.5, and ZGTJNJ 1989 , 726, 742.

[19] That is, because the percentage of women living with their husbands' parents upon marriage remained almost constant from the 1970s through the 1980s one presumes that it is the one-child campaign and not other factors that was most decisive. For estimates of these percentages, see Martin K. Whyte and William Parish, Urban Life in Contemporary China , 155 n.5, and ZGTJNJ 1989 , 726, 742.

tion that economic development accelerates the nuclearization and individuation of family life. Rather, in China in the 1980s households became smaller, but not necessarily more nuclear.

A closer examination of field research on family composition in both urban and rural areas indicates that the trends in household size and composition have not been uniform; rather, they vary by the economic incentives of the local economy or the strictness of the planned-birth campaign. By the mid-1980s the most numerous incentives to preserve extended families were found in (1) villages where, as in industrializing Taiwan of the 1960s, reliance on labor-intensive cash crops or cottage industries rewarded unified budgets with one or more married sons or (2) cities where housing shortages forced newly married couples to remain in their parents' homes.^[20] In short, while it is clear that average household size declined, it is still not clear to what extent the reforms have altered family composition or the meaning of joint budgets and shared property.

This is particularly true in urban areas where, as Unger (chapter 2), Davis (chapter 3), Phillips (chapter 11), and Ikels (chapter 12) indicate, residence in separate households does not necessarily mean functionally separate families. Cooking, childcare, care of the elderly or disabled, and monetary transfers all frequently take place among geographically divided branches of "networked" families. In addition, as is especially clear in Davis's longitudinal study of Shanghai families (chapter 3), the fluidity of household membership is highly sensitive to stages of the family developmental cycle, and if a survey captures families at that point when children are in prime years for marriage, or when parents pass into years with high risk of chronic illness and higher mortality, households may change membership rapidly over a short period. In Davis's Shanghai study, where households were observed once in 1987 and then again in 1990, 60 percent of the reinterviewed households had altered their membership, a rate of turnover that Davis attributes not only to the age of her newly retired respondents, but also to the new prosperity and freer social environment of the eighties. Jonathan Unger's review of the Chinese five-cities survey of 1983 (chapter 2) suggests, however, that the instability of urban households may be a short-term response to current housing shortages, and that in the long

[20] Gu Jirui, Jiating xiaofei jingjixue (The economics of family consumption) (Beijing: zhongguo caizheng jingji chubanshe, 1985), 31; Isabelle Thireau and Mak Kong, "Travail en famille dans un village du Hebei," Etudes chinoises 8, no. 1 (Spring 1989), 7-40; Zhou Ying "Wo guo xiandaihua guocheng nongcun jiating" (Rural families in the transition to modernization), Renkou yanjiu , no. 2 (1988): 17-21. Zhang Yulin, "The Shift of Surplus Agricultural Labor," in Small Towns in China , ed. Fei Xiaotong (Beijing: New World Press), 186; William Lavely, "Industrialization and Household Complexity in Rural Taiwan," Social Forces 69, no. 1 (Sept. 1990): 235-51; Zhao Xishun, "Zhuanye jiating tedian qianxi" (An analysis of characteristics as specialized households), Shehui kexue yanjiu , 1985; Deborah Davis, chapter 3 of this volume.

term one would expect smaller, stable nuclear households, probably still linked in a variety of ways, to be the urban norm for both the newly married and the recently retired.^[21]

In rural areas, housing shortages are not the problem they are in the cities, but there still exist factors that either retard formal family division or encourage links between families already divided. It is important to note, however, that these factors operate differentially in the varying economic circumstances of different rural areas. For example, in the western part of the Pearl River delta, where Graham Johnson has conducted fieldwork since the early 1970s (chapter 5), the new wealth and openness to the outside world have revived connections with overseas relatives and produced new forms of emigrant families, which favor female-headed households and delayed family division. In the central and eastern parts of the delta, by contrast, the main outside influences have come from Hong Kong capital investment, and families there tend to be male-headed and to divide more quickly. Harrell's comparison of three villages in the Sichuan-Yunnan border region (chapter 4) also illustrates the importance of increased economic opportunities. Here, despite the more stringent application of the one-child policy in the richest village, households in that village were still more complex than in villages practicing subsistence farming, because wealthier families postponed household division.

Given the short time that has elapsed since the implementation of the reforms and the still limited access to national-level data, none of the chapters in this volume will conclusively specify the degree to which initial shifts away from collectivist economy and ideology determine the shape of Chinese households. Yet by documenting a range of outcomes, the chapters challenge arguments that posit a uniform or direct process of household simplification. Future scholarship will thus need to scrutinize both the rules of family formation and the shifts in state power and policy that proved so decisive in the past, and to situate findings in specific local contexts of cultural difference, as well as in economic and political change.

Marriage

In the first few years after 1949, the Chinese Communist Party (CCP) intervened directly to change many common elements of traditional marriage practices. The 1950 Marriage Law outlawed concubinage, child betrothal, multiple wives, and the sale of sons and daughters into marriage or prostitution. Village leaders and enterprise cadres exhorted the young people in their charge to delay marriage and in most cases enforced a policy of

[21] See the critique of this aspect of Goode's formulation in Peter McDonald, "Adjustments of Rural Family Systems," unpublished paper, 1987.

late marriage by withholding choice job assignments, welfare benefits, or promotions.^[22] They also condemned traditional rituals and elaborate feasts. Most of the ideals behind such policies were inspired by Engels's The Origin of the Family, Private Property, and the State , but the desired outcomes were also some of the very trends predicted by Goode for industrializing families. Success was mixed: the state succeeded in holding the line against such pre-communist practices as minor marriage and concubinage, but was much more successful in the cities (Whyte, chapter 8) than in the countryside in popularizing frugal weddings or eliminating brideprice and parental control of mate choice.

In September 1980 the National People's Congress passed a new Marriage Law, ostensibly motivated by a decision to strengthen the rule of law and bring the law more in line with practice. But in some areas the law actively encouraged "retrogression" to earlier practices of the precollectivist era. For example, the new law sanctioned minimal ages of marriage (20 for women and 22 for men [article 5]) that were significantly lower than the actual practices of the Cultural Revolution decade, and aside from a single sentence prohibiting "exaction of money or gifts" (article 3) there seemed to be no effort to control wedding expenses.^[23] And in fact, in the decade following the passage of the new Marriage Law, marriages did indeed become decidedly "less modern" and more like those of the precommunist era. In 1978 the average age of marriage for women was 22.4 for rural brides and 25.1 for those in the city.^[24] By 1987 the average age of marriage for all women had fallen to 21.0 years,^[25] and teenage marriages and even child betrothal had begun to reappear in rural areas throughout the country.^[26] The 1980s also saw an explosive increase in the overall expenses associated with marriage, as well as the return of lavish dowries and high brideprice.^[27]

[22] Whyte and Parish, Urban Life in Contemporary China , 113, found an average age of marriage for weddings between 1974 and 1978 to be 28.5 for males and 24.4 for females. William Parish and Martin Whyte, Village and Family in Contemporary China (Chicago: University of Chicago Press, 1978), 163, found average age at weddings between 1968 and 1974 to be 24.8 for males and 21.1 for females.

[23] For an English translation, see Beijing Review , March 16, 1981, 24-27.

[24] Zhongguo Shehui Tongji Ziliao 1987 (Chinese Social Statistics for 1987) (Beijing: Zhongguo tongji chubanshe, 1987), 29.

[25] RMRB , Feb. 19, 1989, 8.

[26] The Hong Kong paper Ming Bao , on April 20, 1990, 8, reported 3.5 million underage marriages in 1988; RMRB , March 31, 1989, 4, summarized a recent survey by the Women's Federation on betrothals of rural school girls.

[27] Gu Jirui, Jiating xiaofei jingjixue , 344; Qian Jianghong et al., "Marriage Related Consumption," Social Sciences in China , no. 1 (1988): 208-28; Beijing Review , Dec. 25, 1989, 32; Sulamith H. and Jack M. Potter, China's Peasants , 209; Burton Pasternak, Marriage and Fertility in Tianjin, China: Fifty Years of Transition , papers of the East-West Population Institute, no. 99 (1986), 28-31; Martin K. Whyte, "Changes in Mate Choice in Chengdu," in Chinese Society on the Eve of Tiananmen , ed. Deborah Davis and Ezra F. Vogel, Contemporary China Series (Cambridge: Harvard University Press, 1990), 203.

Despite these clear trends away from Maoist preferences, however, not all of these shifts occur uniformly among rich and poor, or from region to region. Helen Siu (see chapter 7) suggests, in fact, that the trend follows Goody's general observation that dowry becomes more important in areas where competition for status is acute and parents need to endow their daughter richly in order to achieve a "good marriage." In her study of a Pearl River delta market town and its surrounding villages, she shows how connections to Hong Kong have greatly intensified the more general trend toward commodification and producing for the market. Overseas Chinese have invested heavily, and recent émigrés return frequently with gifts of consumer items and cash. As in the areas visited by Graham Johnson, families are eager for their daughters to marry an Overseas Chinese, and local men find themselves disadvantaged in the marriage market.

In the market town described by Siu, dowry had been a major investment among the precommunist merchant elite, but in the 1980s it became pervasive among many strata, as parents used a daughter's marriage to stabilize and advance family or individual interests in an era when the local political economy was in tumult and the Maoist bureaucratic hierarchies were in decline. In the outlying villages, at the same time, a shortage of female labor, caused partly by pressure on young women to marry into more prosperous town families, drove up brideprice. Thus in Siu's work we find an excellent example of where the immediate results of greater commodification and openness to the West were elaborate multigenerational investments that advanced corporate family interests as well as individual aspirations of the parents and the newly married couple. We also find traditional patterns seemingly revived but, as Siu herself points out, for purposes often quite different from those of precommunist years: village families pay high brideprices as much to keep their children from migrating away entirely as to assure the productive and reproductive contributions of future daughters-in-law.

Similar effects can be seen in the patterns of marriage ages and of village endogamy and exogamy, as is shown in Mark Selden's study of Wugong, a former model village in central Hebei, a region where brides were traditionally three to five years older than their husbands. During the collective era, age at marriage rose, and a long-standing prohibition against intravillage marriage fell into disregard. After only a short experience with economic decollectivization and political liberalization in Wugong, however, age at marriage fell, and village daughters were again marrying at a slightly older age (23.4 years) than village sons (22.7 years), although the traditional pattern of brides two to four years older than grooms did not appear and intravillage marriage continued. In other words, some but not all traditional preferences reappeared after the reforms. A similar pattern obtained in the relatively poor Shaanxi villages studied by Susan Greenhalgh in 1988 (chapter 9), where Maoist ideals of late marriage quickly

crumbled, and the average age for a bride's first marriage fell precipitously from twenty-three to twenty between 1979 and 1987.^[28]

Based on a survey of almost six hundred women in Chengdu, Martin Whyte shows somewhat similar patterns in urban areas. Some but not all precollective wedding practices have reemerged—for example, weddings, simple and cheap under Maoism, have once again become more elaborate in the reform era, but freedom of choice of spouse, introduced in the collective era, has remained under the reforms. In addition, the considerable social class variation in marriage practices in the precollective era was largely eliminated during the period of "high Maoism," but in the reform era, Whyte finds, even though weddings have once again become more elaborate, there has been no reemergence of class-stratified patterns in freedom to choose a mate, elaborateness of weddings, presence or absence of a dowry, the share of the wedding expenses paid by the couple, or the pattern of postmarital residence. Whyte provisionally attributes this to the continuation in the 1980s of a "rank/network/bureaucracy" social system that originally replaced the "status group/class/market" system with the imposition of state socialism in the 1950s. According to this logic, urban wedding behavior had not revived class-based differences in the 1980s, as had rural behavior, because the urban society of the 1980s was fundamentally still what Andrew Walder called "communist neo-Traditionalist,"^[29] basically untransformed from the social structure of the Maoist period.

The diverse paths of change in marriage practices might cause us to reexamine a claim made in the important earlier study of urban marriages in the late Mao era carried out by Whyte and William Parish. In that study, the authors emphasized that after 1949 urban Chinese marriages resembled those of other developing societies. In line with Goode's predictions, there was a steady rise in marriage age, widespread disapproval of arranged marriages, and a convergence in the behavior of mate selection across different economic and educational strata. What was distinctive about China was the speed of the changes, the low cost of weddings, and the absence of a dating culture, characteristics Whyte and Parish attributed to CCP organizational and political resources not present among other rapidly modernizing or industrializing agrarian societies.^[30] Nevertheless,

[28] This pattern, however, was not uniform over the whole country either. In the three villages studied by Stevan Harrell in southernmost Sichuan, marriage age continued to be low from the 1950s through the late 1980s—there late marriage was never rigidly enforced in the collective era, and so there was no perceptible drop with the new law of 1980 or with the rise of the family farming economy. See Stevan Harrell, "Aspects of Marriage in Three Southwestern Villages," China Quarterly 130 (June 1992): 323-37.

[29] See Andrew Walder, Communist Neo-Traditionalism: Work and Authority in Chinese Industry (Berkeley: University of California Press, 1986).

[30] Whyte and Parish, Urban Life , 147-51.

they stressed that the causes of change lay primarily in the economic and social transformation and only secondarily in specific government policies aimed at reform of marriage practices.

Much of the "backsliding" that has gone on in the 1980s, however, makes the attribution of causality much more doubtful. Many of the changes in marriage during the Maoist era fulfilled Goode's predictions because the Chinese state was operating according to a political agenda, derived from Engels, that made behaviors that for Goode were predicted outcomes into explicit policy goals. The state planners probably thought they were just helping the inevitable along, but in fact they were creating a structure of inevitability much more narrowly restricted than the structure that would have been created by economic and cultural transformations. During the decade of the eighties, the state withdrew from close supervision of the local rural economies, and there was no single opportunity structure in the countryside. In the cities, the structure remained much more uniform, but several of its parameters evolved rapidly away from the Maoist period ideal, particularly the cost and lavishness of weddings. In retrospect, then, it seems that state policy had more effect in the Maoist period than Parish and Whyte originally gave it credit for, though the situation during the reforms is much more in tune with their model of primary causality residing in social and economic change. In any case, the situation is clearly more complex than any unilineal model of change would predict.^[31]

Childbearing

The relative importance of state power and economic imperatives is addressed with regard to a different area—fertility—by Susan Greenhalgh and Hill Gates in chapters 9 and 10. Although there is considerable disagreement about the fertility levels of the Chinese population before the early decades of the twentieth century,^[32] it is clear that from the 1920s through the 1950s virtually all Chinese couples considered it essential that they have enough sons to ensure that at least one would grow to maturity and continue the patriline. Given the high mortality of infants and young adults during these war-ravaged decades, the goal of one surviving son encouraged most women to bear as many children as possible.

Throughout most of the 1950s the Chinese government took a decidedly

[31] For example, see Arland Thornton and Thomas E. Fricke, "Social Change and the Family: Comparative Perspectives from the West, China, and South Asia," Sociological Forum 2, no. 4 (1987): 746-79.

[32] See the debate between Arthur P. Wolf, in "Fertility in Prerevolutionary China," and Ansley J. Coale, in "Fertility in Rural China: A Reconfirmation of the Barclay Reassessment," both in Family and Population in East Asian History , ed. Susan B. Hanley and Arthur P. Wolf (Stanford: Stanford University Press, 1985).

pronatalist position and did little to alter traditional expectations of marital fertility.^[33] After 1960 there was a shift, and the government began to promote birth control in cities, and urban fertility began to diverge from national trends (table 1.1). The real revolution, however, occurred during the decade of the 1970s when, as a result of both explicit government birth control campaigns and altered definitions of ideal family size, China's total fertility rate (TFR) fell from 5.8 in 1970 to 2.7 in 1978,^[34] a drop without historical precedent in a society that had not suffered war, epidemic, or famine, and a rate that resembled that of Canada more closely than those of more comparable countries, such as Brazil, Iran, or Mexico.^[35] In short, even in advance of the one-child campaign, Chinese women, in both rural and urban

[33] There is a vast literature on Chinese fertility shifts. In this chapter we have relied primarily on work that has drawn on demographic work in the 1980s. Judith Banister, China's Changing Population (Stanford: Stanford University Press, 1987); William Lavely and Ronald Freedman, "The Origins of the Chinese Fertility Decline," Demography 27, no. 3 (August 1990), 357-67. Burton Pasternak, Marriage and Fertility in Tianjin, China: Fifty Years of Transition ; Peng Xishe, "Major Determinants of China's Fertility Transition," China Quarterly , no. 117 (March 1989), 1-37.

[34] Judith Banister, "Population Policy and Trends in China 1978-1983," China Quarterly , no. 100 (Dec. 1984): 717-41.

[35] Statistical Abstract of the United States 1982-83 , 661.

areas, had reduced their fertility well below that of their peers in other developing nations.

In contrast, however, to the changes demanded by the one-child campaign, the shifts of the 1970s had been achieved without requiring half the population to forgo a son or all couples to risk a childless old age if their singleton died. Balanced against the substantial decreases in infant and child mortality rates, in fact, these early birth-limitation campaigns may not have produced much perceived difference from precommunist expectations of having a son survive adulthood. The one-child campaign unveiled in 1979 marked, therefore, a very radical departure from past efforts and, particularly in light of the new opportunities for considerable return on child labor, created the conditions for direct confrontation between family and state self-interest.^[36]

Based on observations in the late 1980s, however, the impact of the policy is not clear-cut. That is, when one examines the overall trends in fertility and parity for the decade of the eighties, it appears that overall fertility has declined very little (see table 1.1). However, when one examines trends in terms of percentage of births that are third- or higher-order parities, the one-child policy has had a significant impact (table 1.2). The question, therefore, is what were the conditions that produced this record of mixed success. Was it the uneven efficacy of birth control cadres who had to rely primarily on intimidation and coercion, or the more universal pattern of variation by the economic and educational status of childbearing women?

Greenhalgh's and Gates's chapters, both based on 1988 fieldwork, confirm previous conclusions that the two-child norm has been widely

[36] Steven Mosher, "Birth Control: A View from a Chinese Village"; Delia Davin, "The Single-Child Family Policy in the Countryside," 74.

accepted,^[37] but they focus more extensively on the particular balances of socioeconomic incentives and state coercion that favor compliance and very low fertility. Working in Shaanxi periurban villages, Greenhalgh found that by 1987 local cadres had essentially given up enforcing a one-child policy: the fining system had collapsed, second children were spaced even more closely than they had been in 1980, marriage age had dropped to twenty, and more than 8 percent of women married in their teens. On the other hand, a two-child family appeared to be the mode, and if a family had two boys or a boy and a girl, parents were in most cases satisfied that they had achieved an ideal size. In short, Greenhalgh concludes that as a result of successful bargaining with local cadres, village families had "peasantized" population policy: it was not the state's enforcement power but the rational calculations of peasant families that brought rural fertility close to compliance with state goals.

In Gates's study of small-scale entrepreneurs in urban Chengdu and Taiwan, the focus is also on negotiations. But in this case the author stresses negotiations between a wife and other members of her (husband's) family. Interviewing several generations of women who bore children both before and after 1949, and across all decades since 1949, Gates finds that women entrepreneurs rationalized childbearing as a means to meet specific obligations, and that their willingness to have a child depended on their own individual relationship to the market. In the PRC during the one-child era, there is of course little official space in which to negotiate. There is a question, however, about delaying the first birth that can be (and is) subject to negotiation, and also there is a question about the degree to which a one-child norm would ever be acceptable to Chinese women and their families given pronatalist preferences and behavior in earlier decades and in Taiwan. Based on the Chengdu respondents, Gates finds a clear relationship between increased capital assets and what she terms "antiphiloprogenitivism." That is, urban petty capitalists do not desire more than the minimum number of children to meet their family obligations; therefore, in the 1980s when the state limited them to one child, these particular women found no difficulty in accepting that limit.

Certainly Gates's respondents represent only a small minority of urban women, since most urbanites are employed in state enterprises where they have no personal assets and no entrepreneurial ambitions. However, it is also true that the model of kinship that Gates has defined, whereby a woman's childbearing desires can be explained in terms of how her particular household relates to the larger economy, is in no way limited to urban

[37] After completing four rural surveys in Fujian and Heilongjiang during 1987, Joan Kaufman found that enforcement at village level appeared weak and that while the household-responsibility system rewarded higher fertility, a two-child norm appeared to have been widely achieved; Joan Kaufman, "Family Planning Policy and Practice in China."

petty capitalists, but rather establishes a conceptual model that could be applied to all women, whether they live in cities, towns, or villages.

In fact, this conceptual model clearly fits with Harrell's data on the wealthiest of the three villages he investigated, where leading economic roles for women, readier acceptance of childbearing limits, and possibly even a tendency to pool capital, were most pronounced among entrepreneurial households. Given this parallel, it would be prudent to revise the ubiquitous urban-rural dichotomy and replace it, at least provisionally, with a three-way division between subsistence farmers, urban state/collective employees, and a third, entrepreneurial category that spans both cities and villages. And given the direction of China's economy toward an ever-larger private sector, we can expect that entrepreneurial households, a tiny minority at present, will come to constitute a much greater percentage of all Chinese families in the near future.

It is this third category of households that calls most sharply into question the unilineal theories of family change. While these petty capitalist households, as Gates and Harrell both suggest, do exhibit the predicted trends of lower fertility, at the same time they may well tend to greater household complexity because of capital and labor-pooling needs. In addition, as property owners they are more likely to be concerned with dowries, wedding banquets, and other forms of conspicuously displayed wealth. Whether or not the entrepreneurial category gains further prominence, the very fact that their family behavior differs from that of agriculturalists or state-sector employees once again points out that the process of family change, relatively homogeneous in the Maoist period, has become complex and heterogeneous.

Dependents and Family Obligations

One of the explicit rationales for several aspects of the ideal model of traditional Chinese family organization, with its early and universal marriage and strong intergenerational interdependencies, was that such a system not only facilitated the continuation of the patriline but also provided for the frail and the vulnerable. If the family were to modernize in accord with a Marxist blueprint, families would relinquish these responsibilities, and the weak and the needy would turn first to public institutions. To quite an extraordinary degree, the Chinese Communist leadership in the collectivist period realized a large portion of these Marxist ideals. As might be expected in a state where the proletariat is by definition the vanguard, the greatest investments were made in cities, and within ten years of 1949 urban families enjoyed free primary education, free basic health care, and very inexpensive housing. In the countryside, where programs were funded by local communities and per capita fees, services were far more meager

and families continued to be the primary source of aid for the young, old, and disabled. Nevertheless, money from central coffers subsidized construction of schools in small villages throughout the country, colleges sent thousands of new graduates to county and commune schools and clinics, and village cadres created welfare funds that supported cooperative clinics and guaranteed subsistence to all village households.

With the collapse of the commune as a political and economic unit, and the rapid commodification of rural labor, the structural supports for public welfare collapsed in the countryside and fees for services rapidly multiplied.^[38] Tuitions became onerous, village welfare funds served only the most destitute, and most small clinics became privately run.^[39] Even in the cities, where welfare provisions remained structurally intact and fully funded, there were also steps to monetize. Urban parents paid more extraneous school fees, and firms began to exclude some services from the health benefit. Bribes to medical personnel became more ubiquitous, and the cost of medicines not covered by health plans rose very fast as pharmacies and clinics tried to increase profits.^[40] In short, the reforms noticeably reduced collective responsibility for care of the young, the sick, and the needy, and increased the financial burden of illness and disability on individuals and families.

The question here is how these cutbacks in public services have affected family behavior, and in particular whether the retreat of the state has affected family solidarity. Because of the more complete decollectivization of rural services, one would assume that village families have been most immediately and decisively affected, and that individuals who come from the poorest households or have the weakest network of extended kin will find themselves less secure than they had been under the collective. Conversely, given new incentives for individuals to develop family- and kin-based strategies of cooperation, one would also expect a resurgence of such risk-sharing institutions as burial societies, crop-watching associations, or lineage schools and scholarship funds. Reports from Guangdong, where lineage activities were traditionally strong, suggest that kinship associations have begun to assume welfare functions that were previously the responsibility of the village or commune.^[41] But as Michael Phillips, working in

[38] Deborah Davis, "Chinese Social Welfare," China Quarterly , no. 119 (Sept. 1989): 577-97.

[39] Gail Henderson, "Increased Inequality in Health Care," in Chinese Society on the Eve of Tiananmen , ed. Deborah Davis and Ezra F. Vogel, 263-82 (Cambridge: Harvard University Press, 1990).

[40] Ibid.

[41] See chapter 5 in this volume, by Graham Johnson; also Sulamith H. and Jack M. Potter, China's Peasants , 266 and 337; and Helen Siu, "Recycling Tradition," Comparative Studies in Society and History 32, no. 4 (Oct. 1990): 765-94.

Hubei province, noted during discussion at our conference, lineage resources appear to be available for only a tiny percentage of rural families. Thus, at his hospital he has observed that the increased cost of hospital care and the collapse of collective medical insurance programs in the countryside have reduced the number of rural patients and shortened the stay of those whose family members are able and willing to pay the fees. Similarly, studies of rural enrollment rates show marked declines in entry to school and decreased completion rates from primary and secondary school.^[42]

In this volume there are no studies of the impact of Deng reforms on rural families in distress. But in chapters 11 and 12, which analyze caregiving dynamics in urban families, we do find evidence of negative impact. In the families of schizophrenic patients he treated, Phillips observed how the more individuated reward structures and higher medical fees of the late 1980s increased tension between two core values of urban Chinese families: the desire to sacrifice for dependents, and the commitment to "judicious investment of family resources . . . [for the advancement] of the social status of the family." In the Mao years, when many employers were rather lax in their demands of the least skilled workers, families could find a place for their schizophrenic child or spouse. Treatment costs were covered by employers, and although families were often deeply ashamed by their relative's behavior, they still could maintain a veneer of normalcy. Moreover, the loss of this individual's full economic potential did not pose heavy financial costs in the absence of many alternative demands for surplus income. In the more competitive society of the 1980s employers can refuse to employ mentally ill people even for menial jobs, health insurance has reduced payments for medication, and the family, because of its heavier medical expenses and lost income potential, can become impoverished. In general, the cost to families of having a profoundly ill member increased during the reform years, and families with the fewest social and economic resources were hardest hit. Families continued to assume the main obligations, but financial burdens made the disability more immediately costly, and greater competition at the workplace and more approval of individualistic striving undermined the willingness of some family members to sacrifice for another.

In the final chapter of this volume, Charlotte Ikels discusses what the reforms have meant for elderly residents of Guangzhou. Overall, there have been no efforts to undermine financial security; in fact, over the decade of the 1980s pensions grew at about the same rate as wages, and the percentage of urban residents over age sixty who received a pension steadily rose.^[43] However, for the same reasons that Phillips found schizophrenics more of a burden on their kin, so Ikels found Guangzhou elderly and their

[42] RMRB , Feb. 15, 1989, 1; Aug. 7, 1989, 1.

[43] Davis-Friedmann, Long Lives , 107-16.

children increasingly aware of the costs of providing services to the weakest and least socially active. Among her respondents the reforms did not directly reduce security or increase family conflict, but they did intensify anxieties about the future, and Ikels found elderly parents reaching out to make themselves useful or less needy, in anticipation of a day when they would need to draw heavily on the goodwill and filial loyalty of one or more of their children.

Chinese Families in the Post-Mao Era

Through the presentation of these research findings, we hope to establish a baseline, and perhaps even set an agenda, for future research on families in China. We believe that the approach taken here, looking at family behavior as the adaptation of cultural rules to changing and diverse political and economic circumstances, helps us do several things. It takes us beyond either/or debates about culture and political economy, or about state policy versus economic change, to examine the interaction of these factors. It moves us away from sweeping generalizations about the Chinese family, or even the urban or rural Chinese family, to consider different outcomes of adaptation to different circumstances. And it allows us, on the basis of empirical case studies, to begin to make certain generalizations, not so much about outcomes as about processes.

The first generalization is that many key characteristics of family life during the Maoist period—high age of marriage, elimination of polygamy and concubinage, reduced (or absent) dowry, and weak corporate kin groups—which earlier students of family life such as William Goode had expected to find as consequences of industrialization, may in the case of China be best explained by family laws, which Goode had presented as a consequence, not a cause, of change.^[44] Thus we suggest that the Maoist period witnessed a rapid shift away from corporate kin groups, elaborate weddings, concubinage, and early marriage, not because the economic and social transformations of the 1950s and 1960s made such changes irresistible to individual men and women, but because the state drafted laws or regulations that required immediate compliance.^[45] Therefore, it is in the interaction of preexisting "family culture" with both convergent processes of industrialization and particular state policies that we should look for explanations of the complex patterns of family changes in revolutionary China.

The second generalization responds to the resurgence of traditional

[44] Goode, The Family , 186.

[45] This point has also been made by William Lavely and Burton Pasternak in their studies of fertility decline. Lavely and Freeman, "The Origins of the Chinese Fertility Decline," and Pasternak, Marriage and Fertility in Tianjin .

rituals, ceremonies, and family behavior such as child betrothal, bridewealth, lavish dowry, and joint family households, that initially appear to document a "step backward." None of these traditional practices should be interpreted as a simple return to the status quo ante. Rather, the return of precommunist festivals and traditions is first and foremost a response to the post-Mao political economy, and by examining these rituals in their contemporary context one may identify the emergence of new boundaries between kin and nonkin and the emerging character of intergenerational obligations in an era of political uncertainty. Thus, from such material we can perhaps more clearly see the family ideals and the rules guiding household formation that Caldwell and Hajnal found decisive but that currently are still not well articulated in scholarly discussion of China.

The third generalization is that, when explicit state policies do not force homogeneity, the general tendency is for Chinese families to adapt their marital, parental, and welfare strategies to local economic conditions. Recent family behavior does not converge on a uniform pattern of extended patrilocal households with strong corporate links and equal inheritance among all sons, much less on an evolutionarily determined isolated nuclear family. Nor is there an urban form and a rural form. Instead, today's Chinese mainland, like twentieth-century Hong Kong and Taiwan, displays a variegated mosaic of family forms and behaviors, each demonstrating the adaptation of basic principles to differing conditions.

In his introductory comments to Louise Tilly and Joan Scott's study of English and French families in the early years of the industrial revolution, historian Michael Katz remarked:

A close relationship exists between the organization of the family and the mode of production at any given time. Yet [Scott and Tilly] also show that the relationship is very complex. Domestic organization does not change quickly or easily. Families adopt complex strategies which enable them to preserve elements of customary practices in altered circumstances, and the family patterns that emerge represent adaptations, complex compromises between tradition and new organizational and social structures.^[46]

Observations of Chinese families during the economic and political upheavals of the 1980s confirm much of what Katz found to be the case for European families in another era of rapid change. But as we and most of the authors in this volume stress, in China neither the family nor the individual is as autonomous as were Europeans in the eighteenth and nineteenth centuries. Because during the second half of the twentieth century a powerful, intrusive Chinese state frequently rewrote the basic "laws" of economic exchange and terms of trade, the "complex compromises" that Katz

[46] Louise A. Tilly and Joan W. Scott, Women, Work, and Family (New York: Holt, Rinehart and Winston, 1978), xi.

found in Tilly and Scott's discussion of gradual shifts from the household as a mode of production to a family wage economy may be quite different for the PRC.

As is obvious in our description of the several ways in which the Communist revolution served traditional family goals, and in the ways in which families have prospered under the Deng reforms, members of Chinese families, like those of Europe and North America, seek to maximize their resources under different modes of production. Economic constraints are critical in family form and family behaviors, but economics alone cannot capture all the external constraints. Politics also matter, and no study of Chinese families can be complete unless it is as concerned with the state as it is with the economy. And as everywhere, the changes that stem from politics and economics all play themselves out against the background of a Chinese family culture that does change, but only slowly.

PART ONE
HOUSEHOLD STRUCTURE

Two
Urban Families in the Eighties: An Analysis of Chinese Surveys

Jonathan Unger

During the 1980s our knowledge of the shape of urban family life in the People's Republic of China increased several fold. Some quite excellent research, to be sure, had been conducted by Western scholars in the 1970s through interviews with emigrants in Hong Kong.^[1] But it was only in the eighties that Western sociologists, anthropologists, and demographers at long last were able to conduct research inside China. And perhaps more important, it was only in the 1980s that China's own social scientists were able to begin serious research of their own.

Under Mao, scholarship in the social sciences had been sacrificed to the whims and dictates of politics. The government had deemed sociology potentially dangerous, in that it intruded on the Party's desire to hold a monopoly over analyses of society. All sociology departments were abolished in 1952,^[2] and it was not until 1979-80 that three departments—in Beijing, Tianjin, and Shanghai—were reestablished by the government as a first step in rebuilding a capacity to monitor and analyze social problems. Sociology departments soon opened at universities in other cities, with staff hurriedly recruited from other disciplines.

Though the field was still new, a substantial number of Chinese surveys on urban family composition were conducted between the years 1982 and 1985. The most significant of these projects was a coordinated effort in 1982-83 to study family patterns of eight residential districts in five of

[1] See, for example, Martin King Whyte and William L. Parish, Urban Life in Contemporary China (Chicago: University of Chicago Press, 1984), and Deborah Davis-Friedmann, Long Lives: Chinese Elderly and the Communist Revolution (Cambridge: Harvard University Press, 1983).

[2] Martin K. Whyte and Burton Pasternak, "Sociology and Anthropology," in Humanistic and Social Science Research in China , ed. Anne F. Thurston and Jason H. Parker (New York: Social Science Research Council, 1980).

China's major cities: Beijing, Shanghai, Tianjin, Nanjing, and Chengdu. The more than a dozen scholars who cooperated in this endeavor published a 566-page book that not only included essays on their separate findings but also provided appendices with hundreds of tables of local and cumulative data.^[3] A rush of other surveys followed: of young marrieds, of middle-aged women, of the elderly, and so forth. By 1986, however, this boomlet in survey research was on the wane, and the number of new survey findings that were openly published declined precipitously thereafter.

Not all of this research was methodologically sound in sampling techniques, nor did commonsense always prevail in the composition of questionnaire questions. Some of the researchers, without training and new to this mode of social science research, were obviously learning on the job. Perhaps a third of the survey reports from those years are of little use on account of such problems. But the other two-thirds, those that appear to have been reasonably sound methodologically, are invaluable in promoting our understanding of social change in urban China. Cumulatively, they enable us to see the broader outlines of urban family structures and the changes these were undergoing.

Unfortunately, this is not at all true with respect to rural surveys—which is why this book does not contain a parallel chapter on Chinese findings about peasant families. Rural surveys by Chinese sociologists are hard to come by,^[4] and the comparatively few that do exist generally seem to have been conducted hurriedly, using abnormally small samples. Overwhelmingly, Chinese scholars have concentrated instead on the residents of big-city neighborhoods.

This chapter will draw upon some thirty-five of these urban surveys. Since almost all of them date from 1982 to 1986, much of my analysis will concentrate on the shape of urban Chinese families in the first half of the eighties. In a number of places my analysis will differ from those of the Chinese authors, and my citation of their statistical findings does not imply that they are responsible for the interpretations I have placed upon this data. The first sections of the chapter will draw heavily upon the statistics of the five-city survey, to set out the circumstances of families at the opening of the eighties. The latter sections will refer almost exclusively to other, later surveys.

[3] Wu chengshi jiating janjiu xiangmu zu (Research Project Group on the Families of Five Cities), Zhongguo chengshi jiating (Chinese urban families) (Jinan: Shandong renmin chubanshe, 1985). The scholars involved in this project subsequently published two edited conference volumes of papers that discussed the findings: Pan Yunkang, ed., Zhongguo chengshi hunyin yu jiating (Chinese urban marriages and families) (Jinan: Shandong renmin chubanshe, 1987); and Liu Ying and Bi Suzhen, eds., Zhongguo hunyin jiating yanjiu (Research on Chinese marriages and families) (Beijing: Shehui kexue wenxian chubanshe, 1987).

[4] The major exception is the native village of the grand old man of Chinese sociology, Fei Xiaotong. Fei has sent assistants into the field to survey this one village repeatedly.

Making a Choice: Nuclear and Stem Families

First the simple facts. As of the early 1980s, according to the five-cities survey, some two-thirds of all the households in China's major cities were nuclear in composition: that is, they consisted only of parents and their children. We often think of Chinese families as including not just a married couple and their children but also one or more grandparents—that is, stem families^[5] —and the data showed that, while urban China was largely composed of nuclear families, stem families, too, were indeed commonplace, constituting a quarter of all households. The 1982-83 survey of eight neighborhoods in five of China's major cities revealed that nuclear and stem families combined accounted for more than 90 percent of all households (table 2.1).

What was the prevalent trend, though? Were increasing numbers of newlyweds setting up their own independent households rather than living with parents-in-law in stem families? Not if we go by the figures gathered by the five-cities survey. These show that prior to the establishment of the People's Republic of China in 1949, the numbers of urban newlyweds who formed independent households were already on the rise, and that this trend had continued into the 1950s, but leveled off and subsequently dipped sharply. As can be observed in table 2.2, approximately 57 percent of the urban newlyweds who married during the dozen years between 1954 and 1965 had established their own households, but the incidence of this practice had been cut almost in half, to 32 percent, within the next dozen years.

One plausible explanation for the earlier shift into independent households is that the first decade of Communist Party rule had witnessed a large wave of immigration of single young adults into China's cities from the

[5] In the early 1980s, Chinese scholars devised various translations for the term "stem family." The two most common translations were zhugan jiating and zhixi jiating . By the latter part of the 1980s the usage of zhugan jiating had prevailed.

countryside. They had had no families in the cities to fall back on, and when they married had necessarily set up independent families. This inflow of migrants was cut off in 1958, and strict controls against new migrants were introduced thereafter. We may presume that for at least the half decade immediately after the inflow from the countryside was halted, large numbers of the young adults who had arrived before 1959 were continuing to marry, which would explain the persistence during 1958-65 of a relatively high level of independent households of newlyweds (see table 2.2). But from the mid-sixties onward, the vast bulk of the marriage-age young people were from established urban families and, accordingly, could move in with parents after marriage.

Notably, it is also evident in table 2.2 that among the couples who married in the half-decade following Mao's death (the years 1977-82), the proportion establishing independent families was substantially lower than in the prior decade of 1966-76 and scarcely higher than for newlyweds in the years before 1937.^[6] What accounts for the overwhelming preponderance of stem families among this younger generation of newlyweds? It should be remembered that a large portion of China's urban young people had been shipped off to the countryside to settle as peasants in 1968, when the Red Guards were crushed and the Cultural Revolution violence ended. They were joined in the countryside during the 1970s by a large proportion of the new urban secondary-school graduates, under a government policy of having the villages absorb the great bulk of the urban young people who were surplus to the needs of the urban labor force.^[7] By the late 1970s, fully eighteen million urban youths had been forced into the countryside. Very few of them married during this decade of enforced rustication, in the belief that married couples would be less likely to receive permission to return to the cities. The wait was rewarded, for between 1975 and 1979 one province after another abandoned the hated to-the-countryside movement and ordered

[6] Most of the other neighborhood surveys of the early 1980s support the evidence of the five-cities survey. One provides contrary evidence, though. A 1982 survey of an urban neighborhood of Chengdu, Sichuan, claims that in the parents' generation (396 households), 59% of young couples had lived in households separate from their parents; 33% lived with the husband's parents; and 8% lived with the wife's parents. In comparison, among the 333 young Chengdu couples of 1982, a considerably higher percentage, fully 74%, reportedly lived independently; 14% lived with the husband's family; and 11% lived with the wife's parents. See Li Dongshan, "Lun ju zhi" (On systems of accommodation), Shehui diaocha yu yanjiu (Social Investigations and Research), no. 5 (1985): 58. In short, the data for the parents' generation are very much in line with that of the five-cities survey; and the Chengdu data for the present generation are very much at odds with the five-cities data. It seems quite possible that the Chengdu investigators' data for the present generation records as "nuclear" those young couples who initially had resided with parents but who had already moved out of the parents' home.

[7] Jonathan Unger, "China's Troubled Down-to-the-Countryside Campaign," Contemporary China 3, no. 2 (Summer 1979): 79-92.

most of the youths back to the cities. During the succeeding years, covering the period 1977-82, a large number of this horde of returned young people were not able to secure any regular urban employment nor the wherewithal to obtain separate accommodations when, after so many years of delay, they finally married. Such couples crowded into parents' apartments (the bride's if adequate space was not available at the groom's)^[8] until separate housing became available.

That a temporary lack of alternative accommodations was a fundamental reason for the high proportion of stem families among newlyweds is suggested by the fact that most young couples moved out of the parents' home after only several years. A 1985 survey of 419 such multigenerational households in the city of Tianjin nicely illustrates this: 65 percent of the younger couples who had initially participated in stem families had moved out within the first five years of their marriage; and a further 17 percent moved out in the sixth to the tenth year of marriage: that is, in all, more than 80 percent of such stem-family participants moved out to form independent households within the first ten years of marriage.^[9] This phenomenon was true not only for the most recent generation of young couples; the same shift out of parental homes had been true, too, of older cohorts of couples.

The consequence was that a clear majority of middle-aged couples, as of the early 1980s, were living in independent nuclear families, not with parents-in-law. Table 2.3, based on a survey of a Beijing neighborhood in late 1982, shows this very much to be the case for wives aged thirty-three to forty-five.

Demographics alone would dictate that many of these middle-aged couples would by necessity live as nuclear households. In past decades, from the 1940s through the end of the 1950s, urban families very often were giving birth to three to five children; and with the near-abandonment of the joint-family tradition (see below), when these children grew up all but one of them necessarily had to leave the household, either immediately at the time of their wedding or when their next younger brother married. This goes far in explaining the phenomenon, seen above, whereby a high propor-

[8] The proportion of young couples moving into the groom's home as compared to the bride's home remained relatively stable from 1958 through 1982, though with a slightly rising bias in favor of living with the groom's parents, according to the data contained in table 2.2. In the period 1958-65 the ratio stood at 2.26:1; for the period 1966-76, 2.47:1; and for the period 1977-82, 2.58:1.

[9] Pan Yunkang and Lin Nan, "Zhongguo chengshi xiandai jiating moshi" (A model of contemporary Chinese urban families), Shehuixue yanjiu (Sociological Research), no. 3 (1987): 61. (The number of individual respondents was not given; the survey covered 989 households.) This Tianjin neighborhood survey found that when first married, 52% of the newlyweds in the sample lived with the husband's parents; only 4% lived with the wife's parents; 42% lived in an independent household; and 2% other (p. 60).

tion of young couples shifted out of the parental home within the first five years of marriage. The longer-term stem families usually have comprised the elder couple, the youngest son, and his wife and children.

The shift of young married couples into nuclear households may have gone beyond this, however. Chinese statistics suggest that in many cases even the last remaining child moved out, leaving the older couple on their own. Table 2.3, for example, shows that fully 77 percent of the surveyed Beijing women who in 1982 were in the age group thirty-three to forty-five lived in their own nuclear family—and that only 21 percent of the women in that age group lived in a stem or joint family. This latter statistic must have included a substantial number of women in their forties who were living in a stem family with their own newly married children, rather than living with their old parents-in-law. In short, the great bulk of the younger middle-aged women were living entirely apart from their elders.

Are we viewing here new trends in intrafamily patterns, or instead longtime mores? This question can be answered by examining the situation of the households of interviewees' parents, on the eve of interviewees' weddings, as observed in table 2.4. On average, the parents of a groom would have been in their late forties or early fifties (and since brides generally have wed at a younger age than have grooms, the brides' parents would have been a few years younger than the grooms' parents: say, in their early to late forties). By examining the data vis-à-vis people who wed in different time periods, we can derive a picture of the changing status of the households of these two age-sets of middle-aged couples. Table 2.4 is again based on data from the five-cities survey of 1982-83.

It is evident from table 2.4 that as of the early 1980s, the nuclear-family form was more frequent among the middle-aged than in any previous decade, but this had not led to declines in the incidence of stem-family organization. Rather, it can be seen from the table that over these decades both forms of family organization had been the beneficiaries of sharp declines in single-member households and of a sharp decline, too, in the incidence of joint families (that is, families in which two or more married children and their spouses live together, normally with an elderly parent or parents). Similar trends have been evident in Taiwan.^[10]

In urban China, this decline in single-member households is presum-

[10] The data available to me from Taiwan are not entirely comparable, in that the Taiwan statistics are for urban and rural households combined. These Taiwan-wide statistics do, however, exhibit trends similar to urban China:

ably the dual consequence both of lower levels of bachelorhood and spinstership^[11] and, among all the elderly who had married, of a rising survival rate among offspring. That is to say, a considerably higher proportion of elderly widows today presumably can count on having one or more grown children who can take them in if the need arises.

The decline in the numbers of joint families can largely be attributed to the changed class structure of post-1949 China; many of the joint families of earlier times had been built upon and held together by the wealth and property of the paterfamilias. The scarcity of joint families can also be attributed in part to the cramped living conditions that are almost universal in modern Chinese cities, and which preclude the sharing of the parents' meager accommodations by a large number of people. Whatever the exact factors that have been at work here, the incidence of joint families in a 1935-37 survey in Shanghai ranged from 7 percent among working-class families to 19 percent among the upper classes,^[12] whereas quite different figures apply for more recent decades. In a survey by Whyte and Parish of Guangdong urban families as of the mid-1970s, joint families constituted only some 2 percent of households.^[13] That same figure held true for China's large cities as of the early 1980s. As is seen in table 2.1, joint families constituted only 2.3 percent of households in the five-cities survey of 1982-83. (In many cases, I would surmise, these are households containing more

These Taiwan data show a far sharper rise in the nuclear-family form than was true for urban China. One possible explanation is that there was a far more rapid rural-to-urban migration in Taiwan than was allowed in the People's Republic. This rapid urban expansion resulted in the creation of a great many smaller households in Taiwan, with the younger couples in the cities and the older couple on the land.

[11] According to the figures of the 1982 census, only 6.8% of all Beijing men and a mere 3.3% of all Beijing women in the 30-34 age group had not already married, and in the 35-39 age group only 3.9% of the men and 0.8% of the women. Chen Wanzhen, "Wo guo hunyin zhuangkuang de tongji fenxi" (A statistical analysis of the state of affairs of marriage in China), Renkouxue kan (Journal of Demography), no.1 (1987): 19.

[12] Martin King Whyte and William L. Parish, Urban Life in Contemporary China (Chicago: University of Chicago Press, 1984), 154.

[13] Ibid.

than one set of newlyweds who have not yet been able to find suitable accommodations elsewhere.

The Determining Factors

Among what types of families do we find the highest proportion of nuclear as against stem families? Are there any "class" differences?

The answer to this latter question is yes. The most obvious "class" influence here may relate to geographic mobility. People who have moved into a city from the countryside or have moved from one city to another generally have not been able to bring kin with them because of residencycard restrictions. These same government restrictions have also more generally limited geographic mobility among ordinary people—the net result being low rates of geographic mobility but high rates of separation of relatives when mobility does occur. The net effect, according to a survey in the mid-eighties of an industrial district in Wuhan, was that, overall, 39 percent of the interviewees in the district lived in a different city or region than their parents, and 12 percent of the interviewees lived in a different city than all of their own children. University graduates and government officials most frequently have been assigned to jobs away from their place of origin, and this Wuhan survey revealed, not surprisingly, that residence at a distance from one's parents or children or both was highest among intellectuals, then cadres, and was least true of workers. One consequence, the survey indicated, is that a "class" distinction can be found in family composition. The incidence of stem/joint households in the district stood at 40.2 percent of working-class families, at only 33.1 percent of cadre families, and a mere 15.6 percent of intellectuals' families.^[14]

A low level of stem-family organization may not just reflect enforced separation, however. Surveys show that even when relatives reside in the same city, most interviewees report that they want to live in separate nuclear households. The "class" differences in the above Wuhan survey may therefore also partly reflect the varying abilities of people of different professions to obtain the separate housing needed to set up new households. Certainly, surveys of stem-family participants variously show that somewhere between 21-27 percent of such families (based on interviewing parents) and 36 percent (based on interviewing married children) would want to divide the stem household if additional housing became available.^[15]

[14] Cai He, "Chengshi jiating jiegoude bianhua ji qushi" (Alterations and trends in the structure of the urban family), Juece yu xinxi (Policy and News) (Wuhan), no. 9, 1987, republished in Shehuixue (Sociology), People's University Reprint Series C4, no. 6 (1987): 96.

[15] Hu Ruquan, "Cong Tianjin, Wuhan deng chengshide diaocha kan woguo chengshi laonianren jiating shenghuo" (The family life of elderly urban Chinese, as witnessed through surveys of Tianjin, Wuhan, and other cities), Shehuixue yanjiu (Sociological Research), no. 4 (1986): 104, provides tables of surveys of Tianjin in 1984 (21%) and Wuhan in 1985 (27%) of parents in stem families, who stated that lack of housing was the principal reason for the stem family. Similarly, in an eighteen-city survey of married young people in stem families (1986), 36% of them reported that the main reason for the stem-family arrangement was that alternative housing was unavailable. Qian Jianghong et al., "Wo guo da zhong chengshi qingnian jiehun xiaofei yanjiu" (Research into the marriage expenditures of young people in China's large and medium-size cities), Zhongguo shehui kexue (Chinese Social Science), no. 3 (1987): 121. Parallel to this finding, in a 1985 survey of 431 stem households that had already divided up, 35% of respondents cited their obtaining additional accommodations as the principal reason for the family division. (Pan Yunkang and Lin Nan, 62-63.)

Overall, attitude surveys of young unmarried adults in the major cities have shown that a very large proportion would prefer to live separate from parents after marriage. A survey of a Beijing district in the first half of the eighties claimed that 88 percent of the surveyed youths felt this way, and that only 12 percent preferred to continue to live with their elders after they married.^[16] Other surveys show the same general preferences, albeit not so dramatically. A 1985 survey of a Tianjin urban district, for instance, showed that 63 percent of the youths would prefer to live separately once married.^[17]

The Chinese surveys do not directly offer data on any possible class differences in such attitudes, but indirect circumstantial evidence would suggest only a modest difference between working-class youths and other young people in this respect. One illustrative example is a survey among young unmarried female textile workers in the city of Harbin, in which 43 percent of the respondents (N = 127) expressed a preference to live in future with in-laws, and 57 percent a preference to live independently.^[18] This finding among young working-class women differs only moderately from the Tianjin survey results for the populace as a whole. (To be sure, the question of gender differences may also have entered strongly into the Harbin survey findings; it is highly possible that young working-class men , had they been included in the survey, might have been more amenable to living with their own parents.)

City size may be as much a factor as social class in the statistics showing

[16] Pan Yunkang, "Shilun Zhongguo hexin jiating he xifang hexin jiatingde yi tong" (An attempt to discuss the differences and similarities between Chinese and Western nuclear families), Tianjin shehui kexue (Tianjin Social Science), no. 2 (1985): republished in Shehuixue , no. 3 (1985): 111.

[17] Pan Yunkang and Lin Nan, 57.

[18] Zou Naixian, "Qingnian fangzhi nugong shenghuo fangshi xuqiu bianhuade diaocha" (Survey of changes in young female textile workers' demands on life), Qingnian yanjiu (Youth Studies), no. 10 (1985); republished in English in Chinese Sociology and Anthropology 20, no. 2 (Winter, 1987-88): 15. The published data included a high proportion of women who declined to respond one way or the other, and I have excluded them in the figures here.

distinctions in family structure. The available data are not particularly reliable regarding the actual differences in family composition between China's largest, medium-size, and smaller cities, but scattered sources do suggest a higher incidence of nuclear families in China's large cities than in smaller ones.^[19] Attitude surveys show a similar, albeit small, distinction in preferences. For example, a 1985 sample of fifty urban households in a small city in Fujian province, all of the households containing unmarried young adults, came up with a greater balance in the youths' postmarital preferences than was true of either Beijing or Tianjin: that is, only 51 percent of the Fujian youths wished to live apart from parents after marriage, while 41 percent thought it preferable to live with their elders. (In turn, the parents in these fifty households were generally much more favorable toward stem households than were their offspring: 69 percent of the parents thought it best that a married child live with them, and only 28 percent felt it best that married children live separately.)^[20]

Parents and Children: Generations and Attitudes

Should we even expect the preferences of young people to count heavily in determining whether they remain in the home or leave upon marrying? Would a young man and woman or would his parents generally have the greater say in such a matter?

[19] One source that makes this claim is Xu Jingze, "Wo guo jiating jiegoude tedian jiqi fazhan qushi chutan" (Initial investigation into the characteristics and developmental trends in the structure of Chinese families), Wenshizhe (Literature, History and Philosophy) (Jinan City), no. 1 (1985), reprinted in Shehuixue (Sociology), People's University Reprint Series C4, no. 2 (1985): 109-10. As circumstantial evidence, this article provides tables showing, e.g., a higher proportion of nuclear families in a 1982 Tianjin neighorhood survey than in a survey of a semi-industrialized suburban district of a Shandong city.

Such a conjecture is reinforced by the available data on family size. The 1982 census, for example, revealed that the households in China's 239 largest cities averaged, overall, 4.09 members. Households in the 38 cities that had populations of at least one million averaged 4.01 members, while those in the three largest cities were smaller yet: Tianjin, 3.90 persons; Beijing, 3.69 persons; and Shanghai, the largest city, 3.60 persons. Yuan Fang, "Zhongguo Laonianren zai jiating, shehui zhongde diwei he zuoyong" (The status and functions of the Chinese elderly in the family and society), Beijing daxue xuebao, zhe she ban (Beijing University Studies, Philosophy and Social Science edition) 3, no. 1 (1987), reprinted in Shehuixue , no. 4 (1987): 115.

[20] Zhuang Jia, Ren yu ren (The Human Relationship), Guangdong renmin chubanshe (Canton: Guangdong People's Publisher), 1988, 147. A parallel sample of 50 households from villages in nearby counties found, as expected, that the surveyed young unmarried peasants were more amenable to living in stem families than their urban cousins: 54% of them thought it best that children live with elders after marriage (and 84% of their parents thought so). A rather high 41% of the unmarried village youths thought it best to live apart (compared to only 14% of the surveyed rural parents).

Certainly, young adults have gained a somewhat larger say over their own lives than in generations past. Whereas parents once very frequently arranged their children's marriages, that custom and the power it represented have all but disappeared in urban areas. More than that, parents play a markedly diminished role even in introducing their children to potential mates. As is seen in table 2.5, more than 80 percent of young marrieds during 1977-82 either had been introduced to each other by friends or had met on their own. Yet other surveys show, nevertheless, that young people still very seriously take their parents' advice into account in deciding whether to marry a particular person. As one example, a survey of 970 young workers in a district of Wuhan in the early eighties found that only 21 percent had made that decision entirely on their own. Another 63 percent sought out and respected their parents' counsel, and another 10 percent let their parents play the predominant role, asking only that they themselves consent to the parents' choice; 2.5 percent left the decision entirely in their parents' hands.^[21] Other surveys came up with somewhat similar findings.^[22]

[21] Li Minghua et al., "Wuchang qu qinggong lian'ai hunyin xianzhuang diaocha" (Investigation into the current circumstances of falling in love and marriage among young workers in a Wuchang [Wuhan] district), Shehui (Society), no. 6 (1983): 31.

[22] See, e.g., Liu Ying, "Tianjin Huningrong wu chengshi jiating diaocha chuxi" (Initial analysis of Tianjin's Huningrong neighborhood in the five-cities family survey), Shehuixue yanjiu (Sociological Research), no. 4 (1986), table, 80. Also see Zhuang Jia, table 2, 147; and Hu Ruquan, "Family Life," table, 98. Also see Martin K. Whyte's similar survey findings in Whyte's "Changes in Mate Choice in Chengdu," in Chinese Society on the Eve of Tiananmen , ed. Deborah Davis and Ezra F. Vogel, 185 (Cambridge: Harvard University Press, 1990).

One reason why the young still turn to their parents for advice and approval in such matters is that they remain largely dependent on their parents to get on in life, especially financially. The elder couple's earnings normally are considerably higher than the young person's,^[23] and young people usually cannot afford the expenses of getting married and setting up a new family without parental help. A 1983 survey in Tianjin of 121 households found that 84 percent of young couples had obtained financial support from their parents in order to get married; only 16 percent of the marriages had occurred without the parents' financial help.^[24]

Perhaps more important, young people normally are unable to obtain separate accommodations on their own; after marriage they need either to live with parents or to obtain the parents' aid in securing housing. In 1986 an eighteen-city survey of married young people between the ages of 18 and 35 (N = 726) found that only 27 percent had themselves been responsible for acquiring publicly owned housing (this was mostly provided by the husband's work unit, not the wife's).^[25] The great majority either lived with the husband's parents (35 percent), or with the wife's parents (7 percent), or in publicly owned housing obtained for them by the parents (14 percent), or lived apart in private housing (7 percent), which in most cases presumably was owned by the parents.^[26]

The parents, in short, have the wherewithal to play a very major role in determining their children's postmarriage arrangements. Yet, as was observed earlier, a very high proportion of young people in the major cities prefer to live apart from their parents. Is there a high degree of dissension and conflict, therefore, between the generations?

The answer appears to be, not usually: for during the first half of the eighties a shift was occurring in the attitudes of parents. Their views came largely to mirror those of their children. By 1984-85 surveys were revealing that the percentage of parents in the major cities who expressed a wish to live apart from all their children once the children got married was almost exactly as high as the percentage of young people expressing a wish to live apart from their parents (table 2.6).

According to the 1985 survey data from Tianjin, the desire to live apart from all of one's children once they got married depended partly on whether the parents had one child or more than one (table 2.7). Almost half of those with only one child were looking ahead to living together as a stem family, whereas some three-quarters of the parents with more than one growing child looked forward to the peace and quiet of the departure of all of them! (Note, too, that parents with more than two children who desired

[23] Li Dongshan, table, 64.

[24] Hu Ruquan, 109.

[25] Qian Jianghong et al. (n.15 above), 120.

[26] Ibid.

to be part of a stem family—that is, those parents who could be expected to have a choice as to whether to live with a married son or daughter—were unanimous in their preference for the traditional form of stem family: that is, for living with a son instead of a daughter.)

Not only did a majority of both parents and adult children in urban China prefer, by the mid-eighties, to live in separate households; increasingly during the eighties they were able to see those desires fulfilled. Lack of alternative accommodations had held something like a quarter to a third of all stem families together, as has been noted. By the latter half of the 1980s, the severe scarcity in urban housing had been considerably alleviated through a massive program of public-housing construction. (On this, see the beginning of chapter 3, by Deborah Davis.) Whereas urban resi-

dents had had, per capita, only 5 square meters of living space in 1980, by the end of 1987 they enjoyed 8.5 square meters. That increase in living space had been brought about largely through the construction of large numbers of small new apartments rather than through any substantial enlargement in the size of new apartments. The opportunities for the establishment of new nuclear families had been substantially enhanced; the living conditions for stem families had not. As of the close of 1987, urban accommodations, on average, consisted of only 2.44 rooms.^[27]

Networked Families

In urban China today, even those parents who live apart from their married children still tend to maintain very close mutual contact, more so than would be the norm in most Western societies. A multifaceted interdependency between the households of parents and married children very frequently is deliberately sustained even when they live apart. This can be seen particularly clearly in relation to the interchange of gifts of money.

[27] "Changes in the Life-Style of Urban Residents," Beijing Review , November 14, 1988, 27.

Although Burton Pasternak's 1981 study of a Tianjin neighborhood and the 1982-83 five-cities study both found that only some 36-38 percent of all households sent regular remittances of money to close relatives outside their own household, and that only 11 percent of urban households regularly received such remittances,^[28] a later and more detailed study draws a rather different picture. This 1984 survey, conducted in Tianjin, focused on 181 households of younger married couples and the 312 households of the couples' parents. Thus, unlike the earlier studies, the statistics in this 1984 survey referred only to families where married children lived apart from parents, providing a more accurate picture of such families' interrelationships. It found that 84 percent of the married children gave money to both sets of parents and, notably, that 91 percent of the households of married children received financial help from one or more sets of parents. Examining this group of 91 percent, it found that fully 61 percent of them regularly received parental gifts of money to cover their daily expenses.^[29] The survey revealed, moreover, that among these 181 households of married children, 44 percent received noticeably more money from their parents than they gave the parents, while only 27 percent of the children gave more money to their parents than they received in return.^[30] In short, while children often provided small monthly remittances to parents as an accustomed way of showing filial respect, these same married children quite frequently were actually net financial dependents on their parents.

This should not surprise us, inasmuch as the Chinese government for decades has promoted a steeply inclined seniority system in state-sector wages. A 1982 survey of a Chengdu district accordingly found that each of the parents taken separately—74 percent of fathers and 65 percent of mothers—earned more than their surveyed married child.^[31] At the same

[28] On this, see Burton Pasternak, Marriage and Fertility in Tianjin, China: Fifty Years of Transition , Papers of the East-West Population Center, no. 99 (July 1986): 37-38; also Wu chengshi . . . , 428, 432. Pasternak's survey discovered, interestingly, that of the parents who did receive remittances from children, 71% were receiving regular monthly gifts of money from daughters, and only 55% from sons. Similarly, among the households of married children that sent money to parents, 66% of such households were sending money to the wife's parents, more than the 55% who were sending money to the husband's parents (27). Undoubtedly, one factor contributing to these statistics is that a considerably higher number of parents would already be living with the family of a married son than with the family of a married daughter, and such parents would not, by definition, be receiving inflowing remittances from their live-in son, but only from their daughter. But several survey reports have noted that parents frequently cite their married daughters as more filial and more emotionally caring than their married sons.

[29] The survey found, moreover, that 74% of them received financial support from parents at the time of festivals; 8% to buy major consumer items; 11% just when in straitened circumstances (jieju ); and 7% only in emergencies.

[30] Pan Yunkang (n. 16 above), 110-11; also Pan Yunkang and Lin Nan, 65.

[31] Li Dongshan, table, 64.

time, their married children's financial needs, with young children to raise, were now frequently greater than their own. Middle-aged urban parents consequently quite often, as a matter of course, subsidize their married children.

Yet it should be remembered that the gifts of money regularly crossed both ways. It was and is a practice that tangibly symbolizes the mutual affection and sustaining support that bind the families to each other, cutting across the fact that normally the help that is actually needed lies largely on one side.

It is not just in financial terms that the younger couples are often dependent on one or both of their two sets of parents. They also frequently rely on one or more of their parents for daily childcare. Almost invariably, a young husband and wife in urban China both work full-time, and if one of their parents is retired—women in state-sector employment tend to retire at age fifty or fifty-five—they will want to rely on her to take care of their young children. In the 1984 Tianjin survey, 12 percent of the younger couples sent children to the grandparents on the father's side to be minded; and a further 8 percent sent children to the maternal grandparents. Some busy parents even sent the little ones to live full-time with the grandparents.^[32]

More than that, this Tianjin survey found that in fully 86 percent of the younger couples' households, one or more family members regularly went to eat meals at one or both of the elder couples' houses. In examining this group of 86 percent, it was found that only 34 percent of them shared fewer than twenty meals a month at the homes of parents who lived apart from them (these figures, incidentally, include grandchildren minded by grandparents). Another 21 percent shared twenty-one to forty meals a month at the parents' homes, a further 11 percent shared forty-one to sixty meals, and a final 20 percent shared sixty or more meals a month.^[33]

In other words, while living apart, they tended to interact almost like stem or extended families. Chinese social scientists have come up with the term "enwebbed" or "networked families" (wangluo jiating ) to describe this situation, which some researchers claim has become increasingly common during the eighties. One scholar, based on yet another Tianjin study, reports that although young couples and their elders increasingly prefer to live apart, they want to live near each other, in order to be able to carry on such a "networked" arrangement.^[34] Thus, families maneuver to acquire housing that will enable one or more of the married children's homes to be

[32] Pan Yunkang, 111.

[33] Ibid.

[34] Hu Ruquan, "Shitan 'Wangluo jiating'" (A discussion of 'Networked families'), Zhongguo funu bao (China Woman's Paper), March 7, 1986, 3.

clustered near the original home. A different strategy in "networked" living arrangements is for the younger couple to "network" almost equally with both sets of grandparents.

With one-child families enforced by government diktat and now the norm in China's cities, we might expect the incidence of stem families to begin increasing in the future at the expense of a "networking" of separate young and middle-aged nuclear families. After all, no longer will most grown children, as in the past, be obliged to leave the household after their marriage so as to leave room for yet younger siblings. In fact, in future a young couple would have a choice of parents' homes to live in. But the predictions that have been made in China of a sharp future rise in the proportion of urban stem households could be premature.

The percentage may well rise, but perhaps not by as much as one might expect by purely demographic reasoning. For one thing, as we have seen, a clear majority of parents as well as of children today prefer to live in separate housing. This was even true in 1985 with a majority of the parents who had only one ungrown child, as is seen in table 2.7.

In future, such feelings may well be strengthened, moreover, by the very fact that each of the two sets of parents will normally have only one child. Such a circumstance could well promote fears of "losing" their only child to another family after the child's marriage. A "networked" family/household structure would resolve such fears. Given that fact, at the turn of the next century it might well be a strongly favored form of family organization for the majority of the present generation of single-child urban families.

Family Arrangements of the Elderly

Nuclear households, as we saw above, already predominate among the urban middle-aged, often in a "networked" relationship with their married children. As we saw also, such nuclear-household arrangements are also preferred by a large majority of young marrieds. But what of the elderly— all those who are in their sixties and seventies and eighties? As people enter the tail-end of their lives do they prefer to live separately from their children? Did many of them indeed live apart from their children's families during the eighties, or did most of them live in stem-family households? And finally, what type of living arrangement seems most probable for the urban elderly of the decades to come?

First, the facts about living arrangements among the elderly during the eighties: at the same time that a very large proportion of middle-aged couples maintained their own nuclear family, a high portion of the elderly lived with one or more of their married children. Among the very elderly, as we shall see, more than 80 percent were members of stem or joint households.

In short, most urban Chinese, as they have passed from one stage of life

Fig. 2.1.
A Life Cycle in Family Arrangements

to another, have swung between the stem and the nuclear form of household, often more than once. Instead of living continuously with the family of a married child, many members of the elder generation have preferred to live apart and then have joined the home of a married child only after their own spouse dies or when they become increasingly frail in advanced age. This was observed in table 2.3, which showed 64 percent of all women over age sixty living in stem or joint families. Other neighborhood surveys have suggested a lower proportion, but still high. In Shanghai (where more than 7.5 percent of the total population as of 1986 was over the age of sixty, a percentage that has been climbing yearly)^[35] a 1983 neighborhood survey of 613 people aged sixty and older revealed that 52 percent lived in stem or joint families.^[36]

[35] Yuan Jihui, Chou Liping, and Dong Jiahua, "Shanghai shiqu gaoling laoren shenghuo zhuangkuang diaocha baogao" (An investigative report into the living conditions of the truly aged of a Shanghai urban district), Shehuixue yanjiu (Sociological Research), no. 3 (1987): 75.

[36] Zheng Guizhen, "Shanghai shi jiating jiegou yu yang lao shiye" (Shanghai's urban family structure and facilities for the care of the elderly), Zhongguo renkou kexue (Chinese Demography), no. 5 (1988): 136. A similar Shanghai survey from 1982 showed that among 681 elderly aged sixty or over, only 7% lived alone; 35% lived in a nuclear family (e.g., with their spouse); and 53%, again a majority, lived in a stem or joint family. (Liu Ying, "Tianjin . . . ," 79.)

These figures do not reveal the complete picture, however. A higher number of elderly lived with one or more grown children than the above statistics would suggest—for a substantial proportion of them lived with an unmarried child (which would have placed them in the column of "nuclear households"). A 1984 Tianjin survey of 1,032 elderly shows this clearly. The pool of interviewees included only those elderly people who had one or more already married children. Of these elderly, 44 percent lived with a married child, but another 31 percent of them lived with one or more unmarried children. Only 24.5 percent of them lived just with their spouse or by themselves.^[37]

Other surveys suggest, moreover, that as the elderly became increasingly aged, they were ever more likely to enter the family of a married child. A 1986 survey of 835 truly aged Shanghai residents—over eighty years of age—revealed that fully 80 percent of these octogenarians and nonagenarians lived with married children,^[38] far greater than the 52-53 percent of Shanghai elderly sixty years old and over who were living with married children.^[39]

Opinion polls reveal that the preferences of the elderly run parallel to their actual living arrangements. The more aged they are, the more apt they are to desire to live together with married children. A 1985 survey in Shanghai (N = 2,556) found that whereas 52 percent of the elderly aged sixty to sixty-four preferred to live in the vicinity of, but not with, their children (i.e., preferred a "networked" situation), that number had dropped to 25 percent among respondents aged seventy-five and older.^[40]

[37] Hu Ruquan, "Cong Tianjin . . . ," 103. If the elderly people who did not have any married children are added to this pool of interviewees, and if we ask simply what proportion of all the elderly lived in the same household as a grown child, the 1984 data from Tianjin show that fully 71% of the elderly were residing with children. Only 6% were living alone, another 17% were living with a spouse, and 6% were living with a grandchild. (Some in this last category, with or without spouse, were living with small grandchildren sent to them by working parents to be tended full-time.) Hu Ruquan, 97.

[38] Yuan Jihui et al., 76. According to this survey, a further 10% of those eighty and over lived in a nuclear household, in most cases presumably consisting of themselves and their spouse. Only 8% lived alone.

[39] To be sure, the two sets of surveys are not entirely comparable, inasmuch as a sizeable number of the Shanghai elderly over sixty presumably would still have been living with unmarried children; and these children almost invariably would have married by the time their parents were in their eighties. This can be observed in yet another Shanghai survey (N = 3,614), this one from 1985, in which it was found that 30% of the elderly between the ages of sixty and sixty-four lived with unmarried children, compared with 10% for the surveyed elderly over the age of eighty. In this 1985 survey, 47% of the elderly aged sixty to sixty-four were living with married children, compared with 62% of the octogenarians (Gu Quanzhong, Chen Yangming, and Qian Fang, "Status and Needs of the Elderly in Urban Shanghai: Analysis of Some Preliminary Statistics," Journal of Cross-Cultural Gerontology 2 [1987]: 180).

[40] Ibid., 182.

A different 1985 opinion survey, conducted in a Shanghai neighborhood, shows this change in preferences clearly (table 2.8).

Among these interviewees, as one would expect, a rather higher proportion of widows and widowers desired to live in a married child's family than was the case where spouses were still alive. Perhaps more surprising, according to this 1985 Shanghai survey, 87 percent of widowers (N = 37), but only 64 percent of widows (N = 214), preferred to live with a married child.^[41] If this finding from a small sample is true more generally, we can surmise that widowed men may feel more at a loss to cope for themselves, and considerably less self-reliant in terms of their own domestic life, than do widowed women.

In addition to the two factors of very advanced age and widowhood, survey data reveal a third salient factor that strongly predisposes some of the elderly to seek to live with a married child's family—if they do not receive a pension after retirement, or, if they do, if that pension does not amount to enough to make them financially self-sufficient.

Pensions in urban China normally range from 60 to 80 percent of preretirement pay,^[42] and given the steeply inclined nature of state-sector salaries, even a couple in retirement often earns more than a married child's household with two wage earners. But not all elderly people are so fortunate. A great many urban retirees are not entitled to pensions, particularly those who had been employed in "collective" neighborhood enterprises rather than the state sector; and the great majority of these are retired women.^[43]

[41] As is obvious from the above statistics—37 widowers and 214 widows—Chinese urban women, to an overwhelming extent, tend to outlive their husbands. The five-cities survey of 1982-83 (518) found that 13% of all ever-married women are widows—and this in a survey that included women of all ages.

[42] Yuan Fang, 117.

[43] The consequence is that among Shanghai octogenarians, the great majority of whom are women, only 42% receive pension incomes, while 58% do not. (Yuan Jihui et al., 76). An excellent general study of the pension system is Deborah Davis-Friedmann, "Chinese Retirement: Policy and Practice," Current Perspectives on Aging and the Life Cycle 1 (1985): 295-313.

To be sure, a fair proportion of the elderly obtain new employment after retirement: as of the mid-eighties, 27 percent of retirees in Tianjin and 22 percent in Beijing had taken up paying jobs.^[44] But that still leaves a large number of elderly without independent incomes. Since most retired men hold pensions, in most cases nonpensioned women can become dependents of their husbands. A 1984 survey in Tianjin found that fully 55 percent of the elderly without incomes (N = 371) could depend on a spouse's income; but, lacking that, 44 percent depended on children or grandchildren.^[45]

Moreover, not even all pensioners are self-supporting. As of 1984, a survey in Shanghai found that retirees with pension incomes of fifty or more yuan a month (supposedly more than sufficient to support themselves) constituted 48 percent of the survey, and that those with pensions just sufficient for their own needs, between forty and fifty yuan a month, were a further 18 percent—leaving 31 percent of all pensioners with monthly pension payments below forty yuan, which, the author concluded, left them dependent on their spouse or children.^[46] Since the level of pension payouts is determined by length of service and size of final salary, and since women generally are provided with considerably lower salaries than men,^[47] and since they tend to work for fewer years (regulations normally require them to retire earlier), it is not surprising that women constitute the great bulk of these pensioners who obtain totally inadequate payouts.^[48] And from the available data we can deduce that those who were in this situation and who did not have spouses to fall back on tended to move in with the family of a child as a dependent.

Data are more generally available showing a very sharp distinction between households with pensions and those without. In a survey in Wuhan in the mid-eighties, it was found that parents with pensions were four times more likely to live apart from all their children than were the elderly without pensions.^[49] More than that, when attitudinal questions were posed of

[44] Yuan Fang, 120.

[45] Hu Ruquan, "Cong Tianjin . . . ," table, 99.

[46] Zheng Guizhen, 50. A second Shanghai survey, conducted among 6, 715 elderly people in late 1985, found that of those on pensions, 6% received 100 yuan or more; 20% received 71-100 yuan; 27% received 51-70 yuan; 32% received 31-50 yuan; and fully 15%, mostly old women, received 30 yuan or less: in other words, one out of seven pensioners obtained extremely meager monthly incomes (Gu, Chen, and Qian, 173).

[47] A 1982 survey of a Tianjin neighborhood found, for example, that whereas 29% of all employed wives earned less than 40 yuan a month, only 6% of all employed husbands earned that little (Wu chengshi . . . , 69).

[48] On this, see Gu, Chen, and Qian, 173.

[49] Cai He, 96.

married children in stem and joint families, fully 93 percent of the children in such households where the elderly were without pensions expressed a willingness to retain the expanded family over the long term, compared with only 54 percent of those in stem or joint families where the parents were eligible for pensions and other benefits.^[50]

There is a fascinating exception to this general tendency: retired cadres. Surveys during the eighties showed that the married children of retired officials lived at their parents' home at a far higher rate than was the case with other strata. A 1984 survey of 251 households of retired cadres in Tianjin found that fully 84 percent of them lived with their children.^[51] A survey in 1983-84 of 570 households of retired cadres in the northern city of Harbin similarly found that, extraordinarily, fully 93 percent were living together with children.^[52] Moreover, these retired cadres reportedly continued to be described by themselves and by their families as the "family head," to a far greater extent than was true in the stem families of other social strata.^[53] The reason is probably as follows: it is a commonplace that officials can continue to exert a measure of bureaucratic power and influence even after retirement, enabling them to continue to be the pivotal movers within their families, much like the paterfamilias of old.

This is not to aver that the elderly in other families tend to lack much say in family affairs. In a 1985 Shanghai survey of 961 people aged sixty and over who resided in stem families, fully 87 percent felt that their views were listened to in the home, whereas only 10 percent felt that they had no say. Moreover, the great majority responded that they felt their children were appropriately filial.^[54] Only 16 percent reported that their household was disharmonious.^[55]

[50] Ibid.

[51] Hu Ruquan, "Cong Tianjin . . . ," 98. Among the 16% of retired cadre households whose children all lived apart, 14% had children living in the same city; in another 2% of the cases the children all lived in a different city or region; and in 0.4% of the sample the retired cadre had no children (ibid.).

[52] Ibid. (a separate table on the same page); 0.5% of the retired cadres in Harbin were living alone; 5% were living with their spouse; another 6% with a child or children but without a spouse; fully 54% were living together with spouse and one or more children; and 33% were living with spouse, one or more married children, and grandchildren.

[53] Yuan Fang, 117.

[54] Ling Bi, "Shanghai shide laoren jiating" (Families of the elderly in urban Shanghai), Shehuixue yanjiu (Sociological Research), no. 4 (1986): 92. Also Zheng Guizhen, 136-37. Also see Liu Bingfu, "Wo guo jiating zai 'hexinhua' ma?" (Are Chinese families 'nuclearizing'?), Xuexi yu shijian (Study and Practice), no. 5 (1986), reprinted in Shihuixue , no. 3 (1986): 64. Also Hu Ruquan, "Cong Tianjin . . . ," table, 101—a Tianjin survey of 1,064 elderly in stem or joint families, in which 72% reported that their children/grandchildren were very respectful, and 26% that they were comparatively respectful, while a mere 2% related that the younger generation were not respectful (bu da zunjing ). A 1988 survey of several smaller cities in Jiangsu province came up with quite similar findings: 65% of the elderly reported that their children were quite filial; 30% said the children were average in this respect, and only 5% reported that the children were not filial. Among these interviewees, 89% claimed harmonious relations with their children and only 11% reported conflict (maodun ). Ye Nanke and Tang Zhongxun, "Laonianren shenghuo zhiliang chutan" (A preliminary investigation into elderly people's quality of life), Renkou yanjiu (Population Research), no. 6 (1989), republished in Shehuixue (Sociology), no. 1 (1990): 124.

To the extent that their physical abilities allow, in almost all such stem households the elderly pull their own weight by taking care of the housework, minding the grandchildren, doing the marketing, and so forth, while the younger couple is off at full-time jobs. Inasmuch as the elderly handle most of the household's daily purchases during these daylight hours, they also normally end up controlling most of the household's money for daily living expenses; in fact, the 1985 Shanghai survey found that they did so in fully 88 percent of the households.^[56] In other words, the younger couple hands over a good portion of their income, and in many cases a majority of it, to the elderly. The elderly not only feel they get respect; they even hold responsibility for most of the household budget.

In short, while a heavy preponderance of middle-aged couples and even a majority of the elderly prefer to live apart from all their married children, they do not do so because the alternative is in some respect onerous. Most of the elderly who do live in stem families seem to find it a relatively satisfactory and harmonious arrangement. And for those who are truly aged or infirm, it is a needed lifeline.

Here perhaps lies the rub in the government's enforced one-child family policy. It was observed earlier that the "networked" family of separate but linked households provides an increasingly popular living arrangement for the middle-aged parents of married children, and that it perhaps will become increasingly popular in future when two middle-aged couples will have to "share" one couple of married children. But what will be the situation several decades further into the future, when two pairs of quite elderly couples, or two widows, feel a need to move in with their married child's family, in circumstances where the elderly of both sides of the family share only one set of married children? A great many elderly, against their will, may be obliged to retain a "networked" form of family organization into advanced old age. It promises to pose one of the great social dilemmas in urban China's future.

[55] Ling Bi, 92. Of this 16% of the elderly who reported a disharmonious stem or joint household, 43% felt the problem stemmed from tense mother-in-law/daughter-in-law relations, and a further 21% from conflicts between the elderly parent(s) and their married child. All told, these statistics suggest, though, that only some 7% of the elderly in stem households faced acute mother-in-law/daughter-in-law conflicts, a lower proportion than the scholarly literature on Chinese culture would have led us to predict.

[56] Ling Bi, 91. A breakdown of this 88% shows that 70% were women and 18% men. (On the elderly's management of household funds, also see Hu Ruquan, "Cong Tianjin . . . ," 99).

Three
Urban Households: Supplicants to a Socialist State

Deborah Davis

Household size and composition are shaped by demographic, economic, and cultural factors. Mortality and fertility rates determine the number of potential members as well as the relative size and completeness of each generation. Availability of housing and the ease with which individuals move between homes or between cities affect the composition and stability, while culturally specific "rules of family formation" decide which kin are most likely to be coresident.^[1] To the extent that state policies create or alter any of these constraints on living arrangements, government actions and priorities decisively shape family life. In much of the scholarly work on the family, however, the state has been assumed to be a secondary actor, and primary emphasis has been given to how increased urbanization and industrialization alter the economic utility of family relations or how new ideals emerge from long-term secularization or altered market incentives.^[2] In China, and especially in urban China under the governance of the Communist Party, such assumptions do not hold because the state has assumed a central role in establishing the economic, cultural, and even demographic parameters that decide the size and composition of urban households.

In 1955 the government established a national system of household (hukou ) registration as part of a plan to control rural to urban migration and to administer food rationing. As a result, all changes of address or any alterations in household membership were registered at the local police sta-

[1] John Hajnal, "Two Kinds of Pre-industrial Household Formation Systems," Population and Development Review 8, no. 3 (Sept. 1982): 449-94.

[2] Classic statements of these two positions are found in William Goode, World Revolution and Family Patterns (New York: Free Press, 1970), and Gary Becker, A Treatise on the Family (Cambridge: Harvard University Press, 1981).

tion, and any residential move required police approval.^[3] A new Marriage Law required children to support (and house) elderly parents, and only elderly without responsible kin could turn to social welfare agencies to support independent living or institutional shelter.^[4] During the massive push to collectivize the urban economy in 1958 virtually all women under the age of forty-five were required to work full-time outside the home, and as a consequence the career of full-time housewife disappeared among women under the age of fifty, and the typical urban household became one with two working parents.^[5]

Equally decisive was the bureaucratic system of allocation that replaced the markets of the precommunist era and transformed urban residents from consumers and employees into clients of a socialist state.^[6] In the case of housing, the shift from market to plan had immediate and far-ranging consequences for urban households. The building of new housing became the responsibility of municipal housing bureaus and individual enterprises.^[7] Allocation of new apartments and redistribution of old flats depended on bureaucratic decisions rather than on a family's ability to pay. Households that felt they were inadequately housed or preferred a new location applied to the housing office of their neighborhood or their place of employment; then they waited until they were informed of a favorable decision. In rural

[3] In the first years after the CCP victory, citizens could change residence quite freely. But with the strict enforcement of the 1955 household registration (hukou ) system after 1958, residential mobility between and within cities greatly declined. In a study of three residential districts of Shanghai, it was reported that 35 percent of the urban population moved in 1954 and 43 percent in 1955, but by 1958 the rate fell below 10 percent. Christopher Howe, Employment and Economic Growth in Urban China 1949-1957 (Cambridge: Cambridge University Press, 1971), 65-66, 72. In their study of Guangdong in the 1970s Martin Whyte and William Parish found that only 10 percent had changed residence in the past five years. Urban Life in Contemporary China (Chicago: University of Chicago Press, 1984), 21. Comparison of overall rates of urban growth in Beijing, Shanghai, and Tianjin also document a sharp decline after 1957 and before 1982. See Judith Banister, China's Changing Population (Stanford: Stanford University Press, 1987), 302-3. For a description of how the system operated, see Lynn T. White, Careers in Shanghai (Berkeley: University of California Press, 1978), 148-54.

[4] Deborah Davis-Friedmann, Long Lives (Stanford: Stanford University Press, 1991), 34-60.

[5] Whyte and Parish, Urban Life in Contemporary China , 201.

[6] For the fullest exploration of the evolution and consequences of this process, see Andrew Walder, Communist Neo-Traditionalism (Berkeley: University of California Press, 1986). Also see chapter 8 by Whyte in this volume for the consequences for choice of mate.

[7] In the early eighties 15 to 20% of urban families still owned homes purchased in the years immediately before or after 1949; however, the majority of urban residents lived in public housing allocated either through their place of employment or a municipal housing bureau. In December 1981, 54% of urban housing was owned by units, 28.5% by municipal real estate bureaus, and 17.5% by private individuals. Beijing Review , Oct. 11, 1982, 27-28. In Wuhan in 1985, 21% of housing was owned by real estate bureaus and 62.7% by enterprises. Wuhan Nianjian 1986 , 566.

China, where families were responsible for building their own shelter, people wanting more space raised their savings rate as a primary strategy for better housing.^[8] By contrast, urban residents faced with a noncommodified, welfare system of housing developed strategies to present themselves as deserving supplicants.

After 1980 Chinese Communist Party (CCP) leadership under Deng Xiaoping inaugurated a development policy that initially promised urban residents significant departures from the Maoist political economy. Consumer industries were given a new priority, vast sums were invested in residential construction, and there was a revolution in the comfort and convenience of urban life. Average living space rose from 4.96 square meters per urban resident in 1980 to 8.7 square meters in 1988, and the percentage of households that had less than 4 square meters per capita fell from 33 percent to 15 percent.^[9]

Increased availability of consumer goods and more plentiful housing, however, did not signal a renunciation by the CCP leadership of their role as social engineers. On the contrary, with the inauguration of the campaign to universalize a one-child family in 1979, the state assumed the right to dictate the fertility of every urban woman and thereby control family size more completely than any previous regime. In the long term, this decision would produce urban families with no siblings and very high ratios of elderly as a percentage of family dependents. But in the short term, the impact of the one-child campaign, especially as it affected household composition and the processes of household formation, is more difficult to predict. Knowing that a first child is a last child, will young couples remain in multigenerational homes longer than did earlier generations, or will they experience a heightened desire to set up an independent household? How will the absence of a son affect the solidarity of marriages, or the absence of a grandson alter obligations between parents and adult children? Will sustained rates of low fertility foster the rapid proliferation of small households of two and three members, or stabilize multigenerational arrangements?

Drawing on a range of recent Chinese publications and a longitudinal study of one hundred families in urban Shanghai interviewed first in 1987 and then again in 1990, this chapter addresses several of these questions about the impact of the post-Mao reforms on household size and composition in Chinese cities. The first section introduces the neighborhood

[8] Thus between 1957 and 1978 rural families on average spent less than 4% of their incomes on housing, but as their incomes rose not only did the absolute amounts invested in housing increase but so did the percentages. Thus, for example, in 1978 they spent on average 12.4%, in 1986 14.4%, in 1988 14.9%, but only 8.2% in 1989, a year of economic downturn. Zhongguo Tongji Nianjian (henceforth ZGTJNJ ) 1989 , 743, and Beijing Review , March 19-25, 1990, 39-40.

[9] ZGTJNJ 1988 , 806, and ZGTJNJ 1989 , 756.

where I conducted the interviews and describes the general changes in household size and composition that occurred between 1987 and 1990. The second section focuses on one particular characteristic of these households—a strong "tilt" toward joint living with sons—and analyzes the cultural and practical supports for the predominance of this apparently traditional arrangement in an era of rapid economic and social upheaval. In the third and last section, I briefly review the housing policies of the 1980s and then look closely at which families in my sample initially drew the greatest benefits. In this way the chapter explores not only the impact of the reforms on urban household composition but also raises questions about the emergence of new inequalities and class divisions. The underlying theme unifying the chapter is the degree to which the reforms have altered the dependency of all urban families on the state. More narrowly, the chapter focuses on the question, To what extent and in what ways did the process of household formation in large metropolitan cities diverge from that of the late Mao era?^[10]

Household Formation in Urban Shanghai

In 1987 as part of a multiyear study of occupational mobility in urban China, I interviewed one hundred Shanghai women born between 1925 and 1935. In addition to material on work histories of the respondents, their husbands, and their children, I also gathered information on household arrangements at the time of the interview and at the time of each respondent's first marriage. In July 1990 I returned to the original interview site to update the information on the household arrangements of the seventy-three women who could be reinterviewed and to inquire systematically about the housing arrangements of all their married children (N = 188). ^[11]

As a group, the household heads in my survey represented the upper reaches of the working class and the lower rungs of middle management.^[12] Only 5 percent of the fathers and 15 percent of the mothers worked outside

[10] In 1984 the official definition of urban life was relaxed to include suburban and smalltown populations, which had been excluded from urban statistics since 1963. However, in this chapter when I cite changes in urban household figures I am drawing consistently from tables that identify the city core of urban areas, and with my Shanghai survey I am generalizing to family situations typical in the 31 large metropolitan cities with one million or more inhabitants, a population that as of 1986 equaled approximately 46 percent of all nonagricultural population living in the largest 354 cities and towns. Zhongguo Chengshi Tongji Nianjian 1987 , 1.

[11] Funds to support field research came from the Committee on Scholarly Communication with the PRC and Yale University. For professional support and essential introductions, I wish to thank the Retirement Committee of the Shanghai General Union and the Putuo District Government.

[12] The percentage breakdown among respondents and their husbands by type of current or last job was as follows:

the state sector, and 38.5 percent and 18 percent, respectively, were Party members. All families lived in a new residential neighborhood built in 1979, where each apartment had two rooms, a balcony, and a private toilet.^[13] The estate represented good housing, and my respondents should be seen as successful supplicants in the bureaucratic allocation of scarce resources.^[14] When I first interviewed them in June 1987, almost all expressed high levels of satisfaction with their current arrangements, and in only three cases was I able to identify families who were actively looking to move to better apartments.^[15] However, when I returned in 1990 to reinterview the original one hundred respondents, I found that fourteen had moved to a new neighborhood, two had died, and most astounding to me, 60 percent of those who agreed to be reinterviewed had altered household membership.^[16] In 23 percent of these families households became larger, in 34 percent they had become smaller, and in 3 percent they were the same size but included

[13] In 1987 there were 3,016 residents living in 814 households.

[14] Although all apartments in the estate had only two rooms, there were two different configurations of the rooms, so that one gave 27 square meters and the other only 24. At the time of the 1987 interviews families in the larger apartments averaged 5.8 square meters per person, those in the smaller apartments 4.8 square meters. By way of comparison, one can note that in 1985 Shanghai households averaged 5.5 square meters per person, while nationwide the average was 7.46 square meters or 6.1 square meters depending on how you defined city residents. Hu Renxi, "Shiqu gaoling laoren zhuzhu qingkuang chouyang diaocha" (Sample survey of the housing situation of the very old in the urban districts), in Gaoling laoren wenti yanjiu , ed. the Research Group of Shanghai Municipal Committee on Problems of the Elderly (Shanghai: Shanghai shi laonianxue xuehui mishuzhu, 1987), 71-80; ZGTJNJ 1988 , 806; China Reconstructs 38, no. 8 (Aug. 1989).

[15] All three households were headed by professionals (a head engineer, a surgeon, and an import-export executive), and each had only two children. By June 1990 the first two had moved out of the estate and the third had both children working in Hong Kong and appeared to be making plans to emigrate themselves.

[16] Besides those who had moved or died and were therefore not available for the reinterview, three were in the hospital, four were traveling outside Shanghai, and four were at work on the scheduled interview days. A comparison of the household profiles of the Shanghai reinterviewees with the original sample of one hundred respondents, however, indicates no serious problem of selectivity. Those who were reinterviewed had lived in only marginally larger households (4.65 as opposed to 4.57 persons) and in both groups a majority had lived in three- or four-generation homes: 57% among those reinterviewed, 54% in the entire 1987 sample.

different people. Equally noteworthy was that whereas only 11 percent of the three-generation households became generationally "simpler" by contracting to either one- or two-generation households, 31 percent of the two-generation families had become generationally "more complex," expanding to three-generation arrangements through the addition of a new grandchild, an elderly in-law, or a coresident niece or nephew (table 3.1).

Because of their rather uniform age, favorable occupational status, and high rates of Communist Party membership, these Shanghai families cannot represent the urban population, and readers should not conclude that annual turnover rates of 20 percent in household membership were the norm between 1987 and 1990. However, I do think that the experience of these families is typical of urban families at an especially transformative stage of the life cycle. All households in this study were selected because they included a woman born between 1925 and 1935, thereby guaranteeing that most respondents would be able to give information on both a parent and a child who had worked for a full decade after 1949. This focus on households with a mother in her late fifties or early sixties also meant that respondents and their husbands were at that stage of life where they were assuming new roles of in-law and grandparent but were also still fulfilling the role of child to an elderly parent and of parent to an unmarried child. Therefore, while my sample cannot represent the total urban population, it does speak to a universal stage of the life cycle where the processes of family formation and household division are typically most dynamic.

In Unger's table summarizing household composition among Beijing women of five different cohorts, those between ages forty-six and fifty-nine had markedly different households than those between the ages of sixty and sixty-nine (see chapter 2, table 2.3, of this volume). Specifically, while the modal household for the younger women was nuclear, that for the older women was stem. In addition, Unger's finding of greatest levels of coresidence among both the neediest pensioners and the wealthiest high-ranking

cadres coincides with what I observed in Shanghai. That is, when parents were either especially impoverished or particularly privileged, Unger's surveys revealed large and complex households because there were strong expectations that the wealthier or more fortunate would shelter the needy and poorer. This expectation, which Ikels has so aptly called an unwritten contract of entitlement (see chapter 12 of this volume), encourages multigenerational homes when, as a result of real or relative poverty, one generation is perceived as unable to provide adequately for itself. Because my sample targeted families where adult children were in the prime marriage years and elderly parents likely to be frail, the study caught individuals at a point in life when demands for intergenerational reciprocity and entitlement were especially numerous. No national survey will produce the same frequencies of coresidence, but a national survey targeted at this same age group in major metropolitan cities would look quite similar.^[17]

Life Cycle Dynamism and Intergenerational Obligations: Three Cases

Among my respondents, as among those interviewed in Chinese surveys described by Jonathan Unger, the ideal housing arrangement was an independent household for each nuclear family. The housing goal of most of my respondents and their children, therefore, was a separate home for each married couple and shared homes only when elderly parents needed care or companionship. But, as was clear in the pattern of household arrangements among my respondents in 1987 and in 1990 (tables 3.1 and 3.2), few families realized the ideal. Instead, as the following three case histories illustrate, persistent housing shortages and intergenerational obligations

[17] A second consideration about this sample is the uniformity of the neighborhood. Nationwide, depending on the specific city, between 50 and 60% of urban residents live in apartments built by their employers, and only 20 to 30% live in municipal housing. Without parallel samples drawn from enterprise estates or company towns I cannot know the consequences of this difference for patterns of household composition. But given the more anonymous environment of city housing estates and the "entrepreneurial talent" that allowed these families to trade previous housing into what was perceived in 1979 to be excellent housing, we might hypothesize that my respondents were somewhat more aggressive than the average urban resident in mobilizing resources to improve their housing situation. A third concern is the uniqueness of Shanghai. However, in terms of crowding, one of the best indicators of intercity differences, Shanghai by the late 1980s was not very different from other large metropolitan areas. In 1986 the average per capita space in all urban homes was 5.9 square meters, in Shanghai it was 6.0, in Beijing 6.3, in Tianjin 5.9, and in Wuhan 5.7. Average household size in Shanghai also did not vary from the national norm, suggesting that while Shanghai was much more densely populated than average (2,118 persons per square km as opposed to 118 in all of urban China in 1990), in terms of availability and crowding of residential space it was not very different from the other major cities. Zhongguo Chengshi Tongji Nianjian 1987 , 19, 27, 29, 31, 32, 35, 37, 39, 40, 291, 323, 331, and Beijing Review , Nov. 19, 1990, 11.

to less fortunate members repeatedly overrode the desire to establish nuclear households, and the norm was multigenerational homes and persistent complexity.

The Lis (a pseudonym)

When I first interviewed Mrs. Li in 1987, she was fifty-eight years old and just recently retired from her job as a daycare worker in a textile mill. Originally she had come from Ningbo, where her father had been a shop clerk. But after her marriage in 1948, she and her husband had moved to Shanghai. When they were first married, they rented a fourteen-square-meter flat near the Huangpu River. Later three children and her husband's mother joined them, and by 1979 they were designated a "hardship case" qualified for quick resettlement.

When they moved to their new apartment the household was officially a three-generation family of six members: Mr. and Mrs. Li, his widowed mother, and three unmarried children between the ages of twenty-five and thirty-one. In fact, Mr. Li slept in his office, and the daughter had a bedspace at her factory, so the actual household consisted of the two sons in one room, and the mother-in-law and Mrs. Li in the second room. Then, within six months of their move to this apartment, the eldest son married,

and he and his bride took over the room that he had previously shared with his brother, while the younger son set up a cot in the hallway.

Two years later, the older brother, his wife, and their small daughter qualified for their own apartment on the grounds of hardship. But as soon as this couple moved out, Mrs. Li's second son married and immediately moved his bride into the room just vacated by his brother's family.

Mrs. Li's daughter had even greater difficulty in establishing her own household. Married in 1982, she and her husband were unable either to move in with in-laws or find a room of their own. As a result, for more than seven years they each lived in their factory dormitory, while their daughter was raised by the husband's mother. Finally in 1989, at age thirty-five and with seventeen years work experience, the daughter was assigned a two-room apartment with a private bath and kitchen in a housing block built by her husband's employer.

When interviewed in July 1990, Mrs. Li maintained a household of four: herself, her second son, his wife, and their son. Her mother-in-law had died, and her husband continued to spend most nights at his office. However, during the summer the household was even more crowded than the official registration booklet indicated because Mrs. Li was keeping the children of both her elder son and her daughter for the school holidays. Usually six people were sleeping in the apartment, and most evenings Mrs. Li cooked for eight or nine people, depending on which of the adult members of her family were off from work that day.

Ideally, Mrs. Li and her son would like to have their own homes. But at the moment this family has stopped even trying to find an alternate arrangement. By the criteria of hardship, they are comfortably housed. The building is in good order, and although they live on the sixth floor, they have piped gas, a private toilet, and most important, the married couple has a separate bedroom accessible from a central hallway. Seniority is also unlikely to help either the parents or the son. Although a Communist Party member and a manager, Mr. Li is officially retired, and like his wife he has lost the leverage at the old unit that would enable him to advance a claim for new housing. The son, a manager of thirty-seven and also a Party member, would ordinarily be a prime candidate for a new apartment. But in 1988 he moved to a new job in a joint venture, and although the salary is higher than average and in 1989 he received a major promotion, the Hong Kong management has made it clear that it will never build staff housing.

In the case of the Lis a statistical profile would document gradual "nuclearization" of urban family life. Two of the three children maintain small neolocal homes, and only one son is coresident with his parents. The reality of the several intermediate steps, the long periods of intense crowding and dormitory living, however, quickly contradict any such simple interpre-

tation of nuclear predominance. Rather, the Lis' experience graphically illustrates how in the midst of China's longest and biggest housing boom multigenerational living persisted even as the neolocal ideal animated much of the ongoing maneuvering.

When the Lis moved from their fourteen-square-meter apartment in 1979 they gave up all claims on the room they had rented from the city, and there seemed to have been no opportunity for them to keep title to both the old and the new rooms. However, for many of my respondents, even among those who did not own their own flats, relocation had not required them to relinquish their previous apartment, and the family maintained claims to two places. Thus one strategy that became obvious amidst the greater prosperity and openly entrepreneurial culture of the Deng era was that families shifted members between old and new apartments to maximize their leverage, or simply to minimize the crowding that occurred as children married, grandchildren were born, or elderly parents needed nursing care. This was the experience of the Zhaos, a family that in terms of the formal household-registration system had become less extended between 1987 and 1990, but in reality was still actively involved in a complicated process of reconfiguration.

The Zhaos (a pseudonym)

When I first met Mrs. Zhao in 1987, she was a fifty-seven-year-old mother of two grown sons and temporarily employed as a service worker in a small hotel. Her household was a four-generation home, which included her husband, her husband's mother, her second son, his wife, and their toddler. By 1990 her life had been simplified. She had retired from her hotel job to care for her newly retired husband, and her household consisted only of herself, her husband, and his ailing father aged eighty-five. At first glance, it appeared that the household had become markedly less complex. Further conversation revealed a different interpretation.

Mrs. Zhao was born in a Jiangsu village and came to Shanghai before 1949 as a textile worker. She married in 1953 and immediately moved into the two-room home of her husband's parents. Her husband was an only child, and for the next twenty-five years she, her husband, and their two sons lived with her in-laws. In the late 1970s the city tore down their front room, and as compensation assigned the Zhaos two rooms in a new municipal housing estate. However, when Mrs. Zhao moved, her elder son stayed behind with his grandfather (Mrs. Zhao's father-in-law) in the back room, which had not been torn down, while Mrs. Zhao, her husband, her mother-in-law, and the younger son moved to the new home. In 1987 when I first interviewed Mrs. Zhao, the older son, who had married in 1985, was living with his wife and baby in a three-generation household headed by

her father-in-law. Mrs. Zhao maintained the six-person, four-generation home described above.

How had the household reconfigured by 1990 as a two-generation home of an elderly couple and an even older father-in-law? The process reflects both the new opportunities of the late 1980s and the traditional obligations imposed by the norms of intergenerational reciprocity and interdependence that had also been decisive in the family life of the Lis.

In 1988 Mrs. Zhao's mother-in-law died, and soon after, her eldest son left China to study in Japan. At that point, her father-in-law moved to live with her, and her younger son, his wife, and their child moved back to their old house to live with the older daughter-in-law and her son.

For the Lis and the Zhaos family relations appeared harmonious, and the strategies that shaped their household shifts away from the nuclear ideal were animated by the morality of intergenerational entitlement. In other families, relations were tense, and in an era of greater choice, conflicts split some households apart. However, as we will see with the Wangs, even in situations of conflict, obligations to provide for the needy continued to be a major force sustaining and even elaborating already complex households.

The Wangs (a pseudonym)

Mrs. Wang was born in 1932 on a riverboat in northern Jiangsu province. As a small girl she had come to Shanghai with her mother, and before she was ten she was working full-time in the textile mills. She was married in 1950 to a street scavenger, and the couple set up their new home with both his mother and her mother in a timber and stucco hut that he had rented sometime around 1949. In 1979 the building was condemned, and as compensation they received not only a new two-room apartment but also two single rooms in other locations, one for each marriage-age son. At first Mrs. Wang, her mother-in-law, and her younger, unmarried son lived in one place; the elder son, his wife, and two children lived in another, and the two daughters shared the new two-room apartment. At this time, Mrs. Wang's husband was serving a five-year prison term for crimes committed during the Cultural Revolution and was not a member of the household.

In 1987 when I first interviewed Mrs. Wang the family no longer was living as they had in 1979. Mrs. Wang's mother-in-law had died, and her husband, who had returned from prison, had moved with her to the new apartment, which they then shared with their younger daughter, a son-in-law, and an infant granddaughter. The elder son and his family remained in the room he had received in 1979, the elder daughter had married and was living with her husband's parents, and the younger son was living alone in the room he previously had shared with his parents and grandmother.

Three years later the family had redistributed itself once again. Mrs.

Wang remained in the same apartment she had occupied in 1987, but her husband had moved out and now slept in a loft they had built in the home of the second son. This son had married in fall of 1987, and now he, his wife, and their child occupied the main room, although during the day the child was sent to Mrs. Wang while her daughter-in-law was at work. The younger daughter, the daughter's husband, and their four-year-old child continued to live in the same room they had used in 1987, but now they also shared the kitchen, toilet, and hallway with the older sister, who had moved back with her young daughter after her divorce in 1988. The explicit strategy that Mrs. Wang articulated without any prompting was that by registering as many people as possible in this modest two-room flat, she hoped that within two or three years the younger daughter and her family would be allocated a third room in this estate on the grounds of hardship, leaving Mr. and Mrs. Wang, the divorced elder daughter, and her child in this apartment. With six persons in two rooms, the family averaged only a bit more than four square meters per head. However, none of the four adults in the apartment had a work unit with good housing supplies. Mrs. Wang and her elder daughter were from the worker ranks at a textile mill with no staff housing, and the younger daughter and her husband both worked for a suburban collective that neither built staff housing nor gave cash subsidies for home purchase or rental. Because not one of the four had any chance of gaining their own place by moving up the seniority queue at their workplace and Mr. Wang had no responsible unit as a result of his prison sentence, the only realistic strategy was to accentuate their position as especially needy supplicants and hope that the district housing office would view them as a hardship case worthy of solicitude.

The Reforms and Preference for Sons

In the experience of the Lis, the Wangs, and the Zhaos, the primary concern behind their search for new housing was to reduce crowding and improve the location and amenities of their homes. However, families also pursued strategies to achieve socially optimal membership, which among those forced to accept multigenerational arrangements encouraged coresidence with sons, not daughters, or dependency on the parents of the groom, not of the bride. For the Lis this ideal meant providing for needy sons before a needy daughter, and for the Wangs it meant establishing each married son in an independent home while using their own apartment as a refuge for a divorced daughter and as a temporary solution for a daughter who had married a man with poor housing resources.

This tilt toward sons was also explicitly articulated by Shanghai residents with whom I spoke about housing goals, and here I quote from a con-

versation I had in June 1987 with a middle-aged cadre from my research site.^[18] She had just moved from a fairly spacious home within walking distance of her office to a new apartment that required a long commute, and I was curious about what had motivated a move that had created such inconvenience.

Q: But, from what you have said, your new apartment doesn't really sound like an improvement.

A: Well, my husband and I will retire in the next few years, and our son will probably get married soon.

Q: But you have at least five more years in this job, and your husband won't retire until 1997. Is this new apartment especially close to the workplace of your future daughter-in-law?

A: Oh, no, my son doesn't even have a [girl] friend.

Q: Then I really don't understand; why did you take this apartment?

A: It is the layout. In our old place we had two bigger rooms, but you had to walk through one to get to the other. When our son does marry, it will be much better for each couple to have separate doorways. If we didn't take this place while we could, he would have a harder time finding a wife.

Additional evidence of the expectation of joint living with sons was found in the household arrangements of the 188 married children of my Shanghai respondents. Among those living with parents at the time of the interview, sons were more than twice as likely as daughters to be coresident (see table 3.2, lines 6 and 7), and among all married children unable to establish an independent home at the time of marriage, 84 percent lived with the groom's parents (table 3.3). Moreover, in only one year between 1979 and 1989 did more than 25 percent of coresident children deviate from this virilocal pattern.

Yet, as we saw in the Li family, some young couples who lived jointly immediately after marriage later established their own independent homes, and therefore it is possible that the tilt toward sons captured in table 3.3 misrepresents long-term trends. However, a comparison of housing arrangements among firstborn sons at the time of their marriages and again as of July 1990 suggests that the male bias persisted even after couples left a parent's home. For example, if the couple moved into an apartment arranged by their parents either at the time of marriage or in subsequent years, the apartment was far more likely to have been set up by the hus-

[18] This sentiment was also clearly articulated at a seminar in November 1990 at Yale when two former residents of China drew on their own early years of marriage in 1980. In both cases they recalled direct pressure not to move to the home of their wives' parents, saying that their colleagues would consider such a move, even when it meant having more space, a "disgrace." One also said that the patrilateral tilt even extended to avoiding taking an apartment in his wife's unit. In that case, he said, the problem was that he did not want to be a subordinate in her unit, and his unit should not be subordinate to her unit.

band's parents than by the wife's. Or, if their own employer supplied housing, it was the man's unit more often than the woman's that served as the landlord (table 3.4).

While to some observers of Chinese society a concern about the degree of a male preference might seem uninteresting given China's traditional patriarchal culture and patrilineal ancestor worship, to me the predominance of coresidence with sons and the greater reliance on parents or employers of the groom raised central questions about the effect of the Deng reforms on family life. But before addressing these larger questions, let me explain briefly why this "tilt" toward sons in housing arrangements struck me as worthy of further scrutiny. First, from my respondents' reports of their early married life, I discovered that the postmarital households of my respondents who had married in the 1940s and early 1950s differed very little from the arrangements of their sons who had married in the 1970s and 1980s. In fact, to the extent that there was any shift, it was toward greater not less virilocality. At the time of their own marriages 42 percent of my respondents had lived with the groom's parents and only 9 percent with the bride's. Among their firstborn sons, however, the percentages were 52 percent and 2 percent, respectively. In short, based on a comparison of hous-

ing arrangements immediately after marriage, it appeared that after many decades of CCP hegemony in which one might have presumed that ideological supports for traditional preferences would disappear, the onus of providing shelter continued to be a male or patriline responsibility.^[19]

Second, although an examination of household arrangements among these different cohorts of newly married couples revealed continuity with traditional expectations, other material suggested that a preference for sons was not simply a continuity with the precommunist past, but possibly represented a more pronounced "tilt" toward sons after 1980. During a 1979 trip to China, where I interviewed urban elderly in Beijing, Shijiazhuang, and Changzhou, I found strong evidence of departures from past preferences for coresidence with married sons.^[20] In addition, among children of the Shanghai respondents who had married between 1971 and 1978, only half had moved in with in-laws, and of these 34 percent of the coresident couples had departed from virilocal practice and lived with the wife's parents (see table 3.3). Therefore, I was quite struck both by the higher rate of coresidence and by the lower rates of uxorilocal living among the most recently married.

By examining how urban residents have benefited from the Deng economic and political reforms, the possibility that the late 1980s may have seen an increased emphasis on traditional preferences becomes more understandable. Between 1978 and 1988 city residents experienced a dramatic

[19] This was also the bias that Whyte and Parish found in their study of Guangdong families in the 1970s (Urban Life in Contemporary China , 144-45).

[20] Davis-Friedmann, Long Lives , 47-59.

improvement in their standard of living. Urban wages tripled,^[21] rationing of most foodstuffs disappeared, and items that had been scarce luxuries became almost commonplace. But what did this new affluence mean for family life? Most immediately it incited a consumer revolution that dramatically redefined the external markings of social status. After 1980 status came increasingly to be defined in terms of material possessions. In most of the families I visited, older members were less obviously affected by the new materialism, but in decisions about wedding expenditures, middle-aged and newly retired parents were as vulnerable as the young. A big banquet, a large trousseau, and a complete bedroom set were the minimum purchases necessary for a socially acceptable wedding. Anything less not only made the family lose face, it also could cause the children to claim parental selfishness or neglect, and thus justify children's distancing themselves from parents and filial obligations.

During the 1970s the official media had praised young couples who celebrated their marriages with a simple tea party and began married life with nothing more than two new winter jackets, two quilts, and a thermos bottle. In practice, urban weddings of the seventies typically had involved far more elaborate celebrations. Most parents hosted a family feast, and brides expected several sets of new clothes, basic home furnishings, and when possible, a new bicycle or wristwatch. Expenses were variable, but on average the groom and his family spent three to five months' income, and the bride and her parents only somewhat less.^[22] Yet in most cities a family's reputation did not absolutely demand an expensive wedding, and it was possible (and of course politically correct) to marry with not much more than a new set of bedding, a table, and two chairs. Therefore, a young couple could, if they wished, marry without any financial obligations or debts to their parents.

The relative affluence of the 1980s and the influx of new consumer goods effectively eliminated the option of a simple wedding and thereby increased the dependency of young couples on their parents. By the mid-1980s an urban wedding averaged between 5,000 and 6,000 yuan,^[23] but most young adults needed at least 80 percent of their monthly salary of 100-120 yuan

[21] In 1978 average wages in state enterprises were 615 yuan; by 1988 they had risen to 1,853 yuan. Zhongguo Tongji Zhaiyao 1987 , 101; ZGTJNJ 1989 , 125, 130.

[22] Whyte and Parish, Urban Life in Contemporary China , 137.

[23] In a twenty-city survey of urban marriages, average costs in 1985-86 were 5,069 yuan, which included 1,000 yuan for the wedding and 4,000 yuan for presents. Gu Jirui, Jiating xiaofei jingjixue , 343. Based on a 1986 survey in 18 large and medium cities, Chinese researchers estimated that for marriages concluded between 1981 and 1985, a groom's family on average needed to save between 39 and 50 months, and the brides' families needed between 24 and 31 months. Qian Jianghong, "Marriage Related Consumption," Social Sciences in China , no. 1 (1988): 221.

just to cover necessities. Unable to accumulate savings of any size on their own, they leaned heavily on their parents to purchase the color television, refrigerator, and washing machine that had become the essentials for any respectable new home of the late 1980s. In light of the exponential growth in the cost of an acceptable wedding and the high levels of dependency on parents among this generation of young urbanites, it is not surprising that parents frequently expanded their home to include the spouse of a newly married child. The question remains, however, why did they welcome a daughter-in-law more frequently than a son-in-law?

If I were to take my respondents' replies to this question, the answer would be quite simply this: "If a man moves into his wife's house, he will have no status or power." In short, what these urban residents emphasized was that Chinese culture is patriarchal and discriminates against females. But this answer, while true and relevant, does not explain why the affluence and greater ideological freedom of the eighties maintained—perhaps even strengthened—these traditional patrilineal preferences, or why these biases were so pronounced among young urban couples establishing new families during the 1980s. However, if we place the explosion of consumerism and the prolonged dependency of urban youth into the larger framework of the post-Mao political and economic reforms, causal links do emerge.

The Deng reforms championed commodification, decollectivization, and privatization, and in rural areas, as Harrell, Johnson, and Selden have documented in chapters 4, 5, and 6 of this volume, the result was a dramatic departure from the household economy of the Mao era. In the cities, by contrast, the reforms had only a minimal effect on the core economic institutions. Households did not become units of production, and housing supplies continued to be a privilege distributed to deserving clients through the municipal governments or places of work. After 1980, however, there were changes in labor policies that affected the relative earning power of men and women, which in turn shaped a parent's choice of coresident child.

In the 1970s the job assignment procedures for young urbanites often ignored educational or gender differences.^[24] In some years all secondary school graduates regardless of training or preferences would be sent to the countryside; in other years, only those with an older sibling already in a village or a state farm could be assigned a city job. Sometimes families could keep a child at home on the grounds that an elderly parent needed care. Where the parent was relatively young or in good health, the family would argue that the new graduate was too physically frail to withstand the

[24] Susan L. Shirk, Competitive Comrades (Berkeley: University of California Press, 1982), 52-84; Jonathan Unger, Education Under Mao (New York: Columbia University Press, 1982), 164-70.

rigors of rural life. Parents worked energetically to protect the interests of whichever child—son or daughter—was scheduled for a job assignment that year.

In this environment, parents with several teenagers could not predict which one would ultimately be the best source of support in the future, or even which child would be allowed to live in the same city as the parents. Therefore, many parents were as intensely involved with their daughters' futures as with their sons', and strategies that favored sons might be discarded. During the 1970s the massive outmigration of secondary and university graduates to manual jobs in the countryside depressed career prospects of sons and daughters. But because males had historically been the primary candidates for leadership positions and high salaries, the labor policies of the Cultural Revolution years created greater relative losses for males. Therefore, in the Deng era, when job assignments more predictably rewarded credentials and discrimination against females was no longer a focus of policy,^[25] parents could logically expect more material gains by living jointly with a son than with a daughter. In the more relaxed ideological climate of the eighties, urban residents were freer to act on traditional preferences for sons, and given the increased discrimination against women it was also quite rational to value a son's financial contribution more highly than a daughter's.

The depoliticization of work rationalized preferences for sons in another way. During the years when political stigma had created especially severe discrimination, daughters often offered more reliable support because the regime did not police female kin ties as closely as those of males.^[26] Under these circumstances, it made sense for some families deliberately to develop a household with the bride's family rather than the groom's. In the more depoliticized atmosphere of the eighties political blemishes no longer so directly determined family wealth and status. Daughters therefore lost their previous comparative advantage, and elderly parents and young adults would spontaneously, and without embarrassment, tell a curious foreigner, "It is shameful for a man to move to his wife's home."

In the 1980s, as in the 1970s, housing shortages denied many newlyweds the independent home they desired. But daughters faced the same housing crisis as sons; in fact, because many units provided housing for male workers before female workers, the needs of daughters could actually be greater.^[27] Daughters were also as reliable a source of care as a son and daughter-in-law. Thus while necessity and external constraints explain why

[25] Emily Honig and Gail Hershatter, Personal Voices (Stanford: Stanford University Press, 1988), 244-63; Stanley Rosen, "Women, Education and Modernization," in Education and Modernization: The Chinese Experience , ed. Ruth Hayhoe (Elmsford, N.Y.: Pergamon Press, 1992).

[26] Davis-Friedmann, Long Lives , 78-79.

[27] Honig and Hershatter, Personal Voices , 140.

high percentages of urban households included two married couples, cultural preferences and discrimination against women explain the bias in favor of sons.

The Unequal and Uneven Outcomes of Housing Reforms

In China between 1950 and 1975, urban housing policy went through several phases. In years when municipal and enterprise leaders decided to make housing a priority there was a steady increase in the supply of cheap housing. But in years when "more productive" investments were given priority (as was the case between 1961 and 1972), little was spent on new urban housing.^[28] Young couples or single adults could only find new quarters as members of larger, multigenerational households, and even couples with several young children were often required to share a suite with another young family in an apartment designed for one household.^[29] As a result, by the late 1970s crowding was almost as acute as it had been at the end of the civil war, and average per capita living space was actually 20 percent lower than it had been in the early 1950s.^[30]

The post-Mao leaders responded to this crisis with alacrity. In 1980 as much new urban housing was built as had been constructed during the entire Great Leap Forward, and one government source estimated that between 1979 and 1981, enterprises and city governments had rehoused the equivalent to the entire populations of Beijing, Shanghai, Tianjin, Guangzhou, Wuhan, and Shenyang.^[31] Surprisingly, however, given the rhetoric

[28] Reeitsu Kojima, Urbanization and Urban Problems in China (Tokyo: Institute of Developing Economies, 1987), 36-37.

[29] Liu Yirong, "Yinggai zhongshi yanjiu yishihu de sheji wenti" (We must research the problem of one-room apartments), Jianzhu xuebao , no. 5 (1981): 61-63; Shen Kuixu, "Zhengque quli zhujai sheji yu fen pei di guanxi" (Resolve the relationship between housing supply and allocation), Jianzhu xuebao , no. 12 (1981): 59-64.

[30] In August 1955 rents that previously had been set to provide maintenance and replacement were lowered in absolute terms. They were again reduced during the Cultural Revolution, thereby further widening the gap between expenditures and rents. Nicholas Lardy, "Consumptlon and Living Standards in China, 1978-83," China Quarterly , no. 100 (Dec. 1984): 856. One official Chinese source estimated that in 1979 average rents were 0.10 yuan per square meter, while cost to state was 0.38 yuan. "Urban Housing Rental," Beijing Review , Oct. 11, 1982, 27-28. In 1978 in 182 large cities households averaged 3.6 square meters per person; in 1952 the average had been 4.2 square meters. People's Daily , Aug. 5, 1980, 5, translated in FBIS , Aug. 20, 1980, 15-22. On rate of investment 1951-81, see Reeitsu Kojima, Urbanization and Urban Problems in China , 36-37.

[31] For 1953-1984, see Kojima, Urbanization and Urban Problems , 37, and Beijing Review , May 3, 1982, 9. In 1991 the second source estimated that between 1979 and 1989 three times as much housing was built nationwide than had been built in the preceding three decades, and 500 million people moved into new or rebuilt homes. Beijing Review , Jan. 14, 1991, 38.

of commodification and privatization of the Deng reforms, there was no sustained effort to pass on the costs in the form of higher rents. Instead, housing reform, to the extent that it was a reform, consisted of a decision to treble the share of housing in the capital-construction budget.^[32] The low-rent policy and the bureaucratic system of allocation remained in place (table 3.5).

By the mid-1980s, however, new ideas began to circulate, and for a brief period there was a flurry of publicity announcing a fundamental shift away from public housing.^[33] One of the most widely publicized models of change, sometimes called the Yantai plan, was a system whereby each employee was given a voucher to cover the real cost of building and maintaining what was defined as an adequate number of square meters of space (no price appears to have been placed on the land). Those who had more space than was considered necessary for their current household size were then to pay higher rents that approximated market value, and those who had less than adequate space would be given priority for renting or purchasing new housing.^[34] In addition, there were experiments in Yantai, and elsewhere, with outright sales both of newly built homes and of existing enterprise

[32] In 1978 housing took 7.8 percent of the construction budget; in 1981 and 1982 it went above 25 percent. Barry Naughton, "The Decline of Central Control over Investment in Post-Mao China," in Policy Implementation in Post-Mao China , ed. D. Michael Lampton (Berkeley: University of California Press, 1987), 66-67.

[33] Guowuyuan Gongbao (Bulletin of the State Council ), April 20, 1985, 270, and Beijing Review , Nov. 14, 1988, 18-20. A good summary of the several reforms is found in a commercially available neibu publication, Zhou Ying (ed.), Chengshi zhuzhai gaige cankao ziliao (Xinhuashe "Jingji cankao" bienjibu, n.d.; est. date of publication early 1988).

[34] RMRB , July 1, 1988, 5.

facilities.^[35] But by 1989, and particularly after the massacre in Tiananmen, central and municipal-level planners realized that at current wage levels, commercialization in the largest cities would be financially infeasible and politically destabilizing. Reforms were reduced to mild statements in support of gradual rent increases or lower housing subsidies, and plans for massive privatization were shelved.^[36] Housing continued to take a trivial fraction of a family's income, and in most instances one pack of high-quality cigarettes cost more than a month's rent.^[37] In fact, because wages rose steadily over the decade, the actual fraction of monthly income paid to rent fell to under 1 percent (see table 3.5).

In distributing apartments housing officials continued to establish a priority list according to the degree of crowding. However, even with the cutoff points as low as two or three square meters per resident,^[38] more families qualified for relocation than there were new flats available. As a result, housing cadres considered the composition as well as the size of the household in defining the degree of need. If a one-room household contained three generations, or if a two-room home had more than two married couples, the family was given priority. Even with these stringent guidelines, supplies were insufficient, and officials created priorities among the "needy" on the basis of the the family head's rank in the enterprise. Senior workers and staff jumped ahead of the more junior, and families of model workers or those who had suffered particular injury during the antirightist campaigns of the 1950s or the Cultural Revolution were given priority. Although the policy was not designed to encourage multigenerational homes, in practice it did. The most crowded homes were those like the Lis, the Zhaos, or the

[35] RMRB , March 11, 1988, 1; Oct. 11, 1988, 2.

[36] See Guofa , no. 11 and no. 13 (1988) in Guowuyuan Gongbao , 1988, 179-88; RMRB , Aug. 21, 1989, 1; Lu Shiming and Hu Anqi, "Shanghai zhufang zhidu"; Lu Xing, "Shanghai zhumin dui shangpinfang de xianchuan he qianjing de kanfa"; and Jan Middelhoek, "Urban Housing Reform in the PRC," China Information 4, no. 3 (Winter 1989-90), 56-72; Ming Bao (Hong Kong), July 4, 1990, 49, reports that in Guangzhou private building and buying continues, but at a cost of over 2,000 yuan per square meter it is simply beyond the average consumer. Similar sentiments were expressed by most Shanghai and Wuhan residents I met during my 1988 and 1990 fieldwork, although not in smaller Sha City where new blocks had entered the market at 200 yuan per square meter in December 1986, and rose to only 500 yuan by July 1988.

[37] In Shanghai in 1987 out of an average per capita income of 1,147 yuan, only 16 yuan (1.3%) went to rent and 30 yuan (2.6%) to utilities. Shanghai Tongji Nianjian 1988 , 410.

[38] In 1987 in Shanghai city, families with under 2 square meters per person were ranked as families in special need, but in practice, because after 1980 individual units supplied approximately 85 percent of new housing, the criteria varied from one enterprise to another and from one year to the next. In one collective enterprise described to me the cutoff in 1987 was 2.5 square meters, whereas in Watch Factory no. 4 it was 3 square meters in 1986 and 3.8 square meters by 1987, and in Textile Mill no. 6 it rose from 2.3 square meters in 1987 to 3 square meters in 1988.

Wangs, where a middle-aged couple had responsibility for sheltering an elderly in-law and several adult children. Newly married couples could apply to their workplace for a new apartment, but as a two-person household with little seniority they generally were deemed a low priority. Therefore, throughout the 1980s many newly married couples moved into the home of one of their parents, increasing the crowding of an existing household, but also moving the entire extended family further up the waiting list. Under this system, one should note that the household in line for the new apartment was that of the older generation, and thus when new flats did become available, the parents were assigned the new flat, and only through the intervention of the parents did the young couple receive an apartment of their "own." Moreover, if the parents had only one son or if a parent was already widowed, the housing authorities would allocate space with an eye to the parents' housing needs in old age and give priority to a joint or stem household rather than to the young couple. After new apartments were distributed to the most "deserving," the enterprise would then offer the recently vacated old flats to the more junior staff. Certainly some young couples did move into the newest estates immediately after their marriages, but among my respondents most of these lucky families obtained their new apartments through the intervention of their parents (see table 3.4). In a typical situation a parent (usually the father) had been assigned a new apartment as a reward for his seniority, but instead of moving to the new home himself he allowed a newly married child (usually a son) to live in the new apartment while the parents remained in the original place with other unmarried children or an elderly parent.

Urban China is hardly unique in its reliance on bureaucratic procedures rather than markets to distribute housing. Throughout the socialist world, and in segments of the capitalist economies, public housing allocated on the basis of need is commonplace. Moreover, in all these settings some families are more successful than others. Studies of urban housing in Eastern Europe in the 1960s and 1970s suggested that higher socioeconomic status often translated into more space and better amenities,^[39] and that the characteristics that predicted success in the workplace—Party membership, advanced education, urban birthplace—also determined housing outcomes. However, in Whyte and Parish's study of urban China in the 1970s, China appeared a somewhat deviant case. Professionals and Communist Party functionaries reaped few advantages from their social status, and overall urban housing was distributed more in proportion to need than to political position or wealth.^[40] The question is, to what extent did this egalitarian outcome persist during the housing boom of the 1980s?

[39] William Parish, "Destratification in China," in Class and Social Stratification in Post-Revolution China , ed. James L. Watson (London: Cambridge University Press, 1984), 93-96.

[40] Parish, "Destratification," 93-95.

Data from the official Chinese press suggest that the Deng reforms did not noticeably erode the basically egalitarian distribution of rents. For example, in Guangzhou in 1988, per capita monthly income was three times higher for families in the highest decile than in the lowest decile, yet rental burdens were virtually identical: 1.0 percent for families who averaged 84 yuan and 0.7 percent for those who averaged 245 yuan.^[41] Similar egalitarian outcomes were documented in official statistics and independent fieldwork.^[42] However, in the only report I have found that specifically compares distribution of residential space by the occupation of household head before and after the reforms, there is clear evidence of increasing inequality.^[43] In 1978 average per capita housing space for respondents in this study ranged from 4.8m² for families headed by a service worker to 6.3m² for those headed by a factory leader. By 1988 average household space had increased for all categories of staff, but the gains for those previously at the top were greater than for those at the bottom, and the absolute disparity between the lowest and highest had doubled.^[44]

Closer examination of the distribution of living space among the Shanghai families interviewed in 1987 documents similar discrimination against families headed by blue-collar fathers and also suggests how these inequalities in housing were created by the continuing role of enterprises as the brokers for most urban housing. All the families in my original 1987 sample lived in one new housing estate built in the first years of the post-Mao construction boom. Each apartment had the same number of rooms, but the layout and size varied in relation to the placement toward the staircases. Flats that ran parallel to the stairs had two rooms of equal size and contained twenty-four square meters of living space; those that were at the head of the staircase had one large room and one small room and contained twenty-seven square meters. Using the number of residents and the address of each family, I calculated the number of square feet each household occupied at the time of the first interview. And it is on the basis of these calculations that I was able to link the quality of housing and individual characteristics such as occupation and Party membership.

We can see from table 3.6 that despite the many shared characteristics of my respondents in terms of their age and their generally privileged

[41] Guangzhou Tongji Nianjian 1989 , 422.

[42] Guangxi Tongji Nianjian 1989 , 342; Beijing Shehui Jingji Tongji Nianjian 1989 , 564; Andrew Walder, "Collective Benefits and Social Stratification in Urban China," paper presented at the 42nd Annual Meeting of the Association for Asian Studies, Chicago, April 5-8, 1990.

[43] A study in five urban industrial units is summarized in Zhongguo Laodong Gongzi Tongji Nianjian 1989 , 350.

[44] Thus while service worker families gained on average 1.4m , production workers 1.6m , engineers gained 2.5m and factory leaders 3.3m . Zhongguo Laodong Gongzi Tongji Nianjian 1989 , 350.

socioeconomic position, families in this sample did not have equal, or even roughly uniform, shares of housing space. A small elite (N = 10) enjoyed more than twice the average space, while nearly a quarter of the families (N = 21) lived at the hardship level of three to four square meters per capita. In terms of family or individual characteristics that might explain this pattern of inequality, three items immediately seem salient. Those in poor housing had more children, fathers were more likely to be manual workers, and as a group they had the highest percentage of Party members.^[45] At first glance these observations seem contradictory. Upon further scrutiny, the outcome is consonant with what we know about the provision of urban housing in the semireformed political economy of Chinese cities of the 1980s.

The inequalities identified in table 3.6 document a bias against manual workers; they also reflect the impact of unequal institutional resources and the importance of unit-level patronage networks that reinforced class differences. The apartments in the neighborhood where I interviewed were built by the city government. However, the city had not taken sole responsibility for distributing the apartments. Instead, approximately half of the units

[45] I also examined age of mother, history of past political problems, and work status of mother. There was absolutely no difference by age of mother (57 years old in 1987) and thus it appears stage in the life cycle was irrelevant. Similarly the percentage with past political problems ranged between 9% and 10%, indicating no variation on this dimension. Only with percentage of mothers with blue-collar jobs did I find a difference, parallel with that found for fathers; the worst housed had average percentage (71%) while the best housed were far below average (20%).

had been sold to more than thirty different enterprises, which then in turn allocated flats among employees waiting for better housing. Therefore, while all families in my sample moved to these apartments as the result of successful supplication, the focus of entreaty and the criteria to establish a family's priority in the housing queue varied among employers. As a result, some of the inequalities documented in table 3.6 may be as attributable to the different resources of the units that had purchased the flats as to the different needs or prestige of the individual families. That is, it is possible that for the city slaughterhouse or district bottling plant, which had previously had no staff housing, this estate represented the first and best housing option, while for the import-export bureau, which was a long commute from this estate and which had long had its own hostels for its largely white-collar staff, the estate was not highly desirable. As a result, families who received their apartments from the import-export bureau were viewed by their unit as receiving only moderately good housing and did not have to document the same level of hardship to their enterprise housing office to qualify for relocation as those coming from the nearby slaughterhouse or bottling plant.

This is not to say that I found no individual characteristics that patterned the inequalities beyond the disadvantaged position of families headed by manual workers. On the contrary, both number of children and membership in the Party appeared to have an influence. In the case of family size the effect was in the expected direction. Parents with the most children had the greatest burden and therefore were most likely to live in the most crowded apartments. In the case of the disadvantage of Party membership, the result was quite unexpected. Placing this negative return on Party membership in the context of the bureaucratic allocation of housing, however, this anomalous finding becomes reasonable.

In urban China after 1955 most job changes were initiated by supervisors redeploying scarce talent.^[46] Blue-collar workers, therefore, had lower geographic mobility than white-collar, and Communist Party members, who were overrepresented in the professional and managerial roles, were more mobile than non-Party members.^[47] But because in urban China employees needed to cultivate personal connections in order to gain favorable treatment in the distribution of such benefits as new housing, mobility could be costly to the individual and his or her dependents. As a result,

[46] For a fuller description of this situation, see Deborah Davis, "Urban Job Mobility," in Chinese Society on the Eve of Tiananmen , ed. Deborah Davis and Ezra F. Vogel (Cambridge: Harvard University Press, 1990), 85-108.

[47] In the Shanghai sample, for example, 84% of blue-collar parents and 84% of nonparty members had worked in no more than two enterprises between 1949 and 1987; by contrast among the professionals only 40% were as immobile, and among Party members, it was only 39%. By the chi-square statistic both distributions were significant at the .001 level.

Party membership with its attendant higher rates of mobility could have negative rather than positive outcomes. Certainly at this point in our study of Chinese families, this conclusion is offered only as a hypothesis. However, it should be noted that in the case of these Shanghai families there is a clearly negative return to high mobility. Overall, 18 percent of fathers worked in four or more units between 1950 and 1987. Yet among the twenty-one families with the worst housing, 24 percent of fathers were this mobile, while among the ten who lived in the most spacious quarters there was not a single case of the highly mobile.

Conclusion: The Deng Reforms and Urban Families

The economic and political reforms of the 1980s altered the larger environment in which urban residents created and maintained their homes. They gave urban residents an unprecedented opportunity to indulge their desires for a comfortable and, by Maoist criteria, luxurious private life. Freed from political pressure to denigrate individual preferences, urban residents made choices that the Maoist party-state had censored for three decades. Supplied with more plentiful and better-quality housing, they also had unprecedented opportunity to upgrade their accommodations and enjoy more space and privacy. In terms of material conditions the results were more homes with running water, indoor plumbing, separate kitchens, as well as a massive upgrading in terms of the outward trappings of modern households: color televisions, velour bedspreads, refrigerators, and even VCRs. But as we saw in the high rates of multigenerational living at the time of marriage and the tendency to share a home with a son rather than a daughter, the reform decade appeared to have had little effect on traditional rules of family formation. In fact, the comparisons between the experience of parents and firstborn sons, and between older and younger siblings in this Shanghai sample, suggest that the new consumerism and the freedom to fulfill personal dreams may actually have strengthened precommunist preferences for virilocal living.

Close study of these one hundred families also reminds us that despite the new affluence and the freedom to satisfy more materialistic and individual pleasures, the Chinese party-state had by no means renounced its ability to impose binding limitations on urban family structure and behavior. The most extreme example of this persistent power was found in the one-child policy, from which virtually no young urban family could deviate.^[48] But equally important for understanding the general context in which households were created and maintained were the continuities with

[48] In 1988 urban Total Fertility Rate (TFR) was 1.09, and in the city of Shanghai 96.3% of all births in that year were first births. RMRB , Jan. 14, 1989, 3, and Sept. 20, 1989, 5.

the Maoist political economy and the police controls over internal migration. Throughout the 1980s housing continued to be a welfare benefit allocated by employers primarily on the basis of seniority. Domestic space legally still fell under the purview of the police, and it seems telling that the 1982 constitution, which introduced several new safeguards against state intervention into private life, did not restore "freedom to change . . . residence," a right that had been protected by the earlier constitution of 1954.^[49]

Urban families of the 1980s enjoyed new consumer freedoms, but they still were highly constrained in deciding when and where they could establish their homes. As a result of the reforms, there was some realignment in the relative strength of the defining parameters, but overall the prototypical urban household under Deng, as under Mao, could best be described as a supplicant to a socialist state.

[49] Beijing Review , Dec. 27, 1982, 10-29.

Four
Geography, Demography, and Family Composition in Three Southwestern Villages

Stevan Harrell

Since 1979 the political economy of rural China has undergone two fundamental changes, one in the sphere of production and one in the sphere of reproduction. In the productive sphere, economic reforms have both decollectivized farming and allowed for family entrepreneurship; in the reproductive sphere the planned-birth campaign has severely limited the number of children a family may have. These two changes have been so fundamental as to transform both the physical and the social landscape of rural China, with unprecedented material prosperity, the disappearance of large, collective fields, the virtual disappearance of the once all-important production team, and the rise of family enterprise in commercial agriculture and cottage industry. Given these fundamental changes in society at large, we might expect the structure and organization of families and households to also experience widespread changes. At the same time, the societal changes brought about by economic reform and the birth-planning campaign have not had the same effect everywhere in the Chinese countryside, and we might expect changes in family patterns to also show local variation. For example, the economic reforms have permitted entrepreneurship everywhere, but they have actually facilitated it only in those places where geographic and market conditions are otherwise favorable. The planned-birth campaign has been implemented more strictly in some places than in others, so both the plan and compliance with the plan have varied. The result is that there is still, even after the implementation of the reforms and the planned-birth program, considerable local variation in the way Chinese rural households organize for production and reproduce themselves. There is also a difference in the way the birth-planning campaign and the economic reforms have affected family organization. The effects of the birth-planning campaign on the family are direct and easily observable: as fertil-

ity declines so do household size and the ratio of children to adults. In areas where the fertility-control campaign has been implemented more strictly, these changes are more noticeable. The effects of the economic reforms, however, are more indirect. While the reforms have had a visible effect on the ways families organize their labor for production, especially in areas where entrepreneurship has begun to flourish, the effects on the family developmental cycle are less easily proven, though some trends can certainly be detected. This chapter examines household organization in three villages whose economy in the 1980s varied from almost pure subsistence to free-wheeling entrepreneurial, and examines the differential effects of the planned-birth program and the economic reforms in each village.

Variation in Chinese Family Size and Composition

In premodern society, Chinese families everywhere strove toward an ideal model developmental cycle in which a couple's daughters would marry out and their sons' wives would marry in, forming a patrilocal joint family. While this joint family could not be expected to last forever in reality, it was an ideal that ought to be achieved in every generation for at least a few years before the sons divided their household economies and took their individual equal shares of the family property, thus beginning the cycle over again. But in fact even this temporary joint form did not occur in every family in every generation: sometimes a couple had only one son who grew to maturity (or perhaps none at all), or sometimes a family's resources were too meager to support all its sons, in which case one or more might have to leave the family to make a living elsewhere.

And even where demography permitted, the joint phase of the family cycle was not always of equal length. As in so many societies with this and other forms of family organization, there was a series of centripetal and centrifugal factors that united brothers in pursuit of common economic goals or divided them in suspicion of each other's and each other's wives' motives.^[1] When the centrifugal factors were stronger—when there was little economic motivation to hold the joint family together in the face of the psychological tensions among its branches—the joint phase in the cycle was short or nonexistent. When, however, as in the case of pooled labor, pooled capital,

[1] For examples showing the effect of economic change on the series of decisions determining household form, see the several articles in Robert McC. Netting, Richard R. Wilk, and Eric J. Arnould, eds., Households (Berkeley and Los Angeles: University of California Press, 1984); the general introduction by Netting and Wilk, as well as contributions by Wilk, "Households in Process: Agricultural Change and Domestic Transformation among the Kekchi Maya of Belize," and Andrejs Plakans, "Serf Emancipation and the Changing Structure of Rural Domestic Groups in the Russian Baltic Provinces: Linden Estate, 1797-1858," are particularly useful as models.

or a diversified economy,^[2] the potential for common gain outweighed the frustration of common domesticity, the joint phase was likely to last longer.

In looking at variation in Chinese family size and composition, then, we can see two kinds of causal factors that may explain variation over time or between one sample of families and another. Demographic factors will determine the personnel available for the formation of stem and joint families,^[3] while various other social factors will influence the perceived balance between the advantages and disadvantages of staying together or splitting apart.^[4]

In addition, any convincing look at variation from demographic or social factors must consider interaction between these factors. Demographic behavior is constrained by social norms, and particularly by the complex of choices within which members of a family frame and plan their strategies. In explaining the differences among communities, we must look for the effects that recent social changes have had on demographic behavior (fertility in particular, but also age at marriage), as well as their effects on decisions to divide the family or keep it together under certain circumstances.

Recent Social Changes and Their Posited Effects on Family Organization

The importance of the planned-birth program and the rural economic reforms for household organization can be illustrated rather easily. The planned-birth policy has reduced the total fertility of Chinese women from

[2] See Myron L. Cohen, "A Case Study of Chinese Family Economy and Development," Journal of Asian and African Studies , 1968; "The Developmental Cycle in Chinese Family Groups," in Kinship and Marriage in Chinese Society , ed. Maurice Freedman (Stanford University Press, 1970), and House United, House Divided: The Chinese Family in Taiwan (Columbia University Press, 1976). See also Burton Pasternak, Guests in the Dragon: The Social Demography of a Chinese District (Columbia University Press, 1983), and Stevan Harrell, Ploughshare Village: Culture and Context in Taiwan (University of Washington Press, 1982).

[3] For studies pertaining to China, see Arthur P. Wolf, "Chinese Family Size: A Myth Revitalized," in The Chinese Family and Its Ritual Behavior , ed. Chuang Ying-chang and Hsieh Jihchang (Taipei: Academia Sinica, 1985), and Stevan Harrell, "Marriage, Mortality, and the Developmental Cycle in Three Zhejiang Lineages," as well as Ts'ui-jung Liu, "Demographic Constraint and Family Structure in Chinese Lineages," both in Chinese Historical Microdemography , ed. Stevan Harrell (Berkeley: University of California Press, forthcoming).

More theoretical treatments of the interaction of demographic and cultural factors in determining household membership can be found in K. W. Wachter, E. A. Hammel, and Peter Laslett, Statistical Studies of Historical Social Structure (New York: Academic Press, 1978); an interesting discussion of the interaction of these factors in specific cases is Anthony Carter, "Household Histories," in Households , ed. Netting, Wilk, and Arnould.

[4] Like previous authors, I propose to hold culture constant in the initial analysis, assuming that Chinese norms and ideals of family size and composition are the same everywhere. This becomes more problematic the further industrialization proceeds, but it is probably no great worry in the rural areas of Panzhihua city in 1988.

6.5 in 1965 and 6.3 in 1970 to 2.6 in 1980 and 2.3 in 1986.^[5] Because of the discrepancies in both the policy and the practice of its implementation,^[6] however, its effect has not been uniform. The number of children born has an obvious and immediate effect on the size of the family; though it does not affect the family's composition (its structure as nuclear, stem, or joint) immediately, it will have important structural effects in the future, so that communities where the planned-birth policy weighs more heavily will present different statistical pictures of the household size and composition that are reflections of the underlying developmental cycle.

The effects of the rural economic reforms are considerably more indirect and complex, but potentially no less important for understanding changes in family organization. Returning land to individual families for farming, and permitting or even encouraging various kinds of entrepreneurship, have meant that many families who under the collective system did little but contribute to the collective economy and thus try to survive, with very little leeway for entrepreneurship or planned growth, now enjoy a wide range of choices in where and how to invest their resources. An earlier calculus involving little but the contribution of various work-point earners to the family's grain share-out, and perhaps some marginal decisions about what to do with the small private plot, has been replaced by a complex of possible costs and benefits involving subsistence versus cash crops, possibilities of local or migratory employment, and for some families real opportunities for various kinds of investments of labor and capital.

Like the planned-birth program, however, the rural economic reforms have had differential effects in different places. Villages with convenient transport, or those that are close to urban centers, have seen the growth of wide ranges of wage-labor and entrepreneurial opportunities. This has meant the growth of inequalities in income within these villages, in addition to general growth. Remoter villages, however, have often seen a collective subsistence economy simply turn into an individual subsistence economy, and have also experienced considerably less internal inequality.

These differential effects of the economic reforms can be expected to have differential effects on family strategies, at least if we are correct in assuming that family strategies of fertility, marriage, and division are all affected by the perceived costs and gains to be incurred in having a child, in taking in a daughter-in-law or marrying out a daughter, or in dividing the family or keeping it together.

The three villages treated in this chapter lie within about sixty kilo-

[5] Norman Y. Luther, Griffith Feeny, and Weiman Zhang, "One Child Family or Baby Boom? Evidence from China's 1987 One-per-hundred Survey," Population Studies 44 (1990): 341-57.

[6] See Jeffrey Wasserstrom, "Resistance to the One-Child Family," Modern China 10 (1984): 345-474.

meters of each other in Panzhihua city, Sichuan, but they have been affected quite differently by the reforms and the planned-birth campaign. As a result, there are considerable differences visible in their patterns of fertility and household economy, and these, in turn, have affected household size and composition over the last two decades, differences that reflect underlying differences in the demographic and economic processes that constitute the developmental cycle. The relevant figures are presented in tables 4.1 and 4.2.

Although there are no consistent differences in household size among the three villages, notice that household size has contracted considerably more rapidly in Renhe with the effective implementation of a one-child policy there. More important, Renhe and Yishala consistently show higher percentages of stem families, and correspondingly lower percentages of nuclear families, than are found in Zhuangshang. This reflects differences in the processes of family growth and division in the three villages, differences that

are at least partially explicable in terms of the effects of the twin changes of the planned-birth campaign and the economic reforms.

Three Villages in Panzhihua City

Panzhihua city stands at the intersection of the Chengdu-Kunming railroad and the Jinsha (upper Yangtze) River; as an administrative area, it includes the city of 400,000 people, built from scratch after 1965 to process one of China's richest ferrous metal deposits, and three rural, county-sized areas, whose combined population is about the same as that of the city (see map 4.1). The most accessible and developed of these is Renhe qu , or district, a mostly rural area the size of a county. The three villages discussed in this chapter are all in Renhe qu ; our research team spent from eight to fourteen days in each. Questionnaires covering the history of family composition since 1949,^[7] as well as details of the family economy from 1985 to 1987, were administered to 290 families randomly sampled in four communities,^[8] including 80 families in Renhe, 60 in Zhuangshang, and 98 in Yishala. What has become clear from the analyses of the data collected in this survey is that geographic position has made a big difference in the effect of recent policy changes on family process.^[9]

Renhe Cun, a Prosperous, Commercialized Village

The capital of Renhe qu is Renhe zhen , a town of about 20,000 residents about twenty minutes' drive on a smoothly paved highway from the southern edge of Panzhihua city proper. We interviewed in Renhe cun , an area of about 1,200 households stretching eastward from the outer streets of the

[7] All retrospective questions designed to elicit information on family composition reached back to 1949, but my impression is that the data are less reliable before about 1970. For that reason and because the conference dealt with changes since the beginning of the reforms, I have chosen to concentrate here on the period from about 1970, which covers several years under collectivization as well as the period since.

[8] To be precise, in Renhe we selected five of the ten production cooperatives (formerly teams), according to local cadres' descriptions of their economic activities. Thus one of the two grain-growing cooperatives was sampled, one of the two where fish raising was important, and so forth. In each selected team, we took a random sample of all households. Where a household could not be interviewed because there was no competent adult, we selected the household immediately below it in the household registers. In Yishala, we took a random sample of the main village (80 households) and a random sample of the satellite village of Xiachang (10 households). In Zhuangshang, we took a random sample of each of the village's three cooperatives.

[9] Geographic differentials in economic development in rural Panzhihua have also been treated by Li Shaoming in a paper entitled "Panzhihua shi gongyequ de jianshe dui zhoubian nongcun de yingxiang" (The effect of the construction of the industrial district in Panzhihua city on surrounding villages), presented at the Conference on Urban Anthropology held in Beijing in December 1989.

Map 4.1.
Panzhihua and Vicinity

town itself into the foothills surrounding the rich valley where the town is located.^[10] Renhe and the other cun that make up Renhe zhen , along with the neighboring xiang , or townships, of Qianjin and Zongfa, are probably the wealthiest villages in all of Panzhihua city; we chose Renhe partly as a contrast to the poorer, more remote villages that were our other interview sites. The mean per capita income in our sample households in 1987 was 1,024 yuan (with a household median of 4,800, or about 1,000 per person),^[11] which puts Renhe comfortably above the national rural average, though still short of really wealthy villages in the Pearl River delta, as described by Johnson and Siu in this volume, or on the Chengdu plain or the Jiangnan area outside Shanghai.^[12] About four-fifths of Renhe villagers' reported income was in cash. Most houses in Renhe are still built of mud (people say brick housing is too hot in the summer), but modern appliances such as televisions and tape players are ubiquitous, and several families own refrigerators, color televisions, or washing machines.

There has been a market at Renhe since at least the early nineteenth century, but the area remained a sleepy backwater of north central Yunnan until it was given to Sichuan in 1965, when the construction of the Panzhihua steel complex was begun. At that time, the flat land in the valley surrounding the town was converted to vegetable cultivation to serve the markets of the new city. To this day, eight of the ten cooperatives that make up Renhe village are engaged in vegetable cultivation; the ninth and tenth, in the hills to the east of town, cultivate grains.^[13]

[10] The village and the urban area of Renhe district town overlap spatially, with farmhouses interspersed between the apartment buildings occupied by government and construction company employees. The difference between village and town residents lies technically, not in the type of housing, but in the hukou registration: if you have nongcun hukou , or village household registration, you belong to Renhe cun ; if you have danwei hukou , or unit household registration, you belong to the urban population of Panzhihua city. Some families have both; a man, for example, who has a state-sector job in the city will not have his household registration with his family, even though he lives with them. When some of the former agricultural areas of Renhe were taken over by Shijiu Ye (19th Metallurgy, which is actually a giant state construction company) and some residents took Shijiu Ye jobs, the affected production teams lost population, even though nobody actually moved.

[11] The absolute level of these income figures should not be taken too seriously when comparing these villages with communities in other parts of China. We did not have access to the official methods of calculating income in kind (agricultural goods both produced and consumed by the household), and so we devised our own formulas. My impression is that our method yields a figure about 25 percent higher than the methods used in official reports.

[12] I visited a village in Jiangyin county, Jiangsu, where the announced per capita income was over 1,200 yuan in 1985. It did not look much different from the surrounding villages. But officially reported average household incomes for Chinese province-level units showed only rural Shanghai to be slightly above Renhe at 1,059 yuan; the richest provinces were Guangdong at 644 and Jiangsu at 626. See Zhongguo Tongji Nianjian 1989 , 746-47.

[13] Called production "teams" during the collective period, these groups of up to a hundred or more households changed their name to production "groups" in the early 1980s; in 1988 the name was changed again to "cooperatives," though this does not indicate a recollectivization of agriculture.

Vegetable growing, in its own right, can be very lucrative. The figures for the vegetables sown by one family on their tiny plot of .63 mu , or about a tenth of an acre, illustrate why.^[14] Multiple cropping gives them a sown area of about 2.4 mu , on which, in 1987, they grew tomatoes, peppers, eggplant, cucumbers, green beans, cabbage, "Chinese" cabbage, and a kind of greens known as wosun . The family has minimal production costs for seed, fertilizer, and paying someone to plow their land; subtracting these from the total market price they gained for their vegetables, about 2,400 yuan, shows a probable profit of over 2,000 yuan. And they have much less land than certain other families that specialize in vegetables and sometimes make as much as 4,000 per year. There is skill in growing vegetables and in knowing when to sell to get the best prices, and some families do not do nearly as well. But vegetables are far more profitable than grain farming, even with the tiny plots available to Renhe villagers.

Only nineteen of the eighty households we interviewed in Renhe obtained all their income from farming, however. This is a great change from the 1970s and is attributable to two kinds of changing economic opportunities that are a direct result of the rural reforms.

First, there is salaried labor. While only four of our sample families earned all their income from wages, another twenty-six combined wage labor and farming, four wage labor and capital investment, and eight wage labor, capital investment, and farming. Wage labor has become increasingly available as industrial development spreads outward from Panzhihua city proper into the near suburbs such as Renhe. When industrial enterprises take over farmland, they are required to offer jobs to the farmers they displace. In the 1980s this has meant that most of the families in the seventh cooperative, which houses a large plant of the city construction company, have given up farming altogether: only three of the twelve families interviewed in that cooperative still received any income from farming, while eleven had some wage income and six lived partly by entrepreneurship. Wages vary, of course, and the wages of one worker are usually insufficient to support a family of more than two or three. For example, the family income of a man whose 110 yuan per month as a tractor driver is supplemented only by sale of an occasional pig is on the low side at 2,100 per year for four people. Another family, with three wage earners supporting six mouths, manages to make about 4,500 per year, still below average for Renhe, but enough to allow the family to save nearly 1,000 a year after expenses.

[14] It might be useful for nonfarmers to visualize the size of this plot: a square of about 20.5 meters on a side.

It is in entrepreneurial investment of capital, however, that the real opportunities lie. Sometimes this can be something that one person can do alone, as in the chicken-raising business run by the wife in the vegetablefarming family described above. Her husband orders fertilized eggs from Chengdu, or even from Beijing or Shanghai, several times a year; his wife raises them until they are at about the transition from furry to feathery, at which point she sells them for .20 yuan profit each. At 30,000 chickens per year, that is a net income of 6,000 yuan for one person's labor. Added to the aforementioned vegetable gardening and a fledgling fish-raising business, this pushes the family very close to the fabled category of wanyuan hu (ten thousand) household.

A few families in Renhe have made it to wanyuan hu status and beyond; there were five in our sample. As an example, take the leader of the ninth cooperative, a grain-growing community in the area called Longtangou, in the hills to the east of town. His family has three trucks used to haul sand and gravel. His two grown sons drive two of the trucks; he hires a driver for the other vehicle, as well as about three laborers to load and unload the construction materials. After paying wages and operating costs, this business netted the family about 18,000 yuan in 1987; added to about 1,000 from pig raising and another 1,000 from fish farming, plus in-kind income from grain farming, this brought the family's income to about 22,300, the second highest of any family we interviewed. The family head thought he was probably about the eighth wealthiest in Renhe cun . He had just enlarged his family mansion to twenty-nine rooms (six of them currently empty) built in two stories around a spacious courtyard, for a total of about eight hundred square meters of floor space.

From these few examples, it is easy to see how much the possibilities for family economy have changed since the basic reliance on collective agriculture came apart at the end of the 1970s. I should emphasize, however, that not all families have benefited equally, or even benefited much at all from the reforms. The closest neighbors to the family with the twenty-nine-room house live in a one-room mud cottage, and the poorest families in our sample had incomes of 300 yuan or less per person. In addition, families with several televisions and a truck live side by side with people who do not even own a watch or a radio.^[15]

A visitor to Renhe thus goes away with a general impression of a dynamic economy. Families have a variety of economic alternatives, many of them involving investment of capital, and those who are successful in this have not only increased their wealth dramatically since the reforms, they

[15] The topic of reemergence of social classes in prosperous parts of the Chinese countryside ought to be addressed systematically. The official position, cited to us by Panzhihua vice-mayor Tan Huizhang, is that differences in income, because they do not involve exploitation, represent divisions among strata within a class rather than differences between classes.

have also positioned themselves for further growth in perhaps new and as yet unknown directions. There are almost no Renhe families whose major income still derives from subsistence farming; except for some of the families in the ninth and tenth production cooperatives up in the hills, even the farmers are in business for profit through the growth and sale of vegetables.

Yishala, a Model Minority Village with a Mostly Subsistence Economy

When visitors from outside Panzhihua, Chinese or foreign, come to the area and ask to be taken to a minority village, Yishala is often the one chosen. It lies in Pingdi xiang at the southernmost tip of Renhe qu , high above Jinsha River at an altitude of about 1,680 meters. From Panzhihua city it is an hour-and-a-half drive on the paved highway to Pingdi town, which has a periodic market, and another twenty minutes or so on a good gravel road to Yishala.

Yishala is in many ways typical of Pingdi,^[16] in that its population is composed entirely of farmers, most of whom belong to the officially designated Yi minzu . They call themselves Lipuo and their language Libie ; they use this language in all everyday interactions, but almost all of them are fully bilingual and use the Han language with outsiders, as well as within the community for formal occasions such as weddings and funerals. Their customs, with a few minor exceptions that serve as ethnic markers, are the customs of the local Han; their family, lineage, and inheritance systems are likewise those of northern Yunnan peasants generally.

At the same time, however, Yishala is somewhat atypical, even of Pingdi, in that it is a model minority village for Panzhihua. This means that the city government, along with the provincial Office of Minority Affairs (minwei ), has provided funds for investment in village improvements. The most important of these are the cement factory, whose dormitories housed us for our stay in Yishala, and the cultural center, a brick building that stands locked and unused next to one of the village irrigation ponds. For village families, however, model status makes only a minor difference: a few younger members are employed at the cement factory, which may raise the family cash income by the amount of wages (about 50 yuan a month for an average worker), but not all families take advantage of this opportunity, and the idle cultural center does not seem to do anybody any good.

[16] There has been at least one previous survey of families in Yishala, conducted by the Panzhihua Office of Artifact Management in cooperation with the Anthropology Department of Zhongshan University in Guangzhou. See Chen Qinghua, "Pingdi Yizu de jiating bianqian" (Changes in the Yi family of Pingdi), unpublished paper, and Li Xiaorong, "Panzhihua shi Yishala Yi cun shehui jingji diaocha" (An investigation of the society and economy of Yishala Yi village in Panzhihua city). Aside from their greater emphasis on the Yi ethnicity of Yishala villagers, the findings of both Chen and Li are in general and specific agreement with mine.

The major crops in Yishala are winter wheat and barley, summer rice and corn, supplemented by draft cattle, as well as goats, pigs, and chickens raised both for consumption and for sale. Vegetables are grown for family consumption only. A comparison of family income sources in the three villages is found in table 4.3.

Superficially, the occupational structure of Yishala bears some resemblance to that of Renhe, in that only twenty-one families of the ninety-eight we interviewed in Yishala and its small satellite hamlet of Xiachang made their living entirely by farming; this is an even smaller percentage than in Renhe. The largest category (forty-two households) combined farming and wage labor of some sort, while another seventeen combined farming and small-scale sideline production (construction lime in Yishala; incense in Xiachang), and eighteen derived income from all there—farming, sidelines, and wage labor. But in another sense, Yishala is like the subsistence-oriented village of Zhuangshang: in neither village was there a single family that did not farm, while in Renhe 15 percent of the families had given up farming altogether.

When we look at the economic nature of nonfarm production in Yishala and Renhe, however, we find a fundamental difference. Many people in Renhe are true entrepreneurs—they invest capital in the hopes of a return, which in turn will bring greater capital. They also innovate; there is a wide range of activities from hauling sand to raising chickens, and families are tied, not to one enterprise or another, but rather to the principle of investing in anticipation of profits. For vegetable farmers such as those described above, even farming is an entrepreneurial activity.

Nonfarm production in Yishala, in contrast, is not really entrepreneurial

in most cases. Though thirty-four families in our Yishala sample engaged in selling a product for a profit, only eight of these realized a profit of more than 500 yuan in 1987, and none earned more than 2,500 from such activity. In addition, most of the petty industry of Yishala families consists of quarrying limestone and making lime for sale, an activity that requires only a small capital outlay (one has to own or have access to a small tractor to haul the heavy rock) and realizes only a small profit, too inconsequential to provide much capital for reinvestment. In other words, this is not the building of a financial base independent of farming, as is the case with Renhe entrepreneurs; it is merely the supplementing of income.

Zhuangshang, a Poor Village of Subsistence Farmers

It is possible to drive to Zhuangshang, but it is faster if you park your car at the neighbor village of Jingtang and walk: by trail from Jingtang to Zhuangshang is about thirty minutes for a fast hiker.

Zhuangshang lies just above the banks of the Jinsha River, in the township of Pinjiang (see map 4.1). Parts of Pingjiang are prosperous (it overlaps with the urban district of Geliping in the same way that Renhe cun overlaps with the urban population of Renhe town), but the southwestern corner of the township, where Jingtang, Kuqiao, Laluo, and Zhuangshang lie, is dry, undeveloped, and poorly served by transport. Zhuangshang itself is the poorest village in this corner of the township, with a per capita income in our survey of only about 424 yuan, not low enough to qualify it for state aid as an officially impoverished (pinkun ) area, but so low that it has no electricity, no bicycles (some families could afford them, but they could get nowhere on them, so people use donkeys as their main means of carrying things), many people dress in patches, and all houses are built of unplastered mud walls.

Zhuangshang is officially a minority village; the people consider themselves to be members of the Shuitian minzu , a group not officially recognized by the state, which places them in the larger category of Yi.^[17] But the Shuitian people in Zhuangshang and in neighboring Futian township and Huaping county practice very little that would enable an outsider to distinguish them from local Han Chinese without specifically asking. Their customs, language, and social organization seemed from our investigations to be identical to those of the neighboring Han villages. The only practical difference that may be connected to their official Yi identity is state permission to have two children,^[18] a factor that, as we will see below, has had noticeable effects on family size and composition in the 1980s.

[17] For a detailed description of ethnicity in several villages in the Panzhihua area, including Zhuangshang, see Stevan Harrell, "Ethnicity, Local Interests, and the State: Yi Villages in Southwest China," Comparative Studies in Society and History , July 1990.

[18] It is not even clear how big a factor ethnicity is in the Zhuangshang villagers' permission to have two children. Birth quotas in Pingjiang xiang are determined by relative wealth and prosperity, rather than exclusively by ethnicity, so that villagers in Geliping and Dashuijing, two urban-overlap villages, may have only one child; people in Jingtang and other moderately prosperous places may have two if the first is a girl; people in Zhuangshang may have two, and Miao people in a couple of remote mountain areas may have three if the first two are girls.

In contrast to the variety of family economies found in Renhe, and the typical Yishala pattern of farming supplemented by wage labor, Zhuangshang villagers are all primarily farmers. In our sample of sixty households, we found only two families with any members engaged in any sort of entrepreneurial activity—both made small amounts of cash by buying miscellaneous goods cheap in one local market and selling them slightly dearer in another. One of these households also contained a wage laborer, as did nineteen other households in the sample, leaving thirty-nine households dependent entirely on farming (or in a few cases also fishing in the Jinsha River or gathering medicinal roots for sale).

Even here, not all agricultural activity is purely subsistence-oriented. One family, whose income of 7,304 yuan (over 3,000 in cash) in 1987 was the highest in our sample, may nonetheless provide an example of a slightly more successful version of the typical Zhuangshang family economy. On their allotted 4 mu of irrigated land and 2 mu of dry land, they produced 6,530 jin of rice, 400 of wheat, 2,000 of corn, and 6,000 of sweet potatoes. In addition, their banana groves netted them about 100 yuan profit (much less than some families with more trees), and three months of fishing in the Jinsha River by the household head and two of his sons brought in another 1,500 in income. In addition, they sold two pigs for a total of 800 yuan, and about twenty chickens for a total of 160. With this income, this family was extraordinarily prosperous by Zhuangshang standards. They owned a black-and-white television (useless since electricity to the village was terminated a few years before), a radio, a sewing machine, and five watches. Their house had floors of lime plaster and contained four bedrooms, two common rooms, two kitchens, and three storerooms, in addition to numerous stables.

More typical, though in fact somewhat below average, was another family consisting of the household head, his wife, and two teenaged sons. On their two mu of irrigable land, they grew 2,000 jin of rice in the summer seasons, and 200 of wheat in the winter;^[19] dry land produced 800 jin of corn

[19] The low wheat-yields are due to the lack of irrigation water in the winter season. Ever since a state-owned coal mine diverted water in 1980 from an irrigation channel used by Zhuangshang and Jingtang to a coal-washing facility, it has been difficult to grow a winterwheat crop in this area. Villagers are still angry about this: they have attacked the coal mine physically and submitted several formal complaints to city authorities and asked us to bring the matter to the authorities' attention (which we did), but so far they have not gotten their water back. As a result, disputes over irrigation are extremely common, especially during the nighttime, when people accuse others of stealing or diverting their water. Cadres in both villages told us they spend most of their time in the late winter and early spring mediating irrigation disputes.

and 400 of sweet potatoes. Except for the compulsory quotas sold to the state at low prices, they ate these food crops. Banana and papaya trees brought most of their cash income; together with the sale of a few chickens, these fruits netted them about 160 yuan cash. Since they had no income from wages, fishing, or entrepreneurial activity, the rest of their 1,200 yuan for the year was income in kind—the aforementioned subsistence crops plus one pig slaughtered at New Year's. This left them with a shortfall of about 400 yuan to meet their basic expenses; they borrowed this in small amounts from a variety of people.

This latter family, obviously, was poor. Their house, built in 1968 and not renovated since, had dirt floors; they had none of the appliances or amenities asked about on our questionnaire—not even a watch or a battery-operated radio. They were not starving—nobody we saw in Zhuangshang had any signs of malnutrition or lack of food—but theirs was truly a subsistence economy.

These two families represent something of the extremes of Zhuangshang villagers, but they are in fact not very different from each other—the only real distinction is the first family's rather profitable fishing activity. But even this activity is not really entrepreneurial, in the sense that it involves little capital investment (fishnets are about all), and seems unlikely to grow except by the marginal addition of another laborer or two. And most of the families in Zhuangshang, although they fall quantitatively somewhere between the incomes of our two examples, are qualitatively more like the second—only about a third of Zhuangshang families' total income is in cash, and even most of their cash income comes from primary production (bananas, papayas, wild grasses, and fish from the river). And aside from the marginal gains in agricultural productivity that might come from a more secure supply of irrigation water, there is little visible chance that they will be able to increase their income in the near future. They seem bound to remain subsistence farmers, with perhaps a slight increase in income from migratory wage labor, for the next few years at least.

If the general agricultural regime and the low level of income make Zhuangshang family economies seem not very different from those of Yishala, there is still a difference between them. Per capita income in our Yishala families was 430 yuan, essentially the same as we found in Zhuangshang, but it was distributed differently, with about one-half of the typical family's income in cash, mostly from wages or the manufacture of lime or both. This was reflected in the different standards of living in the two

places—people in Yishala had considerably more consumer goods on the level of bicycles, sewing machines, and black-and-white televisions, though none of them had a car, a refrigerator, or a color television.

Demographic Changes: the Fertility Decline

These differential economic patterns have had considerable effects on family size and structure. In order to understand the changes in family in general, however, we also need to look at the other kind of factors influencing family process: the differential decline in the birth rate during the same period.

The basic facts are indicated in tables 4.4 and 4.5, which present various kinds of evidence pertinent to changes in fertility over the last fifteen years in the two villages.

All three villages can be seen to have undergone a sharp decline in fertility in recent years, but the decline in Renhe is sharpest, followed by Yishala and then Zhuangshang. These differences in fertility among the three villages, indicated both in the declining size of recent birth-cohorts and in the declining parity of recent marriage-cohorts, hint at an explanation of the more precipitous decline in the size of Renhe families, even though those families are more complex than those of Zhuangshang. In Yishala, family size has not declined so rapidly, in spite of fertility decline, because families have remained even slightly more complex than those in Renhe. But what accounts for the more abrupt fertility decline in Renhe and Yishala?

One factor is simply the difference in birth quotas, under what appears from our data to be a very effective regime of implementation. The birth quota for Renhe, as a Han village and a periurban one, is a single child, whereas Zhuangshang and Yishala families are allowed to have two, because of either their poverty or their ethnicity, or perhaps both (the township governments in Panzhihua set their own quotas; the standards vary

considerably from one township to another). In the first years of the planned-birth campaign, corresponding roughly to the 1978 to 1982 birth cohort, Yishala seems to have implemented a policy stricter even than was allowed by its two-child quota, causing its birth rates to drop temporarily to levels resembling those in Renhe, but since then implementation has loosened considerably. The question thus becomes one of how Renhe has managed to keep the lid on births even in the more relaxed atmosphere of the late 1980s.

The low birth quota may not be the whole story. The number of children born in Renhe dropped more suddenly in 1978 than even during the famine years of the late fifties and early sixties,^[20] and has remained low until the present. This has not been the case everywhere, even in our own study—note the slight rise in the Yishala birthrate between 1978-82 and 1983-88^[21] —and it has also been reported from many parts of China that enforcing the single-child policy is particularly difficult in the rural areas, where children are seen both as potential labor-power and as necessary old-age insurance.^[22] In fact, Greenhalgh's chapter (9) in this volume argues that the state and the local community have in essence negotiated a two-child policy in the Han villages she studied in Shaanxi, which like Renhe were vegetable-growing communities. Enforcement has worked, though, in Renhe, and I think the reason has partly to do with three interconnected factors that distinguish Renhe's experience in the reforms. These are physical and economic proximity to the city, entrepreneurship, and the status of women. Renhe, of all the rural areas in Panzhihua at least, is most like an urban area and thus most susceptible to effective population control. The many families that live partly from state wages will have pensions, as have urban workers; this lessens the value of children as old-age insurance. Also, the relatively high standard of living expected in Renhe these days (considerably higher than in the villages surveyed by Greenhalgh, which seem comparable to Yishala, though lower than in the Pearl River delta villages surveyed by Johnson), along with the cash basis of the economy, in which children's labor contribution declines relative to their consumption needs, also makes children seem less profitable, and the pressure to have

[20] A Renhe resident whom I interviewed about the famine of 1960-62 told me it was not particularly serious in this area. People were hungry, he said, for six to eight months in 1960, but a good winter harvest in 1961 saved the area from starvation, and harvests after that time were reasonably good. Our figures support this interpretation, but I did not pursue this point systematically with other people.

[21] There is little increase in the General Marital Fertility Rate in Yishala between these two cohorts, indicating that the rise in the birth rate is due to a change in the age structure of the female population. However, if we consider that village birth quotas are based on total population, rather than on the population of married women, to the cadres of the plannedbirth campaign this would look like a slackening of enforcement.

[22] Wasserstrom, "Resistance."

fewer of them thus less onerous. And even entrepreneurial families, which are often larger than their farming or wage-earning counterparts in other places, are usually larger because they tend to retard family division, not because they are particularly fertile.

Zhuangshang and Yishala share none of these characteristics that Renhe has in common with the nearby city. The standard of living itself is lower, and with the continuing battles over irrigation water, Zhuangshang villagers are ever mindful of the possibility of food shortages. What food they do manage to produce will come from hard labor, and hard labor belongs to the young and middle-aged; this along with the impossibility of any savings and the unlikelihood of the implementation of any sort of pension plan (nobody interviewed in Zhuangshang earns wages in the state sector; in Yishala we know of only one state-sector wage earner, a cadre in the xiang government at Pingdi), means that old-age insurance in the form of children is still a crucial consideration. All but one of the thirteen women who married into Zhuangshang families from 1980 to 1984 already had their legal maximum of two children by the beginning of 1988; although Yishala brides seemed somewhat slower to reproduce, about half of the comparable cohort there had already borne a second child by 1988.

The other two factors contributing to the relative ease of enforcement of the one-child quota are entrepreneurship and the status of women, and the two are closely interconnected. Entrepreneurship in its own right undoubtedly releases the pressure to have more children; those who replace a subsistence economy with an entrepreneurial one trade a strategy of survival for a strategy of advancement, and advancement relies on the accumulation of capital rather than of labor, especially since successful entrepreneurs such as the sand-hauling family described above can actually afford to hire laborers.

It is important to distinguish nonentrepreneurial productive sidelines, such as the ubiquitous lime production in Yishala, from true entrepreneurship as found in Renhe. Making lime is a welcome source of income for Yishala villagers; five hundred yuan in cash is a considerable supplement to a family's average of two or three thousand total, and enables not only saving but the purchase of bicycles, tape players, and in some cases televisions or small tractors. Zhuangshang villagers, of course, do not even have these meager opportunities, and indeed they are poorer than their counterparts in Yishala. But the presence in Yishala of a cash income source of this kind does not alleviate the pressure to build old-age security through children; it is a supplement rather than an alternative strategy.

In a similar manner, the status of women is a factor in the calculus of fertility strategies. Hill Gates suggests in her chapter (10) in this volume that for entrepreneurial women in Chengdu, "anti-philoprogenitivism," or restricting the numbers of one's offspring, is a strategy pursued by women

who want to shepherd available capital for reinvestment and also wish to reduce the burdens of childcare so they can devote more time and energy to business. But if fertility restriction is a women's strategy, women must have enough control or leverage in family decision making to be allowed to pursue this strategy.

There is considerable evidence that the status of women is higher in Renhe than in Zhuangshang or Yishala. One of our more surprising findings concerns the amount of schooling given to boys and girls in the three villages, as shown in table 4.6. As the table indicates, despite the considerable differences in income and urban cultural influences in the three villages, the average years of schooling for males differ very little. The Communist Party's program to bring basic education to villagers seems to have succeeded rather spectacularly in the case of Zhuangshang, which is otherwise a poor and remote village.^[23] What does differ, however, is the imbalance of education between girls and boys. In Renhe, girls born since 1950 have received slightly more education on the average than have boys, though the difference is not significant.^[24] In both Zhuangshang and Yishala, boys have been to school for almost two years more, and the difference is fairly consistent across all cohorts. Only one female in eight from Zhuangshang, and one in nine from Yishala, has gone beyond elementary school, while at least a quarter of Zhuangshang and Yishala males, and Renhe residents of both sexes, have had more than a primary education.

[23] Zhuangshang is particularly noteworthy here, since it got such a late start. In Yishala, the first school was established in 1906, and most males received some primary education in the decades before 1949. In Zhuangshang, however, there was only a short-lived attempt at a school in the 1930s, and most men born before 1940 are still illiterate.

[24] There is, of course, some right-censoring in the data for the later cohorts, since many of the people in those cohorts are still in school.

This male-female difference in education seems to reflect the difference between the subsistence pursued by the Zhuangshang villagers and the partly entrepreneurial economy now flourishing in Renhe. The subsistence economy usually entails the retention of the patrilineal, androcentric social system, in which livelihood comes from the property-and-labor-managing patriline, in which men and women play the traditional roles—even the few people we did find in Zhuangshang who engaged in production for profit were men, as were all the participants in the grueling labor of lime manufacture in Yishala. In Renhe entrepreneurship is a departure from the traditional system and centers around individual initiative. This initiative can be taken by either men or women, and women are thus seen to be just as productive and to have just as much potential as men; perhaps this is why people send their girls to school for at least as many years as they send their boys.

But it is not merely the potential for entrepreneurial activity that is shared by men and women in Renhe; many of the actual prime movers in Renhe's entrepreneurial families are women. I have already mentioned the vegetable farmer's wife with her chickens; there is also the entrepreneuse who took out a contract on a formerly collective fishpond in 1987; though she barely recovered her investment that year, she expected to net over 7,500 in 1988. Ren Hai and I interviewed her because we were interested in talking to someone who had recently contracted for one of the fishponds; we spent a couple of hours discussing her balance sheet and her plans for expansion. When we had taken enough notes she insisted on leading us out to the pond and giving us a tour; in particular she wanted to show us the brick shed she had recently had built, which not only housed her gear but was planned as a little office for the rent-a-rod business she was planning to open for Sunday visitors from the city.

In addition, many of the vegetable-growing enterprises in Renhe are run and managed by women. This is particularly true in a large number of families whose male members hold salaried jobs; the women plant, grow, and sell the vegetables. This was brought out strongly to me in an interview with a man who was a professional cook (a state employee); he was hardpressed to come up with any figures on acreage, fertilizer, yields, or profits from agriculture. His twelve-year-old son, who helped his mother out in the fields after school, knew much more than the father, but I had to wait until his mother got home to get the figures straight.

This is not to suggest that the women of either Zhuangshang or Yishala act retiring or submissive. In Yishala, which we knew better because we lived there, women were publicly very visible on all sorts of formal and informal occasions, and they often took the initiative in helping us with our interviewing and other tasks; our limited time in Zhuangshang showed the same sorts of patterns. But in neither place do women control capital; as

Gates points out in chapter 10, this seems to be the crucial difference, not just in granting them some independence, but also in permitting a strategy that would allow low fertility. Even if a woman were fully in control of the family economy in a place like Zhuangshang or Yishala, her rational strategy, despite the possible hardships involved, would still be to have enough children to ensure a comfortable old age.^[25]

This still leaves open the reasons for the relative success of the planned-birth campaign in Yishala, where a two-child policy was negotiated only later, and where it was very effectively implemented. I would suggest provisionally that this has something to do with Yishala's model-village status; it may have been possible to acquire and preserve such status because the villagers were behaving in model ways, including controlling their fertility. This supposition was strengthened on my visit to Pingdi in September 1991, where township cadres told me Yishala has become a provincial-level Civilized Village, preserving its model-community status.

Family Division and Family Complexity

All these factors, I propose, combine with the single-child birth quota to contribute to the dramatic diminution in the size of Renhe families between 1983 and 1988. As urban proximity, entrepreneurship, and the status of women have made this decline in family size more dramatic than the corresponding declines in Zhuangshang and Yishala, we must remember a rather contrary fact expressed in the opening tables of this chapter—Renhe families, although the same size as those of Zhuangshang, are nevertheless consistently more complex, while Yishala families are as complex as those of Renhe and larger than those in either Renhe or Zhuangshang. Stem families have accounted for from 39 to 46 percent of the families in Renhe over the last fifteen years, and from 41 to 48 percent in Yishala, while in Zhuangshang they have been as little as 26 percent and have never exceeded 37; the current percentage figures at the time of our survey were 46 in Renhe, 48 in Yishala, and 30 in Zhuangshang. In addition, though the figures are probably too small to carry much weight, Zhuangshang has had no joint families since the late seventies, whereas Renhe and Yishala have always had a few. What is the reason for this?

It might be possible to explain different proportions of stem and nuclear families by demographic variables. For example, if the eldest generation is living longer, there will be a higher proportion of stem families, with everything else held constant, since there will be a higher probability of three generations being alive at once. For the villages discussed here, however,

[25] Sons are preferred, here as elsewhere. But a daughter will do; uxorilocal marriage is a continuing option in all three villages and has always accounted for about 10 percent of marriages in each place.

there seems to be no difference—the average age at death for members of our Renhe families who lived to age 20 was 55.3 for women and 57.7 for men, in Zhuangshang it was 56.5 for women and 55.2 for men, and in Yishala it was 56.8 for women and 57.1 for men.

Age at marriage would also affect the proportion of joint families in a population, since earlier marriage would also mean earlier childbearing and a consequently greater chance that three generations would be alive at once. But again, we find no significant differences in the two villages—for male age at marriage, the crucial variable in a patrilocal family system, the average for Renhe is 23.2, for Zhuangshang 22.2, and for Yishala 22.1; certainly no significant difference.

Finally, the size of the brother-set might affect the proportions of stem and nuclear families. If there are more brothers, and if most families divide before the death of the parents, then there will be more nuclear families (including brothers who do not stay with the parents) for each stem family (the brother who does stay with the parents). I have calculated numbers of surviving children for mothers in the three villages who married from 1951 to 1960 (whose sons grew to manhood and faced family division in the 1970s and 1980s), and indeed the numbers were slightly larger in Zhuangshang than in Yishala or Renhe, but only slightly so—4.7 children in Zhuangshang and 4.4 in both Renhe and Yishala. This is not enough to account for the differential proportions of nuclear and stem families.^[26]

To explain this difference, then, we must turn away from demographic factors and look at the causes of variation in the pattern of family division. We might at first suspect that Renhe and Yishala families are dividing later than those in Zhuangshang. The late division of Renhe families would go along with previous findings that entrepreneurial families tend to stay together longer in order to pool labor, capital, or both. But if this were the case, we would expect to see a high proportion of joint families in Renhe, and we would also expect that a higher proportion of entrepreneurial families would be of the more complex, stem or joint type. Figures do not bear

[26] If we figure that in every brother-set, one will stay with the parents upon marriage, while the others move away to form nuclear families, then if the brother-set averages 2.35, as in Zhuangshang, 42% of the families formed upon marriages of the sons will be of the stem type, while an average brother-set size of 2.2, as in Yishala and Renhe, will produce 46% stem families. The actual percentages, as indicated above, diverge much more widely than that.

If parity makes little difference for family complexity now, it is bound to matter more in the future. If Renhe people continue to have only one child, every person with living parents at the time of marriage will form some sort of three- or four-generation family; perhaps many of these will be of the double-grandparent type, in which both the husband's and the wife's parents live with the couple. The alternative is a large proportion of old people living alone with none of their children, and the kind of "networked family" described by Ikels and Unger in this volume. In Yishala and Zhuangshang, however, there ought to be just enough children for everybody to form a stem family.

this out. There have been a few joint families in Renhe in the last ten years, but always less than 5 percent of the population. And the relationship between family economy and family composition is not much more convincing: in a cross tabulation of family complexity and family economy, we find essentially no relationship.^[27]

But in fact there is an important difference between Renhe and Yishala, on the one hand, and Zhuangshang, on the other, with regard to the process of family division. We can see this if we classify family divisions into two types. One, which I call type A, is family division after the brothers have already formed a joint family. This is the classical fen jia of Chinese ethnology (as described by Cohen in House United, House Divided ). The other, type B, consists of brothers either hiving off one by one soon after their own marriages, or in some cases, even before marriage. Table 4.7 shows the prevalence of type A and type B family divisions in the last three five-year intervals.

It is apparent that type A divisions, those in which brothers wait until they have formed a joint family for at least a few months or years before they divide, have increased as a proportion of all family divisions in Renhe in the years since the economic reforms; whereas they were less than a fourth of family divisions from 1973 to 1982, they constituted almost half the divisions in the more recent interval. In Yishala, this type of family division has always constituted a significant proportion of all divisions. If there is no visible difference between the timing of divisions within Renhe or

[27] Since no significant results were found, this tabulation has been omitted from this chapter.

Yishala, from one category of family economy to another, there is still a very real difference between these villages and Zhuangshang, where essentially all family divisions come piecemeal at the time of each brother's marriage; this explains the complete absence of joint families in that poor village in the 1983 and 1988 cross sections.

Moreover, this pattern has emerged in Renhe with the economic reforms, with the transition from a subsistence economy to an entrepreneurial economy, while the families of Zhuangshang continued to be constrained to an almost purely subsistence-farming economy. In this kind of economy, it matters little whether brothers stay together after their marriages. There are no economies of scale in subsistence grain farming, and there are few if any opportunities to save and invest in an economy as poor as Zhuangshang's. So the stresses of joint family life are avoided, and each brother and his family go their own way when they marry.

But why the lack of differences between entrepreneurial and other families in Renhe? Perhaps this is because almost all families in Renhe are entrepreneurial to an extent; even those not currently engaged in entrepreneurial activity face the possibility not only of entrepreneurship but of diversification in general, which in today's economy is a strategy of advancement. It may be best not to make too much of the current differences in each individual family's economy when all families are caught up in an expanding, petty-capitalist race for income and status. And as such they are not automatically dividing their households as soon as brothers marry.

The problem for this kind of argument, of course, is Yishala. The families there are in no sense entrepreneurial; though they supplement their farm income with lime manufacture and have enough cash to invest in a few modern conveniences, they are not plunging headlong into the pursuit of profit, as are so many families in Renhe. Nevertheless, for the last twenty years at least, nearly half their families have grown to the joint phase before dividing. It is possible that the additional labor requirements of lime production may retard family division somewhat, but it seems doubtful that this is a major consideration. In addition, the typical family cycle does not seem to have changed from the pre-1978 period to the period of the reforms. Perhaps it is simply local tradition; certainly there need not be a rational economic calculus behind every change in family composition.

Nevertheless, no family in Renhe has remained in a joint state for more than two years, at least since before 1968. In Yishala, the longest recorded period of joint-family organization is six years, with another family having stayed together four years before dividing. It appears that even in Renhe's entrepreneurial environment and in the face of Yishala's long-held customary practices, the centrifugal pressures still outweigh the perceived advantages of staying together. Or perhaps really significant entrepreneurship,

the kind where capital-pooling confers a perceptible advantage in expansion, is just now under way in Renhe, and we will see more families remaining into the joint phase in the next few years. Of the two joint families in our survey in 1988, one was the transport, fish-farming, and stockraising enterprise of the family with the twenty-nine-room house; perhaps they will remain together longer. However, at least three entrepreneurial joint families divided between 1983 and 1988.

Conclusion

It is clear that family process in the rural areas around Panzhihua city is changing and that the changes have something to do with the twin policy initiatives of the post-Mao government: strict population control and liberalizing economic reforms. Population control has been important everywhere, but has had a differential effect on the three villages studied, with Renhe effectively reduced to one-child families and a consequent drop in household size, Yishala showing a more moderate drop, and Zhuangshang, where people are allowed to have two children, seeing family size decline more slowly.

The other factor, the economic reforms, seem to have had a less universal effect on family process and composition. The cycle of growth and division in Yishala and Zhuangshang has changed little because of the reforms; in Yishala before the reforms, people often allowed their families to grow to the joint form; this tendency has remained. In Zhuangshang, the prereform tendency not to form joint families has also continued. That neither of these villages has seen much change in this aspect of family process is not surprising, since the transformation from collective to individual farming in these villages has not resulted in a massive shift away from subsistence farming. The villagers of Yishala have experienced some prosperity with the reforms, but no great transformation in the physical rhythm of their lives or in the expenditure of their labor. In Zhuangshang, even the increase in prosperity has been quite modest.

Renhe is another kind of place altogether; there the entrepreneurial, petty-capitalist economy is in full swing, and there is some suggestion that it has prompted some families to pool their resources by remaining undivided longer than they would previously have done. It will take another decade of such entrepreneurial expansion, however, before we can determine with any certainty whether there has been a fundamental change in family process as a result of the economic reforms. As it now stands, the most important factor influencing changes in family process is the birth-limitation campaign.

Five
Family Strategies and Economic Transformation in Rural China: Some Evidence from the Pearl River Delta

Graham E. Johnson

The arrangement of familial economic roles such as redistributor, manager or worker and the pooling and redistribution of resources are characteristics of [Chinese] family life adaptable to a great variety of economic activities .
MYRON L. COHEN, HOUSE UNITED, HOUSE DIVIDED

The great majority of the Chinese population lives in rural settlements, and their economic activities are largely agricultural. Rural policies in China were shaped by distinct ideological directions after 1949 and had profound implications for the operation of rural households and rural kinship organization. Chinese rural cultivators were first caught up in land reform and then in the process of the collectivization of production. The fundamental attributes of a peasant economy relate to land, the allocation of household labor to work the land or to meet cultivation responsibilities, and the link with nonpeasants through an elaborate network of marketing. Land reform affected cultivation rights. Collectivization compromised the household allocation of labor. From the mid-1950s, planned purchase and supply of major agricultural commodities transformed marketing behavior. The consequences for the economic organization of rural households were farreaching. In 1979 China embarked on a new developmental course, which was first charted in the countryside. There quickly emerged different possibilities for rural Chinese households and their members to pursue economic and other goals different from those that had prevailed during the first three decades of the People's Republic of China.

There is a consensus that the Chinese peasant household is a corporate entity in which its members cooperate to meet (economic) goals.^[1] That co-

[1] Thoroughly described in Cohen, House United, House Divided (New York: Columbia University Press), 57-85.

operation is reinforced by religious and ritual solidarities. The intimate linkages between economic requirements and domestic social organization give rise to the pooling of labor and the allocation of labor by household units. In south China, especially Guangdong and Fujian, localized lineages elaborated principles of Chinese kinship and were central features of rural social structure. Administrative change and political leadership after the formation of the People's Republic of China compromised well-understood methods of family and kinship organization. The active administrative interventions from the cooperative movement in the early 1950s up to the formal demise of the commune system in 1983 disrupted traditional methods of household economic organization. Control over decision making by households for the three decades after land reform was lost, not to world economic forces, but to a bureaucratic state and its intervention.^[2] Lineage organization in south China was compromised by land reform and was thereafter ideologically suspect.

This chapter will examine aspects of the characteristics of Chinese households and kin groups in rural areas of the Pearl River delta in Guangdong province in the mid-1980s. The Pearl River delta is a distinctive and well-described cultural setting for Chinese households and kin groups. The economic transformations as a consequence of policy initiatives since 1979 have been substantial in all the coastal provinces of China. They have been especially marked in Guangdong and the Pearl River delta.^[3] The Pearl River delta, more than any other region of China, has become firmly incorporated into the global economy and has assumed a particular role in the new international division of labor.^[4] Major rural economic change has occurred as a consequence of incorporation into the world system. There has been a shift from a relatively insulated, state-dominated economic system to one that is open to outside influence, one in which rural households and the communities of which they form part have assumed, or reassumed, control over production activities. Peasant households in the Pearl River delta have responded with alacrity to opportunities in the wake of reform. In certain regions of the Pearl River delta, and especially those where the Overseas Chinese presence is marked, lineages have been revived and elaborated.

[2] The broad dimensions of the process are most provocatively explored by Vivienne Shue in The Reach of the State: Sketches of the Chinese Body Politic (Stanford: Stanford University Press, 1988). Some of Shue's arguments are also explored in Helen Siu, Agents and Victims in South China: Accomplices in Rural Revolution (New Haven: Yale University Press, 1989), which are placed in the historical context of developments in Xinhui in the Pearl River delta. See especially 189-211.

[3] E. Vogel, One Step Ahead in China: Guangdong under Reform (Cambridge: Harvard University Press, 1989).

[4] David Thorns, "The New International Division of Labor and Urban Change: A New Zealand Case Study," in Pacific Rim Cities in the World Economy , ed. M. P. Smith, 68-101 (New Brunswick: Transaction Books, 1989).

Within Third World rural development, the critical social stratum is the peasantry. The sociology of the peasantry is not part of an extensive literature and has assumed major significance only in the modern period.^[5] The analysis of the role of Chinese peasants in development is compounded by three decades of policies that were formulated to create "purposive change."^[6] The best-documented accounts of the transformation of traditional peasantries, and yet ones that have become subject to considerable controversy, are European examples.^[7] The transformation of the peasants of European Russia constitute one example directly comparable to the Chinese case, in which their "awkward" character has been clearly demonstrated.^[8]

The issues in the transformation of Asian peasantries have been explored in the debate between Scott and Popkin.^[9] How peasants react to their incorporation into the world-system is a critical question raised by those discussions. Scott has most recently provided a finely worked case study,^[10] which has gone some way in meeting Keyes's comment that "theories . . . can shape significant questions but answers to those questions if they are to be good answers must take into account not only objective conditions that constrain social action but also the historically situated 'social space' within which action actually occurs."^[11]

Households and the communities of which they are a part are key elements in the "social space" of a peasant social stratum. It is a generally held view that peasant households are adept at meeting threats to the integrity of their basic cultural structures. Peasant households can be

[5] Eric Wolf, Peasants (Englewood Cliffs, N.J.: Prentice-Hall, 1965), remains one of the best general statements. See also his "The Vicissitudes of the Closed Corporate Peasant Community," American Ethnologist 13 (1986): 325-29.

[6] W. L. Parish and M. K. Whyte, Village and Family in Contemporary China (Chicago: Chicago University Press, 1978), 8-15.

[7] See T. H. Aston and C. H. E. Philpin, The Brenner Debate: Agrarian Class Structure and Economic Development in Pre-Industrial Europe (Cambridge: Cambridge University Press, 1985).

[8] See T. Shanin, The Awkward Class: Political Sociology of a Peasantry in a Developing Society: Russia 1910-1925 (Oxford: Oxford University Press, 1972), and Robert Edelman, Proletarian Peasants: The Revolution of 1905 in Russia's Southwest (Ithaca: Cornell University Press, 1987).

[9] James C. Scott, The Moral Economy of the Peasant: Rebellion and Subsistence in Southeast Asia (New Haven: Yale University Press, 1976); Samuel I. Popkin, The Rational Peasant: The Political Economy of Rural Society in Vietnam (Berkeley: University of California Press, 1979). See also the provocative arguments in Daniel Little, Understanding Peasant China: Case Studies in the Philosophy of Social Science (New Haven: Yale University Press, 1989), and the study by Wan Hashim, Peasants Under Peripheral Capitalism (Bangi: Penerbit Universiti Kebangsaan Malaysia, 1988).

[10] James C. Scott, Weapons of the Weak: Everyday Forms of Peasant Existence (New Haven: Yale University Press, 1985). See also the examples in Forest D. Colburn (ed.), Everyday Forms of Peasant Resistance (Armonk, N.Y.: M.E. Sharpe, 1989).

[11] Charles F. Keyes, "Peasant Strategies in Asian Societies: Moral and Rational Economic Approaches—A Symposium: Introduction," Journal of Asian Studies 42, no. 4 (August 1983): 754.

highly flexible in their responses to new economic and social opportunities. Equally, the communities that groups of peasant households form may become part of a collective strategy in the response to new opportunities implicit in incorporation into the world-system.

It is often assumed in the contemporary period that rural peoples are overwhelmed by the forces unleashed as a consequence of incorporation into the global economy.^[12] Instead of their becoming passive victims of world forces, a community-led response by these peoples to new economic opportunities may create protection against global forces that threaten to disrupt the delicate social fabric of rural areas. Such responses may reveal the resilience of long-established social units, not least family and kin groups, that major theoretical paradigms have argued are readily compromised by major economic transformation.^[13] Peasant kinship structures may not be weakened by new economic forms. On the contrary, they may become spatially more ramified to take advantage of new economic opportunities. Further, the arguments proposed by Skinner that markets and marketing create a major focus for Chinese peasant social existence are sustained as market areas themselves undergo a domestic renewal and form the major link with the global economy.^[14]

The Chinese Case, Guangdong and the Pearl River Delta

The rural sector has been a key element in China's development strategy. The record up to 1979, however, was an uneven one.^[15] In the thirty years after 1949 a collective approach to rural production was created in which the traditional link between peasant households and the cultivation of privately owned land was compromised, and a large state-organized planning apparatus intervened in the Chinese peasant economy.

In 1979 major changes in China's policies began in the countryside. Theories and practices of a thirty-year period were significantly modified. There was significant decentralization of responsibility for agricultural pro-

[12] Peter Worsley has described this as "the undoing of the peasantry." See his The Three Worlds: Culture and World Development (London: Weidenfeld and Nicholson, 1984), 61-167; see also William E. James, Seiji Naya, and Gerald M. Meier, Asian Development: Economic Success and Policy Lessons (San Francisco: International Center for Economic Growth, 1987), 157-90.

[13] See the arguments, for example, in Neil J. Smelser, "Mechanisms of Change and Adjustments to Change," in Industrialization and Society , ed. Bert F. Hoselitz and Wilbert E. Moore, 32-54 (Paris: UNESCO and Mouton, 1963).

[14] G. William Skinner, "Rural Marketing in China: Repression and Revival," China Quarterly , no. 103 (September 1985): 393-413.

[15] Nicholas Lardy, Agriculture in China's Modern Economic Development (London: Cambridge University Press, 1983).

duction, and peasant households once again became the focal point of the agricultural sector of the rural economy. After 1979 Chinese peasant households assumed major management responsibility over agricultural production, including both subsistence grain production and economic crops, which became increasingly subject to market forces. At the same time, rural development policy encouraged diversification of the rural economy and promoted a shift away from a central reliance on agricultural production.^[16]

While Chinese peasant households could contemplate new forms of (private sector) economic participation, the major transformation of local economies in some of China's regions was led by a dramatic growth in collectively managed enterprises, the lineal descendants of the commune and brigade industries, whose growth had been severely limited in the period before 1979. The formal demise of the system of the People's Communes is linked to "decollectivization" and the increased ability of peasant households to determine the allocation of household labor.^[17] It is nonetheless paralleled by an intense reorganization of local administration in which the entrepreneurial and business skills of local cadres have become more critical in their judgment of performance than political and ideological accomplishments, which had prevailed for much of the previous thirty years.

Community control of nonagricultural (and especially industrial) activities was vigorously pursued, which augmented household control over strictly agricultural production. Economic opportunity in nonagricultural production expanded throughout rural China and had major consequences for peasant households, which could deploy their members into an array of decidedly new economic activities. Major changes in China's rural development policies and their impact in a variety of regional contexts since 1979 have been extensively documented.^[18] The Pearl River delta in Guangdong has been favorably located to take advantage of new directions in policy.

Guangdong, located strategically along China's south coast, is a large province of substantial geographic, economic, and cultural diversity. Its economic core is the Pearl River delta, with Guangzhou as its regional center. The delta also contains Hong Kong and Macao and two of the three

[16] Jeffrey R. Taylor, "Rural Employment Trends and the Legacy of Surplus Labour, 1978-1986," China Quarterly , no. 116 (December 1988): 736-66.

[17] A point made by Vivienne Shue, "The Fate of the Commune," Modern China 10, no. 3 (July 1984): 259-83.

[18] See the essays in W. L. Parish, Chinese Rural Development: The Great Transformation (Armonk, N.Y.: Sharpe, 1985); K. Griffin (ed.), Institutional Reform and Economic Development in the Chinese Countryside (London: Macmillan, 1984); E. Perry and C. Wong, The Political Economy of Reform in Post-Mao China (Cambridge: Harvard University Press, 1985), 1-194; Robert F. Ash, "The Evolution of Agricultural Policy," China Quarterly , no. 116 (December 1988): 529-55; Jorgen Delman, Clemens S. Ostergaard, and Flemming Christiansen (eds.), Remaking Peasant China: Problems of Rural Development and Institutions at the Start of the 1990s (Aarhus: Aarhus University Press, 1990).

Special Economic Zones (SEZ) that were located in Guangdong after the promulgation of the new "open door" policies in 1979. Outside the delta, the province is mountainous and therefore relatively poor, with inadequate communications and a pronounced ethnic diversity. Guangdong is one of China's great rural regions, ranking third in terms of the total value of rural production in 1988. Although deficient in grain production, it produces over 40 percent of China's sugarcane and is the largest producer of fruit, fish, and forest products. The province's heavy industrial base is relatively small, although it is an important center of light industry. Only Shanghai produces more sewing machines, and the province leads the nation in the production of electric fans. It is a major producer of bicycles, refined aluminum products, hand tractors, and certain food products, of which refined sugar is most significant. Since 1979 there have been sharp increases in the production of domestic appliances, electronic goods, and cameras. Overall economic growth rates have been especially marked since the middle of the 1980s. Guangdong has consistently attracted the bulk of China's foreign investment,^[19] which has had consequences for the province as a whole but which has been especially marked in the Pearl River delta.

Aggregate statistics for Guangdong province, although impressive, tend to mask the performance of the delta region which, since early 1985, has been designated an "open economic region" (jingji kaifang qu ).^[20] The delta region occupies a commanding position in the Guangdong provincial economy. With less than a fifth of the provincial population, it has built on its advantageous location, closely proximate to Hong Kong and, through it, to the world economy. Guangdong was unable to take full advantage of its long tradition of commercialized production before 1978 but, nonetheless, was the major provincial producer of cash crops such as fish, silk cocoons,

[19] Yang Dali, "Patterns of China's Regional Development Strategy," China Quarterly , no. 122 (June 1990): 248.

[20] The core of the delta is composed of the two "municipalities" (shi ) of Foshan and Jiangmen, which were created out of the former Foshan prefecture, in the early period of reform. The western part of the delta is composed of the four xian (Enping, Taishan, Kaiping, Xinhui), from which the bulk of the North American Chinese population traces its ancestry, Doumen and Heshan; the central delta extends from Guangzhou south and includes Nanhai, Punyu, Shunde, and Zhongshan and also mountainous Gaoming and Sanshui. The eastern delta consists of Dongguan and Bao'an, adjacent to the Shenzhen Special Economic Zone and Hong Kong. In 1987 the administrative definition of the "Open Economic Area" was expanded up the West River to include Zhaoqing shi , north of Guangzhou to include Huaxian and Conghua xian , and eastward to include much of what had formerly been Huiyang prefecture. Most of these new additions are not strictly delta, are mountainous and Hakka-speaking. They had begun to share in the prosperity of what is known as "the small delta," the original, essentially Cantonese-speaking, core. For a discussion of the "small" and "large" delta, see Guangdongsheng Nianjian 1989 (1989 Guangdong Yearbook), 252-53; 492; 496.

fruit, and vegetables. In the wake of rural reform its agricultural sector flourished under the newly liberalized policies. While sericulture in the central delta declined, the production of fish, fruit, and vegetables expanded dramatically, destined for the domestic market as well as Hong Kong and Macao. The major transformation in the rural sector, however, was not in agriculture. Rapid economic growth occurred as the rural sector shifted to nonagricultural production. Entrepreneurial energies were released in a flurry of industrial growth, concentrated, initially, in certain regions of the delta. The central delta from Guangzhou south to Macao and the Zhuhai Special Economic Zone and the eastern corridor to Hong Kong and the rapidly expanding Shenzhen area comprised the major areas of new economic activity. These developments were possible partly because national policy allowed the province considerable autonomy to pursue innovative measures to seek capital and to retain foreign exchange earnings. The delta was able to mobilize the extensive links it possessed with its expatriate kinsmen who live in Hong Kong, Macao, and overseas.

The entire delta has long been an area of out-migration. From the mid-nineteenth century large numbers of émigrés from south and southeast China met some of the developmental needs of the burgeoning global economy and worked as unskilled laborers in the Americas and in Southeast Asia. The western reaches of the delta have provided the great majority of Americans and Canadians of Chinese origin. Under different circumstances, a large number of delta residents left for Hong Kong in the late 1940s and 1950s, and many contributed to Hong Kong's economic transformation after 1950.

In the 1980s the entreprencurial skills and capital of these émigrés were actively sought. Kinship connections and local loyalties have become a central part of local development initiatives. In the process the delta has become firmly linked to the global economy through its Hong Kong connections. It has thus begun to share with other parts of East and Southeast Asia some of the developmental characteristics that McGee has described as the "desakota process," in which an intense mixture of agricultural and nonagricultural activities stretch along linear corridors between large city cores. The process typically occurs in regions characterized by high population densities and which were formerly dominated by wet rice agriculture.^[21] The Pearl River delta constitutes one of those regions.

[21] T. G. McGee, "Urbanasasi or Kotadesasi? Evolving Patterns of Urbanization in Asia," in Urbanization in Asia , ed. L. Ma, A. Noble, and A. Dutt, 93-108 (Honolulu: University of Hawaii Press, 1989); G. E. Johnson, "The Political Economy of Chinese Urbanization: Guangdong and the Pearl River Delta," in Urbanizing China , ed. Gregory Guldin, 185-220 (Westport, Conn.: Greenwood Press, 1992).

Map 5.1.
The Pearl River Delta

Research Sites

I have been examining the process of rural development within the Pearl River delta since the reforms began. Building on work begun in the 1970s,^[22] I have worked in five sites across the delta. I chose the sites in the 1970s, when each was administratively a commune, because each had some distinguishing characteristics. Duanfen, located in Taishan xian in the western delta, was a center of out-migration to North America. Leliu in Shunde xian , in the central delta, had a distinctive cash-cropping economy and possessed some unique cultural characteristics.^[23] Luogang and Renhe are close to Guangzhou, located in Baiyun qu (formerly the "suburban district"). Although Luogang is mountainous, it is notable for fruit production. Renhe, to the north, was a grain specialist area, known for its radical political character in the Cultural Revolution and coincidentally a major source of migration to Canada. Fucheng, in Dongguan shi (formerly, xian ) is a major source of migrants to Hong Kong, where I had worked in the 1960s.^[24]

The economies of the five units and the villages that compose them have changed substantially since reform got under way in 1979.^[25] One index of the change and the broadly successful response to the reform initiatives that they have enjoyed is substantial growth in per capita income (table 5.1). The opportunities for economic transformation are a consequence of internal factors, of which the indigenous economy and leadership are key, and location. Since initiation of the reforms linkages with expatriates in Hong Kong or Macao or in Overseas Chinese communities have become of increasing importance.

[22] Elizabeth and Graham Johnson, Walking on Two Legs: Rural Development in South China (Ottawa: International Development Research Centre, 1976).

[23] The distinctive ecological features of Shunde (and of Leliu in particular) are described in Kenneth Ruddle and Zhong Gongfu, Integrated Agriculture-Aquaculture in South China: The Dyke-Pond System of the Zhujiang Delta (Cambridge: Cambridge University Press, 1988).

[24] For much of the 1980s I interviewed cadres at a variety of administrative levels associated with the units. In the summer of 1986 I selected five villages from among the units, based on my general knowledge of them. Using the household records I drew random samples of village households and administered a standard questionnaire. Much of the information that follows comes from this survey. The survey was funded by a research grant from the Social Sciences and Humanities Research Council of Canada. The Chinese Academy of Social Sciences gave me every assistance. In Guangdong I was greatly helped by the Guangdong Academy of Social Science and by the Department of Sociology at Zhongshan (Sun Yat-sen) University. I express my thanks to Li Ruichang and Chen Daojin and to Professor He Zhaofa. Mr. Tan Xiaobing assisted me in collecting the survey data and doing a great deal of coding, data cleaning, and analysis in Canada. His efforts are greatly appreciated.

[25] Some general findings are indicated in the following: G. E. Johnson, "The Production Responsibility System in Chinese Agriculture: Some Examples from Guangdong," Pacific Affairs 55, no. 3 (Fall 1982): 430-52; "1997 and After: Will Hong Kong Survive? A Personal View," Pacific Affairs 59, no. 2 (Summer 1986): 237-54; "Rural Transformation in South China? Views from the Locality," Revue Européene des Sciences Sociales 27, no. 84 (1989): 191-226.

Duanfen, located in southern Taishan, has benefited the least of the five units from the opportunities of the reform period. It is distant from Guangzhou and Hong Kong and its prospects for growth and development were, until recently, hampered by inadequate communications. As an Overseas Chinese area it suffered disabilities from almost the moment of its incorporation into the People's Republic of China. The overseas connections and linkages were political liabilities for much of the three decades from 1949 to 1979, which were especially marked during the period of the Cultural Revolution. It remained a subsistence economy with a substantial dependence on remittances from abroad. Per capita income in the middle 1970s was below a hundred yuan.

After 1979 the overseas linkages were viewed in a positive light, and many of the harsh political judgments that had colored local policy options were reversed. Although the economy has remained agricultural and firmly based on grain production, its base has diversified. Cash crops, such as fruit and sugarcane and the raising of poultry, have become widespread. Duanfen has prospered since reform, but its incorporation into the world-system has not occurred on the same scale as other units in central and eastern portions of the Pearl River delta. The question is partly one of location but also one of the particular nature of Duanfen's linkages to the global economy.

Leliu lies in Shunde in the central region of the delta. It enjoyed a distinctive economy based on fish farming and silkworm cultivation. There are no rice paddies in Leliu, and its commercialized (and export-oriented) agricultural production allowed its peasant cultivators to enjoy relative prosperity even before the reform period. It was an area of extensive out-migration to Hong Kong before 1949, and the linkages with its expatriates were vitalized in the wake of reform. A young and dynamic leadership sought to in-

volve the entrepreneurial energies of Hong Kong in its economic transformation. Its agricultural economy, especially fish farming, has remained buoyant, although sericulture has been eliminated as a consequence of pollution and low prices for cocoons. Dramatic growth in nonagricultural production has overwhelmed its agriculture. The market towns of Shunde have become the centers of light industrial production, which has allowed its households to enjoy the highest per capita incomes in rural China.

Luogang is located off the main highway that links Guangzhou to eastern Guangdong at the eastern edge of Baiyun qu of Guangzhou municipality.^[26] Formerly part of Panyu xian , it has a long history of specialization in fruit production. Changes in pricing policies and release from the highly bureaucratized purchase and supply of fruit has allowed peasant cultivators to increase household income. The region did not have a history of extensive out-migration, and the local economy has seen no growth in its small-enterprise sector since the reform period began. Its villages remain firmly agricultural, although their proximity to major markets (and especially Guangzhou) has allowed them to take full advantage of its new commercial opportunities and profit from high prices paid to fruit growers. Grain production remains important. Proximity to centers of labor demand (especially in construction) has allowed members of peasant households to seek employment opportunities outside agriculture.

Renhe is located at the northern edge of Baiyun qu . The large and dense population is largely engaged in grain production. Renhe lies beyond the zone of intense vegetable cultivation that has developed to meet the needs of the Guangzhou market, and has relatively little involvement in cash-cropping, although it has had some success in raising poultry. Outmigration was substantial in the past, and there are significant numbers of its natives overseas and in Hong Kong. Its industrial capacity at the zhen , village, and individual levels has grown since the reform period, often in cooperation with Hong Kong entrepreneurial interests. Yet, the degree of Renhe's incorporation into the broader economy of the delta is less extensive than that of other units in the central delta or of those closer to Guangzhou or Hong Kong. The capacity of the local economy to absorb surplus labor is limited. Large numbers of men, especially those under forty years of age, have left the area to seek work in Guangzhou and elsewhere in the delta, leaving the management of the agricultural economy in the hands of women and older men.

Fucheng extends in a wedge to the south and east from Dongguan city. The economic growth of Dongguan since the reform period got under way has been remarkable. Fucheng has fully shared in the transformation of the

[26] For a broad account of Baiyun qu (formerly the "suburban district"), see Guangzhou Nianjian 1989 (Guangzhou Yearbook) (Guangzhou: Wenhua chubanshe, 1989), 487-89.

entire shi .^[27] Dongguan has benefited from its close proximity to the Shenzhen Special Economic Zone (SEZ) and Hong Kong. It has fully utilized investment resources and skills of its expatriates in Hong Kong who have provided much of the capital for an extensive array of enterprises, both industrial and agricultural, which have transformed the rural economy of Dongguan. Fucheng has over four hundred enterprises, whose foreign exchange earnings in 1987 were over $HK50 million. They include plastics, garment manufacture, electronics, and toys, virtually all destined for the would market as subsidiaries of Hong Kong-based manufacturers; they also include large shipments of vegetables for the Hong Kong markets and fruit for both Hong Kong and domestic markets throughout China.

The five units are broadly representative of the Pearl River delta. They extend across the delta, representing differing production regimes, which are historically determined but also a consequence of different leadership strategies and different developmental possibilities in the wake of rural reform after 1979. The villages within the five units and their constituent households have responded with alacrity to the new possibilities. I will first outline the consequences for peasant households in general and then discuss peasant household strategies in the central and eastern delta (what Vogel calls "the inner delta").^[28] I will finally give an account of household strategies in the distinctive western reaches of the delta, in which overseas-Chinese connections are a dominant characteristic.

Pearl River Delta Households: A Descriptive Profile

Households in the sample population had a mean size of 5.1, not dramatically different from household size in other national rural samples in the 1980s,^[29] from the 1970s estimate of 5.0 for Guangdong by Parish and Whyte,^[30] or from the 1930s figure for Guangdong of 5.2.^[31] Household size ranged from single individuals to an enormous household of 26. There were

[27] The performance of Dongguan is detailed in Zhonggong Zhongyang Bangongting, Dongguan Shi'nian: 1979-1988 (Dongguan Ten Years: 1979-1988) (Shanghai: Shanghai renmin chubanshe, 1989). Fucheng's performance is detailed at pp. 219-20.

[28] Ezra Vogel, "Guangdong's Dynamic Inner Delta," China Business Review , September-October 1989, 56-62.

[29] Household size from a national sample of rural households in 1986 was 5.07, cited in Zhongguo Tongji Nianjian 1988 (Statistical Yearbook of China 1988) (Beijing: Zhongguo tongji chubanshe, 1989), 822; Guangdong samples indicate a decline from 5.74 in 1978 to 5.12 in 1985, cited in Guangdong Tongji Nianjian 1988 (Statistical Yearbook of Guangdong 1988) (Beijing: Zhongguo tongji chubanshe, 1989), 415.

[30] Parish and Whyte, Family and Village , 134.

[31] Irene B. Taeuber, "The Families of Chinese Farmers," in Family and Kinship in Chinese Society , ed. M. Freedman, 71 (Stanford: Stanford University Press, 1970).

26 single-person households (5.0 per cent), 18 (older) women and 8 men. The modal household as in the 1970s was nuclear in structure. There were relatively few joint households, though a large number of stem or broken stem households (table 5.2). Family relationships in the sample of households are similar to the patterns detailed by Parish and Whyte in their study of Lingnan villages in the 1970s (table 5.3).

Certain kinds of household relationships that existed in the past—and were familiar in the studies of household and family in the peripheries of Hong Kong and Taiwan in the 1960s^[32] —either are nonexistent in the contemporary period or are extremely rare. Polygyny became illegal in 1950 as a result of the Marriage Law, and its economic possibility was compromised by land reform and later collectivization. Plural wives are a characteristic of the past. "Minor" forms of marriage, such as "small daughters-in-law," once common throughout the delta, become impossible in the context of a legally enforceable age of marriage,^[33] and certainly the logic for most forms of minor marriage is less compelling than in the pre-Liberation context.^[34] Households are no smaller than they were in the past, although they were likely to be somewhat larger in the late 1970s, but they are less complex. There are fewer members of senior generations than there were in the past, and there is an almost total absence of father's brothers and father's brothers' wives. Similarly, there are only few siblings (and their wives) and a virtual absence of kin outside direct lines of descent.^[35]

One of the more obvious elements of continuity is patrilocal marriage.^[36] The dominant form of postmarital residence is firmly patrilocal. A close adherence to well-defined areas from which to draw brides is clearly appar-

[32] Summarized most effectively in Maurice Freedman, Chinese Lineage and Society: Fukien and Kwangtung (London: Athlone Press, 1966), which inspired the first generation of anthropological fieldwork in Chinese settings after the formation of the People's Republic of China.

[33] This is not to suggest that all unions occur only upon the assumption of legal age. There are ways that socially recognized marriages can occur regardless of formal legality. See Parish and Whyte, Village and Family , 164-66.

[34] See, for example, the arguments in James P. McGough, "The Domestic Mode of Production and Peasant Social Organization: The Chinese Case," in Chayanov, Peasants and Economic Anthropology , ed. E. Paul Durrenberger, 183-201 (Orlando, Fla.: Academic Press, 1984).

[35] There is a virtual absence of unrelated kin. There was one (legal) household of unrelated men (in the Shunde village). The villages did contain substantial numbers of individuals who were unrelated by either kinship or marriage to the indigenous people of the villages. They were not living in legal households with formal registrations. I could not discover any way to interview such individuals, who generally appeared to be non-Cantonese-speaking males working as day laborers. The situation had changed by the end of the decade when non-Cantonese-speaking households began to contract land for two- and three-year periods and build houses and small hamlets in styles distinct from those of the indigenous delta villages.

[36] It is not easy in survey situations to distinguish minor marriage or distinctive local customs. Divorce appears to be very low. The three cases of divorce in Naamshui (Leliu, Shunde) were sought by men from women who had refused to consummate their marriages and move from their natal households. See Janice Stockard, Daughters of the Canton Delta (Stanford: Stanford University Press, 1989), for a discussion of marriage patterns in Shunde, which were anything but "major." In the Leliu village of Naamshui I also came across a household headed by an unmarried sister well past marriage age, who was keeping house for her younger brothers, had engaged special tutors for them, encouraged them to study hard, and was delighted to find two of them at university. The commitment of Shunde women to their natal homes is a very strong one.

ent. Village endogamy is uncommon in all but the large multiple-surname village of Wantong in Dongguan. The marriage area may be somewhat more tightly drawn than in a previous generation. More spouses and daughters-in-law tend to be from the same xiang than are mothers. The most common provenance for wives is within the bounds of the production unit (the former commune, now typically called zhen ) and rarely extends beyond the adjacent administrative entity (table 5.4).

The coincidence between former commune boundaries and standard markets was very close.^[37] Marriage relations were most intense within the bounds of the market area. It was a pattern reinforced by administrative practice during the commune period and has not been compromised by new administrative arrangements, which in the Pearl River delta have seen few shifts in administrative boundaries, although much confusion in changes of name.^[38] Administrative and other changes in the 1980s have only reinforced traditional boundaries that have salience for rural households.

[37] This was first detailed in G. W. Skinner, "Marketing and Social Structure in Rural China: Part III," Journal of Asian Studies 24, no. 3 (May 1965): 363-99.

[38] Rural Guangdong has a distinctive administrative structure, which is a consequence of its particular characteristics and historical developments. The basic unit of local government is called the zhen (market town) and is equivalent to the former commune. The former brigade, which was based on the pre-Liberation xiang (village alliance) became a unit of government for only a brief period in the mid-1980s and paralleled the creation of "township governments" (xiang zhengfu ) in most other Chinese provinces. It is now officially referred to as the "village" (cun ). The former production team, which may have formed a natural village or a village neighborhood during the commune period, is called the "village small group" (cun xiaozu ).

Marriage is central to family continuity. The patrilocal system of marriage has been little affected by the dramatic changes of forty years. Brides are sought from other villages, which have tended over a long period of time to be linked in a system of bridal exchange. Administrative practice during the commune period likely contained and intensified the long-established system of marriage exchange. The renewal of the system of periodic markets has only reinforced the long-established contexts within which marriage occurs. To be sure, marriage is no longer "blind," and individuals have choice in the matter of whom they will marry. The decision to marry, nonetheless, occurs as part of clearly understood family consequences and is arranged in consultation with members of the kin group. Go-betweens make connections between families with eligible children. Their provenance is the area served by the periodic market, and their sphere of operation is the market town on market day.^[39]

Demographic changes over the past forty years have had some major significance for rural household structure. Infant mortality rates have fallen progressively since the 1950s and especially after the Great Leap Forward.^[40] Hence, rural households have more surviving children than was the case for rural households in the republican period. This likely has had effects for adoptions and other forms of incorporation into the household of individuals outside the direct line of patrilineal descent.^[41]

There are more female heads of household (16.9 percent) than at earlier periods. If single-person households are excluded (the majority of these are elderly widows) the proportion of households headed by women is 14.1 percent. There is some regional variation. In the Overseas Chinese areas, households are much more likely to be headed by women. Women are least likely to head households in the two wealthy areas of Shunde and Dongguan.

The incidence of (recent) household division is substantial. While 30.0 percent of households remain undivided, 41.2 percent have divided since the beginning of reform. Households are most likely to have remained undivided in the Overseas-Chinese-dominated village of Kongluen in Duan-

[39] Margery Wolf makes similar arguments. See her Revolution Postponed: Women in Contemporary China (Stanford: Stanford University Press, 1985), 167-68.

[40] Judith Bannister, China's Changing Population (Stanford: Stanford University Press, 1987), 103-20.

[41] Parish and Whyte, Family and Village , 136, reported a virtual absence of adopted sons (except among households headed by women) and an absence of adopted daughters and "small daughters-in-law." This is also the observation of Sulamith H. Potter and Jack M. Potter, China's Peasants: The Anthropology of a Revolution (Cambridge: Cambridge University Press, 1990). They note in the Dongguan village where they worked, which is adjacent to the one I surveyed, but in a different administrative unit, that "adoptions are rare" (220).

fen (Taishan). Division has been most marked in the fruit-growing village of Tsimkong in Luogang, Guangzhou. It was also high in Wantong in Fucheng (Dongguan) as the reform period got under way. I will offer an explanation below.

The past four decades have been characterized by dramatic shifts in economic and political organization in rural China. Rural households in the Pearl River delta are caught up in late-twentieth-century change. Although the context has changed, household structure at first appears to be characterized by continuities. Households are not immune to what occurs beyond their bounds. Change is, however, kept at arms length, and peasant families have devised strategies to deal with the economic and political threats to their integrity. The patrilineally structured rural household has not merely maintained its corporate character but has also used its intrinsic flexibility to meet new possibilities in the 1980s.

Changing Economic Characteristics of Households

The elaboration of the global economy and the creation of a new international division of labor has had sometimes catastrophic consequences for economic management by peasant households. There is a general assumption that in the twentieth century rural cultivators (peasants) have become less able to meet their own (or local) needs as their production activities become overwhelmed by the need to meet the demands of markets they do not control and participate in only indirectly.^[42] As rural production becomes increasingly capitalized and incorporated into the world-system, nonsubsistence production and wage labor become increasingly important. Maclachlan has commented on the consequences of global economic change for rural households in the Third World. He has noted:

In most cases of the capitalization of the countryside, the grounds for divorcing the needs of all or part of the peasantry from production decisions depend upon the character of intruding market forces, while the settlement, the division of costs and benefits, depends heavily on property institutions which prevail at the time of divorce. The role of judge in the whole process is played by government policy. . . . It makes a great deal of difference whether the agricultural sector is commercialized to produce export crops, or whether it is used to provide food and a source of labor for an urban industrial sector. It also makes a great deal of difference whether land tenure features powerful landlords or freeholding peasants. And it likewise makes a difference to know

[42] This is an explicit part of the argument in Eric Wolf, Peasant Wars of the Twentieth Century (New York: Harper and Row, 1969), esp. 276-302.

if a government's chief aim is the acquisition of foreign exchange, or import substitution, or the feeding of its people.^[43]

The economic responses of peasant households in the Pearl River delta to the domestic and international pressures to which they have become subject in the 1980s are varied. There would seem to be three broad responses for those households that are economically active: (1) to remain in essentially subsistence production; (2) to become predominantly cash-cropping households; (3) to endeavor to place one or more members of the household in nonagricultural production. Peasant households will therefore fall into the following categories: subsistence, cash-cropping, multisector. The three types of economic households are to be found in all the villages I surveyed. Given the characteristics of the Pearl River delta, it would be expected that households engaged in other than subsistence production will be most numerous. The region has a highly commercialized agricultural sector, and the growth of the rural-enterprise sector has been marked in the 1980s. Table 5.5 indicates the distribution of household types, classified by their production orientations, in the five villages I surveyed.

Characteristics such as traditions of local production and location are important elements in determining why a village and its constituent households will likely shift from a subsistence to a nonsubsistence orientation. Given policy shifts, the cash-cropping tendencies of villages with well-established traditions of commercial production are intensified, and the proportions of cash-cropping households will therefore likely increase. Similarly, villages close to urban settlements will be subject to more extensive flows of information and capital, and both cash-cropping and multisectoral

[43] Morgan D. Maclachlan, "From Intensification to Proletarianization," in Household Economies and Their Transformation , ed. M. D. Maclachlan, 15 (Lanham, Md.: University Press of America, 1987).

households will be greater. Finally, political leadership, the ability of leaders to establish connections and create entrepreneurial opportunities (for village- and community-based enterprise), will have consequences for economic diversification, a feature of rural development throughout rural China during the reform period.^[44]

Both the size of household and its internal structure, in particular the number of adults, have consequences for labor-force attributes and the ability of households to meet productive goals. The resources a household has, such as the amount of land or available capital, determine the ability to expand productive horizons. Social (class) status and the assumption of administrative or political roles, as cadre or Party member, by household members can augment the ability of households to respond to the changing economic and social options.

Chayanov argued that peasants were adept at devising coping strategies and could utilize control over the deployment of household labor to meet new opportunities. Opportunity for Chinese peasant households was substantially restricted under conditions where collective structures took the major decisions for labor allocation. In the reform period, the devolution of responsibility for production decisions generated greater flexibility and greater possibilities to deploy household resources. If land parcels become small and the number of adults is large, households may pursue a strategy of moving some of their members into other sectors and, if it seems essential and possible, away from the local economy and into economic niches elsewhere.^[45]

The capacities of peasant households to create strategies to meet (changing) economic and political challenges will be determined by internal and external factors of the kind outlined above. In the wake of reform policies, opportunities for delta peasant households have varied from locality to locality. The reform process has turned on "openness," and the external connections that villages possess have had a critical kinship dimension. The presence or absence of kinsmen outside of the village and the broad communities that villages form have been decisive for the development process in general and the kinds of strategies that village households have adopted. Kinsmen have become sources of investment and entrepreneurial skill. Yet, it is important to distinguish between the effects of kinsmen who are resi-

[44] Jean Oi, State and Peasant in Contemporary China (Berkeley, University of California Press, 1989).

[45] The very largest household in my sample had seven marital units: parents, five married sons, and a married grandson. In addition to contracting fishponds and cane lands in the natal village, the family corporation had contracted fishponds in the adjacent zhen , had further contracted fishponds in Dongguan, 100 kilometres away, and had assumed management responsibility over a brick kiln in Shaoguan, in northern Guangdong. Family income was in excess of 40,000 yuan in 1985.

dent in Hong Kong or Macao and those who are Overseas Chinese. Unquestionably, the areas in the western delta have benefited from extensive Overseas Chinese linkages. The substantial (and growing) funds that have flowed into the region from North America and elsewhere have been channeled, however, into projects such as roads, bridges, and public buildings, of which schools and hospitals are most common. Investment funds and entrepreneurial energies that have derived from Hong Kong and Macao, by contrast, have been directed toward the creation of productive enterprises and have had a major and direct effect on local systems of production. The consequences for household strategies in the eastern and central delta, where the presence of kinsmen based in Hong Kong and Macao is proportionally large, can be readily contrasted with those in the western delta, where the proportion of Overseas Chinese kinsmen and Overseas Chinese households is larger. Fundamental economic change, as a consequence of the entrepreneurial activities of Hong Kong—based kinsmen, has been greater in the central and eastern delta. The greater the degree of incorporation into the global economy, the greater will be the effect on household strategies and the greater the implications for changes in household structure. The extent of change will vary from village to village, depending on the extent and nature of the expatriate involvement in the local economy. The corporate kinship structures of the delta are the buffers that allow a highly flexible response to the dramatic possibilities of the reform period.

The Central and Eastern Delta: Hong Kong Connections

The central and eastern delta areas have been most advantageously located in the 1980s to take advantage of new policy options. The region was historically highly commercialized and open to a variety of external linkages. The assumptions of development policy for much of the thirty years after 1949 compromised the operation of local economies, which turned inward under intense pressure from the Chinese state.^[46] Peasant resistance was substantial, but the degrees of maneuverability were highly constrained.^[47] The possibilities for household economic behavior changed dramatically after 1979.

[46] Helen Siu, Agents and Victims , decribes the consequence of state policies from the 1950s for citrus and fan-palm production in Xinhui. Xinhui is on the highly commercialized eastern edge of the western delta. Her field sites are in and around the xian city. The arguments that she makes for Huancheng can be appropriately made for much of the central and eastern delta. Even where cash-cropping was extensive, its development was compromised by the emasculation of peasant marketing after 1956, and especially after the Great Leap Forward.

[47] See David Zweig, "Struggling Over Land in China: Peasant Resistance After Collectivization, 1966-1986," in Everyday Forms of Peasant Resistance , ed. Forrest D. Colburn, esp. 153-62 for an account of peasant resistance to collectivist policies in the Lower Yangzi region.

The large multisurname village of Wantong in Fucheng, Dongguan, which had been a brigade during the commune period, responded with alacrity to the changed circumstances after 1979. It had responded well to collectivization, and a dynamic and highly respected leadership had organized major land reclamation during the Great Leap Forward, which resulted in high grain yields and relatively high per capita incomes even before 1979. Its Party secretary was transferred to the commune level shortly before the reforms got under way. Working closely with a xian leadership committed to exploring new economic relationships, he argued for mobilization of the capital and entrepreneurial resources among the substantial number of expatriates in Hong Kong. The success has been dramatic. It was helped by its proximity to the growing Shenzhen SEZ and the dramatic improvement in road communications with Hong Kong.

The consequences of local economic transformation have been major. There has been little mobility of local labor away from Fucheng as the village economies have intensified.^[48] Only 34 percent of the indigenous labor force is engaged strictly in household-managed agricultural production. The rest works in enterprises, the bulk of which are run by the village as collective entities. There is no single lineage base to the village (as in neighboring Zengbu),^[49] but there is an intense local loyalty, a loyalty that was furthered during the period of intense agricultural collectivism in the 1960s and the 1970s. It has been heightened with the extensive involvement of Hong Kong—based expatriates in the village economy in the 1980s. The links to the global economy are direct. Much of the production of the village-run enterprises, such as garments and plastics, is destined for Hong Kong and for reexport to Europe and North America. Not all the village-run enterprises are industrial. Vegetables are cultivated year round for the Hong Kong market by an enterprise that works jointly with a Hong Kong—based expatriate; it employs wage workers, the bulk of whom are villagers. The local economy is thus closely integrated with the world-system through Hong Kong.

The consequences for village households are various. In terms of economic strategies, none of the economically active households are engaged in subsistence production, and most are multisector households. Rapid economic advance has given rise to increased demand for household division, hence the number of nuclear households is higher than in all other villages,

[48] On the contrary, growth in the number of enterprises in the zhen as a whole (there were 346 enterprises in 1987, of which 118 were wholly engaged in the processing of materials for the international market [wailai jiagong ]) has resulted in the recruitment of several thousand workers from outside the area, many of whom are not Cantonese speakers.

[49] See Potter and Potter, China's Peasants , esp. 251-69. Zengbu is across the river from Wantong, the village I surveyed in Fucheng, and draws the bulk of its brides from the village.

except for the fruit-growing village of Tsimkong (in Luogang), where rapid economic growth has also been marked.^[50]

Management of the household (dangjia ) is least likely to be in the hands of the household head (40.3 percent) than in the other sample villages and is almost as frequently (38.7 percent) in the hands of the spouse of the head. There are proportionately fewer female household heads in Wantong than in the other villages. Women in south China have long been noted for a high degree of economic independence. It is tempting to conclude that increased economic participation of Wantong women in wage labor has enhanced their importance as family managers. There is a common argument that women are increasingly marginalized as incorporation into the global economy grows.^[51] The impact of incorporation into the world-system has been very dramatic for Wantong. It has not been at the expense of marginalizing women indigenous to the area.^[52]

The fruit-growing village of Tsimkong (in Luogang) has prospered in the reform period but not as a consequence of global incorporation. Its major comparative advantage is a long history of growing fruit (oranges, lichee, and pineapple). Reform of the price system and the resurgence of private marketing have caused rapid increases in household income. Household economic strategy is less elaborate than in Wantong; it involves maximum allocation of household labor in highly commercialized agricultural production. There are only a few multisector households and none involved in subsistence production. There is little out-migration and a virtual absence of kin abroad or in Hong Kong or Macao. Household division in the wake of heightened economic opportunities has been marked, and the number of nuclear households is proportionately large. Household management, however, is typically in the hands of (male) household heads who both allocate household labor and retain control over its financial resources. The intensification of agricultural-sector activities, driven as they are by domestic

[50] Margery Wolf, Revolution Postponed: Women in Contemporary China , discusses the issues of household division based on observations she made throughout China in the early 1980s, which, given earlier work in rural Taiwan, are instructive. See esp. 224-26.

[51] The "classic" study is Esther Boserup, Women's Role in Economic Development (London: Allen and Unwin, 1970); Lourdes Beneria and Gita Sen, "Class and Gender Inequalities and Women's Role in Economic Development: Theoretical and Practical Implications," Feminist Studies 8, no. 1: 157-74; Also, J. C. Robinson, "Of Women and Washing Machines: Employment, Housework and the Reproduction of Motherhood in Socialist China," China Quarterly , no. 101 (March 1985): 32-57.

[52] The same cannot be said of wage-workers who are recruited from outside the area. Large numbers of young unmarried females have been recruited to work in the Fucheng enterprises. They are paid well by local Chinese standards, and their accommodation, though spartan, is adequate. They are provided with subsidized food and have access to modest medical services. They are simply wage-workers. They do not participate in management or decision making. It is expected that they will return to their native places when they reach marriageable age.

market demands, seems to have done little to enhance roles for women in the household.

The other two villages in the eastern and central delta are different. Naamshui is less involved in the economic transformation of Leliu than other villages in the zhen . There has been a persistent inability, however, to develop industrial enterprises and thereby diversify the village economy. As a consequence, a large portion of the labor force remains in agricultural production, but significant numbers (40 percent), and most of the men under forty, have left the village and often Leliu itself. They work as carpenters, decorators, and temporary workers in construction and in commerce and transportation. They regularly return to the village, but they have no continuing role in village-based production activities. The village agricultural economy is dominated by women, despite the demise of sericulture, and by older men who have contracted fishponds, either individually or (which is more common) in cooperation with men from other households.

Village households make a steady living, if not spectacular by local standards (per capita income is about 1,100 yuan, slightly below the level of the zhen ). One reason for the failure to diversify the village economy is that it is in a remote corner of the zhen . Its Hong Kong expatriates are few in number, and the village lacks the "connections" (guanxi ) that have been so critical for the success of other villages in Shunde. There is an intense village solidarity, and the organization of production does not differ dramatically from its practice in the 1960s and 1970s. Fish production is organized about cooperative principles, and income from the contracted fishponds is distributed (fen hong ) on an annual basis. The collective health care system works well, there are pensions, and only two "five-guarantee" households.^[53]

Household economic strategy is to maximize opportunities in the cash-cropping skills that Naamshui has long developed, while placing its younger male members outside the village economy. Household division has been less pronounced than in other villages, which have experienced more dramatic internal change in the 1980s. Leliu has the fewest nuclear households and rather more stem households (35 percent) than others in the sample villages in the central and eastern delta. Household management is disproportionately in the hands of household heads, and only 7.2 percent of spouses assume these duties. It has the largest proportion of households with unusual structures, including a "collective" household of men, two headed by older unmarried women who keep house for their (as yet unmarried) brothers, and the only case in the sample of a household in

[53] Five-guarantee households are those (normally elderly) with insufficient family resources to provide food, clothing, shelter, medical care, and a burial, which are provided (guaranteed) by the collective.

which the wife of the head is working elsewhere. Shunde in the past was distinguished by unusual domestic arrangements, of which delayed-transfer marriage was the most notable. Such distinctive practices are still clearly discernible.^[54]

Ngawu, in Renhe, has also changed less dramatically than some other villages in the delta. It, too, is firmly agricultural and has only a small enterprise sector. A significant proportion of its (male) labor force is working outside the village. Ngawu has large numbers of expatriates abroad. They are, however, Overseas Chinese, many resident in Canada. The impact of Overseas Chinese kinsmen on local economies is distinctive and does not translate into economic transformation of the kind experienced in those parts of the delta where Hong Kong expatriates are numerous. There is a further complication in Renhe: it is a grain area with little commercialized crop production; it is too far away from the Guangzhou market to develop fully a specialization in vegetables; its local leadership also had many reservations about the responsibility system and was very tardy with its implementation.

Ngawu has not been able to capitalize on the opportunities to transform its system of production. A large proportion of the households remain in subsistence production. Household strategy therefore relies heavily on meeting basic grain requirements and in placing household members in nonagricultural production outside the village to augment household income. Only a few village households have members working in the zhen -managed enterprises, and there are only a few enterprises at the village level to absorb local surplus labor. The out-migration of household labor is therefore substantial but predominantly male.

Household division is not marked, and household structure reflects patterns that may have been more typical of a previous generation.^[55] Ngawu has the largest proportion of joint households (9.0 percent) and also the lowest proportion of isolates (7.0 percent). Perhaps because of the frequent absence of household heads, there are relatively more spouses (21.1 percent) or individuals other than the head or spouse of the head (15.8 percent) who assume the role of household manager. The village economy has not been affected by incorporation into the global economy. Out-migration was important in the past as substantial numbers went abroad. Its émigrés,

[54] The distinctive marriage system, which many attribute to its underlying economic base, is outlined in Janice Stockard, Daughters of the Canton Delta , esp. 1-30 and 134-66. See chapter 7 in this volume, by Helen Siu, which deals with this form of marriage and its implications in adjacent Zhongshan.

[55] A 1930s study in the delta indicates rather more complex households (63 percent) and fewer isolates (3 percent) than appears to be the case in the contemporary period. See Lewis S. C. Smythe, "The Composition of the Chinese Family," Chin-ling Hsueh-pao 5, no. 1 (1935), cited in Parish and Whyte, Family and Village , 134.

however, have had little effect on village economic change in the contemporary period. Out-migration, on a temporary basis, is still important as household members find work in the burgeoning economy of the delta. They remain involved in household economic tasks and return at periods of heavy seasonal labor demands in their household responsibility fields. Their contribution of labor and earnings from outside employment is important for the well-being of village households. New economic opportunities appear to have no major effects on household structure in Ngawu.

The Western Delta: The Overseas Chinese Connection

Finally, I want to discuss some of the features of the distinctive western delta, the homeland of North Americans of Chinese origin.^[56] Overseas Chinese have always been deeply conscious of their membership in elaborate kin groups and have remained fiercely loyal to their ancestral points of origin. Chinese communities abroad were constructed, in the past at least, about a set of organizations that had at their core principles of fictive kinship (family or surname associations) or common origin. It is the loyalty to ancestral points of origin on the part of expatriates that calls forth a willingness, even after decades abroad, to donate to homeland projects. It is this understanding that has been so important for local leadership in production units in the Pearl River delta to encourage the involvement of its expatriates in local economic development in the 1980s.

The great majority of households in Duanfen have relatives overseas. They are described as "Overseas Chinese dependents" (qiaojuan ). A small proportion of households do not have kinsmen abroad. Among the qiaojuan households there are those whose kinsmen abroad are distant and others who have close kinsmen with a direct relationship, either through marriage or descent, such as husbands, wives, parents, or children. The closeness of relationships will determine the amount of support that relatives abroad will provide and, in the modern period, the likelihood of obtaining an immigration visa and the possibility of joining kinsmen abroad. In the western delta in the 1980s, family and kinship loyalty has resulted in the creation of distinct household strategies among villagers with extensive overseas link-

[56] I have benefited from numerous conversations with Dr. Woon Yuen-fong of the University of Victoria about the western delta region. She has generously shared her findings from Chikan in neighboring Kaiping xian . See her "Social Change and Continuity in South China: Overseas Chinese and the Guan Lineage of Kaiping County, 1949-1987," China Quarterly , no. 118 (June 1989): 324-44; and her "International Links and Socio-economic Development of Modern China: An Emigrant Community in Guangdong," Modern China 16, no. 2 (1990): 139-72; "From Mao to Deng: Life Satisfaction Among Rural Women in an Emigrant Community in South China," Australian Journal of Chinese Affairs 25 (January 1991): 139-69.

ages. It has also resulted in the extensive revival of lineages, which had seemed forever compromised by land reform, collective forms of economic management, and firmly held ideological positions.

In the western delta it is important to distinguish the proportion of households with relatives abroad and, among these, whether their relatives are directly or more distantly related. It is possible to suggest that there are different strategies employed by households in the region, which are determined in large measure by the extent of their links with kinsmen abroad.

The consequences of extensive out-migration provide a possible explanation for the large (17 percent) proportion of "isolate" family types in the Duanfen village complex of Kongluen. Many are older women who have not yet left to join family in North America or have decided to stay. There is also a disproportionately large number of household heads who are widows (38.5 percent).^[57] The high proportion of widowed household heads is a possible consequence of the high rates of out-migration to overseas settlements. Some men left and never returned or broke contact with their homeland villages. Some may have remarried abroad and established separate families. Contacts between the homeland and the points of migration were frequently disrupted and were severed for almost a decade during the Japanese occupation. Mortality was also very high in Taishan during the Japanese occupation.

Delayed household division (and the out-migration of recently divided families) results in a very small proportion of nuclear households (31 percent). There is a correspondingly large number of stem families (46 percent). Data on responsibilities of household management reflect some of the distinctive features of the consequences of extensive out-migration. There are more heads of households responsible for economic management in Kongluen than in any other village (69.8 percent) and, correspondingly, the lowest incidence of spouses taking on these responsibilities (4.6 percent). This can be partly explained by the large number of isolates and widowed household heads. Equally, household management by other than the household head (or spouse) (26.9 percent) is matched only in Ngawu, where the incidence of complex household forms, as in Kongluen, is high. In both, the numbers of Overseas Chinese households are large.

Among Pearl River delta villages, remittances have a marked significance for household income only in the Overseas Chinese homeland. This is true among households in Kongluen, where remittances are still the major determinant of income disparities. Household income per capita is

[57] For the total sample, household heads who were widowers or widows constituted 19.0 percent. The proportions in the other sample villages are as follows: Naamshui, 13.8 percent; Tsimgong, 18.2 percent; Ngawu, 9.1 percent and Wantong, 16.2 percent. The number of widowers was less than 10 percent of the total number of widowers and widows.

highest among those with relatives abroad and lowest among those who do not have relatives abroad.^[58]

Non-qiaojuan households have the fewest resources and must rely therefore on their own abilities. As a result, these households are more deeply involved in wage labor than are qiaojuan households. Their male members, for example, are the backbone personnel of the construction teams that have become so important for Taishan (and neighboring Kaiping and Enping) in meeting the labor needs of the Pearl River delta construction boom. Such activities take workers away from the villages for long periods, and they may return only infrequently but, as in Ngawu, typically for the busy season of transplanting and harvesting in midsummer. Their wives tend to bear a disproportionate share of agricultural tasks. Those who do remain in the villages and concentrate on agricultural production tend to contract larger amounts of land and to expand into lucrative sidelines, either within the agriculture sector or into transportation or commercial activities. Such households often become "specialist households" of one kind or another.

Distant qiaojuan households have remittances but have little or no opportunity to contemplate out-migration. Such households are therefore also deeply involved in local economic activity. As qiaojuan households they have some advantages when compared with the non-qiaojuan households. They have remittances that can become working capital. They have connections with relatives abroad who can become the source of capital inputs by either bringing in (tax free) or sending items that are unobtainable or in short supply, such as machinery, electric motors or generators, building materials, trucks, or motorcycles.

Changes in policy toward direct qiaojuan families have resulted in the return of confiscated property. In addition, since 1979 direct qiaojuan families have been leaving in large numbers as part of family reunification policies in such countries as Canada and the United States. As a consequence, there is a considerable amount of property standing empty, which is often managed by the remaining (distant) qiaojuan kinsmen. Such properties can be used as an economic resource. Thus, while non-qiaojuan families are typically engaged as wage laborers or on their own account in farm production, distant qiaojuan households are typically engaged in enterprises in the private sector or are in partnership with zhen -run enterprises. The access to capital by distant qiaojuan households has allowed them to contract formerly collectively managed facilities (such as orchards, general stores, or repair facilities) at the village level and therefore augment their participation in agriculture with private production activities in the tertiary sector.

[58] This was also the pattern of income distribution even before the reform in the rural economy got under way, despite the obstacles that sometimes appeared for those households receiving remittances. It appears, however, that income differences have narrowed in the 1980s.

Direct qiaojuan households were the most advantaged in the 1980s in terms of both remittances and their potential ability to leave. As direct qiaojuan they were most severely affected by ideological hostility and the lack of labor power before 1979. Often bitter memories of the past and lack of confidence about the future increase their desire to leave the homeland. This group of households expects to leave, and, despite enjoying all the advantages that their distant qiaojuan kinsmen have in the villages, they are less committed to long-term economic benefit in the homeland. Many, particularly the women, who were left behind at an earlier period, are older, and their only economic activities relate to food production for themselves and whatever dependents remain. Older men who may be waiting for immigration visas spend a great deal of time socializing with their friends in the teahouses. The major concern of younger members of such households is preparing for emigration, and they spend much of their time acquiring a foreign language, typically English, or some other useful and marketable skill. Their levels of education are often high.

The possibility of family reunification abroad and migration away from the homeland has some marked effects on the social structure of villages throughout those parts of the Pearl River delta where direct qiaojuan households are concentrated. In Taishan, as well as other parts of the siyi area of migration to North America,^[59] while the architecture stands as a testimony to foreign remittances in the past, many houses stand empty, some smaller villages are virtually deserted, and there is the constant movement of population away from the area. Over ten thousand people have left Duanfen since 1979, including at least three Party secretaries. There is a curious contradiction. On the one hand, some of the more prosperous residents abroad have returned to the homeland and donated to schools, hospitals, libraries, and other public monuments, often renewing their association with their native places after an interval of decades. On the other hand, there are households who are leaving to join their kinsmen in North America permanently, leaving behind splendid houses, familiarity, comfort, even a modicum of prestige, for an uncertain future in an alien environment.

In the late twentieth century the desire to go abroad is as intense as it was in the earlier part of the century. There is far more contact with a wider range of relatives living abroad, and a detailed knowledge of life outside China is certainly present. There are also visitors, with their obvious affluence and their ability to dispense largesse. There is also television with its seductive images. The direct qiaojuan households are awaiting their

[59] Cantonese: sze-yap , literally "four districts." The four contiguous and culturally distinct xian (Xinhui, Kaiping, Taishan, and Enping) that constitute the ancestral points of origin for the great majority of Chinese in the Americas.

chances to leave, knowing that there is a certain inevitability. The distant qiaojuan and non-qiaojuan households are anxious to marry their daughters to direct qiaojuan or, better, an actual Overseas Chinese. In this way, over time, the entire household can be sponsored and take up residence abroad.^[60]

Lineage Revival

A dominant and distinctive feature of villages throughout the Pearl River delta is that they are lineage villages. Chinese lineages are property-holding corporations that socially integrate groups of households whose male heads trace their descent to an apical ancestor. They may be simple, small, and of only shallow genealogical depth, occupying a hamlet or a village fragment. They can also be large and complex, have a history of many tens of generations, and have a multisettlement character. In Duanfen a third of the households are surnamed Mei. Their development before 1949 was not dramatically different from that of the Guan in Chikan in neighboring Kaiping, as is described by Woon.^[61] In the 1930s they struggled with the Jiang (Cantonese: Kong) who occupy eight adjacent villages (Kongluen) over the control of a major market. In the center of the main Jiang village stand three enormous ancestral halls. There are lesser halls and "study halls" (shushi ) in the satellite villages.

During land reform in Duanfen, as well as other parts of south China, lineage properties were redistributed. At certain points in the history of the People's Republic of China, lineage rituals were attacked as feudal remnants, genealogies were ideologically suspect and hidden away, and ancestral tablets were destroyed—if not in 1958, most certainly in 1966. Ancestral halls, the most imposing structures in many villages, became team headquarters, warehouses, or workshops, and in some instances they were pulled down. Their granite floors and lintels became raw materials to repair village pathways and sometimes to build piggeries. After land reform, the operation of the south China lineage was compromised and its property basis was dismembered. At the same time, collectivization and rural

[60] Some data on American Chinese who take brides from Taishan can be found in Wu Xingci and Li Zhen, "Gum San Haak in the 1980s: A Study of Chinese Who Return to Taishan County for Marriage," Amerasia 14, no. 2 (1988): 32-34. The article emphasizes the desire to leave and the high educational achievement of brides. It also indicates that the SES of the grooms is not high.

[61] In Chikan, the Guan lineages occupied forty natural villages and competed, before 1949, with the powerful Situ lineage to control Chikan zhen , the intermediate market town. See Woon Yuen-fong, Social Organization in South China, 1911-1949: The Case of the Kuan Lineage of K'ai-ping County (Ann Arbor: Center for Chinese Studies, 1984).

development policies kept intact groups of families who were socially integrated by common descent. Old relationships and old conflicts could readily operate even in a changed political environment.^[62]

Village architecture is a constant reminder of the principles of lineage organization, despite the efforts to dismember the lineage as an economic and ritual entity. The carefully crafted policies to destroy lineages in the Overseas Chinese areas (and elsewhere) reflected a sophisticated understanding of the operation of a Chinese lineage. In the revival of many aspects of lineage organization in the 1980s, in Duanfen and elsewhere, the impact of the foreign links are key.

Since 1950 the issue of how to attract investment by ethnic Chinese abroad has been an important and troublesome issue in the formulation of Overseas Chinese policy,^[63] and it readily fell victim to conflicting domestic political currents in the past. After 1979 investment was actively sought abroad, and the involvement of both Overseas Chinese and émigrés from Hong Kong and Macao in the new economic program were important. In the traditional areas of Overseas Chinese migration, the willingness of Overseas Chinese to commit substantial funds to public projects depended on the resolution of some outstanding grievances. The fate of their lineage organizations was one of these.

While the full elaboration of all aspects of the operation of lineages cannot be contemplated, many of them are once again practiced. Graves have been repaired, rituals are performed at the graves of apical ancestors, ancestral halls are being restored, ritual feasts occur in the halls once more, lineage libraries are being refurbished, and lineage officers have begun to act as agents for members of the lineage, similar to the way the administrators of lineage trusts intervened on behalf of members in the period before 1949.

The restoration work that occurs is made possible by donations from local and external contributors. The major donations are from abroad. Funds are remitted through a network that is worldwide but energized from Hong Kong. The network is maintained by a sophisticated and widespread communications system. From the western delta there flows a large stream of local publications that focus on the reliance of local areas on the contributions of kinsmen and fellow countrymen overseas in maintaining local

[62] Anita Chan, Richard Madsen, and Jonathan Unger, Chen Village (Berkeley: University of California Press, 1984), esp. 16-40; the point is also argued by Potter and Potter, China's Peasants , who state, "A core group of close patrilineal kinsmen has now replaced the collective as the major group larger than the household in the organization of agricultural production" (266). There is no evidence from their account that lineage revival has been carried to quite the same degree in Dongguan as in the western delta regions.

[63] Stephen Fitzgerald, China and the Overseas Chinese (Cambridge: Cambridge University Press, 1972), 121-26.

integrity. Local history and local tradition are emphasized, and while economic advances receive careful attention, they are seen to be in harmony with a local cultural base that is firmly rooted in kinship.

In the Overseas Chinese areas, schools and education funded by the Overseas Chinese speak of upward mobility and lineage integrity. Before 1949 in Taishan, class and kinship were linked to overseas connections. In the thirty years after land reform, those connections were often a liability, although they continued to be of importance. In the past ten years, the liabilities have become advantages and have contributed to a developmental process that reflects national policies but, as a consequence of intense and distinctive patterns of overseas linkages, has developed some differentiating features of its own.

Conclusion

I have outlined in this chapter some of the features of Pearl River delta households and kin groups in the contemporary period. The delta has become caught up in a major rural transformation in the wake of reform efforts that began in the late 1970s. This process of economic change has proceeded further and faster than in any other region in the People's Republic of China.

Part of the explanation for the dramatic economic developments in the delta is its proximity to Hong Kong, which experienced its own transformation beginning in the 1950s. Hong Kong shifted from an economy based on entrepôt trade to become a major manufacturing center and a crucial part of the world system of banking and trade. A population largely of Pearl River delta émigrés, in concert with others from the Lower Yangzi and eastern Guangdong, led in establishing the links with the global economy. When reform initiatives were attempted in China after 1979, the entrepreneurial energies of Hong Kong were actively sought. Guangdong province was given distinctive authority to establish linkages with the world-system. These were developed through Hong Kong. By the late 1980s, Guangdong, and the Pearl River delta in particular, had become firmly incorporated into the world-system.

The consequences of that incorporation for Pearl River delta households and kin groups has been the focus of this chapter. Long an area of intense commercialization, the rural economy of the Pearl River became constrained by the assumptions of development policy for much of the three decades after 1949. The cellular character of rural China was reinforced by a high degree of bureaucratic control. The strategies of rural households in the Pearl River delta until the reform period were compromised by administrative practices that linked them firmly to their villages and to agricultural pursuits. Certain forms of kinship organization, notably lineages,

and their ritual expression, were targets of official hostility. There appeared to be a commonality of structural form for peasant households across the Pearl River delta after land reform, despite some variation in local economic possibilities.

I have examined the effects of recent economic change for rural households in five villages across the delta. I have suggested that the ability of local administrative units to adapt to policy changes has increased economic opportunities for households. The ability to establish linkages with kinsmen abroad has been especially critical in generating local-level economic change and moving a village economy away from a reliance on subsistence agriculture. In those villages where economic changes have been widespread, peasant households have taken full advantage of the local transformation and have allocated household labor among the variety of new opportunities that are contained within the reformed village economy. In the Dongguan village of Wantong, where extensive incorporation into the global economy through the entrepreneurial activities of Hong Kong—based expatriates has been particularly marked, household economic strategy has been characterized by an ability to allocate household labor in a wide array of economic activities. In the Luogang village of Tsimkong, where there are few overseas connections, the impact of global incorporation has been slight. Economic growth has been substantial as a consequence of reform policies. An intensification of the village's commercialized agriculture has seen households committing their energies and maximizing returns by expanding the cultivation and sale of fruit and vegetables. The sources of economic growth in the two villages differ. In Wantong, joint economic enterprise between the village and Hong Kong—based entrepreneurs has added to the economic inventory. In Tsimkong, existing economic capabilites have been intensified, and its increased production meets the needs of the domestic market. In both instances, however, the consequence of rapid local economic growth for household structure has been substantial and has contributed to increased household division, which explains the large number of nuclear households. The issue here is one of a flexible response to economic opportunity, which was articulated by Cohen and specifically addressed by Harrell in this volume. Structural change in a village economy may have important consequences, not merely for the productive role of women but also for their position as household managers. The firm incorporation of Wantong into the global economy has augmented the role of women in household structure. In Tsimkong, where growth has not been a consequence of structural change, the dominance of male household heads as household managers has been maintained.

Where local structural change and economic growth have been less marked, the options for peasant households and the adaptive strategies that they have adopted have been fewer and different. In such contexts, village

economies have been unable to absorb the labor that newer production arrangements in agriculture have released. In the rapidly growing economy of the Pearl River delta in the 1980s, economic opportunities have generally increased. Peasant households in villages in which structural change and economic growth have been modest have deployed their younger male members outside the village to take advantage of economic opportunities that the home villages cannot supply. The degree of change in household structure in such contexts is lessened. Household characteristics such as the incidence of household division or the allocation of responsibility for family management are determined less by change within the village economy than by the inability of the village economy to generate change and sufficient growth to absorb the labor surplus.

In the areas where Overseas Chinese ties are widespread, there are distinctive external linkages. Extensive out-migration has had profound structural consequences for peasant households. Linkages overseas only rarely translate into local economic transformation. Overseas Chinese have close and intense relationships to their ancestral homeland. Gifts by Overseas Chinese to their native places in the 1980s have contributed significantly to the infrastructure but not to major structural changes in the local economy. In contrast to the past, entrepreneurial activities by Overseas Chinese in the homeland are not extensive. The economic impact of Overseas Chinese monetary transfers is indirect. The impact of remittances for households in an Overseas Chinese area is, however, very great. Household strategies among peasant households in Overseas Chinese areas appear to be determined by the presence or absence of overseas linkages and the closeness of kinship to overseas relatives, when they are present. The absence of men has major effects for household form, household division, and household management. The very considerable sentiment for the homeland that exists among Overseas Chinese has led to regeneration of lineages to a degree unparalleled in those parts of the delta where the Overseas Chinese presence is less marked.

Chinese rural households in the Pearl River delta have been caught up in a change of great proportions during the 1980s. The change has been uneven in its effects across the region as a consequence of varying local characteristics, one of the most important of which has been the nature and intensity of relations with their kinsmen abroad. There are many continuities with the past. There is also a striking capacity on the part of peasant households in the Pearl River delta to respond creatively to opportunities and formulate strategies to maximize their possibilities. Households in the villages I surveyed are aware of the implications of the reform period for their well-being and have taken steps to deal with them. Some have been bold and innovative in their response. The local contexts have determined what the nature of their response can be. For delta households the past ten years

have seen their material standard of living improve significantly. They have benefited directly or indirectly from the incorporation of the broad region into the global economy. They do not appear at this time to have been overwhelmed by the negative consequences of globalization, which has led to the undoing of peasantries in some other contexts. Their corporate kinship structures with their instrinsic flexibility have buffered the impact of change. They have been able to hold the world-system at arms length to an important degree and to protect themselves from some of its most pernicious effects.

PART TWO
MARRIAGE

Six
Family Strategies and Structures in Rural North China

Mark Selden

Many of the ideas explored here are the product of collaborative research with Edward Friedman, Paul Pickowicz, and Kay Johnson. For background on Wugong and rural Hebei, see Friedman, Pickowicz, and Selden (1991). I am indebted to Richard Ratcliff, who introduced me to computer design and helped conceptualize computer elements of the survey, as well as to Zhu Hong and especially Shih Miin-wen for research assistance in the design and implementation of computer manipulation of the data. Cheng Tiejun, a sociologist, and a native of Raoyang county, who lived and worked in Wugong for several years in the 1970s, has been a source of knowledge, sources, and insight into the issues explored here. Deborah Davis and Stevan Harrell provided insightful critique of successive drafts, as did Edward Friedman, Gail Arrigo, Joan Smith, and an anonymous reader of the penultimate draft.

This chapter explores familial strategies and structures in response to socioeconomic and political changes over two generations in a peripheral region of rural North China. The primary data are household surveys conducted in Wugong village, Hebei, in 1978, at the end of the era of mobilizational collectivism; in 1984, two years after the reform agenda began to transform the village collective structure; and in 1987, when the household contract system had taken root. The surveys, which shed light on familial, demographic, and marital patterns from the 1920s to the 1980s, are supplemented by interviews conducted in the course of more than a decade of fieldwork in this and other villages in Raoyang county and throughout central Hebei. The focus of the study is on familial responses to macro political and economic change over the last half century, notably to the antimarket collectivism and the hukou population control system implemented by a powerful party-state from the 1950s forward, and the contractual and market-oriented reforms and birth control measures of the 1980s.

This study advances and assesses three hypotheses concerning changing familial strategies and structures in rural China in the era of mobilizational

collectivism (the 1950s through the 1970s) and contractual and market-oriented reforms (since the 1980s):^[1]

1. Collectivization, by weakening the household as a productive unit, undermined the economic rationale for the extended family. The result was a sharp reduction in average household size and in the number of stem and joint families as smaller nuclear families became the norm in the years 1955-1980.

2. Collectivization, the suppression of private markets, and restrictions on population mobility from the early 1950s through the 1970s, reinforced village involution in two distinct but intertwined senses. First, the ensemble of policies associated with mobilizational collectivism reversed the historical tendency of progressive expansion of the economic and social world of the peasant from the village to the standard marketing community and beyond. Beginning in the 1950s, the world of rural residents contracted back toward the natural village (see Skinner 1964-65). Second, with extravillage income-earning opportunities reduced or eliminated, with rapid population growth continuing through the 1970s, with expansion in the number of labor days per capita, and with collectives absorbing virtually unlimited supplies of labor, labor input increased substantially while marginal productivity of labor declined (see Geertz 1963; Huang 1985, 7-14; Huang 1990, 11-18; Chayanov 1986). This chapter explores some of the manifestations of village involution at the level of the family, particularly the pronounced tendency toward intravillage marriage, shrinking familial size, and simplification of structure associated with mobilizational collectivism in the years 1955-1980.

3. The resurgence of the household economy and the market in the 1980s reversed a number of trends of the collective era, giving rise to a bimodal pattern in household structures. On the one hand, rural household division accelerated as contracts took effect and the central economic role of households was restored. On the other hand, we observe signs of a resurgence of extended families, notably among entrepreneurial households emphasizing nonagricultural activities, capital accumulation, and expanded access to labor power, in ways that invite comparison with pre-land-reform patterns including the relationship between social class and family size and structure. Both cases suggest the growing autonomy and economic role of the household. At the same time, however, the continued strength and penetration of the state is revealed in the enforcement of the one-child (later the tacit two-child) family planning policy.

These theses are examined in relation to Wugong village, with comparisons to patterns observed in other rural regions and communities.

[1] My approach to mobilizational collectivism and the political economy of reform is spelled out in Selden 1992.

Wugong Village

Wugong, a village founded more than thirteen hundred years ago during the Sui dynasty, is located in Raoyang county in south central Hebei, one of the poorest areas of the North China Plain. Although the village is just 120 miles south of Beijing and Tianjin, throughout the twentieth century Raoyang has remained a poor and peripheral region, whose saline and alkaline soil and dearth of above-ground water sources together with periodic eruptions of drought and flood have proved inhospitable to agriculture. Equally important, its primitive transportation and communications—no railroad, poor highways, and lack of waterways^[2] —left the region beyond the pale of dynamic metropolitan centers of industry, commerce, and modern thought as they experienced capitalist development and incorporation in the global economy in the first half of the twentieth century.

In the 1940s the region was incorporated in the networks of power that eventually brought the Communist Party to power. Its very peripheral character proved advantageous in insulating the locality from Japanese control and forging bonds of nation and community that linked Wugong and Raoyang to the Party and the army. Raoyang was located in the Central Hebei base area, one of the few plains regions to sustain organized resistance throughout the anti-Japanese war. Beginning in 1943 with a four-household land-pooling group, Wugong villagers embarked on a path of cooperative transformation that would eventually transform a marginal and poor community into a thriving provincial and even a national model of cooperation, before receding into anonymity with the resurgence of the household economy and market forces in the 1980s (Friedman, Pickowicz, and Selden 1991).

In 1978, when research began, Wugong was a large and relatively prosperous wheat- and cotton-growing village with a registered population of 2,552. With per capita distributed collective income of 190 yuan, it was more than 50 percent richer than the second most prosperous of the ten villages that constituted the Wugong Commune, and three times richer than the poorest village. It was also approximately 50 percent higher than China's average 1978 rural per capita income of 133 yuan (Changes and Development in China , 242). Team Three, the largest and most prosperous of three teams and the focus of our household surveys, had a 1978 population

[2] Until 1954 the Hutuo River was Raoyang's lifeline to Tianjin for several months each year. Damming the river eliminated this connection to the major commercial and industrial center in the region. The state's attack on the market and its controls on population movement reinforced the county's isolation. Raoyang was a major casualty of the general prioritizing of irrigation over transportation and of self-reliant production over market ties that characterized national development from the 1950s through the 1970s. A new Beijing-Guangzhou railroad, in the initial stages of construction, is scheduled to pass through Wugong.

of 232 households and 996 people: 514 females, 480 males, and 2 whose gender was not recorded. This was comparable in size to many brigades in rural China, the size being emblematic of the high levels of collectivization ("advanced socialist production relations") maintained since the 1950s. In 1984, as household- and market-oriented reforms were beginning to take root in a region and a village whose leaders long resisted reform, the registered population of the team was 244 households with 939 people: 480 females and 459 males.^[3] In 1987 its population had dropped to 845 people in 250 households.

As the village that boasted the oldest continuous cooperative in Hebei, the home of national model peasant Geng Changsuo, and, beginning in 1953, the site of a provincial tractor station, Wugong attained national prominence as a large-scale mechanized cooperative. It sustained its status and special access to state resources through the twists and turns of national policy until the early 1980s. In the late 1970s and early 1980s, as national policies changed, Hebei province remained a bastion of opposition to contractual and market-oriented reforms. Wugong too stood in the forefront of defenders of the collective road that had brought fame and fortune to a once peripheral community. Thus it was not until 1982, following the ouster of its provincial patrons, that Wugong adopted the contract system, and even then the collective continued to play a prominent role in both agricultural and sideline production. Like many other model villages, such as Dazhai and Long Bow, just across the border in Shanxi, Wugong was the last village in its commune to ratify fundamental changes in the collective regimen (Hinton 1990, 31-47; 124-39). Our data capture familial dynamics spanning the decades prior to land reform, the collective era, and the early years of the contractual reforms of the 1980s.

Changing Patterns in Family Size and Structure

Land reform, collectivization, and market closure led to village involution and to important changes in household size and structure. The most important of these changes were the reduction in average household size and

[3] The 1978 officially registered Team Three population was 232 households with 1,053 people. This number included, however, 53 women who had left the team, in most cases by marriage, and in many instances 30-40 years earlier. Our data indicate that they neither contributed financially nor had husbands or children residing in Team Three. We have excluded such individuals from our sample, while retaining others who lived and worked or studied outside the village. These include spouses who work outside and provide financial support to family members in the village, and youth who were studying or serving in the military. All of these maintain important local ties and most are expected to return to the village. In 1978 there were 49 individuals working outside the village, 32 holding jobs in factories, as teachers or cadres, 11 in the army, and 6 studying in universities or technical schools. All of these are included in our Team Three sample.

number of generations, and the rise to dominance of smaller nuclear families. Data on other communities and regions suggest that many of the demographic changes described below for Wugong reflect broad tendencies found in villages across North China. Our discussion differentiates outcomes specific to a model village from more representative patterns.

Rural surveys of four Hebei villages in the vicinity of Beijing compiled by Sidney Gamble, a large North China sample of 17,581 households studied by John Buck, and Chi-ming Chiao's North China sample of 28,738 households, provide baseline data on household size and structure in rural North China (table 6.1). Gamble found average household size to be 5.1 members, Buck 5.3 members, and Chiao 5.5 members in the 1930s. In Gamble's 1933 sample, 26 percent of households had 7 or more members and 9 percent had 10 or more members (Gamble 1954, 26-27). Twenty-

seven percent of Buck's North China households had 7 or more members, including 8 percent with more than 10 members (Buck 1937, 368). Chiao found that 29 percent of households had 7 or more members, including 10 percent with 10 or more members (Chiao 1933, 17).

In the 1930s, household size correlated closely with landownership and wealth. The large household was economically advantaged, and it was only the prosperous that could sustain large households. Greater size facilitated division of labor, diversification, and accumulation. In his 1927 survey of 400 Dingxian farm families, Sidney Gamble found that those with farms of less than 10 mu had an average of 3.8 members, those with 21-30 mu averaged 6.2, and families with over 100 mu averaged 13.5 members. Per capita landownership and wealth increased with size of farm. For example, there was 1.8 mu of land per person in farms of less than 10 mu, 4.1 mu per person in farms of 21-30 mu, and 9.1 mu per person in farms of over 100 mu (Gamble 1954, 84). In his 1929-33 national survey of 16,786 farms, Buck found that in the winter-wheat gaoliang areas that included Hebei province, the small farm had an average of 4.7 household members, whereas medium-sized farms had 6.0 members, large farms 8.5 members, and very large farms 10.7 members (Buck 1937, 278).

Comparing Wugong's Team Three in 1978 (table 6.2) with North China in the 1930s, we note the much smaller average household size, just 4.3 members, compared with approximately 5.3 members in the earlier period. By 1978 the number of large households with 7 or more members was small despite the fact that a prosperous village like Wugong could sustain relatively large households compared with poorer communities. In 1978, 11 percent of Team Three households had 7 or more members, the largest being 9 members. By contrast, 26-29 percent of North China households in the 1930s samples had 7 or more members, including nearly 10 percent with 10 or more members. By 1978 household size no longer correlated with wealth measured by per capita household income. Size was above all the product of life-cycle timing. Neither labor power, land, capital, nor special skills were any longer decisive factors in determining household opportunity and income. This meant that there was a close relationship between the ratio of labor power to dependency and per capita household income, but virtually no correlation between household size and per capita income. Households with large numbers of young children had low per capita incomes, but as these children moved into the labor force per capita incomes rose. With marriage and household division, the cycle repeated. We hypothesize that this pattern was widespread throughout rural China in the era of mobilizational collectivism.

In 1978, 25 of Team Three's 232 households boasted three generations, and 2 others had four generations. The 5 largest households had nine members, and just 6 others had eight. The four-generation households were

those of the deputy Party secretary of the commune and the leader of Team Three. These four-generation households reveal leading Party families living out a contemporary version of the big-family Confucian ideal.

The dominant trend from the 1950s forward, consistent with the logic of collectivization, was from larger to smaller and from complex to nuclear families. The social and economic logic favoring stem and joint families was weakened in a collective milieu. Collectivization reduced the power of the family head over sons once control over land and labor passed from the family to the collective. Stated differently, after 1955 the fortunes of adult males no longer rested as heavily on the goodwill of a father who could assure their incomes and their future. Moreover, the advantages of the large household for achieving prosperity, by facilitating both accumulation and a complex household division of labor, were irrelevant in the collective era. Young men working in the collective earned as much as their fathers through work-point allocations from their teenage years, and there were few outlets for household investment. In short, the economic logic of household accumulation and the extended family was irrelevant in the era of collective agriculture.

To be sure, the collective paid income earned by all members in a lump sum directly to the household head, not to individual earners, perpetuating certain forms of intrahousehold dependency on the family head.^[4] The household also remained the primary consumption unit. But sons no longer depended on fathers to provide a share of household land and other property or to organize the household division of labor. Dependence on parents to assure a marriage and provide a home was also reduced if not eliminated in the collective era. In addition, a portion of welfare for the elderly and ill passed from household to collective responsibility, thereby further weakening intergenerational bonds. All these factors strengthened tendencies toward smaller and less complex households, toward reduced intergenerational dependence, and toward earlier household division. This was particularly true in a model village where household production for the market was tightly restricted and the range of collective economic and cultural activities was broad. The fact that model Wugong achieved notable success in birth control by the early 1970s, a decade before most rural communities, strengthened this tendency to reduce household size and complexity.^[5]

How widespread was the tendency toward smaller and less complex household structures in rural China in the collective era? Fruitful comparisons in the changing patterns of household size and structure can be made between Wugong in the 1970s and Parish and Whyte's 1973 findings for Guangdong. First, no significant change in household size occurred in Guangdong as a result of collectivization or other factors. Average household size in South China in 1930 was 5.0 members compared with 4.8 members in Guangdong in 1973. By comparison, we have hypothesized that significant reduction in household size in Wugong by the 1970s was a product both of collectivization and of unusual rigor in implementing birth control guidelines in the model village. Why the difference? One important reason is that birth control was not widely propagated or practiced in Guangdong and in much of rural China at the time of Parish and Whyte's 1973 investigation. This situation would change dramatically across the countryside from the late 1970s, bringing north and south into close alignment with respect to fertility. What had changed in Guangdong by 1973 was the fact that both the nuclear family (50%) and the single-person family (12%) had increased sharply in number, improved health care and nutrition reduced mortality, and the number of stem and joint families

[4] A fruitful area for future research is control of the family purse along both gender and generational lines, including savings to assure proper marriages and funerals.

[5] The population of Wugong village increased from 1,700 in 1955 to 2,413 in 1969, a 42 percent increase, or 2.8 percent per year. Between 1970 and 1975 population increased at a rate slightly less than 1 percent per year, from 2,426 to a peak of 2,567, before declining slowly but steadily to 2,533 in 1979. Wugong's 1984 population of 2,552 remained below the 1975 peak level, and population has subsequently remained stable.

plummeted. Compared with 63 percent in the 1930 South China sample, in 1973 in Guangdong only 39 percent were stem and joint families (Parish and Whyte 1978, 132-36).^[6] As in Wugong, we hypothesize that the shift to collective agriculture and the declining role of the family as an economic unit was the primary reason for the change in family structure leading to earlier division and reduction in the number of stem and joint families. At the same time, household size did not change significantly, because of continued high birth rates together with mortality rates substantially lower than those prevailing in the 1930s.

Between 1978 and 1987 significant changes occurred in Team Three household size and composition as household division accelerated following the implementation of household contracts. The 1984 survey gives years of household division for 83 of 244 Team Three households. Forty-four households divided in the five years 1979-83. By comparison, just 10 of the surveyed households indicated that they had divided in the decade 1966-75. The high rate of household division in the years 1979-83 constituted a response at the household level to the preparations for, and then the actual implementation of, contractual and market-oriented reforms.

Changes in household size and composition in the wake of partial decollectivization were particularly visible and significant among the elderly. Between 1978 and 1984, the average Team Three household size fell from 4.3 to 3.8 persons, and the number of two-person households increased from 9 to 32. Among two-person households, couples aged sixty-two and older increased from four to nineteen. By the mid-1980s, it was common for both spouses to remain alive, active, and self-sufficient long after their fifties, the time when most households customarily divided.

In 1978 fifteen people over age sixty-five lived alone, often in close proximity to sons or other family support networks. These included three women, ages seventy-nine, eighty, and eighty-three, two of whom continued to earn significant work-points permitting self-support with dignity. Their living alone was nevertheless a sensitive issue. Questionnaires repeatedly recorded the unsolicited information that sons performed household tasks and otherwise supported aged parents who lived alone. Responsibility for the aged remained a family affair. But whatever the assistance rendered, if we look at the utterly bare hovels in which some elderly people resided alone, and observe that such people frequently lacked even tea leaves to add to boiling water when serving a guest, it is difficult to escape the conclusion that in rural China as elsewhere the transition to the nuclear family imposes a heavy price on the rural elderly. By 1984 the number of single-person

[6] Sulamith Potter and Jack Potter, China's Peasants: The Anthropology of a Revolution , 217, found a smaller percentage of Zengbu brigade (Guangdong) households living in stem families in 1980, 25 percent. The joint family form did not exist in this community.

households of those sixty-two or older had dropped to nine, including one seventy-nine-year-old woman and an eighty-four-year-old man.

Between 1978 and 1984 the number of three-person households virtually doubled while the number of five- and six-person households was halved and those with seven to ten members dropped from 26 to 19. The great majority of household divisions in these years took place within two years of marriage, most often at the time of birth of the first (and in these years only) child. Earlier and more frequent household division, together with effective birth control, gave rise in the 1980s to the new archetypal three-person household composed of parents and one child.^[7]

These tendencies toward small, simpler, and more homogeneous households were carried further in the years 1984-87. By 1987 there were only two single-person households, and the largest household in the team had only seven members. Between 1984 and 1987 the number of households with seven to ten members dropped from seventeen to just one, and the size of the largest household fell from ten to seven. Household division occurred throughout the 1980s more frequently and earlier than in the 1930s or in the collective era. We see this clearly in a pattern that had no earlier twentieth-century precedent in Wugong or, to my knowledge, in most of rural China. Of the sixty-eight two-person households in 1987, twenty-three consisted of childless couples who married in the years 1980-86. This constitutes a break in two important ways. First, prior to the 1980s, it was extremely rare to establish one's own household before the birth of the first child, and frequently it was only after the birth of a second child, if at all, that the household divided. Our 1978 survey indicated no household composed of a recently married childless couple. Two such households existed by 1984. By 1987, twenty-three childless couples who had married within the previous five years had established independent households.

Prior to land reform, the economic logic of the household economy prevented early household division. At the heart of the family compact was the exchange between the care of aged parents by male offspring and the eventual transfer of land.^[8] Collectivization eliminated both the element of land transfer and the household as the organizer of productive activity. The

[7] A return visit in 1991 revealed that Wugong remained a leader in birth control, but with the new relaxed guidelines, more households took the opportunity to have a second child following the birth of a daughter. A second child was permitted four years after the birth of a daughter (or the death of a first child) on condition that one parent accept sterilization. In all but one known case, it was the wife who accepted sterilization.

[8] This formulation of a compact excludes daughters. I tentatively suggest that a weaker and more ephemeral but not insignificant compact rested on parental obligations to nurture and to arrange a suitable marriage for a daughter in exchange for services rendered prior to marriage, after which the allegiance and obligation of a daughter was expected to be to her husband and his family.

household contract system of the 1980s offered young men and women equal and immediate access to their own land—if they established independent households. Moreover, with the village providing land for home construction, growing numbers of young people chose to free themselves from the constraints of parental and mother-in-law authority as quickly as possible following marriage.^[9]

A second significant departure from earlier norms of both the republican era and the collective years is that by the late 1980s many young couples were not producing offspring in the first several years of marriage. By contrast, prior to the 1980s, virtually every couple produced a child within the first year or two of marriage. The changes noted here with respect to small nuclear families, autonomous households, and delayed reproduction bring the family in Wugong in the late 1980s closer to prevailing urban patterns in China, and indeed to norms prevailing in industrial societies elsewhere.

Important dimensions of the 1980s economic and social reforms impinge heavily on household size and structure: The household contract system weakened the former collective structure and strengthened the role of households as productive units both in agriculture and in sideline production and market activity. Between 1978 and 1984 the total number of those listed as working or studying outside the village or serving in the military dropped by nearly half to twenty-seven. Outside employees declined by one-fourth from thirty-two to twenty-four. The largest changes centered on outside study and the army. In 1984 no Team Three member was studying outside the village, and the number of those in the army had fallen from eleven to just three. The changes primarily reflect Wugong's loss of status as a model village coinciding with the opening of alternative channels of mobility associated with the private sector and the market, changes that Wugong people were relatively slow to act upon. The drop in the number of youth joining the army may also be a function of the declining prestige and shrinking size of the military in the 1980s.

The 1984 figures do not, however, suggest a trend toward village isolation, involution, and subsistence farming. Quite the opposite. In 1984, 107 households, nearly half of the households in Team Three, listed one or more members who were engaged in nonagricultural economic pursuits, including teachers, factory workers, and contractors in diverse industrial, craft, commercial, and handicraft activities. In 1978, when the collective

[9] Did accompanying policy changes in, for example, access to land and housing for young newlyweds occur at the village level that facilitated the greater autonomy of young people in the 1980s? This is a subject for future research. Here I wish to observe the limits of the present research, which moves between national policy and household behavior but is unable to clarify the important intermediate range of local-level policies, particularly those at the village and county levels, affecting family decisions. Joan Smith drew my attention to these issues.

still held sway, 50 households, just over one-fifth, listed such activities exclusive of unpaid team and brigade political positions. Compared with 1978, many more people engaged in craft and industrial production, provided services, and produced for and sold in the market. While many of these were listed as living and working in the village, in 1984 their work regularly took them outside for commercial and other economic activities. As in the first half of the twentieth century, the Wugong economy was more fully embedded in a wider economic world extending from the standard marketing community to other parts of North China, including Beijing, Tianjin, and beyond to Inner Mongolia and the Northwest.

By 1984, two years after the implementation of household contracts, we observe a bimodal pattern in household strategies. On the one hand, there was continued reduction in household size from 4.3 people in 1978 to 3.8 in 1984. This was the product of large numbers of household divisions in the early 1980s and the continued effects of birth control. Most people lived in nuclear families, and by 1984 the number of four-generation households had further declined from two to none. In 1978, four- and five-person households predominated, accounting for 48 percent of the 232 households; approximately equal numbers of three- and six-person households made up an additional 31 percent. By 1984 three- and four-person households had become the norm, accounting for 134 (55 percent) of the 244 households. The number of households with six or more members plummeted from 64 (28 percent) in 1978 to 32 (13 percent) in 1984. Likewise, the percentage of households with zero or one registered child increased from 45 to 77 percent while the percentage of households with three or more children dropped from 43 to just 16 percent. The number of single-person households fell from 15 (7 percent) to 12 (5 percent) while the number of twoperson households tripled from 9 (4 percent) to 32 (13 percent).

On the other hand, we observe apparently contradictory tendencies at work. Specifically, there was a resurgence of stem, joint, and multigenerational structures as a significant group of households reasserted themselves economically and socially and the reforms stimulated new and diverse individual and family strategies (table 6.3). The striking fact is that in a very brief time, and simultaneous with a sharp drop in numbers of births and in household size in response to the birth control program, the number of stem and joint families nearly doubled from 31 in 1978 to 53 in 1984. Among stem and joint families, the somewhat higher concentration of nonagricultural contracts and members in state-sector jobs as workers, cadres, or teachers underlines a degree of congruence between diversified cashearning economic activity and extended family arrangements. Twenty-seven out of 53 stem and joint families (51 percent) recorded the existence of nonagricultural contracts, extravillage income, or members earning state

salaries. The comparable figure for nuclear and one-person households was 80 out of 191 (42 percent).

Our findings concerning the resurgence of extended families in the early 1980s, consistent with the Davis-Harrell hypothesis in the introduction to this volume, run directly counter to the modernization hypothesis that leads one to anticipate that the growth of the market, commodification, and growing cash income would all contribute to strengthening the nuclear family (Goode 1982).^[10] Rather, with the opening of the household sector and the market, significant numbers of Wugong people, particularly those with an entrepreneurial bent, chose to reconstitute joint and stem families, reversing the pattern of their sharp decline following collectivization. In just six years, from 1978 to 1984, the number of stem and joint families increased from 31 to 53, even as the average size of family dropped from 4.3 to 3.8 persons. Team Three demographic experience in this respect lends weight to the suggestion by Davis and Harrell in this volume that for the 1980s and 1990s we replace the urban-rural dichotomy with a tripartite division between subsistence farmers, urban collective and state employees, and entrepreneurs in both city and countryside.

One interesting example of the trend toward extended families is provided by the sole Team Three family classified as "landlord." Li Maoxiu, the youngest son of a small landlord, was a resistance activist who rose to head the district children's organization in wartime while his brother joined

[10] William Parish and Martin Whyte, in Village and Family in Contemporary China , 137, observe that, contrary to another modernization hypothesis, in Guangdong in the 1970s, no correlation existed between higher education and incidence of nuclear families.

the Party and the army en route to a government career in Shandong. With his father dead and his brother away, Li Maoxiu was branded a landlord in the 1947 land reform. Isolated, stigmatized, and barred from joining a cooperative, in 1952 Li Maoxiu committed the crime of flight, seeking, with the aid of a seal carved out of a turnip, to begin a new life as a teacher in the Northeast. He was caught and jailed for five years, and following his return to the village lived the life of a pariah. The Lis' only child, a son, married in 1964 at age nineteen. During the Cultural Revolution, to insulate the younger generation from the harassment heaped upon "class enemies," the family divided. Fifteen years later, in the early 1980s, the landlord cap having been officially removed, not only did Li Maoxiu initiate several family enterprises, including long distance trade and the opening of a bakery and a plastic workshop, but he also reconstituted a three-generation stem family. In 1984 the family included his wife, his son and daughter-in-law, and their teenage daughter and two sons. In 1990 the family redivided. Li Maoxiu and his wife lived alone in their original, dilapidated home, a grim reminder of their days of opprobrium. Their son and his wife and children moved into a new home at the other end of Team Three.

The 1987 data indicate that some trends already evident in 1984 have gone further. Average size of household continued to drop from 3.8 members in 1984 to 3.4 in 1987. The drop was particularly precipitous among households with 6 or more members. By 1984 such households had already declined to just 13 percent of the total. By 1987 they had dropped further to just 5 percent, and the number of households with 7 or more members fell from nineteen to one.

At the same time, the hypothesized bifurcating pattern in which the majority form small nuclear households while an entrepreneurial minority favors stem and joint households, appears less pronounced in 1987 than in 1984. While the number of stem and joint families increased significantly between 1978 and 1984, the trend reversed between 1984 and 1987 as the number of extended households dropped from 53 to 46 and their size was reduced. Unfortunately, our 1987 data do not permit clear differentiation between entrepreneurial and nonentrepreneurial households. In any event, the hypothesis concerning correlation between multigenerational extended families and entrepreneurship will require testing against the performance of other and larger samples.

Intravillage Marriage

In the decades following collectivization, the shrinking of the world of the villager from the standard marketing community and beyond to the natural village was manifested in the closing off of extravillage economic, social, cultural, and marital ties. Most rural people found their world restricted to

the confines of the natural village. According to the 1978 survey, of the 985 Team Three residents whose birthplace is known, 771 were born in Wugong and 214 were born outside the village, nearly all of the latter being brides who married in. Famine, political turmoil, war, population pressures, and the hope for better opportunities had led millions of Chinese to flee ancestral villages in search of land, food, and work in the century prior to the founding of the People's Republic. Beginning in 1955, and with particular force after 1960, however, powerful institutional mechanisms associated with collectivization, market controls, and rigorous state enforcement of the hukou system of population control kept rural residents in place in the village of their birth or marriage. Village communities attained extraordinary levels of stability. Only a handful of households settled in Wugong after 1945, all of them prior to collectivization, and in every known case they had familial connections. Of 480 males registered in Team Three in 1978, only 11 were born outside the village, including 3 sons born in the Northeast where both parents were factory workers prior to their forced repatriation to Wugong. The repatriates were among the twenty million industrial workers and family members sent back to their villages of origin in 1962 following the failure of the Great Leap Forward. The combination of collectivization, market curbs, and household registration controls put a brake on the dynamic movement of the Chinese population both within China, where millions had migrated to the Northeast and other areas during the late Qing and the Republic, and abroad. This involutionary pattern too was a facet of the growing isolation of village communities associated with collectivization and population controls. In this respect Wugong's experience was representative of that of rural China.

One important change in marital practice in Wugong concerns intravillage marriage. There were deep structural reasons for the trend toward intravillage marriage following collectivization and market closure. Prior to the 1950s, the standard marketing community not only was the primary locus of trade and cultural activities but also defined the terrain within which most marriage contracts were negotiated. Team Three brides of the previous half century who are listed in the 1984 survey came from some forty villages in Raoyang county and sixteen villages in neighboring counties, nearly all within a ten-mile radius of the village. The combination of collectivization and market suppression substantially changed these patterns. With villagers prevented from buying and selling in the few remaining markets, with the closure of rural fairs and traditional cultural events after 1957, with the penetration of party-state networks deep into village life, and with attacks on old culture, customs, and ideas, including the purchase of brides, most people found their world restricted to the natural village. At the same time that extravillage social and economic ties were severed, with activities largely restricted within the team and brigade, the most valuable

alliances for families struggling to survive shifted to within the village. From work assignments to income, from opportunity for higher education, to a place in the army or the right to build a house, villagers depended heavily on the goodwill of team and brigade cadres who ruled the village. This was the socioeconomic and political basis for the rise in intravillage marriage in rural communities during the collective era, a pattern that spread throughout much of rural China in the 1970s. Village involution was reflected in marital patterns as more and more brides were from Wugong.

There were, in addition, special reasons why Wugong, in this as in many other areas, led the way. As Wugong reaped the fruits of its status as a model village, per capita income soared above the levels of neighboring communities. In the process, marital strategies changed and taboos restricting intravillage or same-surname marriage began to erode.^[11] Parents became reluctant to send daughters to live in much poorer villages that traditionally exchanged brides with Wugong. Moreover, with women sharing in the rapidly expanding educational opportunities that made primary school education the norm in Wugong in the 1950s, junior high education in the expanded village school the norm in the 1960s, and even enabled many to graduate from high school in the 1970s (in each instance Wugong was approximately a decade ahead of neighboring communities), young people in the model village had numerous opportunities to meet. Other villages shared in the expansion of education and accompanying social practices, but at a slower pace.

Participation in collective labor and group activities organized by the youth league, militia, and the county government provided important venues in which young people could meet and court. These opportunities were especially numerous in a model village whose youth enjoyed disproportionate opportunities to participate in state-sponsored cultural, social, and political activities. Growing numbers of young people took marital choice into their own hands, often enlisting the services of a matchmaker to legitimate and sanctify their choice after the fact.

There were limits to the related trends toward autonomous and intravillage marital arrangements throughout rural China, particularly constraints imposed by poverty. The poor were at times still forced to enter paired marriages. One young woman in a poor Raoyang village in the 1960s, for example, was forced to abandon her lover, an only child, to accept a blind-exchange marriage that simultaneously assured her brother a bride. Such exchange marriages were illegal, and the Women's Federation sporadically sought to eliminate them. Larger multiple marriages sometimes involved

[11] In Wugong and other multilineage Hebei villages, certain kinds of intravillage marriage, including same-surname marriage, had long been acceptable.

package arrangements for six, eight, or more young people. The attractiveness of paired and multiple marriages lay in the fact that they often eliminated or greatly reduced costly brideprice and dowry, thus making it possible for the poor to wed. While frowned on by the state, this remained a recourse for poor rural households, for some their only hope to arrange a marriage.

Our 1984 survey reveals significant changes in intravillage marriage patterns related not only to collectivization and market closure, but also to Wugong's prosperity and prominence vis-à-vis neighboring communities. The year 1970 marked the divide with respect to intravillage marriage, and this tendency grew more pronounced with the passage of time (table 6.4). While examples of Team Three intravillage marriage are found as early as the 1930s, just 31 (21 percent) of the 151 brides marrying prior to 1970 were born in Wugong. By contrast, 48 (45 percent) of the 88 brides who married between 1970 and 1984 were born in the village, and 26 of 38 brides (68 percent) who married in the years 1984 to 1986 were born in Wugong. The pattern of intravillage marriage grew stronger throughout the 1980s, despite the fact that villagers again participated in a wider regional economy.

Many other Chinese rural communities followed Wugong in the shift toward intravillage marriage in the collective era. The literature abounds with examples from South China. Parish and Whyte (1978, 171) found that intravillage marriage in Guangdong increased from 3 percent prior to 1949 to more than 20 percent in the 1950s through the 1970s. Potter and Potter (1990, 217), in their Zengbu brigade research (Guangdong), noted an increase in intrabrigade marriage from 10 percent of the total in 1964-68 to 21 percent in 1979-81. In this single-lineage community, intrabrigade marriage had to overcome the taboo against same-surname marriage, which was considered incestuous prior to the 1950s.

Chan, Madsen, and Unger describe a "marriage revolution" in Chen village, Guangdong, during the Great Leap, when several youths dared to defy the taboo on intralineage (and intravillage) marriage and took local

brides of the same surname. In the post-Leap famine, intravillage alliances became the marriage of choice, involving 70-80 percent of the total (Chan, Madsen, and Unger 1984, 188-91). Jan Myrdal provides an anecdotal example from the Northwest, noting that in the 1960s intravillage marriage was commonplace in Shaanxi's Liulin village (Myrdal 1965, 21).

In the 1980s, with the rise of the household contract system and the market, Wugong, slow to shift gears in step with the times, lost its model status. Within a few years neighboring communities rapidly overcame the income and productivity advantages that the model village had long enjoyed. The relative attractiveness of marriage into Wugong under these circumstances declined. Young people nevertheless continued their efforts to secure control over marriage and choose partners from their circle of local acquaintances.

New marital patterns in Wugong were not confined to intravillage marriage. Some Team Three marriages even began to take place within the narrow confines of the team. In 1978 a local schoolteacher and women's activist married the deputy battalion leader of the militia. Both were natives of Team Three. Some Wugong people privately worried over possible adverse genetic consequences of close inbreeding. Wugong had no historical taboo against intravillage marriage or same-surname marriage; indeed this multilineage village had a long history of such marriages. But the village was divided into four incest-taboo neighborhoods, and strict codes had long prohibited marriage within the neighborhood. Two of these areas eventually formed teams one and two, while the third and fourth, separated by a lane, combined to form Team Three.

Geneticists and demographers with whom I have consulted indicate that intermarriage within a community of 600 households (Wugong village), or even 300 households (Team Three), poses no serious threat of birth defects provided close relatives do not marry. Yet China has serious problems of birth defects, and a rash of articles in the Chinese media in the years 1988-90 warned darkly that rural inbreeding had reached serious proportions in the countryside. On December 6, 1988, and January 2, 1989, the People's Daily (overseas edition) cited surveys by sociologist Gao Caiqin to warn of the debilitating consequences of intravillage marriage. On the basis of a study of 1,441 peasant families in six provinces, Gao concluded that evercloser consanguinity in single-lineage villages was "causing population degeneration and undermining social development." Gao found that 30 percent were marrying within their native villages and 51 percent in the township. On January 30, 1989, People's Daily (overseas edition) reported the existence of more than thirty million Chinese with congenital defects and warned that an important source of the problem in remote mountain areas lay in inbreeding. The Impartial Daily , a Party organ published in Hong Kong, presented even more alarming figures in its edition of April 12,

1990. It claimed that China had some fifty million people who were mentally or physically handicapped and again pointed to intravillage and intratownship marriage as a cause. What none of these sources mentioned was the fact that China's hukou system of population control, together with the antimarket collectivism that prevailed until the late 1970s, had strengthened the tendency toward intravillage reproduction.

By early 1989 the state began to act on findings like those mentioned above. Chen Muhua, chair of the Women's Association, told a symposium that the cost of raising children with congenital defects, which many attributed to close kin marriage, was seven to eight billion yuan per year. Chen announced the promulgation of a eugenics law, "Regulations to Prevent Mental Defectives (chidai sharen ) from Bearing Children" in Gansu province (People's Daily , overseas edition, January 30, 1989). The method: forced sterilization. The May 21, 1990, Taiwan newspaper World (Shijie) cited a People's Daily report that Gansu had already sterilized 5,500 of the province's 260,000 mental defectives. Officials stated that most of the rest would be sterilized within the year, and it seems likely that the Gansu sterilization program is a pilot project preparatory to national implementation. It is difficult to imagine a state program more susceptible to political manipulation and human tragedy. What criteria for "mental defectives" will be applied, and by whom?

Neither the broad marital policies of the state nor youth initiatives to take control of marriage challenged the patrilineal, patrilocal marital traditions that were particularly entrenched in North China. (Uxorilocal marriage had long been acceptable for families without sons in the southeast, east central, and southwest but not in the north; Wolf and Huang 1980, 11-15; 94-107; Goody 1990, 106-7.) Virtually the sole state initiative calling patrilineal, patrilocal practices into question in the People's Republic was the 1974-75 campaign pressed by Jiang Qing, with Xiaojinzhuang village in suburban Tianjin as its model. The campaign encouraged intravillage marriage and marriage into the family of a daughter who was an only child. One young man from neighboring Shen county did marry into Wugong at that time, enabling an only child to remain at home to care for her aged parents. The husband benefited by the move to a prosperous model village, doubtless a step up the economic ladder, but entailing some loss of face. With the end of the campaign, the issues disappeared from the political agenda.

The resurgence of the household economy and the market in the 1980s, and the weakening of restrictions on population movement, should reinforce trends toward extravillage marriage as the social and economic world again opens to marketing communities and beyond. Whether these effects will overcome other factors conducive to intravillage marriage, particularly changing patterns of youth courtship, remains to be seen.

Marriage Age

As the place of origin of Wugong brides changed, so too did age of first marriage. The earliest recorded age of marriage in our 1978 survey is that of an eighty-two-year-old poor peasant who married his sixteen-year-old bride when he was fourteen; one woman then in her fifties also married a middle peasant when she was fourteen.

Gamble's study of Dingxian in the 1930s produced numerous examples of much earlier marriages, particularly for boys, than those recorded in Wugong at that time.^[12] In a sample of 5,255 rural families, 169 males and 33 females under age fifteen were married, including 21 boys age eleven and under. Girls married as young as age twelve. All but 5 women in Gamble's sample had married by age twenty-one, but as many as 10 percent of males had not married by age thirty-eight. In another study of 515 families, Gamble found the range of ages for first marriage for males to run from seven to fifty-one and for females twelve to thirty-eight (Gamble 1954, 37-41; 58-59). In Buck's 1929-31 North China rural sample of 1,600 married men and 1,760 married women, 58 percent of the men and 86 percent of the women married by age nineteen, including 12 percent of men and 13 percent of women who married prior to age fifteen (Buck 1937, 380). Comparable results were recorded in C. M. Chiao's 1929-31 survey of 12,456 farm families. He found that in North China 62 percent of men and 80 percent of women married before age nineteen (Chiao 1933, 28, 31).

Gamble's 1930 Dingxian sample of 515 families makes clear the high correlation between male marriage age and wealth, using landownership as a gauge for wealth. Among households with less than 50 mu of land, 33 percent of married males wed by age fourteen, and 11 percent first married between the ages of twenty-nine and thirty-five. By contrast, 81 percent of males in households with 100 mu or more of land married by age fourteen and all but five percent married by age seventeen (Gamble 1954, 59).

In another Dingxian sample of 5,225 families, Gamble found that more than 99 percent of females had married by age twenty-one. By contrast, just 60 percent of males married by age twenty-three, and it was not until age thirty-nine that 90 percent of males married (Gamble 1954, 39-40). The great majority of the unmarried 10 percent, virtually all of them males from poverty-stricken families, would never marry. They would face old age without familial support, and their family line would die out, thus violating the filial injunction to maintain the family line.

One measure of the degree of state penetration of village life during the collective era is the success of the campaign to postpone marriage for substantial numbers of rural youth, from the late teens to the early and even,

[12] In the 1930s Wugong was the extreme southwest part of Dingxian prefecture.

for a period, to the late twenties. By the 1960s, early and child marriages were largely eliminated and very late marriages (over age thirty) as well as the number of men who were never able to wed, were sharply reduced in most of rural China. From the 1950s through the mid-1970s, most Team Three men and women married shortly after age twenty as both earlier and later marriages declined in keeping with state directives and in response to changes associated with land reform and collectivization (table 6.5). The pattern shifted dramatically in the late 1970s. Marriage age for women and men increased on average from twenty-one in the late 1960s to twenty-five in the late 1970s. The explanation for the shifts in marriage age in the 1970s lies in the intense campaign to postpone marriage beyond age twenty-five, one of the pillars of China's birth control strategy at that time.

In the late 1970s campaign to delay marriage, as in so many others, Wugong more scrupulously enforced state goals than did most communities. For model communities, their "capital" was political capital. The advantages models enjoyed in higher income, access to state resources, and prestige were mediated through ties to the state, not the market, although the advantages of model status included higher incomes and privileged access to state resources. We hypothesize that the delay of marriage by five years to the late twenties was unusually pronounced in Wugong and other model villages, although the campaign had an impact far beyond such communities. After 1982 the state relaxed efforts to postpone marriage until the late twenties while continuing to control fertility. With partial decollectivization and the resurgence of the household economy, the state's ability to enforce marital postponement—and much else—generally weakened. The result in Wugong was that the norm shifted back to the early twenties, where it had been since the 1960s.

Older women and men in Team Three experienced a much broader range of ages at first marriage than did younger generations. Of 56 Team

Three women who married between 1920 and 1949 and were alive at the time of our 1984 survey, 20 married between age fifteen and nineteen, 26 between twenty and twenty-four, 9 between twenty-five and twenty-nine, and 1 in her early thirties. Out of 60 Team Three married men whose marriage between 1910 and 1949 is recorded in our 1984 survey, 1 married at age thirteen, 13 between age fifteen and nineteen, and 4 in their thirties. By contrast, every one of the 100 women and 100 men who married between 1970 and 1984 married between age twenty and twenty-nine. The data for the years 1970 to 1984 reveal with particular clarity the effectiveness of state pressures to defer marriage in a model village. In the years 1970-74, 15 men and 14 women married between ages twenty and twenty-four, and just 3 men and 4 women married between twenty-five and twenty-nine. But between 1975 and 1979, as the state called for delayed marriage to age twenty-eight for men and twenty-five for women, the pattern shifted sharply. In the years 1975 to 1979 just 5 men and 8 women married between ages twenty and twenty-four, while 25 men and 22 women married between twenty-five and twenty-nine. In the years 1980-84, state pressures for delaying marriage to the late twenties eased, and the focus of birth control shifted from delayed marriage to contraception and abortion. The pendulum then shifted back toward somewhat earlier marriage. Of the 38 men and 38 women who married between 1984 and 1986, just 4 men and 4 women did so at ages twenty-five to twenty-six while the remaining 34 men and 34 women wed between age twenty and twenty-four.

Our surveys of Team Three contain no examples of destitute males who never married. In Wugong village, from the 1940s, apparently every male could afford to marry, and all, even the physically handicapped, did so. It is not of course the case that following land reform and collectivization all Chinese rural males were able to marry. In the early 1960s, males in poorer households and destitute communities faced continued difficulty in securing a spouse. A vivid example is provided by Duankou village, a poor community a few miles north of Wugong that was devastated during the Great Leap. In the early 1970s just 17 out of 80 men age twenty-five to forty-five had been able to marry. In frustration, unmarried males, led by a former war hero, organized a gang that set out to prevent other village males from achieving the marriages they had been denied. Their method was to launch a rumor war, warning of dire consequences for families of prospective brides who were contemplating marrying into Duankou.

Throughout rural China, land reform and collectivization not only eliminated extremes in inequality of landholding and economic opportunity, but also initiated processes that compressed wide variations in marriage age. This pattern of homogenization in which early and late marriages were virtually eliminated, illustrated with particular clarity in model Wugong, may

be seen, if less sharply etched, throughout much of the countryside during the collective years.

The wide dispersion of marriage age, the inability of significant numbers of poorer men to marry at all, and substantial age differences between spouses, characteristic of rural China in the first half of the twentieth century, gave way in the decades following land reform and collectivization to the homogenization of marriage age in the early twenties. However, the hardcore poor, predominantly those living in chronic deficit areas, remained outside this process.

One other distinctive marital pattern in Hebei counties near Wugong bears mention. A popular local saying went: "A wife three years older is like a brick of gold; a wife five years older looks older than your mother." Sidney Gamble documented the predominant pattern of older wives for Dingxian in the 1930s. In a survey of 766 couples in 515 families, he found that wives were older in 70 percent of cases, husbands in 25 percent. In the case of older wives, the largest number were two to three years older, with a maximum difference of eleven years. Older husbands, however, averaged eight years older, up to a maximum of thirty-one years. Most of these cases involved poor males unable to marry until late in life (Gamble 1954, 45-46).

In Team Three, the greatest age disparity revealed in our 1978 survey was that of a male who took a bride eighteen years his junior in a 1932 union. Six other males, ranging in age from sixty-two to seventy-three, had wives ten to fifteen years younger. An additional thirty-one couples differed in age by five to nine years, with the male being older in twenty-six of these cases. In all but three of the forty-three instances in which the age difference of partners was five or more years, the male was above forty-five years of age, indicating that by the 1950s such marriages had become rare. Since the 1960s, all Team Three marriages have joined partners of roughly equal age; in no case did the difference exceed four years. The trend to marriages among people of comparable age coincided with homogenization in the age of first marriage.

Gamble (1954, 41-43) documented for Dingxian and rural Hebei in the 1930s a class-based pattern of differentiation in which the favored approach for the prosperous was to marry off a son at a very young age, averaging thirteen years for families owning more than 100 mu. Brides in such cases averaged 3.6 years older, to assure that they were ready both to bear children and to care for their in-laws and their child husbands. Wugong data likewise reveal class differentiation in marital patterns, although few men married as young as did Gamble's wealthy scions. The 1978 Team Three survey showed that among surviving couples, twenty-five men who married prior to 1955 had married women five to eighteen years younger. All but

three of these men were classified as poor peasants in land reform (the others were middle peasants). Many of them first married when they were in their thirties or forties, and in one case the groom was sixty years of age. Eleven women married men four to seven years younger. In these marriages, favored by the more prosperous, eight of the husbands' households were middle peasants, and one a landlord. Just two were poor peasants.

Our 1984 survey of Team Three showed that the practice of taking a bride ten or more years younger, which persisted in Team Three into the 1950s, has long since disappeared. On the other hand, forty-three women were two to four years older than their husbands, and thirty-nine men two to four years older than their wives. The pattern of wives two to four years older continued in some relatively recent marriages, including twelve out of twenty-two cases in the years 1974-84. However, in none of the thirty-eight marriages consummated between 1984 and 1986 did the age differential exceed two years. The traditional preference for brides two to four years older appears to have declined. The emerging pattern in the mid- to late 1980s is one of close age parity and a preference for intravillage marriage, with husband and wife of similar age in their early twenties.

Marriage Age and Class

The Team Three survey data provide striking confirmation of class-based differentiation in marital patterns prior to land reform. Wugong was part of the central Hebei base area during the anti-Japanese resistance. Landlord power in this region, with initially very low tenancy rates, was further reduced as a result of wartime reforms prior to the 1947 land reform. Consequently, when class labels (chengfen ) in this village were fixed in 1947, the Party based them on politicized memories of landownership and social relations that purportedly existed in 1936, more than a decade earlier. Only in that way could the Party assure the existence of class enemies to target for struggle and expropriation. The labels nevertheless provide certain clues concerning differential marital patterns among the poorer and more prosperous in the years prior to land reform.

The sample is of twenty-seven males classified as poor peasants, twelve as middle peasants, and one as landlord who were either age sixty by the time of the 1978 survey or had married by 1947. The marital data for the middle peasants and landlords displays the preferred regional pattern of early marriage: eleven of thirteen men married between ages fifteen and twenty to women who were two to five years their senior. By contrast, only four of twenty-seven poor peasants were able to marry by age twenty. Thirteen others were unable to marry until age twenty-five or later, including eight who first married after age thirty. Only four of twenty-seven poor peasant men married an older woman. Sixteen out of twenty-seven of these

poor peasant men married women five or more years younger, including four wrose brides were ten to eighteen years younger.

The 1978 survey data reveal the expected correlation between class designation and marital patterns in Wugong in the years prior to land reform. By contrast, those who married in the 1970s and 1980s display no significant differences based on class designation, for all hewed to state guidelines for marriage in their twenties regardless of class origin or income.

Marriage Age and Party Membership

Marital patterns for male Communist Party members differ significantly from those of non-Party members before and during the era of mobilizational collectivism. The 1978 survey reveals that all fifty-nine male Party members married between ages fifteen and twenty-nine. By contrast, the range of non-Party marriage ages spans the entire spectrum from age fourteen to sixty, with twenty-nine taking place after age thirty. The crux of the issue is the generation that included people who were in their sixties to their eighties at the time of our 1978 survey, since it was among them that later marriage, or no marriage, frequently occurred. Eight of the eleven male Party members age sixty and above had married by age twenty-four, and the remaining three married between twenty-five and twenty-seven. By contrast, the thirty-seven non-Party males for whom marital data is available included ten who married for the first time in their thirties, two in their forties, two in their fifties, and one at sixty. The fact that male Party members who reached marriage age prior to the 1947 land reform all married by age twenty-nine strongly suggests that in Team Three the Party did not recruit extensively from the very poorest strata, that is from marginal households whose sons could only marry very late if at all. Historical research documents the significant role played by prosperous peasants and landlords as well as independent cultivators in the anti-Japanese resistance in Wugong and throughout central Hebei. Such people had the resources to marry at relatively early ages. The Party also recruited among the poor, but rarely from the ranks of the utterly destitute and marginalized.

Party members responded only slightly more loyally than others to the campaign to delay marriage in the years 1975-84. Five of the six Party members who married in the years 1975-80, including two women, waited until ages twenty-five to twenty-eight. The remaining Party member, who was serving in the army, married at twenty-four. Of the eighty-nine non-Party members who married in those years, fifteen men and seventeen women, over one-third of the total, married prior to age twenty-five. Within the context of significant marriage delay for all, Party members tended to hew slightly more closely to state guidelines. But in a model village, nearly everyone conformed to state demands to delay marriage.

Conclusion

This chapter has highlighted changes in family size and structure, marriage age, location of marital partners, and the locus of marital decision in response to powerful state initiatives. Of particular importance are the following: collectivization and population control leading to village involution in the era of mobilizational collectivism, and market-oriented contractual reforms and rigorous birth control measures since the 1980s. The changes associated with the collective era in Wugong and in much of rural China led to homogenization in age of marriage in the range of twenty to twenty-four, a marked shift toward intravillage marriage, shrinking family size, and predominance of nuclear families long before the state succeeded in reducing fertility. Comparative data strongly suggest that these patterns, documented here with respect to Wugong, were widely shared throughout the Chinese countryside. Many trends, particularly delayed marriage and later reduced fertility (the one-child family), constitute direct responses to state policies and priorities. But others reveal the striving of rural inhabitants, and particularly rural youth, to expand opportunities to make autonomous decisions concerning marriage, family, and family division. Certain outcomes, such as the unusually early and high response rate to the state's delayed-marriage campaign, the early and rigorous implementation of birth control, and the trend to intravillage marriage, reflect Wugong's model status, a factor that peaked in the 1950s to 1970s before declining in the early 1980s.

The survey results suggest that the household contract system and expanded opportunity for market and mobility since the 1980s produced significant changes in family and individual marital strategies, including further shrinkage of family size, the existence for the first time of substantial numbers of new couples forming independent households prior to the birth of a child, and the decision to postpone conception for at least several years after marriage. Finally, we note signs of the emergence of a bimodal familial pattern that can best be understood in terms of new class divisions. At one pole we observe small nuclear farm families and at the other a significant number of joint and stem families, many of them engaged in significant entrepreneurial activity.

Seven
Reconstituting Dowry and Brideprice in South China

Helen F. Siu

This chapter is based on fieldwork conducted periodically from 1986 to 1990, funded by the Committee on Scholarly Communication with the People's Republic of China and Wenner Gren Foundation for Anthropological Research. I am grateful for the comments from Deborah Davis, Jack Goody, Stevan Harrell, and participants in the conference on Family Strategies in Post-Mao China, June 12-17, 1990, Roche Harbor Resort, San Juan Island.

The recent decade of reforms in the 1980s brought drastic changes to Nanxi Zhen, a market town with a rural hinterland.^[1] Situated at the heart of the Pearl River delta, connected by easy road and water transport to Guangzhou, Macao, and Hong Kong, it has enjoyed an unprecedented boom.^[2] Rows of new houses have mushroomed on the landscape. Some are funded by relatives living in Hong Kong and Macao. Many are built by local entrepreneurs themselves. There is also the ever-growing number of rural enterprises employing young migrant workers on the town's outskirts, accompanied by the bustling businesses on the roads and in shops and restaurants, by the common sight of color television sets, washing machines, hi-fi systems, and videos in private homes, and by the lavish rituals at funerals and weddings.^[3]

Family dynamics have taken a sudden turn. Nearly everyone I encoun-

[1] All the local place names in this chapter are pseudonyms. Nanxi Zhen went through a few administrative changes after the revolution. It was an administrative town after 1923, headquarters for the third district of Dagang county. It maintained a market-town status in the 1950s. In 1963 it was made into the Nanxi Zhen Commune. In 1987 it merged with the surrounding Nanxi Rural Commune to become a zhen .

[2] This prosperity has been partly brought about by investments from natives who emigrated to Hong Kong, Macao, and Southeast Asia in the first half of the twentieth century and who have become successful businessmen and industrialists. The particularly successful counties in the delta are Nanhai, Panyu, Shunde, Dongguan, and Zhongshan.

[3] See Helen Siu, "Socialist Peddlers and Princes in a Chinese Market Town," American Ethnologist 16, no. 2 (May 1989), for the economic changes.

tered in town with grown children complained that the young people of today are drunk with the new wealth. They are desperately energetic, but they are also said to be brash, uncaring for their parents, vulgar in their conspicuous life-styles, and lacking moral restraint. Surprisingly, the young people who show little knowledge of the Chinese cultural tradition eagerly participate in lavish funeral and wedding rituals and shoulder most of the expenses themselves.^[4]

Town residents also complained that the escalating dowries, some reaching over 10,000 yuan, are ruining those who have daughters to marry off, and that sons and daughters keep most of their wages and bonuses for their own future homes instead of paying for their keep. With an average monthly income of 200-300 yuan, young workers give at the most 40 to 80 yuan per month to their parents. Although the payments from the groom's side averaged 1,000 yuan in cash, the demand on the groom's family to provide a new house for the young couple puts a considerable strain on relationships within the family and leads to intricate maneuvers with town cadres to obtain building sites and materials.^[5]

In the villages surrounding the town, which represented an impoverished area in the prerevolutionary and the Maoist period, a similar lavishness is now displayed in the provision of new houses by the groom's side and large banquets.^[6] Dowry from the bride's family remains small compared with that of the town, but contributions from the groom's side have increased several times, reaching an average of 2,000 yuan. Poor males are said to have resorted to taking in migrant women from Guangxi province because the amount demanded for marriage to local women has become unaffordable.^[7]

Have the terms of marriage gotten out of reach and out of control during

[4] See Helen Siu, "Recycling Rituals: Politics and Popular Culture in Contemporary Rural China," in Unofficial China: Essays in Popular Culture and Thought, ed. Richard Madson, Perry Link, Paul Pickowicz (Boulder: Westview Press, 1989), for the analysis of why young people are actively engaging in ritual activities.

[5] In this area where properties belonging to emigrants had been confiscated in the Maoist era, one strategy today is for the emigrants to claim the properties back for their relatives. Cadres in Nanxi have expressed concern for such a trend because the town government has no resources either to compensate for these properties or to relocate those who are made to vacate.

[6] See also the chapter on marriage practices in a rural community in the eastern part of the Pearl River delta, in Sulamith and Jack Potter, China's Peasants: The Anthropology of a Revolution (Cambridge: Cambridge University Press, 1990).

[7] In the 1980s, Nanxi Zhen hired thousands of migrant laborers from other provinces, especially Jiangxi, Guangxi, and Hunan. They filled the lowest level of town factories and also became sharecroppers in the villages. Employment peaked in 1988 with over 15,000. For the pattern of migration, see Helen Siu, "The Politics of Migration in a Market Town," in China on the Eve of Tiananmen , ed. Deborah Davis and Ezra Vogel (Cambridge: Harvard University Press, 1990).

the recent decade of prosperity? Why do people feel compelled to pursue these ends? What do the marital transfers today tell us about family strategies four decades after the revolution? In addition, the different emphasis on contributions from the bride's and groom's sides in town and village is striking and requires our analytic attention. In pursuing these questions, I aim in this chapter to address theoretical debates in anthropology with regard to marital exchanges, and to use the debates to highlight the complexity of family dynamics and cultural exigencies as several generations of local residents experienced major transformations in the political economy during the last half-century.

Most scholars who work on dowry and brideprice would agree that marital transfers affect and reflect relationships between the generations and between families. Jack Goody starts out arguing that dowry, which establishes a conjugal fund, is a form of "diverging devolution" common to complex stratified societies in Europe and Asia. For purposes of maintaining economic standing, families find it important to advance the status of daughters as well as sons, and so to allocate them a share in the parental estate. He contrasts dowry with bridewealth, a circulating fund common in classless societies. Bridewealth is an exchange among senior men used to establish future marriages, especially for the sibling of the bride. Wealth goes one way, and rights over women another.^[8] In Chinese studies (as elsewhere in Europe and Asia), the term "brideprice" applies to that portion of the marital transfers provided by the groom's kin; it is usually part of a wider set of transactions that includes the dowry, the contributions made by the family of the bride and normally destined for the daughter or for the married couple. Goody suggests an alternative phrase, "indirect dowry."^[9] Here the term "brideprice" is retained for transfers from the groom's side, whatever their destination. What usually changes is the different weight given to one element as against the other, with higher status groups tending to stress direct dowry and lower ones brideprice, or indirect dowry.^[10]

[8] The term "brideprice" has been used to cover two different types of marital transfer. In Africa, it was earlier used for the transaction that passed from the family of the groom to that of the bride, where it was available for the marriage of her brothers or other male kinsfolk; the word was generally abandoned in favor of "bridewealth," since nothing like "price" in the usual sense of the word is involved.

[9] Whether the former gifts are passed on to the daughter or used in the marriage festivities themselves, they constitute even less of a "price" than African bridewealth. However, among some lower-status groups, part or even all of these transfers may be retained by the bride's kin, possibly in compensation for the gift they have provided, possibly as a reserve on which the daughter can draw, possibly for their own use. See Jack Goody and S.J. Tambiah, Bridewealth and Dowry (Cambridge: Cambridge University Press, 1973).

[10] Although I have sometimes indicated the element of endowment, it is essential to recognize the above complications and bear in mind that "brideprice" in the wider sense is rarely an alternative to direct dowry, but rather a counterpart.

Stressing prestige building rather than status maintenance, Harrell and Dickey refine Goody's original formulation by underscoring the social meaning of dowry.^[11] They expect dowry to be a prominent feature of socially differentiated but openly competitive societies with urban commercial wealth. While recognizing that dowry also involves an economic transaction between families that reflects how woman's labor is calculated and valued, it is more important as part of a cultural, symbolic complex, a social statement for the upwardly mobile classes.

Before the revolution the marriage transfers of rural China usually consisted of indirect dowry; that is, the groom's family made a contribution, which was returned with the bride as part of the dowry. Only among the very poor would it be retained by her kin.^[12] Maurice Freedman suggests that brideprice payments signify the superiority of the groom's family vis-à-vis the bride's, while Patricia Ebrey sees dowries in historical times as used by the upwardly mobile for purposes of status enhancement.^[13]

Both claims are true in specific circumstances, but dowry is used in a much wider range of situations than social mobility alone, and brideprice is used for more purposes than affecting superiority, since they are elements in virtually all marriages. This point is made by Goody in his recent book where he refines his earlier analysis by arguing that in traditional China, the incorporation of women by their husbands' families was never abrupt or complete. She and her natal family maintained multiple ties through life. He sees only a limited role for marital transfers as a payment for the right to women's labor in the strict sense, but allows analytical room for the role of dowry as a means of prestige building. In the economics of status maintenance, he stresses the cultural and social importance of a woman to both her natal and her affinal kin in providing for "the continuation of the house," to which process the creation of conjugal funds was central. This was evident in China, where most scholars intuitively assume otherwise. Both families contributed to this fund. Direct dowry was the main instrument used by high-status families to endow the new conjugal units, and indirect dowry (or brideprice in the wider sense) by lower-status families for similar purposes.^[14]

This theoretical material is useful for understanding the nature of mari-

[11] See Stevan Harrell and Sara Dicky, "Dowry Systems in Complex Societies," Ethnology 24, no. 2 (1985): 105-20.

[12] William Parish and Martin Whyte, Village and Family in Contemporary China , 180-92 (Chicago: University of Chicago Press, 1978).

[13] Maurice Freedman, Chinese Lineage and Society: Fukien and Kwangtung (London: Athlone, 1966), 55; Patricia Ebrey, "Early Stages in the Development of Descent Group," in Kinship Organization in Late Imperial China, 1000-1940 , ed. Patricia Ebrey and James Watson (Berkeley, University of California Press, 1986).

[14] Jack Goody, The Oriental, the Ancient and the Primitive: Systems of Marriage and the Family in the Pre-industrial Societies of Eurasia (Cambridge: Cambridge University Press, 1990).

tal transfers in Nanxi Zhen and their recent developments. Local historians are eager to point out that the differences in terms of marriage between Nanxi town and the surrounding villages had existed over several centuries: The town elites used dowries to enhance their status, whereas the poor tenants in the villages "sell" their daughters to finance their sons' marriage by asking for a large brideprice. This view is of course debatable, but if it were true, it would confirm the analyses of Harrell and Dicky on a general level, and also the China-specific formulations of Freedman.

The issue is complicated, however, by the question of whether a similar set of factors can be used to explain the differences in the decade of the 1980s. Are the divergent terms of marriage observed today similar to those in the prerevolutionary period, revived after a hiatus of forty years? Or does the explosion of energies in the 1980s indicate how fundamentally family relationships had been changed by the Maoist state in the decades before, and how the post-Mao reforms have further restructured family life in both town and village?

Emphasizing again the social prestige—building element in marriage transfer, as Harrell and Dicky would, one may argue that the escalations of dowry in town today are due to a sudden prosperity combined with a largely untouched cultural tradition. Revived by liberalized policies in the 1980s, the ideal forms of transfer finally become a reality for the majority of people. In the town, the new wealth generates competitive display; parents want to provide for their children in their new households in order to confirm the family's social statuses. It is the same with the younger generation. Nanxi town consists of a large number of young workers, private entrepreneurs and cadres whose fortunes have exploded in the decade of reforms and who are eager to flaunt their newly acquired wealth for the purpose of social networking. Large dowries and at times equivalent contributions from the groom's side are ways to reinforce status. It is also shrewd business strategy, as with the financing of other rituals. This analytical position assumes that the socialist interlude in the past forty years has done little to change cultural expectations and strategies for the perpetuation of family wealth and status.

If one stresses the rights to women's labor in marriage transfer, analytical attention is drawn to the escalating brideprices in the villages and their relationship to labor needs among the farming families. Could the increased pace of economic activities in the 1980s have exacerbated these needs, which are reflected in the ways families strategize and negotiate marriage payments? Social statuses in the villages are changing with increased mobility and rapid out-migration. With an acute labor shortage, it seems logical to assume that families are paying a high brideprice in order to secure the rights of women to augment long-term field labor. That was a concern in the past, and it remains one today. This position regarding the role

of marital transfers assumes that the traditional village-town divide, as well as the gender and generational division of labor, continues to structure the area's ongoing development in much the same way as before.

Observing the divergent terms of marriage in Nanxi town and the surrounding villages in the 1980s, I would caution against viewing these terms as representing two different political economies with their own cultural logic, which have sprung back to haunt the socialist government. On the surface, the town-dwellers' stress on dowry and the villagers' on brideprice seems to indicate that the strategies of prestige building and of the recruitment of women's labor have prevailed in the different sectors. But I suggest that these marital transfers arose in a single culturally constructed political economy, which had undergone drastic transformations in the Maoist period at one level and which had persisted at another. At the level of change, one can view the extravagant dowries in town today as public statements and shrewd networking in an era of entrepreneurial vigor; the peculiar form such vigor takes is based on the popular assumption that the party-state can exert its power against any formal civic organization. I would also show that in both town and village today the marital transfers negotiated involve not so much the exchange of material goods and prestige between the families of the bride and groom as the intense and rapid devolution of property to the conjugal couple at the time of marriage itself. This trend represents a very different intergenerational and interfamilial dynamics in the context of a much-changed political economy. On the issue of persistence, I shall argue that the involvement of women in "the continuation of the house" remains intact in both town and village after forty years of socialism. The peasants who face the rapid emigration of both sons and daughters can meet labor demands by making sharecropping arrangements with migrant laborers; they choose instead to pay a high brideprice in order to anchor much-valued family commitments in the village. Could it be that rather than the family's merely surviving the Maoist interlude, its importance may have been intensified in the last few decades by Maoist politics to the point that the older family members, confronted by a decreasing commitment from their children, are now pursuing family strategies with inexplicable fervor? Goody's analytical views on the importance of the conjugal fund may seem applicable in these circumstances.

Therefore, while appreciating the major roles of both cultural tradition and economic logic in shaping marriage transfers, I suggest that four decades of socialist politics have underlain the mutual impact of marriage and the economy, changing and at times preserving cultural expectations, restructuring economic needs, and at times reinforcing family strategies.^[15]

[15] This is close to the observations of Parish and Whyte for the postrevolutionary period up to the late 1970s. For north China, see Kay Johnson, Women, the Family and Peasant Revolution in China (Chicago: University of Chicago Press, 1983), 208-14.

That history highlights how much and in what ways family relationships have been transformed, and it is a crucial analytic element for understanding marital transfers in Nanxi Zhen today. Using a random sample survey of three hundred households collected in 1986 and on subsequent trips in 1987 and 1989, I explore these issues of the reconstitution of brideprice and dowry in Nanxi Zhen and the surrounding villages in relation to the process of decollectivization or privatization.^[16]

Prerevolutionary Marital Transfers in Nanxi

The Pearl River delta is a complex, ever-expanding social landscape. As the river continues to flow along a southeastern direction and forms marshes at its lower reaches, it attracts migrants who come to settle from different parts of south China. This process has given rise at various periods to communities with vastly different economic resources and cultural configurations. Nanxi town is one of a line of market towns situated between a well-populated part of the delta and these vast river marshes. Known locally as sands (sha ), the marshes were reclaimed and converted to rice fields by lineage and merchant estates based in the town. In this way, Nanxi grew from an outpost in the sands during the Ming to a dominating center of wealth, power, and ritual by the late eighteenth century. At the turn of the present century, the residents consisted of a dynamic commercial class, together with its functionaries, who controlled the landed estates in the sands and the trading of grain, mulberry, silk cocoons, wine, fruit, and vegetables.^[17]

Although the economies of Nanxi town and the villages have been interlocked, they are worlds apart in many ways. The inhabitants of the sands were mobile, boat-dwelling farmers and fishermen who were treated as an underclass by town dwellers. Generally referred to as the Dan, the sands people worked as hired hands and tenants when lineages reclaimed river marshes, and they engaged in the meaner trades, such as carrying coffins and digging graves. Prejudice against them was reinforced during the years of bad harvests when they fled the area in their boats with whatever they could glean from the fields. When they had gathered strength in numbers, they were portrayed as bandits and pirates.^[18] Until the 1940s there were

[16] The survey covers all fifteen of the town's residential neighborhoods and three villages at different distances from the town.

[17] See Helen Siu, "Recycling Tradition: Culture, History and Political Economy in the Chrysanthemum Festivals of South China," Comparative Studies in Society and History 32, no. 4 (1990): 765-94.

[18] See Dian Murray, Pirates of the South China Coast, 1790-1810 (Stanford: Stanford University Press, 1987). See Helen Siu, Agents and Victims in South China: Accomplices in Rural Revolution (New Haven: Yale University Press, 1989), chaps. 2 and 3, on the historical development of the sands in the Pearl River delta.

seldom any real villages. The occasional straw huts on the dikes were in sharp contrast to the 393 ancestral halls and the 139 temples in the town before the 1949 revolution. Even today, the sands consists only of small villages between which lie extensive fields.

Nanxi was a place of residence for an unusual concentration of the wealthy and upwardly mobile until the land reform. An official document of 1971 records that the town then had 482 persons labeled as landlords and 57 as rich peasants. Most of them belonged to the town's four major lineages, whose members married each other. Before 1949 they and their families constituted a significant social group among the town's some 12,000 residents. For the women in these families, there was little extradomestic employment and no need for it. In fact, a married daughter of the wealthy might not touch the food of her husband's household for at least a year, being supplied with foodstuffs from her natal home. The women were also provided with lavish dowries, which might include elaborate jewelry, double sets of suanzhi furniture, delicate porcelain dishes and vases, silverware, together with annual provisions of grain and large tracts of river marshes.^[19] Brideprice was negligible. In 1986, older women whom I interviewed generally agreed that before the war the brideprice for a "respectable family" was about a hundred silver dollars, a few hundred wedding cakes, and banquet food. Whatever was received was largely consumed in the wedding feast. Both the bride's and the groom's family would hold a dinner for friends and relatives, each making up ten or more banquet tables. They insisted that except for the very poor, families accepted without argument whatever the bridegroom's family would offer. That custom remains unchanged today. Negotiating brideprice would have been a tremendous loss of face. "Only the sands people would act in such a manner," I was told.

To attribute this emphasis on dowry only to a strategy for status acquisition among rich families is to neglect the fact that a similar situation of low brideprice in relation to dowry existed in ordinary town families. It might be suggested that the terms of marriage had different meanings for households in various circumstances. But the life-styles of members of the wealthy lineages and what they provided for their daughters represented the cultural ideal in the area.^[20] The smallness of brideprice might also be

[19] I refer to those that controlled the development of the sands from the Ming dynasty on. See the genealogies of the Zhao of Sanjiang in Xinhui county, the Long and Luo of Shunde county, the Huo of Foshan. See Siu, "Where Were the Women? Rethinking Marriage Resistance and Regional Culture in South China," Late Imperial China 11, no. 2 (December 1990): 32-62, for details of marriage customs in the area.

[20] The same was true for Shawan in Panyu county, another center of wealth and power at the edge of the sands. A large percentage of the town households in Shawan literally lived off the focal ancestral estate, Liugeng Tang, which owned 60,000 mu of the sands. As in Nanxi, women learned sewing and helped with family chores until they settled with their husbands. For the historical evolution of lineages in Shawan, see Liu Zhiwei, "Lineage on the Sands: The Case of Shawan," a paper presented at the panel "Lineage Power and Community Change in Republican South China," annual meetings of the Association for Asian Studies, April 1990.

related to the practice of buluojia , the delayed removal of the bride to her husband's house.^[21] The richer the families, the longer their daughters stayed at home. Ties with a married daughter continued to be important in various ways. As long as she remained at her home, the fruits of her labor were shared with her natal family. The significant break with her parents came when she was about to have a child, not when wedding gifts were negotiated and exchanged. When she joined her husband, provisions for continuing support were often made. If brideprice was used at all for acquiring the rights to a woman's labor, the custom of buluojia would have kept the payment low.

Town residents, whether from families that were formerly rich or poor, insisted that their terms of marriage and those of the sands were very different. The inhabitants of the sands did not practice buluojia ; a woman settled with her husband's family immediately after the wedding. Among the fishermen, the mobility of their boat-dwellings made it difficult to do otherwise, since it was not a question of husband and wife being in adjacent villages. At the same time, their dowries were modest, consisting of several sets of clothes, a blanket, a mosquito net, washbasins, bowls, and plates; but the brideprice was large, usually twice that of the town. In my 1986 interviews I came across old men in Zhixi, a village in the sands, who had been married in the late 1920s and early 1930s. They quoted brideprice amounts of one hundred to two hundred silver dollars, which seemed extraordinarily high given the standard of living in the sands at the time. One part of the brideprice consisted of banquet food for the relatives and friends of the bride. In the town only selected members of households (other than those of close relatives) were invited to attend wedding banquets, but in the sands such occasions involved every aquaintance far and near, young and old. Feasting lasted for three whole days. Meals were served for those guests who had moored their boats the night before. They were feasted again on the wedding day, and on the third day meals were served before the guests left the area. As in Africa, the remainder of the brideprice was

[21] Historical documents through the Qing recorded it as common in parts of the Pearl River delta both among the elites and commoners. It involved women marrying at an early age but continuing for a few years to reside at their natal home. They would briefly visit their husband's home for important ritual occasions and festivals, but not until they were about to give birth to their first child would they settle there permanently. See Siu, "Where Were the Women?" for a summary of the relevant literature.

often used as a marriage fund for brothers,^[22] a use that was scorned by town residents, who asserted that the sands people were "shamelessly aggressive" in negotiating brideprice and that "they had to marry off a daughter in order to get a daughter-in-law."^[23] It seems that their strategy for coping with life at the bottom of a social hierarchy was regarded as a stigma in the eyes of the townsfolk, who used it as a way of dissociating themselves from the people of the sands.

The War Interlude

The domination of the town over the sands in the late imperial period continued through the republican period until the decade of the 1940s, when war and endemic disorder finally broke the power of the town-based ancestral trusts. Many tenant farmers in the sands were able to acquire land from their landlords by manipulating cash rents, grain prices, and taxes. In addition, local bosses who rose to the height of their power during the Japanese occupation took over vast estates by force.^[24] Although built with straw and river mud, small villages with more permanent dwellings mushroomed in those places where local bosses set up their territorial bases.^[25] Interviews suggested that in this particular decade marriage strategies changed with the shift in power. The rich households in the villages started to arrange marriages with elite families in town whose fortunes had declined in the years of war. In doing so, they adopted the prestigious custom of buluojia and displayed lavish dowries. My interviews show that several local bosses succeeded in marrying their daughters to the families of merchants and rich peasants from the neighborhoods adjacent to the town.^[26] As for the ordinary people of the sands, the situation did not appear to change, although the new situation provided the basis for them later on to redefine their circumstances. In 1986 I interviewed a dozen of the old residents in one of these villages settled by migrant laborers, Zhixi. Although

[22] See Jack Goody and S. J. Tambiah, Bridewealth , on the functions of brideprice and dowry in African and Indian societies. A circulating fund increases interdependence among siblings and between generations within a family.

[23] The demand of high brideprices in the sands was told to me by many old women in town and confirmed by a marriage ritual specialist who attended many weddings in the area.

[24] For the rise and fall of local bosses, see Helen Siu, Agents and Victims , chap. 5. For Nanxi, see Helen Siu, "Subverting Lineage Power: Local Bosses in the 1940's," a paper presented in the panel "Lineage Power and Community Change in Republican South China," AAS meetings, April 1990.

[25] A richer household might have a tile roof. These houses are known as chan ding ya (wooden beams and bark holding up tiles).

[26] For example, the Liu family of Zhixi village married a daughter into a prosperous Mai family of Yongding in the late 1940s. A great deal of jewelry was given as dowry. The father of the bride was the most powerful local boss in the village.

it was difficult for them to name any ancestor beyond their grandfather's generation, they nevertheless claimed that their families had originated from the older part of the delta where the ancestral halls stood. They insisted that as long as they could remember, they had practiced buluojia . They also disassociated themselves from the boat-dwelling inhabitants of the more recent river marshes further southeast, an area they termed xia sha (the lower sands).^[27] But data from my survey contradicted their claims. Their families at the time demanded extraordinarily high brideprices of between one hundred and three hundred silver dollars. Families of the bride and groom both provided banquets that lasted for several days. Dowries were far below what was given by elite families in town and consisted of kitchen utensils and bedding, although in some cases a set of table and chairs and a chest were added.

Contrast the rural areas with the circumstances of households in town for that period. During the eight years of war with Japan, Nanxi town was ruled by an uneasy alliance of local bosses with the Japanese army that was stationed nearby. Everything was scarce owing to the blockades imposed both by the Japanese military and by what remained of the Nationalist government. The fortunes of landowning families and merchant houses declined rapidly, their estates being divided and sold off to the military bosses. The ideal marital payments were hard to come by. As for ordinary households, the ideal was even harder to imagine. Old women I interviewed claimed that ordinary households provided little dowry to speak of. Brides settled immediately with their husbands because family members were scattered or had perished. The most the couple got was enough food for a dinner with close relatives, together with some basic items of furniture. Those who were refugees from other counties were grateful to have any shelter.^[28] Brideprice, if given, was often in the form of baskets of rice, ranging from one to ten dan .

Except for the local bosses who had risen rapidly during the war, families in town found lavish dowries for daughters harder to reach, and the embarrassment of asking for a brideprice became a necessity. In other words, the dislocations of war and the reversals of fortune for the area's elite families had eroded the traditional terms of marriage before the Communists arrived on the scene. It is this period that provides a realistic baseline for an evaluation of the relationship between the terms of marriage and the transformed social hierarchy after 1949.

[27] Some of the residents of Zhixi might have been poor urbanites who migrated to the villages. They attended the lineage ceremonies in the ancestral halls in town and continued to observe town customs.

[28] In my household survey of 1986, I found many heads of households, men or women, who had come from other counties in Guangdong during the war and who settled in Nanxi afterward.

Socialist Transformations

Since family continuation, social prestige, the devolution of wealth, and the labor needs for households were important factors in marital transfers, family strategies were bound to change during the Maoist period from the 1950s to the late 1970s because the consolidating party-state had made a serious attempt to redefine the ideological principles and to restructure the political economy and the social hierarchy. In Nanxi the step-by-step transformation started immediately after the Communists took power. During the land reform of 1952, the former elites of the town—managers of lineage estates, merchants, administrators of the Nationalist government, and local bosses who came to power in the 1940s—were killed, imprisoned, or had their properties confiscated. Those who had set the standards of wealth and power were then reduced to a caste of untouchables. Not only were the means no longer available for continuing such a life-style, but the policy of overturning the hegemony of "tradition" was actively pursued during political campaigns when activists were brought in from outside the community. Those activists who had often been poor tenant farmers in the sands set about destroying ancestral halls and important public buildings, and carted the materials off to the villages. The political fervor extended to social life itself. Through the 1950s and 1960s, lavish dowries and wedding feasts were periodically stigmatized as "feudal extravagance." Brideprice was condemned as "buying and selling in marriage," and large payments were prohibited. But the uneven ebb and flow of politics meant that local cadres were sometimes compromising and permissive. The uneven pattern of policy implementation was quite similar to the one found by Parish and Whyte in rural Guangdong.

People interviewed in the late 1980s repeatedly expressed resignation about the stripping down of traditional practices in Nanxi town through the Maoist decades. The ideals of traditional marriage payments lingered on after the land reform. It was impossible for the town residents to afford the lavish transfers of the former elites, but dowries did consist of moderate provisions of household furniture, utensils, and jewelry. People recalled carrying brides on sedan chairs and attending wedding feasts. At the time, agriculture and commerce had recovered somewhat from the war. Family fortunes and private property were inheritable. But the subsequent collectivization of town enterprises followed by the famine years of 1958-61 left little for wedding festivities. My friends remembered how they sent their children around the neighborhood streets to gather leaves for fuel, and how fried dough wrapped in banana leaves became a delicacy. When the economy recovered in the early 1960s, local cadres had to brace themselves for the siqing campaigns.^[29] For fear of corruption charges made by

[29] "The four clean-ups" campaigns preceded the Cultural Revolution. The major targets of attack were village cadres who had risen with the Party bureaucracy.

work teams sent from outside, local cadres asked the inhabitants to behave with moderation. Guests could attend wedding dinners only in small groups because the town leadership explicitly limited the number of banquet tables to four per gathering.

The Cultural Revolution tried to kill off whatever remained of the old practices in town, giving rise to severe factional disputes among the town's leadership. Although actual fighting was rare, groups of Red Guards were formed, cadres were challenged by activists recruited from the sands, former landlords were paraded around the town, and ancestral tablets were burnt. The official model for wedding celebrations was to hold tea receptions instead of dinner banquets. Interviews show that because of the concentration of cadres, Party members, union workers, and teachers in town, the official line for simple wedding rituals was imposed more effectively than in the villages. Furthermore, the concentration of ancestral and merchant estates there in the prerevolutionary period meant that there were many families with bad class labels. Reduced to being a group of untouchables, members of these families had to marry among themselves or else "marry down" to whoever would take them. In the latter case the husbands were mostly poor petty traders who could not afford any marriage prestations.

By the mid 1970s those who remained unconvinced by socialist rhetoric were too afraid of the Party's organizational power to express their sentiments publicly. A new generation came of age under the shadow of the Maoist ideological purity who had been denied the appreciation of cultural depth and historical sensitivity. In the eyes of their parents, they had little sympathy for tradition.

Nevertheless, despite economic hardships and the government's explicit restrictions, dowry and brideprice remained an important concern for the local residents. The highest sum for brideprice quoted to me was about a hundred yuan, balanced by a few dowry items, such as bedding, household utensils, chairs, and a table. Such payments were not substantial enough to be viewed as means for prestige building or as shrewd calculations to obtain a woman's labor. But their persistence did reflect the continual importance of family commitments in everyday life under the changing socioeconomic conditions.^[30]

The economy of Nanxi town had been drastically restructured and removed from the sphere of local agriculture. Instead of a place of residence for rentier landlords who lived off the grain from the sands, residents were

[30] The reinforcement of family and patriarchy by socialist transformation has been dealt with by Judith Stacey, Patriarchy and Socialist Revolution in China (Berkeley: University of California Press, 1983); Margery Wolf, Revolution Postponed: Women in Contemporary China (Stanford: Stanford University Press, 1985); Emily Honig and Gail Hershatter, Personal Voices (Stanford: Stanford University Press, 1988). See also the works of Kay Johnson and Elizabeth Croll, and most recently, Deborah Davis's chapter in this volume.

employed in collective enterprises, which increasingly entered into subcontracts with state factories and commercial units at the county capital. While private resources were stripped away, wages in these collective factories were low, and the security or social services offered could not be compared with employment in the state sector. Nevertheless, town residents anxiously held on to their meager earnings and unflattering status, hoping that the government would not look for an excuse to send them to the sands and demote them to the status of peasants. Their fear of such arbitrary political reversals was not unjustified. Their status as town dwellers had been redefined when the adjacent rural commune was administratively merged with the town between 1958 and 1963. Ironically, it was their urban status that also led their children to be sent off to the sands in the wake of the Cultural Revolution.

Factory employment meant that the next generation worked with collective town enterprises rather than with their families. With wage employment, young workers might be expected to develop more individualized marriage strategies highlighting the concerns of the conjugal pair. But because of the relatively stagnant town economy and the post-1949 wage structure favoring seniority and the practice of dingti , young workers continued to be dependent on their parents and were expected to contribute to the family budget.^[31] In the collective town enterprises workers received a great deal less of social services than those in the state sector, and the family continued to provide important support. My survey shows that one-generation families were rare. Many households consisted of the parents, a married son and his children, and unmarried children. Because of the continued practice of buluojia , it was not unusual for a married daughter to eat her meals at home. In other words, intergenerational and sibling ties remained strong despite wage employment outside the home and despite government efforts to replace loyalties to the family with those to the Party and the state.^[32]

Indeed, families relied increasingly on an informal network of relatives because many traditional civic organizations had been eliminated.^[33] This

[31] Dingti is a system of job assignment in which a retiring person's job can be taken over by his or her children.

[32] On age cohorts and structured life chances, see Deborah Davis, Long Lives , 1991 paperback ed., chap. 8; "Intergenerational Inequalities and the Chinese Revolution," Modern China 11, no. 2 (April 1985): 177-201; "Unequal Chances, Unequal Outcomes: Pension Reform and Urban Inequality," China Quarterly , 1989, 223-42; on dependent relationships in the urban industrial structure, see her review of Andrew Walder's book Communist Neo-traditionalism: Work and Authority in Chinese Society (Berkeley: University of California Press, 1986), in "Patrons and Clients in Chinese Industry," Modern China 14, no. 4 (October 1988): 487-97.

[33] I refer to marketing networks, temple and credit associations, neighborhood shrines, and ritual specialists. For example, residents relied more on their relatives for funerals when ritual specialists and coffin-bearers were not readily available.

dependence was compounded by the general bureaucratization of life—for example, in the allocation of housing, for the establishment of quotas in various neighborhoods to send educated youths to the sands, or for the recruitment of soldiers and students based on class background.^[34] Maneuvering the power of the state became a feature of everyday life. It showed even in household registration: married children delayed the registration of their new household until it was strategic to do so. A paradox existed for the town residents: while they relied heavily on the family for services not available through state channels, coping with the power of the state had become a cultural given in domestic decisions. The terms of marriage were negotiated within the framework of this politicized environment.

Maoist politics brought about drastic economic transformations in the sands as well as in the town. Nevertheless, marriage transfers continued to be significant for the local population. Instead of mooring their boats along a wide network of rivers, the inhabitants of the sands were forced to settle on land, where they were increasingly confined to their cellularized villages.^[35] But under the policy of the Maoist period, settlement rights had strings attached. An exclusively grain-growing economy tightly controlled under the state pricing system kept their incomes low and their lives isolated from the towns and cities. Their daughters, whose children would automatically be assigned rural status, were no longer an available choice as marriage partners for those outside the sands. By the late 1950s the traditional strategy of achieving upward mobility by marrying into town mainly as second wives was closed off. Despite government claims to the contrary, the social gap between rural and urban residents widened.^[36]

Brideprice negotiations continued in the Maoist era. Rural women described the marriage market and the transfer of payments at that time as unpredictable. Families with many daughters were fortunate to get them married off and had to choose from whoever was available in the village. At times, they took whatever brideprice was offered. In other villages men complained that there were not enough women around even when their families offered a good brideprice. A big change came about when people took wives from within the village. Daughters naturally remained nearby

[34] For the bureaucratization of urban life, see Martin Whyte and William Parish, Urban Life in Contemporary China (Chicago: University of Chicago Press, 1984); Jonathan Unger, Education Under Mao: Class and Competition in Canton Schools, 1960-1980 (New York: Columbia University Press, 1982); Gail Henderson, The Chinese Hospital: A Socialist Work Unit (New Haven: Yale University Press, 1984).

[35] Cadres in the sands of Nanxi Zhen complained that it was difficult to make the sands people settle in houses clustered in a village, which was achieved only in the 1970s. See Helen Siu, Agents and Victims , on the step-by-step cellularization of rural communities in south China.

[36] On the village-town divide, see Helen Siu, "The Politics of Migration in a Market Town."

even after marriage, and they continued to help out at home. When a family was short of labor, daughters stayed longer at home, like the town people. The collectivization of the economy of the sands and the corresponding cellularization of rural communities paradoxically intensified interdependence within local families as well as between them.

Villagers in Zhixi, Yongding, and Jiuji celebrated their weddings with banquets of a dozen or so tables throughout the 1950s until the late 1970s, except for the very lean years 1958 to 1962 and during the high tide of the Cultural Revolution, when officials explicitly demanded ritual simplicity.^[37] Still, a brideprice of a hundred yuan plus wedding cakes was not uncommon. As for the dowry element, it was not until the 1970s when "the three treasures," a bicycle, a watch, and a sewing machine, were added to the ideal dowry list. They were largely "luxuries" reserved for the families of rural cadres.

The Reconstitution of Dowry and Brideprice in the 1980s

Is the renaissance today a reinstatement of earlier practices spurred on by new prosperity? Not entirely. If both the terms of marriage and the political economy on which they were based were redefined by the state, their expansion today is further shaped by the entrepreneurial energies in town and in the sands. In two earlier papers, I interpreted the upsurge of seemingly "traditional" ritual practices not simply as a revival but also as having undergone great changes as a result of the socialist period. A similar argument can be applied to dowry and brideprice.

In Nanxi Zhen, the ten years of liberalization since 1980 have brought much prosperity for both town and village. Two categories of residents have made remarkable fortunes. First, party cadres have used official connections to make business deals; they have accepted "bribes" to facilitate those transactions, and they have anticipated the new policies of privatization, moving ahead of others to get the best resources for their families. The large modern villas that have sprung up in town are mostly theirs. Second, there are the entrepreneurs whose interests conflicted with the monopolizing powers of officials but who also colluded with them to gain a share of the profits. Many inhabitants of Nanxi town have been able to make contact with relatives in Hong Kong and Macao, who contributed to the building of houses, bought them imported consumer goods, or invested in joint ven-

[37] Yongding, a village adjacent to Nanxi, is a rural settlement older than Nanxi town itself and containing established lineages. Jiuji, three kilometers southwest of Nanxi town, was a sizable village by the late Qing. Zhixi became a permanent settlement in the republican period. Traditional customs in the villages did not disappear until the late 1950s. A village cadre in Zhixi said that when his daughter married in 1953, she was carried in a sedan chair.

tures; these residents are known as having "a southern window" (nanfeng chuang ). Those relatives who contribute to the enterprises are duly rewarded, but meanwhile the local entrepreneurs make profits and are asserting their independence. This assertion of independence is particularly prevalent among the younger entrepreneurs in town, where the life-styles and preferences of these two categories have become models for emulation and envy among the society as a whole.

The marriage transfers for these privileged families are elaborate. In town, brideprice demands are still not explicitly made, but new calls are made on the groom's family, and the expenses incurred by them are large. It has become essential for the groom to provide a new house for his bride.^[38] Because of the shortage of building sites in town, the cost of a square meter of land is about 180 yuan. Together with building materials, decorating, and labor, a standard house can easily cost 70,000 yuan. Cadre families are particularly under pressure to make such contributions because, as one said, "people expect us to use administrative means to secure the prime sites, the imported tiles, and the color television. When we take the lead, others eagerly follow, claiming that present policies allow everyone to be prosperous."^[39] According to another cadre, such transactions not only assure people that wealth is politically acceptable, but also enable them to convince themselves and others that they have not lost control of the situation in an era of ideological redefinition.

Similar pressure was felt by the nouveaux riches, because to spread the impression that "they have the means" is necessary for business. These families are now ignoring the official norms for the age of marriage. In one case, a marriage was contracted and a house built for a twenty-one-year-old, the only son in the family, because the parents, who were former landlords and who had migrated to Macao a few years earlier, were anxious to have male progeny to receive the family wealth.^[40]

Besides the expenses of the house, the groom's family also pays a substantial brideprice, which has reached a norm of 800 to 1,000 yuan. The sum is mainly used to help pay for the wedding feast at the bride's natal

[38] The provision of a house for the married couple illustrates what Jack Goody speaks of as "indirect dowry," which is generally stressed by poorer families, whereas direct dowry is stressed by richer social groups (such as the former landlord and merchant families).

[39] The price for building land shot up in the town as well as in the villages adjacent to the town. In 1988, the price reached 200 yuan per square meter, compared with about 80 yuan for the villages farther away. The head of Nanxi was notorious for having secured some prime sites cheaply for his children and friends because he was in control of town planning and construction. A modern five-bedroom villa in town was 40,000 yuan in 1986 and rose to 150,000 yuan in 1989.

[40] I attended the wedding of this young man, who had two older married sisters. He seemed not to know what he was doing throughout the entire affair. Earlier, I had interviewed him in my random sample survey as the head of a household.

home.^[41] The standard in town is now at least thirty banquet tables at the local restaurant, at a cost of about 6,000 to 9,000 yuan.^[42] Of course, the wealthy give more, as in the case of a successful private entrepreneur where the parents of the groom offered 1,200 yuan as brideprice, another 50 yuan as lucky money for the bride's brother, together with the cost of thirty-two banquet tables for the bride's family, who resided in a village on the outskirts of town.^[43] They also provided thirty banquet tables on the wedding day for their own friends and relatives. Their banquet food included imported beer and soft drinks from Hong Kong, expensive dried mushrooms, squids, scallops, pork and chicken, and six hundred wedding cakes.^[44]

It is difficult to insist that these prosperous families are attempting to "compensate" for the bride's labor. More appropriately, it has become a strategy for the upwardly mobile to win a bride of appropriate background, to acquire prestige in the local community, and to strengthen personal networks in a volatile system where formal organizational ties are still unreliable. For the common folk, the new standard is making it hard for their sons, because they have neither the wealth nor the political means to meet these expectations and would have to wait to get married until the richer households had chosen their brides. One might argue that this state of affairs is no different from the old days. But in the old days, no poor household would be competing with a rich one for a wife, in accordance with the saying "Bamboo door matches bamboo door, and wood door matches wood door." Now it is a tougher game when the woman one fancies in the workplace finally decides to marry a co-worker whose family has the "southern window" and can provide a house and thirty banquet tables for the marriage. A new hierarchy is in the course of defining itself through marriage and marital transfers.

In town, the expenses for the bride's family can be just as high if not higher. Among the wealthy, dowry items include modern furniture, a bicycle, electric fans, a sewing machine, imported color television, a washing machine, a hi-fi system, totaling over 15,000 yuan. It seems natural to argue that there are many upwardly mobile families created by the sudden opening up of economic opportunities in this area, who use the marriage of their sons and daughters to make an impression in town. However, there is

[41] The sum is about four to five times the average wage of a young worker.

[42] In 1986, a standard wedding gift from friends was 5 yuan. In 1989, I was told to give 25 yuan whenever I was invited to a wedding.

[43] The family used to work on bamboo handicrafts. In the 1980s the head of the household worked as a manager for foreign trade in the largest factory in Nanxi Zhen. His annual salary and benefits easily reached 60,000 yuan, according to friends. One of the sons opened a small factory for metal parts and had his own prosperous business. In the previous three years, the family had built two three-story houses, costing over 100,000 yuan each.

[44] In traditional marriages, wedding cakes and banquet food were listed together with the brideprice and other items and presented to the bride's family before the wedding.

an added dimension. Even for the very rich families, the bride and groom both contribute to the sum by saving up over the preceding few years. The self-accumulating dowry has arrived, and the couple are no longer entirely dependent upon their parents. The objects will be publicly transported and displayed during the course of the wedding day when they go to fill the new house assigned to the young married couple.^[45] Dowry may still consist partly of the inheritance due a daughter, but it increasingly takes on the meaning of a conjugal fund created by and for the newlyweds themselves. Among the young workers, technicians and private entrepreneurs, who now earn a great deal more than their parents because of bonuses and shared dividends, the creation of their own independent conjugal fund is both possible and preferred.^[46] Their eagerness to start the union in a new house symbolizes this concern. They take great care and pride in building and decorating their house according to the styles that are promoted in the television programs beamed from Hong Kong and in magazines for home living.^[47]

Although the younger generation in Nanxi have taken the initiative in building their own conjugal funds, parents who have the means eagerly contribute, for both their sons and their daughters. Given their economic situation, it is not difficult to see the logic behind their efforts. Old age security becomes an important concern when the changing political winds and the subsequent competition from private entrepreneurs have shaken the town's collective economy in which they have participated. In fact, the town's enterprises experienced a great deal of uncertainty in the mid-1980s when faced with fierce competition from the mushrooming industries in the nearby villages. When businesses folded, workers were laid off with little prospect of a pension. Retirees have also found their pensions too meager to catch up with the severe inflation. If one argues that the function of the family as the basic provider persisted during the Maoist era, it is reasonable to expect that those with the means contribute to the consolidation of their

[45] The huiyou , close companions of the groom, often collectively give large decorative mirrors for the new couple to hang in their own house. In the case of the successful entrepreneur mentioned earlier, the groom asked the bride's family to buy whatever dowry items it wanted and then have them billed to his family. When asked why they did that, the answer was the concern for "face."

[46] I am not sure how far these attitudes are due to the impact of the television and media of Hong Kong, which promote the individualistic energies of young entrepreneurs. In fact, young people in Nanxi copy the Hong Kong life-style; for example, they buy imported clothes and electrical goods, smoke American cigarettes, drink Hong Kong beer, and use the vocabulary of Hong Kong television.

[47] Two of my good friends, both young entrepreneurs with a good education and contacts in Hong Kong, live apart from their parents, although the latter come to care for their infant daughters. The interior design of their houses, each of which is equipped with a color television, a hi-fi system, and video, matches that of upper-middle-class families in Hong Kong.

children's marriages, with the hope that some measure of intergenerational dependence can be maintained under such conditions. In one case, two retired teachers saved up enough to buy a color television for their married son who had taken a job in a research unit in Human. When the young couple had a baby, the mother-in-law went to Hunan for several months to take care of the newborn, leaving her husband behind with an unmarried son. She eventually brought her grandson back to Nanxi. I asked why she did all that, well knowing that their resources and energies were strained. Her reply was straightforward: she would maintain a good relationship with her sons in the hope that they would reciprocate when needed, Whether her strategy works or not remains to be seen.

The sands have also been rapidly transformed in the last decade. Wedding banquets have grown even more elaborate, and it is common to have thirty or more tables. According to town residents, the sands people have adopted the custom of the town by sitting around tables, where courses were brought out one by one. This differs from their habit of serving food in large bowls all at once, with the guests squatting around to help themselves. New houses are provided by the groom's family for the newlyweds, but these belong mainly to cadre families.

If marriage practices were only mainly a means for families to work their way up the social hierarchy, one would expect the people in the sands to reduce their demand for brideprice, inasmuch as this has been such a stigma in the eyes of the town residents. Surprisingly enough, dowry remains moderate, but the monetary demands for brideprice have escalated sharply, reaching an average of 2,000 yuan, twice that of the town and more than the annual income of an adult worker in the villages.

To understand this phenomenon, it is necessary to trace the history of the socialist transformation in the sands. Today wedding festivities provide a sharp contrast to the normally quiet life in the villages. These rural areas are devoid of young workers, either men or women, with only old men and women doing light farmwork and carrying grandchildren on their backs. Most of the younger ones have migrated to the periurban neighborhoods, women to work in factories, men on construction sites. Farm labor is in such short supply that farmers increasingly make sharecropping arrangements with migrant farmers from the provinces of Guangxi, Jiangxi, and Hunan. Sometimes they hire workers outright by giving them room and board plus 80 to 100 yuan a month during the busy season. Not all of these migrant workers are good farmers. In 1986 the municipality in which Nanxi situates turned back 20,000 illegal migrants because, according to a municipal cadre, the region had suffered poor harvests for the last few years as a result of the poor quality of their labor.

One may wonder why many families continue to enter into contracts for large areas of land for growing grain despite such acute labor shortages.

The reason is simple; grain prices are high in the open markets, often over 100 yuan per dan , in contrast to the negotiated state price of about 25 to 30 yuan. The high price is due either to the fact that many families have switched to the growing of turfs, fruit, and sugarcane, or to the fact that their children have gone into the towns and cities to work.^[48] This forces them to buy grain in the markets in order to fill the delivery quotas that are part of the process of contracting land from the village.

Attachment to the land is also promoted by concerns of a cultural and historical kind, which have been sharpened by the migration of the young. During interviews, older farmers repeatedly expressed a heightened anxiety about maintaining a base in the village because of the past history of settlement rights, especially as their children do not share such sentiment and would often like to emigrate from the rural areas. It took centuries for the people of the sands to be recognized as permanent inhabitants on the land. During the socialist period, attachment to land was seen as a burden that tied families to rural poverty; only with difficulty did the government convince them to build permanent brick houses in order to form "villages." When land again became a valuable household resource in the 1980s, they felt they must defend settlement rights, which entitled their sons to build houses in the village, and from which later arrivals are in their turn excluded. Waves of migrant workers from Yangjiang in southern Guangdong, from Guangxi and Hunan provinces, have become the new underclass of the sands. The verdicts of criminal cases that came before the people's courts in the last few years show that many of the victims of extortion, robbery, rape, and murder in the sands are these recent migrant workers. History repeats itself; the recent migrants take up the social space formerly occupied by the people of the sands.

The increased mobility of the rural population in general has created a dilemma for the earlier families who want to maintain their rights in the sands. They complain that it is increasingly difficult to find brides for their sons because local women, just like the men, prefer factory employment near Nanxi town, with the added intention of finding spouses there. Families are reluctant to let their grown daughters be married off too soon, and when they take in a daughter-in-law with an ever-increasing brideprice, they try to keep her at home.

It is true that where there is ample land and where agriculture continues to rely on manual labor, labor becomes a prized commodity. The crucial issue is why, given the availability of cheap migrant labor, families in the sands choose to pay high brideprices in an effort to keep their women at

[48] In the villages closer to Nanxi town, turf is grown to supply the offices and hotels in the coastal cities and in Hong Kong. In the farther corners of the sands, banana plantations are common. They also produce for both the internal and the export markets.

home. Today settlement rights in the sands are highly valued, especially for the older generation, who were discriminated against before and who finally feel that they have gained a productive foothold in the 1980s. The "continuation of the house," as Goody describes it, made possible by commitments from the parent's generation toward the new conjugal units, is the culturally acknowledged basis for claiming these settlement rights. It also secures for the family the much-needed long-term farm labor. Furthermore, the demand for a high brideprice distinguishes these families from the new underclass of migrant workers, who cannot afford such payments and would not be able to make such a demand. As before, marital strategies serve a political function by excluding those at the lowest end of the social hierarchy.

Conclusion

The analysis of marital payments in this chapter takes into account the history of political transformation of the regional economy. The resources available to families and the way they choose to maneuver within the constraints imposed by the state are intertwined with this political history, which shapes the available options as well as the perceptions of them. In general, marital transfers in Nanxi Zhen have to do with strategies of family continuity and advancement. In the early twentieth century, these transfers took place in a highly stratified socioeconomic context. Apart from endowing daughters, the giving of a lavish dowry in town was a strategy for building prestige among the elites; for the commoners, it was more a way to distinguish themselves from the sands people. The inhabitants of the sands and the town had regarded themselves as worlds apart precisely because their lives were connected in multiple and unequal ways to an evolving regional economy. There were fierce contestations in material relationships and symbolic arenas. Ethnic labels or terms of marriage among families were political statements in this expanding part of the delta. As in dialogues on lineage, community, and settlement history, they were means of exclusion, differentiation, and acculturation among the local population as they sought their respective places in the emergent polity.

It would be naive to expect the "equalizing" measures of the socialist regime to have reduced the inequality between Nanxi town and the sands. There were less status competition and fewer prestige-building strategies among former elite families in town who were condemned to be a caste of untouchables. This process might explain the drastic drop in the dowries in town. But the life chances of the residents of town and sands continued to differ. Their respective positions within the socialist state reinforced and even enlarged the town dweller's traditional discriminations against the people of the sands. Unlike the urban families in the large cities as de-

scribed by Whyte in chapter 8 in this volume, whose livelihoods have been based on an established state sector, these families lived in a small market town dependent on a shaky collective economy near the margin of the rural collective sector. Their cultural strategies to distinguish themselves from the people of the sands remained significant, although the strategies were subdued by a general lowering of living standards of the elites who defined them, and although the motives behind the strategies were based on the realities of a different rural-urban hierarchy.

It would be equally naive to assume that collectivization from the late 1950s to the late 1970s had dwarfed concerns for inheritance and for the continuation of the family, especially in view of the fact that the family has once again become a focal point of social life in the 1980s, which has in turn led to a revival of marital transfers. Interdependence among family members remained strong if not intensified in the Maoist era for both the town dwellers and the villagers; hence the continual negotiation of marital payments despite the attempts of the government to restrict and undercut them. This "preservation" of family commitment was a way for people to cope with the transformed relationship of state and society, another unintended consequence of the policies in the Maoist era.

Family dynamics and the related terms of marriage today cannot be seen as a restoration of what had been put on hold since the late 1950s. Instead, as Whyte observes in Chengdu and Huang in rural Shangdong, they resulted from several contradictory trends that had changed the meaning of the family at one level and reinforced it at another. At the level of the institution of the family vis-à-vis the larger society and the state, relationships have changed because other mediating aspects of the civil society had been stripped away. At the level of commitments among family members, the concerns for continuity have remained strong precisely because other institutions were redefined by political ideologies and were no longer viewed as reliable for the individual.

The economic opportunities of today, which have been created by a decade of liberalization, pose a new social reality for the residents in Nanxi town as well as for those in the sands. The opportunities allow young urbanites to create viable conjugal funds on their own initiative by means of exaggerated dowries and brideprices, largely self-accumulated. Their efforts dovetail those of their parents, who are eager to gain some old age security in a time of uncertainty and who contribute with this end in view. Rural youths, both men and women, are anxious to shed their peasant status and eagerly fill the gaps at the lower end of the town's employment hierarchy, an opportunity that was denied them for over three decades. For the older generation, with the drastic loss of their young at a time when agriculture becomes profitable, settlement rights reemerge as an important issue. Their concern is often not shared by their children. Mustering the cultural re-

sources at their disposal, they try to control marriages for reproduction and for continued commitment to the family. By maneuvering marriage payments, they hope to tie women to the household economy. This new concern increases the demand for a large brideprice in the sands. The "success" of these strategies pursued by the older generation remains to be seen. With the productive links between the rural generations broken by migration and urban employment, and in town by increased entrepreneurial opportunities for the young, generational conflicts are numerous and explicit. Prospects of retaining the young are not high, although some may be persuaded to stay.

In sum, factors leading to different payments of brideprice and dowry are complex. They involve intergenerational dependencies within the family, as well as the fortunes of individual family members in relation to the larger society and state. From the early twentieth century to the post-Mao era, three crucial political turning points—the war interlude, the Maoist era, and the post-Mao liberalization—have continued to create different life chances for the residents of the town and the sands and have triggered divergent strategies. In analyzing these terms of marital transfer in Nanxi town and the sands through this time, I hope I have given a historically grounded and meaning-focused account of the ways the transformations of political economy intertwine with the cultural, symbolic resources people use to make sense of their lives.

Eight
Wedding Behavior and Family Strategies in Chengdu

Martin King Whyte

Marriage decisions are central to any discussion of family strategies. Normally, marriage is a central step in the age-old Chinese effort to continue the family line.^[1] It should also be obvious that not just any marriage will do. Which individuals sons or daughters marry may have important consequences for the future prospects not only of those sons and daughters, but also of the larger families in which they grew up. Which partners are chosen and how the wedding is celebrated may affect the social status and links to useful patrons and allies of those involved. And whether particular marriages result in an expansion of an existing family (by the couple's moving in with what I call an "old family") or the setting up of an independent household (which I call, for the sake of simplicity, a "new family") may influence the opportunities Chinese have to succeed in the world that confronts them. For these and other reasons it is obvious that marriage is a central focus of family strategic thinking. In this chapter I analyze survey data from one Chinese city in an effort to gain insight into how family strategic behavior in regard to getting married may have changed over time, and particularly in the post-Mao or reform period (i.e., since 1978).

With colleagues at Sichuan University I collected survey data in Chengdu, the capital of Sichuan province, in 1987.^[2] Our interviewees were a probability sample of 586 ever-married women between the ages of twenty

[1] Concubinage, adoption, and other steps may be taken to this end, but for most Chinese, and particularly since 1949, marriage has been an essential prerequisite to family continuity.

[2] My Chinese collaborators are Yuan Yayu and Xu Xiaohe, both affiliated with the Sociology Research Office of Sichuan University. Our research was supported by the U.S.-China Cooperative Science Program of the National Science Foundation and also by the United Board for Christian Higher Education in Asia, Sichuan University, and China's State Education Commission.

and seventy who lived in the two main urban districts of that city. Since we asked our respondents a variety of questions about how they first got married, and since the dates of those first marriages ranged over the fiftyfive years from 1933 to 1987, we are able to say quite a bit about how the marriage experiences of these women have varied across time periods.^[3] Given a particular interest in the reform period, we can examine whether women in Chengdu who married after 1978 did things differently than women who married in earlier periods.

Here I intend to go beyond simply examining whether wedding behavior has changed over time in Chengdu. A number of more complex questions are of interest and may tell us something about shifts in family strategies. For example, how did social groups within Chengdu differ in their wedding behavior? Which types of individuals or families were most likely to have experienced arranged marriages, to have had elaborate weddings, or to have set up new households after marrying? Have the "social correlates" of different kinds of wedding behavior changed over time, and particularly in the post-1978 period? Is there any sign in our data, for instance, of a group of nouveaux riches families arising as a result of the reforms and distinguishing themselves by holding particularly elaborate weddings? Did families of Party members "hold the line" against the revival of elaborate weddings in the reform period, or did such families lead the way in this trend? These are ad hoc examples of the kinds of questions we can address with our Chengdu data. Since much evidence on how wedding behavior has changed over time has already been presented,^[4] after reviewing that evidence briefly here, my main focus will be on the social sources of variation in wedding behavior in Chengdu and how those sources may have changed over time.

I should clarify at the outset that, although like my fellow authors I am interested in family strategies, the evidence I have to work with—the

[3] The response rate for the survey was 87.7%. Chengdu is, of course, not "typical" of all of urban China, but it is not especially unusual in comparison with other major cities in China. (Chengdu is China's tenth largest city, with an estimated population of 2.6 million in 1985.) Inferring change over time from this kind of cross-sectional sample is, of course, not an easy matter. Many of the older women in our sample were not living in Chengdu at the time they first married, and many of the women who were getting married in that city in the 1930s and 1940s have either died or moved elsewhere. So the data presented here cannot, properly speaking, tell us how the wedding behavior of representative members of the Chengdu population varied across time periods. Instead, they should be interpreted as telling us how the wedding experiences of a representative sample of women living in Chengdu in 1987 differed depending upon the time at which they married.

[4] See Martin King Whyte, "Changes in Mate Choice in Chengdu," in Chinese Society on the Eve of Tiananmen , ed. Deborah Davis and Ezra Vogel, 181-213 (Cambridge: Harvard University Press, 1990).

Chengdu survey data—does not contain any direct information about such strategies. We asked respondents questions about such things as who made the decision about whom to marry, how much was spent on the wedding, who paid for the celebration, and where the couple lived after the wedding. We did not, however, ask any questions about why things had been done in a certain way or about how the decisions were arrived at and with what family goals and concerns in mind. To address questions such as these properly would require a different research technique, involving intensive, in-depth interviews rather than a roughly one-hour-questionnaire survey. Even if we had used this alternative technique we would have faced difficulties, since the older respondents in our sample might have difficulty remembering and providing accurate reconstructions of family strategic thinking three or more decades earlier, and many of them (particularly those who experienced strictly arranged marriages) were not consulted at the time.

In the absence of direct information about family strategic thinking, I am forced to rely on inference here. My approach is to look for overall patterns in the data to see what those patterns might indicate about family strategies and changes in such strategies. I contend that if family strategies have been operating with any degree of success, we should see evidence of the operation of such strategies in our data on wedding behavior. If we do not find such interpretable patterns in our data, then we have evidence either that family strategies were not successful or that they did not exist in a coherent way in the first place. Inferring family strategies from such patterns in the data is not an easy or exact process.

Changes in Mate Choice in Chengdu

How did wedding behavior change over time in Chengdu, and what do any trends tell us about changing families and family strategies in that city? In dealing with these questions, I focus on aspects of wedding behavior that tell something about changes in intergenerational relations.^[5] This focus is important because it may help us deal with a central question in the study of family strategies: What is the proper unit of analysis? To put the question in a different way, when we want to study the formulation and negotiation of strategies for success in urban China, does it make most sense to look at individuals, at what I call new families, or perhaps at old families? Who, after all, is involved in these kinds of decisions? And might it not be that the locus of strategic thinking in regard to marriage has changed in

[5] A more general examination of changes in a broad range of mate choice and wedding customs has already been published. See Whyte, "Changes in Mate Choice in Chengdu."

fundamental ways over time, from parents and family elders (the leaders of old families) to potential brides and grooms (those who will form the new families)? By examining what wedding behavior tells about shifts in relative power, social ties, and interdependency between generations, we may be able to draw inferences about basic changes in the nature and locus of family strategic thinking in contemporary China.

There are good reasons for assuming in advance that basic changes in the nature of intergenerational relations may have occurred in twentieth-century China. All that we know about the nature of marriage in late imperial China suggests that for most this was a relatively extreme form of arranged marriage in which the old family was the locus of power and the new family had little if any voice. Parents or other senior kin decided when it was time for children to marry and whom they should marry, and in many cases the young couple was not consulted and did not even meet until the day of the wedding. Sometimes a child betrothal or adoption of a future daughter-in-law as an infant was involved, making even clearer the total subordination of the younger generation to the wishes of their elders. Negotiations over brideprices and dowries and the financing of a wedding feast were also matters of family decision, rather than issues left for the young couple. Usually there was an expectation that the new family would be incorporated into the old through patrilocal residence, but even if not, the decision about where to live and how to earn a living were often matters determined more by the old family than the new. In some accounts from earlier times we learn of families making decisions to have one son study for the imperial examinations, another remain in the household and take over the family farm, and another go off to work in a market town or city, all in an effort to maximize chances for family success.^[6] The power of the older generation and the role of that generation in making decisions about family strategy were reinforced and expressed in multiple ways in marriage behavior in late imperial China.

Changes that weakened the power of the old family in marriage matters were already under way before the Chinese Communist Party (CCP) came to national power in 1949. Both growing industrialization and commercialization and the spread of Western ideas about marriage and family life fostered trends toward greater independence of the young in the first half of

[6] This scenario is somewhat idealized and would have applied mostly to families already well above the margin of subsistence. But in general a picture of the old family fairly explicitly dictating the lives of the young is a commonplace in descriptions of social life and social mobility in earlier times. See, for example, Ho Ping-ti, The Ladder of Success in Imperial China (New York: Science Editions, 1964); Yung-teh Chow, Social Mobility in China (New York: Atherton Press, 1966); Lin Yueh-hwa, The Golden Wing (London: Kegan Paul, 1947). The best-known fictional portrayal of the conflict over the power of the elders to control the fate of young family members is Pa Chin, Family (Garden City: Anchor Books, 1972 [originally 1933]).

the twentieth century.^[7] After 1949 the CCP used its impressive power to campaign against arranged marriage and other manifestations of parental power. Both economic development and the socialist transformation of property relations carried out by the CCP should have weakened further the power and resources parents could wield in their efforts to dictate marriage matters.^[8] Further economic development and the return of Western cultural influence during the reform period could be expected to accentuate these trends even more. How does the evidence from Chengdu fit these expectations?

The first aspect of marriage behavior to consider is the degree of freedom of mate choice—whether parents (or other family elders) or the young couple play the dominant role in deciding whom to marry. In this realm there have been clear changes, as the data presented in table 8.1 reveal. (This table and table 8.2 divide years of first marriage into periods that correspond roughly to important turning points in the history of contemporary China.) In general, there has been a shift from the old family to the new family in terms of who makes the selection of marital partners. (Other trends shown in the table have already been discussed in previous writings and will not be focused on here.) A more fine-grained analysis, making use of a summary scale and finer time divisions, reveals a somewhat more complex pattern. In figure 8.1 I display a three-year moving average of a summary Freedom of Mate Choice Scale across years of marriage.^[9]

Figure 8.1 confirms that the change toward increasing freedom of mate choice was already under way prior to 1949 and that this trend continued during the 1950s. However, the figure also reveals that after about 1957 this shift more or less "stalled," with little net change toward greater freedom of mate choice during subsequent periods. Neither the storms of the Cultural Revolution nor the frenetic changes of the reform period had much effect on the freedom of young people in Chengdu to choose their own mates. To summarize briefly, I argue that the 1950s involved a massive structural change—the socialist transformation of the urban economy. That change

[7] These trends are documented in Olga Lang, Chinese Family and Society (New Haven: Yale University Press, 1946). See also Marion Levy, The Family Revolution in Modern China (Cambridge: Harvard University Press, 1949).

[8] The classic work on how various aspects of "modernization" produce changes that give young couples more autonomy is William J. Goode, World Revolution and Family Patterns (New York: Free Press, 1963).

[9] The summary Freedom of Mate Choice Scale was constructed from the mean of the standardized scores of each respondent on six questions—those numbered 1, 2, 3, 6, 7, and 9 in table 8.1. A three-year moving average is constructed by taking the average of the scores for the years before and after a particular year. For example, the value for 1957 is an average of the values for 1956, 1957, and 1958. This procedure is followed to compensate for modest numbers of cases in each marriage year, and the result is a smoother and more readily interpretable graph.

reduced the power of parents dramatically, since they no longer controlled property or access to adult jobs, but urban young people gained increased autonomy only up to a point. The real winner was the bureaucratic system constructed by the CCP, which controlled education, employment, health care, housing, and all other important resources. Young people found that some of the freedom gained from parents was lost to the bureaucratic gatekeepers of the state (in schools, factories, etc.), individuals who were decidedly not indifferent to when and how young people married.^[10]

[10] This central role of the bureaucratic state is the decisive element left out of Goode's account of modernization and family change in capitalist countries. In the capitalist world, both governments and employers and other bureaucratic gatekeepers are largely indifferent to the private lives of the young, so that power lost by parents is gained by the young.

Fig. 8.1.
Freedom of Mate Choice, Moving Average

A variety of measures of wedding celebrations and wedding finances are displayed in table 8.2. Also included in that table are responses to a question about where the couple lived after the wedding. There are dramatic changes across time visible in these figures, but the pattern is quite different from the one seen in table 8.1 and figure 8.1. Broadly speaking, what we see here is a curvilinear trend. Initially there are sharp shifts away from "traditional" patterns. By the late 1950s and 1960s fewer couples were having wedding ceremonies and banquets, fewer people were being invited to such events when they were held, marriage finance transactions (e.g., brideprices, dowries) were less common, less money was being spent, more often the bride and groom were paying for any festivities themselves, and it was increasingly common for the new family to reside separately from either his or her parents.^[11] In more recent times, however, all of these trends have been reversed. By the final period, the era of the reforms, wedding ceremonies, and banquets are once again increasingly common, more people are being invited to them, more is being spent on them, and it is less and less common for the young couple to finance their weddings themselves. By the same token, postmarital residence in a new household has become less common during the reform period, with moving in with the groom's or the bride's parents (most commonly still the former) once again on the rise (see also the figures presented in chapter 2 by Unger).^[12] So in these realms we see not a stalling of a pattern of change, but an apparent reversal and return toward more "traditional" patterns.^[13]

[11] A variety of other aspects of change in wedding customs are not displayed in table 8.2. In general many couples who married between the 1950s and 1970s in urban China followed the models espoused by the CCP by holding a simple celebration with close friends and workmates in the work unit, with tea and candy distributed to participants. In a few cases these "revolutionary" weddings included explicit political statements, such as a speech by the Party secretary encouraging devotion to the revolution, presentation of a set of Mao's writings to the couple, or having the couple bow to a bust or portrait of Mao Zedong. The role of the family is obviously strikingly deemphasized in such wedding ceremonies, which contrast very clearly with the traditional wedding, in which the family-run feast is the central event. The spartan nature of these "revolutionary weddings" helps explain why so much less was spent and why couples could finance things themselves. It might be noted that we asked a question about another officially espoused model, the "collective wedding," in which large numbers of couples are married at a single time in a public ceremony presided over by CCP officials. We found very few cases in our sample of such collective weddings in any time period.

[12] Richard Barrett (personal communication) has found evidence from the 1985 In Depth Fertility Surveys that shows a similar trend of neolocal residence declining and patrilocal residence increasing during the 1980s in urban areas of several provinces (Beijing, Shandong, Liaoning, and Guangdong).

[13] There are a number of "new" elements also visible in Chengdu wedding customs in the 1980s. For example, wedding banquets in restaurants are increasingly common, something that fits the Hong Kong more than the traditional Chinese model. Similarly, the wearing of a Western-style white wedding gown for a formal wedding picture is increasingly the norm, and more and more couples are taking honeymoon trips. My claim is not that all aspects of wedding customs are reverting to traditional forms, but that those aspects of wedding celebrations that have implications for intergenerational relations are moving in that direction.

Fig. 8.2.
Wedding Elaborateness, Moving Average

As earlier in regard to freedom of mate choice, we can examine whether a more fine-grained analysis will reveal further details about these trends in wedding elaborateness and expenditures. I constructed a mean scale of wedding elaborateness from a variety of the available measures, and in figure 8.2 I graph a moving average of scores on that scale against years of marriage.^[14] Figure 8.2 reinforces the conclusions already reached about a curvilinear time trend, but adds a few extra details. According to this measure, the elaborateness or costliness of weddings in the reform period appears to be even greater than in the prerevolutionary period. Also, the decline in wedding expenditures is visible even before 1949, whereas the revival of wedding costs begins prior to Mao's death, rather than after 1978.

Put very simply, I think a variety of forces are at work in these reversals of wedding behavior. In part we see reflected here a return to something closer to normalcy after the Cultural Revolution, as lessened CCP monitoring of private life made it possible for families once again to follow cherished practices in celebrating weddings. Also at work, at least after 1978, is the new official encouragement of conspicuous consumption and the rapidly increasing incomes of urbanites, factors that made possible and respectable the sorts of lavish weddings that would have gotten people into trouble only a few years previously. (On this score the CCP continued to advocate spartan wedding celebrations while saying it was good to get rich and that envy provided a powerful incentive for economic improvement. Encouraging restaurants, photography shops, and other service facilities to cater to popular needs and make profits also helped to fuel the revived "wedding trade.")

Even though the post-1978 improvements in consumption standards applied to housing as well as to food, clothing, and other areas, the increases in urban housing space were not translated into improved chances for young couples to get a place of their own. My interpretation of this apparently contradictory finding, admittedly speculative, is that in the context of housing demands suppressed for so long, with many older people with stalwart work and political records waiting on housing lists for years, newly marrying couples had less chance than in previous periods of getting allocated housing on their own. By continuing to live with either her parents or his they might be able to move into new housing as it became

[14] Included in the scale were seven items: those numbered 1, 2, 3, 4, 7, and 8 from table 8.2 and responses to a question about the monetary value of the dowry items (with zero used in cases with no wedding, banquet, or dowry). As with the Freedom of Mate Choice Scale earlier, this scale was formed by taking an average of standardized scores of each respondent on these seven items. This scale can be interpreted as measuring the elaborateness and costliness of the proceedings. Note that the postmarital residence measure is not part of this scale.

available during the 1980s.^[15] (See the related discussion in the chapters by Davis, Unger, and Ikels.) In any case, in these realms we see a fairly consistent pattern that constitutes a reversal of the patterns that occurred in earlier years.

How are we to interpret the evidence presented so far? What do the data reviewed here tell us about shifting patterns in the relationship between generations in Chengdu families? What we do not see in this evidence is a pattern of "creeping individualism," of continuing growth in youth autonomy. Nor is the expectation of a linear trend toward more "modern" weddings borne out. Young people seem to have gained some power relative to their elders, at least when it comes to deciding whom to marry. However, their ability to live separately and rely on their own resources has actually diminished in recent years. We do not know, unfortunately, how in recent marriages it was decided how much to spend, how many to invite to the wedding, or where the couple would live afterward. But we can see that, even though the dominant role in choosing whom to marry is now in most cases played by "new families," in order to actually complete the process they have to rely on their "old families" in multiple ways. This continued and, indeed, revived intergenerational dependency may not mean that family elders are once again dictating matters to the young, but it would seem to require some sort of negotiation between generations over the nuts and bolts of how the wedding will be conducted and where the young couple will start out married life (a point also made by Davis in chapter 3). These results, taken as a whole, qualify my earlier discussion. This is not a situation in which powerless old families look on helplessly as the lives of their children are determined by an omnipotent and inflexible bureaucracy. Instead, continued family involvement and support appear to be vital to survival in the bureaucratic environment that is urban China. The trend toward increased youth autonomy from earlier in this century has been halted and in some respects even reversed.

Social Background and Wedding Behavior

Now I turn to the questions of how subgroups within the Chengdu population differed in their wedding behavior and whether the sources of variation in such behavior changed across time periods. In the pages that follow I review a wide range of evidence bearing on these questions by examining the correlations between various background traits of our respondents (and

[15] Figuring out how the housing supply affects patterns of postmarital residence is tricky. Young couples might end up with a place of their own for quite contradictory reasons—either because housing is so short that there is no space for them to move in with either set of parents or because housing in the city is so ample that young people can obtain their own place.

their old families) and their actual wedding experiences in various time periods.^[16] From the large number of measures available from the Chengdu survey, I selected a limited range of important measures to examine, and consider only five basic aspects of wedding behavior: the degree of freedom of mate choice, the elaborateness of the wedding, whether or not there was a dowry presented by the bride's parents,^[17] how much of the total wedding expenditures were borne by the young couple themselves, and whether the couple set up a new residence after the wedding. In terms of social background characteristics, I use a select subset of the available measures of three types: background characteristics of the individual bride and groom, background characteristics of their "old families," and "other" characteristics (i.e., traits that are not social status traits either of the bride and groom or of their families—for example, the year of the wedding, the age of the bride at the time, etc.).

Displayed separately here, the correlations between background characteristics and wedding behavior in three broad time periods—1933-57, 1958-77, and 1978-87, are collapsed further from the five categories used in previous tables. Each of these time periods has enough women in it (178, 198, and 210, respectively) to make analysis of correlations possible, but the main reason for these divisions is substantive. The first period involves the prerevolutionary era as well as the years after the CCP came to power but before the new socialist system was fully consolidated. The middle time period might be regarded as the years of full socialism or even "high Maoism"—they begin with the year in which Mao Zedong's visionary Great Leap Forward was launched and include the Cultural Revolution and its aftermath. The final time period is, of course, the era of the reforms initiated under the leadership of Deng Xiaoping.

What can we expect to learn by comparing correlations between social background traits and wedding behavior across these time periods? Of

[16] The correlation is not the most appropriate statistic to use in examining many of these relationships, since most of our variables are ordinal or even nominal, rather than interval. However, most nominal variables can be rearranged to make them ordinal (for example, with occupations ranging from low status to high), and the use of interval statistics, such as the Pearson product-moment correlation, to summarize findings is claimed by most experts to be an acceptable shorthand that in most cases leads to similar conclusions as relying on cross-tabulations and ordinal statistics. (A large number of cross-tabulations were examined as a preliminary to preparing table 8.3, and that examination confirmed this impression.) Also, full tables on so many variables would be completely unwieldy.

[17] The dowry is included as a separate measure even though this is one component part of the Wedding Elaborateness Scale. The reason for this "double counting" is that much previous literature suggests a particular importance for dowry exchanges in reflecting such things as continuing ties with the woman's natal kin and competition for high status. See the discussion in Rubie Watson and Patricia Ebrey, eds., Marriage and Inequality in Chinese Society (Berkeley: University of California Press, 1991).

course, we can consider a wide range of specific predictions about whether well-educated individuals, families with bad class backgrounds, or women who marry late (for example) had different types of weddings than others, and in all periods or only in one or two. However, my main interest in presenting the figures that follow is not to consider such specific predictions, but to look for overall patterns, and for any variation in those patterns across time periods.

To oversimplify somewhat: If we consider that our first broad time period captures marriages that occurred under basically presocialist conditions, then we can see what types of individuals and families had the most freedom of mate choice, the most elaborate weddings, and so forth, under those conditions. (We can also consider whether the distinctions found are similar to those uncovered in research on other nonsocialist societies.) If we designate the second period as the full socialist or high-Maoism period, then we might expect the predictors of variation in wedding behavior to have been somewhat different in those years (in comparison both with the pre-1957 period and with other societies). For example, the heightened emphasis on socialist values and vastly expanded bureaucratic control over people's lives under "high Maoism" might have produced a situation in which education and occupational status lost their predictive power in regard to weddings, but Party membership and class-origin labels gained predictive power.

If we find some clear differences between these first two time periods in the overall pattern of differentiation in wedding behavior, then we have a template against which to judge things in the final, reform period. Have the reforms sufficiently altered urban social structure in China and restored markets, material incentives, and meritocratic competition for personal advantage so that those social background traits that affected wedding behavior in the pre-1957 years reemerge as important after 1978? Or does the pattern of differentiation in wedding behavior in the reform period look much like that under high Maoism, suggesting that the dynamics of urban life have not changed as much under the reforms as many people suppose? Or is there some sort of completely new pattern of association between social background traits and weddings emerging in the reform era, unlike either the presocialist or high-Maoism eras? These are the sorts of ideas that motivated the analysis that follows.

Because the analysis of wedding behavior is spread over tables 8.3-8.7 (one table for each aspect of behavior considered), it may appear difficult to notice general patterns.^[18] However, by looking across all five tables, the

[18] Additional complexities concern how the various measures were constructed. Briefly, the Freedom of Mate Choice and Wedding Elaborateness Scales are interval measures whose construction was described earlier in the chapter; the dowry measure is a simple dichotomy (see no. 7 in table 8.2), the measure of the couple's contribution is a 7-point measure from 0 to 100% (more detailed than is shown in table 8.2), and the postmarital residence measure is again a simple dichotomy of all other arrangements vs. neolocal (i.e., less detailed than the measure used in table 8.2). The various education measures are all 5-point ordinal measures ranging from illiterate to at least some college; the Party membership measure is actually 1 = masses, 2 = Communist Youth League, 3 = CCP; the occupation measures are 1 = industrial worker, 2 = service worker, miscellaneous, 3 = cadre or professional occupation; the parent urban origin measure is 1 = neither parent an urbanite, 2 = one parent an urbanite, 3 = both parents urbanites; the family class origin measure is 1 = landlord or capitalist, 2 = peasant, 3 = worker, 4 = poor urbanite, handicrafts, etc., 5 = cadre, employee, professional; and the parental class label measure is 1 = landlord, capitalist, rich peasant, 2 = petty bourgeois, 3 = peasant, 4 = worker, 5 = cadre or martyr.

main pattern I want to draw attention to should be clearly visible. Generally speaking, there are a number of fairly consistent background correlates of wedding behavior of those women who married during the years 1933-57 (see the first column in each table). In that period, it was common for brides and grooms who were well educated and who had well-educated par-