State-Sponsored Inequality
The Banner System and Social Stratification in Northeast China
Shuang Chen

BUY THIS BOOK


Chapter One

Social Formation Under State Domination in Modern China

Eleven days after the Chinese New Year, in 1824, a procession of fifty-three wagons left Beijing, the capital of the Qing dynasty (1644–1911), and headed northward to a remote settlement in Manchuria1 called Shuangcheng (map 1.1).2 Riding in the wagons were sixty men, fifty-four women, and seventy-two children—a total of fifty-three households.3 These travelers had begun this journey because the government had told them that they were returning to their ancestral homeland, where fertile land, clean and spacious houses, and assistance in farming the land awaited them. Fifteen days later, they went through the Shanhaiguan Pass—the easternmost pass on the Great Wall—and entered Manchuria. Escorted by local officials, they traveled farther north, taking another sixteen days to cross the border into Jilin. After another ten days, they finally arrived at their destination. These fifty-three households were the first Beijing pioneers to settle in Shuangcheng, but others soon followed.

These settlers from Beijing would become the state-designated elites in Shuangcheng. They were descendants of the warriors who had helped the Manchu rulers of the Qing conquer China proper in 1644. As early followers of the emperor, those warriors had been organized under a system called the Eight Banners, and were referred to as bannermen. Because of this distinguished status, they and their descendants were to serve the state as soldiers and receive stipends from the state in perpetuity. But in the early nineteenth century, amid a fiscal crisis surrounding the support of the banner population, the state decided to relocate the bannermen living in Beijing back to Manchuria, substituting state land for the state stipends. Over a period of two decades, a total of 698 such banner households arrived in Shuangcheng. To help them settle in their new homes, the state, between 1815 and 1820, also relocated three thousand households of bannermen from elsewhere in Manchuria (map 1.1) to Shuangcheng. The government’s settlement of the banner immigrants triggered private migration into the area as well. By the 1860s, a total of 5,300 registered households had settled in Shuangcheng,4 establishing a rural society divided into two segments: the haves—that is, the jingqi, or “metropolitan bannermen,” from Beijing and Rehe and the tunding, or “rural bannermen,” from other parts of Manchuria, who were supported by the state with land grants—and the have-nots, that is, the fuding, or “floating bannermen,” and minren, or civilian commoners, who had moved into the area without the state’s permission and thus did not receive land allocations. Among the haves, the metropolitan bannermen received land grants that were twice as large as those of the rural bannermen. These asymmetrical entitlements continued until 1906, by which time Shuangcheng had become a county with more than sixty thousand households, containing 440,000 people (SCXZ 1990, 829).5 This legacy of social segmentation persisted in Shuangcheng far beyond the fall of the Qing in 1911.

Map 1.1. Sending communities of banner immigrants to Shuangcheng.
SOURCES: Information of the sending communities is drawn from the CMGPD-SC. Location of the Willow Palisade is from Edmonds (1979, 602, fig. 1). Data for provincial boundaries and rivers are from “Datasets for 1820,” CHGIS, Version 4, January 2007.

This book explores the social and economic processes of inequality under a state-dominated system by providing a holistic, comprehensive view of the formation of the social hierarchy in Shuangcheng in the nineteenth and early twentieth centuries. The state created a social hierarchy from the top down by classifying people into distinct categories, each associated with differentiated entitlements to land. Under this system, the state directly intervened in wealth distribution and the determination of people’s social status in order to fulfill its administrative goal: to maintain the elite status of the metropolitan bannermen. Eventually, the state-designated social hierarchy played out on the ground at the intersection of state policies and local practices.

As a well-documented case, the Shuangcheng story offers a historical perspective on the phenomenon of states using resource allocation to create structural inequality. Scholars who study social stratification pay increasing attention to the persistence of structural inequality in modern societies (Baron and Bielby 1980; Blau 1994; Diprete et al. 1997; Tilly 1998). In his thesis of “durable inequality,” Charles Tilly (1998, 1–40) points out that inequalities measured by acquired individual attributes that range along a scale from low to high—such as income level or education level—are open, fluid, and easy to change. However, when ascriptive characteristics, such as gender and skin color, become the basis of social differentiation and access to resources, the inequalities arising along these categorical boundaries interact with acquired individual attributes to become “durable” (ibid.). This is because boundaries between categories based on ascriptive characteristics are hard to change. In his study of the rising inequality in post-socialist China, Wang Feng (2008) reveals that the major sources of social inequality are the various categories created by the socialist state, based on household category, ownership type, industrial sector, and geographical location. The durability of structural inequality in a society raises two questions: Among the various ascriptive characteristics of an individual, how do certain ones become the basis of social differentiation? And, how are the boundaries corresponding to these characteristics created and maintained? The study of the Shuangcheng case shines a historical light on the role of one important actor—the state—in creating such boundaries, and on the social construction of these boundaries in an agrarian society.

This book focuses on structural inequality created by the state because such inequality is not only the foundation of the social hierarchy in nineteenth-century Shuangcheng but also a feature of the Chinese society after the foundation of the People’s Republic of China ([PRC]; 1949–present). Among the many categories created by the socialist state, those created by the hukou household registration system most resemble the population categories in Shuangcheng. With its goal of industrialization, the PRC, established the hukou system and divided the population of the entire country into the “rural household” category or the “urban household” category (Wang 2005). This system has created nationwide inequality between the two household categories, with the urban households enjoying better entitlements to state employment and the associated benefits, such as housing and education. Numerous studies have highlighted the profound impact of the hukou system on social inequality in China and its connection with the socialist revolution (Solinger 1999; Wu and Treiman 2004; Wang 2005; Whyte 2010; Brown 2012). Yet people rarely recognize that the state-sponsored inequality in post-1949 China was not a socialist extension but a statecraft that had existed in history.6 The Eight Banners registration system in Qing-dynasty Shuangcheng has many parallels with the hukou system in the PRC. Placing this phenomenon in the broader perspective of state-building and its social consequences renders an in-depth understanding of state-sponsored inequality in the past and present.

In a broad sense, both the state-dominated system in Shuangcheng and the hukou system in post-1949 China are state-initiated projects of social engineering, in which the state implemented policies to proactively design or plan the social order. The efforts to penetrate and control society go hand in hand with the emergence of the state. From time to time, these efforts culminate in projects of social engineering. In the twentieth century, government social-engineering projects garnered considerable attention because they were often carried out in an entire country and impacted the society on a massive scale. James Scott (1998) examines some of these large-scale social-engineering projects, such as Soviet collectivization and the Tanzanian forced villagization. He points out that the ways these projects were carried out are closely associated with the political system and ideology of the respective state; all the large-scale projects he studies were carried out by an authoritarian state with a “high-modernist ideology”: “a strong version of self-confidence based on the development of science and technology and the expansion of production” (ibid., 4). Yet little attention has been given to social-engineering projects undertaken in the past. Although the Shuangcheng settlement and the hukou system were independent projects carried out under different ideologies, both grew out of a state belief in using resource allocation to create a social hierarchy among the population, and both were backed up by a political system that enabled the state to implement such policies. These parallels connect socialist China to its imperial past. By reconstructing the story of social formation under state domination in nineteenth-century Shuangcheng, this book offers a unique perspective on how historians and social scientists typically conceive modern phenomena.

In addition to the implications for the social stratification system in contemporary China, the Shuangcheng case also offers a special setting for the examination of some key issues in nineteenth-century China. The state-dominated system in Shuangcheng was special in two respects. First, the Eight Banners institutionally distinguished Shuangcheng from many other local societies of the period. Although the Qing institutionalized the bannermen and created the Shuangcheng society from top down, it had only limited reach into local civilian society. For most of the Qing dynasty, the state did not establish formal government institutions below the county level. Only in the early twentieth century, the last decade of Qing dynasty, did the state begin to build government offices at the district level. State landownership and the level of state control in Shuangcheng were only seen in the Eight Banner farms in Manchuria, the manor lands in the areas surrounding Beijing, and Xinjiang (Wang, Liu, and Guo 1991; Diao 1993; Diao and Yi 1994; Guo 1997; Qiu 2014). Second, the ability of the state to carry out the Shuangcheng settlement and to design the immigrant society counters our general understanding of Qing history in this period. The received wisdom about nineteenth-century China portrays it as an age marked by the devolution of political power, a vicious cycle brought about by endless internal rebellions and foreign intrusions. At the societal level, the numbers of local elites greatly expanded, and they became more and more important in organizing local society.7 Thus, the Shuangcheng society seems on the surface to be very different from other local societies. However, as this book reveals, Shuangcheng was still part of the China of the nineteenth century and experienced all the social and political changes of this period. Most importantly, banner officials in Shuangcheng adopted a style of local governance similar to that of the civilian system elsewhere in China. Therefore, despite the institutional difference, the ways in which the Shuangcheng immigrants responded to state policies, organized themselves, and carried out daily activities can serve as references to social behaviors of the time. The basic social dynamics—economic and demographic differentiation, social mobility, and social reproduction at the local level—are informative in understanding rural communities in nineteenth-century China.

By showing how the interaction of state policies, local politics, and customary practices gave rise to disparities in socioeconomic status among the Shuangcheng immigrants, the book illustrates in vivid detail the transformation of banner society in Manchuria in the late Qing. As the institutional foundation for Manchu’s rule of China, the development of the Eight Banners has been key to understanding the history of the Qing. The political, socioeconomic, and cultural status of bannermen are important indicators of the vigor of this institution. Traditional views on the Eight Banners in the late Qing emphasize its decline, when it could no longer provide stipends to bannermen, and when the banner people lost their linguistic and military traditions. However, studies that have been done in the last thirty years shed new light on the ramifications of the fiscal and cultural crises of the Eight Banners. As their institutional support gradually dissolved, the banner people became embedded in the broader social fabric of Chinese society (Crossley 1990; Ding et al. 2004; Enatsu 2004; Qiu 2014). During this process, some bannermen were able to maintain their socioeconomic status and even became local elites (Enatsu 2004). At the same time, the banner people developed and preserved a distinct ethnic identity (Crossley 1990). While existing studies focus on either a select group of individuals or on a single aspect of the life of bannermen, the Shuangcheng story provides a holistic view of the banner people living in 120 villages in the last hundred years of the Qing dynasty. It confirms findings in the existing scholarship that “banner people” was not a simple, unitary category but a diverse group who interacted with and adapted to the larger society. It reveals that the Eight Banners significantly impacted the society of Manchuria, not only in terms of the identity of bannermen but also in terms of the social hierarchy in the region.

Moreover, the Shuangcheng story illuminates in particular the questions of to what extent and how the state was able to reach into and transform local society. Over the course of the development of the modern state, scholars have provided abundant documentation and theories about the complexity of state-society interactions and their perplexing consequences. Scott (1998, 2–3) highlights the inherent tension between state-building and local customs. Since myriad local customs are illegible to outsiders, state measures to govern a society, or make a society “legible,” are ways to simplify that society. As such, the interactions between state policies and local society have consequences going both ways: on one hand, local people had the capacity to modify and subvert the state-imposed policies; on the other hand, despite resistance to the state’s simplifications, state policies did shape social institutions and transform the society (ibid., 47–51). At the same time, the boundary between the state and society in everyday life is often blurred. As Joel Migdal (2001, 11–16) synthesizes in his “state-in-society approach,” although the state has an image as “a coherent, controlling organization,” in practice, both the state and society have multiple representatives, with different interests. These representatives formed “shifting coalitions and contended with one another over rules for everyday behavior” (Ibid., 11). This is also true in China. Because of the absence of a formal government structure below the county level, the state in the late imperial period relied on many agents to carry out its rule in local society (Hsiao 1960; Ch’ü 1962; Huang 1985, 219–48; Reed 2000; Zheng 2009). Some local elites helped to promote state policies and disseminate state ideologies to reinforce their own power and socioeconomic status (Wong, Huters, and Yu 1997; Faure 2007). Therefore, scholars studying local society in the late imperial period identify the influence of state ideology and policies everywhere. Whereas existing scholarship on state-society relations looks at how the state penetrates local society, the Shuangcheng case illuminates the boundary between the state and society by asking the question the other way: in a society created by the state, to what extent are local people able to exercise their agency, and how is this agency practiced?

Finally, by counting wealth as an indispensable variable in the process of social formation, and by using a non-Western case to illuminate the levers of wealth distribution in a noncapitalist setting, the book also contributes to comparative studies of wealth inequality. By revealing extremely unequal distributions of wealth, recent studies on global wealth distribution have sparked growing popular interest in this field (Lindert 2000; Davies et al. 2011; Piketty 2014).8 Yet, most of the studies on wealth distribution that have been done so far focus on Western countries, where the patterns of wealth distribution were shaped by the interplay of the development of capitalism and state fiscal policies. Because of the scarcity of systematic data preserved from the early periods, our understanding of wealth distribution in preindustrial societies remains limited. The book offers one of the first empirical studies on the distribution of landed wealth in an entire county to show how wealth inequality in early modern China was produced and maintained in an agrarian society under a state-dominated system.

Above all, the socioeconomic process of inequality in Shuangcheng can be situated in a conceptual framework that encapsulates the question, how does a state-dominated system of social formation influence life opportunities? This framework integrates theories of both state-building and social stratification to provide a holistic view of the consequences of state-initiated projects of social engineering. It also illuminates the key factors that contribute to social stratification in agrarian societies in the nineteenth-century China. Theories of social stratification maintain that structural inequality is created in two steps: in the first, categorical boundaries are clearly defined; in the second, the various actors participate in the social construction of these boundaries to make them durable (Tilly 1998; Wang 2008). The Shuangcheng story illustrates both steps in detail. The institution of the Eight Banners and the frontier setting of Shuangcheng enabled the state to use registration and resource allocation to implement a social hierarchy according to its will. Yet, the customary practices associated with the property regimes in late imperial China and the style of local governance in the Qing gave immigrants considerable room to exercise their agency to pursue their interests. By creating their own patterns of upward and downward wealth mobility, local people simultaneously challenged and reinforced the state-mandated social hierarchy. Eventually, the strategies employed by both the privileged and underprivileged groups reinforced the boundaries defined by the state. Further, the interplay of state policies, the agency of the local people, and the economic and demographic conditions shaped the social hierarchy.

The Eight Banners

The Eight Banners constitute the foundation of the Manchu state, the regime the Manchu rulers founded prior to their conquest of China proper. In the early seventeenth century, the founders of the Qing—Nurhaci and Hong Taiji—created the Eight Banner institution to organize all the people living in Liaodong,9 transforming the various tribes into a bureaucratic state (Meng 1936; Elliott 2001; Liu 2001). Each banner represented an administrative division, and the banners were distinguished by eight different combinations of colors and patterns: plain yellow, bordered yellow, plain white, bordered white, plain red, bordered red, plain blue, and bordered blue. Bannermen, adult males organized under the Eight Banners, were farmers in times of peace and soldiers in times of war. On the eve of the Manchu conquest of China, three sets of the Eight Banners—Manchu, Mongol, and Han-martial (hanjun)—organized populations of their respective ethnicities.

After the Manchu conquest of China proper in 1644, the Eight Banners became an administrative system for organizing the conqueror’s elite population, and the people of the Qing were divided into bannermen and civilian commoners. The emperors not only moved the Eight Banner population into the inner city of Beijing, but also established a national banner garrison system, stationing banner troops in garrisons located in important cities in China proper to police the civilian commoners (Crossley 1990, 47–73; Ding 2003). By serving the state, bannermen enjoyed economic and political privileges. The state provided banner officials and soldiers with two forms of material support based on their ranks: stipends, paid in silver and grain, and property grants, consisting of land and housing that was exempt from taxes and rent. Moreover, state regulations that reserved certain positions for the Manchu and Mongol bannermen increased their occupational mobility by giving them more opportunities to enter the state bureaucracy relative to those of civilian commoners.10 Throughout the 268 years of Manchu rule, the Eight Banners maintained sharp boundaries between bannermen and Han civilian commoners (Rhoads 2000; Elliott 2001).

Although all bannermen belonged to the elite strata of society, the state created categories among them to define hierarchical entitlements to state support. The state classified the bannermen based on multiple sets of criteria—such as ethnicity and service location—and established a hierarchy along with each set of criteria. In terms of ethnicity, the Manchu bannermen, who shared the ruler’s ethnicity, and the Mongol bannermen, who had an affinal relationship with the Manchu, enjoyed better material support and occupational mobility than did Han-martial bannermen, who were ethnic Han people.11 In terms of service location, the court also treated the bannermen in Beijing, those stationed in the garrisons in China proper, and those serving in Manchuria differently. The bannermen in Beijing, the metropolitan bannermen, were at the top of the hierarchy, receiving larger state stipends and land grants; whereas those in Manchuria were on the bottom and received the least material support.12 Compared to the metropolitan bannermen, the bannermen serving in the garrisons in the provinces had much more restricted occupational mobility.13

While the Eight Banners institution distinguished between bannermen and civilian commoners, over time, maintaining the system became more and more difficult because of changes in bannermen’s lifestyle and the growing fiscal burden on the state of supporting them. The first challenge was to the state’s banner-land system, one of the measures that maintained the economic privilege of the bannermen. In the beginning of the Qing dynasty, the state controlled a large amount of land in the areas surrounding both Beijing and the banner garrisons in the provinces, which it allocated to banner soldiers, officials, and nobles (Zhao 1999, 2001a, 2001b; Qiu 2014). The court granted bannermen permanent usufruct on their allocated land but prohibited them from selling their land to civilian commoners (Zhao 2001a). However, because the state stipends to bannermen in the form of cash and rice salaries weakened the importance of banner land as a material support, the majority of bannermen in Beijing rented their land out to civilian commoners. By working on the banner land as tenants, the civilian commoners gradually came to occupy the land. Private land transfers among bannermen and between bannermen and civilian commoners soon became frequent. These private land transfers not only impoverished some bannermen but also jeopardized the state’s control of banner land and thus the Eight Banners as an elite institution. Moreover, as the banner population increased, the state could no longer provide a government post with a state stipend to every adult bannerman. The Qing court used the term “xiansan” to refer to all the bannermen who were not able to obtain a banner post and therefore unemployed. In the eighteenth century, the group of xiansan bannermen grew in size in both Beijing and the garrisons and across the Manchu, Mongol, and Han-martial Banners (Wei 1995; Liu 2008, 721–22).14

Since the 1730s, court officials had been embroiled in heated discussions over policies to solve the bannermen livelihood problem and save the Eight Banners. Among the various solutions that were suggested, moving the banner population to the state-owned lands and rekindling their attachment to the land was especially appealing, because it would relieve the state of the fiscal burden of supporting the bannermen with stipends, on the one hand, and ensure them a means to earn a stable income, on the other.15 Since the state’s banner land around Beijing was already occupied by rich bannermen and civilian commoners, many officials proposed relocating bannermen to frontier regions. In 1742, following a long-term policy discussion, the Qianlong emperor (r. 1736–1795) ordered the relocation of the metropolitan bannermen from Beijing to Lalin, a site southeast of Shuangcheng (Ding 1985; Diao 1993, 198–205; Wei Y. 2010). Between 1742 and 1758, the government settled three thousand banner households in Lalin and allocated land to them equally. However, having lived in Beijing for a century, these bannermen had become acclimated to the urban lifestyle and could not adapt to rural life. Many households failed to settle in, and soon abandoned their land to return to Beijing.16 The failure of the relocation project during the Qianlong reign initially discouraged further attempts to relocate the metropolitan bannermen. However, as the Eight Banners crisis worsened, the new proposals for rusticating bannermen became increasingly attractive to the court.

By the early nineteenth century, relocating bannermen to the frontiers became both a practical solution and a symbolically meaningful move. Between 1796 and 1804, the White Lotus rebellion swept through Sichuan, Shaanxi, and Hubei provinces. The huge military expenditure needed to quell this rebellion created a new fiscal crisis for the Qing state. Moreover, in the process of suppressing the rebellion, local military forces had consolidated under the organization of local elites, which challenged the norm of centralization (Kuhn 1980). Solving the livelihood problem of the banner people became imperative for the Qing court. Practically, the central government relocated bannermen to frontier regions to ameliorate its fiscal crisis. Symbolically, relocation represents an effort by the state to restore the banner tradition and to sustain Manchu ethnic identity as a conquest elite.

Shuangcheng

Located on the alluvial plain of the Songhua River in today’s Heilongjiang province, Shuangcheng in the early nineteenth century was a sparsely populated area with fertile land. In an area of approximately three thousand square kilometers, the relics of seventeen cities built during the Jin dynasty (1115–1234) were scattered on the peripheries (SCXZ 1990, 800–803).17 The name Shuangcheng, meaning “twin cities,” derived from two Jin-dynasty cities located in the southeast. After the fall of the Jin, the area was depopulated, and the cities were abandoned. In the Qing, the region belonged to a forbidden zone. Because the Qing rulers considered Manchuria to be the cradle of the Manchu people, for almost two hundred years, from 1668 through the 1850s, the government forbade free immigration beyond the Willow Palisade (map 1.1).18 The purpose was to ensure Manchu privileges in this area. Even though the Qing government organized China proper under provinces, it always insisted on maintaining Eight Banners military control in Manchuria. Three banner generals (jiangjun)—Shengjing, Jilin, and Heilongjiang—administered Manchuria.19 Shuangcheng fell under the administration of the general of Jilin. In addition to the military function, the general’s office (yamen) in Jilin also handled civil affairs and collected taxes from the few registered residents and on products from the land.20 Consequently, in the early nineteenth century, only a few civilian commoners entered the region without government permission and cleared land there.

The frontier setting and the state control in Shuangcheng allowed the Qing government to draw the blueprint of this immigrant society according to its ideal. Institutionally, the banner system in Shuangcheng had three features: the state intervention in land distribution, the assignment of differentiated entitlements across population categories, and an egalitarian distribution of land within each population category. These were based on principles that not only existed in the Eight Banners but also grew out of the long history of China. The historical precedents included the equal-field (juntian) system, which existed between 485 and 780, and the military farms (juntun) of Ming dynasty (1368–1644).21 In fact, these features exemplified the ideal of imperial state’s control over land and population. Practically, the egalitarian distribution of land at the household level ensured the absolute power of the emperor; it prevented the rise of large landholders who would compete with the emperor for control of the population and taxation. Yet, in most regions of Qing China, private landownership was the norm, and so the state was not able to maintain these ideals. Because Shuangcheng was sparsely populated, it provided the state a place to experiment with an “ideal” institution. For the Qing, the assignment of differentiated entitlements across population categories and egalitarian land distribution within each population category were equally important in preserving the political and economic privileges of metropolitan bannermen and, eventually, restoring the vitality of Eight Banners.

Yet, if the Qing was able to design an “ideal” institution, the political and social transformations in the nineteenth and early twentieth centuries challenged this system. The abundance of land and the shortage of labor in Shuangcheng attracted more and more unofficial immigrants, who moved there despite the state’s prohibition. These unofficial immigrants—including both civilian commoners and bannermen—entered the area to make their livings as tenants or laborers. The growth of this unregistered population created a problem of control. Furthermore, to solve the deepening fiscal crisis and pressing border issues, in 1861, the central government eventually allowed free migration into northern Manchuria (Lee 1970; Diao 1993; Fan 2007). This move initiated mass immigration to Jilin; its population increased by a factor of thirteen, from 327,000 in 1850 to 4,416,300 in 1907.22 Because most of the new immigrants were civilian commoners, their arrival dramatically changed the demographic and ethnic composition of the province.

In the second half of the nineteenth century, some of the banner generals’ proactive efforts at state-building in Manchuria further challenged the Eight Banners; the generals initiated institutional reforms to transform the banner administration into a civilian government—with the goal of integrating Manchuria into China proper. The Manchu court continued to try to protect Manchuria as its backyard, but these banner generals already viewed the region as the northeastern provinces of China, and sought to protect it from invasion by Russia. Instead of considering civilian commoners the enemy of the Eight Banners, the banner generals advocated using them to strengthen the borders. To accomplish this, they also urged the court to develop a civil administration to enforce social control in the frontier regions. Consequently, in 1881 the court established an Intendant of Circuit (fenxundao) in Jilin to take charge of civilian affairs, and organized five local civilian governments in the province.23 This institutional transformation was completed in 1907, as the state abolished the banner garrison system and established three provincial governments in Manchuria—Fengtian, Jilin, and Heilongjiang (Diao 1993, 274–79).

The reform resulted in a dual government system consisting of both banner and civilian administrations in Shuangcheng, which weakened the institutional support to metropolitan and rural bannermen’s land entitlements. A civilian government was established in Shuangcheng in 1882.24 In a related move, the state downgraded the local banner government, preserving it as the authority of the 120 banner villages and more than one hundred natural settlements called wopeng. Although the state planned to administer the bannermen and the civilian commoners separately, in reality, the two governments were not completely distinct because administrative areas overlapped. Both governments often adjudicated disputes that involved both bannermen and civilian commoners; and some bannermen also directly filed lawsuits with the civilian government. Because they were unfamiliar with the banner institution and its land-allocation policies, the civilian officials sometimes justified practices that transgressed the boundaries between the population categories. Moreover, in the first decade of the twentieth century, the Qing privatized the banner land and did away with the institutional support of the hierarchical land entitlements in Shuangcheng. In 1902, the court made the decision to open all land in Shuangcheng to private owners.25 In 1906, the state allowed the free transfer of banner land between bannermen and civilian commoners.26

Along with these political and social transformations, Shuangcheng also developed into an economic center in northern Manchuria. As one of the earliest settlements in the region, Shuangcheng has been a major site of grain production since the nineteenth century.27 The soil in northern Manchuria was rich in organic material, and both crop yields and the quality of the grains exceeded those in southern Manchuria (MMTKK 1909, 13). Located at the intersection of several transportation routes, Shuangcheng also became a center for grain distribution (ibid, 144). The high crop yields in northern Manchuria produced considerable surplus for farm households. In 1909, about 80 percent of the wheat and 70 percent of the beans produced in northern Manchuria were exported from the region. Merchants gathered grains purchased from villages of Shuangcheng and the surrounding counties to the Shuangcheng seat and then transported them to Yingkou, a treaty port in Fengtian (today’s Liaoning province; ibid, 55-59). Shuangcheng’s strategic position became even more prominent at the end of the nineteenth century. Because Shuangcheng is close to Harbin, in 1899, Russia selected it to be the site of a station of the Chinese Eastern Railway and started to build the railway across the county (SCXZ 1990, 20). In 1903, the railway was put in operation. Thereafter, the grains were also transported from Shuangcheng to Siberia (MMTKK 1909, 57–58).

However, Shuangcheng remained a largely agrarian economy, with a low level of economic development.28 Even in the early twentieth century, agricultural production in Shuangcheng and other parts of northern Manchuria was not specialized (Reardon-Anderson 2005, 169–251). Households were still the basic unit of production. Farmers, large landholders and tenants alike, made very limited investment in capital and technology. They used traditional technologies to cultivate land, alternating four types of grains: beans, wheat, sorghum, and millet. Harvests were used first for the consumption by family members and for the following year’s production. Farmers then sold the surplus at the market (MMTKK 1909, 54–57). In other words, the commercialization of agricultural products did not advance the way of agricultural production.

At the same time, amid all the social and institutional changes, the banner administration and banner society in Shuangcheng showed considerable resilience. Even after the establishment of the civilian government, the banner government in the Qing remained powerful in that it administered the majority of rents and taxes from the land.29 The banner government remained active even after the fall of the Qing in 1911. In 1912, the Office of the Banner Affairs (qiwu chengban chu) was established under the republican county government; it continued to be part of the local administration even under the Manchukuo (1932–1945).30 Yet, the banner administration could no longer ensure the privileges of the banner people. The socioeconomic status of the bannermen, as indicated by the size of their landholdings, was especially crucial in determining their fate and fortune in the Republic of China (1912–1949). As Enatsu’s (2004) study on the local elites in Fengtian shows, many Han-martial bannermen, who were already the de facto landlords of the Qing manor lands, purchased land during this political transformation. This also happened in Shuangcheng. The persistence of the social hierarchy in Shuangcheng after the fall of the Qing indicates that the state-designated elites were able to successfully convert the political and economic privileges granted by the Qing into social and economic capital that withstood the political revolution. Moreover, this transformation was accomplished through the long-term interactions between state policies and the efforts of local people during the one hundred years that followed the settlement.

State Registration and Social Engineering

Government registration of the population and property has been an indispensable component of bureaucratic states. Registration systems developed along with the state’s growing capacity to gather information, organize its population, and perform routine tasks such as taxation. By registering people and property, the state established its authority over and defined the rights and obligations of the registered population. In Western Europe, the development of registration systems contributed significantly to the rise of the modern state (Tilly 1975; Szreter and Breckenridge 2012). In China, however, this “modern” feature of the bureaucratic state had emerged as early as the sixth century BC (Liang 1980; Song 1991; Xin 2007; Von Glahn 2012). By the first century, household registers in China recorded biographical information about not only the household head and his dependents—including name, occupation, residence, rank of nobility, and age—but also household property, including the house, land, slaves, and livestock. From then on, successive regimes continued to register their populations (Von Glahn 2012). Registration established a contractual relationship between the state and the people: the registered people paid taxes and provided labor to the state; in return, the state granted them entitlements and rights which the unregistered people did not have.

At times of social engineering, registration becomes an important tool for states in implementing the designated social hierarchy. Theories of social formation hold that for structural inequality to become durable, the boundaries between social categories should be clearly defined (Tilly 1998; Lamont and Molnár 2002; Wang 2008). Registration makes this possible because the systematic recording of the various ascriptive characteristics of individuals and households involves defining and standardizing these categories. As studies on registration and census show, local people’s understanding of categories, such as ethnicity and race, vary dramatically from that of the government (Kertzer and Arel 2002; Mullaney 2010). Therefore, the act registering these categories is an effort at state-building. When carrying out social-engineering projects, the state standardizes the characteristics that best manifest its desired social hierarchy and classifies the people into categories. For example, in Shuangcheng, a combination of institutional affiliation and place of origin was the basis of the social hierarchy. By recording these ascriptive characteristics and compiling separate registers for metropolitan, rural, floating bannermen, and civilian commoners, state registration institutionalized the boundaries between these groups and made them hereditary categories.

Once the boundaries of population categories are defined, the state assigns differentiated entitlements to each category to establish the hierarchy. This creates a system of exclusion in which people outside certain groups are excluded from access to resources.31 The registers document an individual’s membership in a certain category and serve as the official proof of eligibility to the entitlements associated with that category. In this sense, registration helps to produce what social psychologists call a “group-based social hierarchy,” which means that an individual’s power and privilege is derived from his or her ascriptive membership in a socially constructed group (Sidanius and Pratto 1999, 32). The unequal assignment of entitlements also renders people a sense of ranking associated with their group boundaries.

The above statecraft lays out the foundation for the social construction of the categorical boundaries among people. While the state registration and resource allocation define the boundaries that generate structural inequality, these boundaries are not clearly drawn until they go through the process of social construction by which individuals in a group develop a sense of belonging and agree on a collective identity (Wang 2008). Only then does the structural inequality become durable. The social construction of categorical boundaries is a complex process that involves the actions of various actors (White 1992, 127–28; Verdery 1994; Barth 1998), and the actions are based on the actors’ perception of their socioeconomic statuses and interests. As the registration categories and the unequal entitlements largely shape people’s understandings of their status and interests, actors consciously adapt their actions to the state-designated social hierarchy. The privileged groups use the power and privilege associated with membership in that group to maintain their status. At the same time, members of the disadvantaged groups, even when they manage to overcome categorical boundaries and promote their socioeconomic status, are constantly reminded of the barriers the state set on their access to resources. The long-term interactions between the registration categories and state policies give rise to collective identities, thereby transforming these registration categories into socially constructed groups. The Shuangcheng case illustrates the entire process of social formation, from the creation of boundaries by the state to the social construction of boundaries by the immigrants.

Landownership and Local Practices

In Shuangcheng, as in other agrarian societies, the pattern of land distribution and the actions surrounding it are key in understanding the process of social formation. In preindustrial rural societies, land was the most valuable family asset of farm families. Thus, the landholdings of a family are an important measure of its socioeconomic status in the community. Moreover, under the state-dominated system in Shuangcheng, land, especially, was central to both the state’s effort to define the social hierarchy and the actions of immigrants that contributed to the construction of social boundaries. The state used land allocation to differentiate immigrants’ entitlements. Yet, despite the state’s land-allocation policies, immigrants created their own upward and downward wealth mobility, outside the state’s purview: capable families increased their landholdings by clearing or buying more land, and incapable metropolitan and rural banner households lost their state-allocated land. While the patterns of land distribution per se reveal the consequences of this statecraft and social construction, the mechanisms that enabled immigrants to exercise their agency and influence land distribution also illuminate the question of how the state-designated social hierarchy finally played out on the ground.

To better understand the dynamics of wealth distribution, it is important to look at the incomplete nature of landownership in imperial China. In contrast to the “full and free” landownership in the Western capitalist world (Engels 2001, 205), in imperial China individual owners never had exclusive rights to their land.32 Three parties—the landowner and the members of his immediate household, the state, and the owner’s kin and fellow villagers—had ownership claims. Ownership was not evenly distributed among the three parties; there was a general trend toward granting a greater portion of the ownership to private owners and less to the state.33 Understanding this provides a framework for interpreting complicated property regimes in the local societies. In imperial China, with only a few exceptions, property rights belonged to the household, not to individuals (Zhang 2002b; Zelin 2004; Yang 2009, 1–10). Even though land was registered under the name of the household head, his immediate family members and kin living in the household also had legitimate claims on the land. Second, people living outside the household, the kin and fellow villagers of a landowner, also had considerable influence on landownership. When selling land, the seller was required to give priority to buy to his kin and then to his fellow villagers whose land plots bordered the plot being sold. This custom recognized kin and fellow villagers’ rights over particular land plots (Zhang 2002a; Yang 2009). This phenomenon has led Yang Guozhen (2009, 4–5) to conclude that “land ownership in historical China is a combination of private ownership and the ownership of state, village community, and kin groups”

The incomplete and divisible nature of landownership resulted in complicated property regimes in rural communities. On the one hand, officially registered landownership served as the most important indicator of the socioeconomic status of farm households. On the other hand, as a practical matter, tenants working on rented land enjoyed use rights that often eventually developed into partial ownership. Existing scholarship has revealed the prevalence of a two-tiered form of landownership during the Ming and Qing dynasties.34 By investing their labor in increasing the value of the land, tenants as the de facto farmers and managers of land secured their use rights. They were able to preserve their control over land across several generations, even renting out their plots to other tenants. This custom is so prevalent that, in many regions, landownership was divided into ownership of “topsoil,” which belonged to tenants, and ownership of “subsoil,” which belonged to landlords.35

The fact that usufruct can develop into a form of landownership has two important implications for interpreting wealth distribution in Shuangcheng. First, the rights the official immigrant households had over their land was a form of ownership. Although the state owned all the land, metropolitan and rural bannermen had the right to manage their land and pass it down to their descendants. In theory, they could keep their rights so long as they maintained their registration status. Thus, in this book, I use the term ownership to describe the rights immigrants had over their registered land; the officially registered landownership is used to measure wealth status. Second, immigrants without registered land still had opportunities to enjoy partial landownership through customary practices. Despite the state control, local practices of land rental and conditional sale (dian), in which landowners sold their land at a price that reflected only the plot’s partial market value and preserved the right to buy back the plot at this price within a stipulated period of time, were common in banner communities in Manchuria (Sudo 1941; Diao 1993, 133–38, 254–66; Isett 2004, 2007). In Shuangcheng, due to the low labor to land ratio and the metropolitan bannermen’s limited knowledge of farming, a large number of “have-nots” worked on the land as tenants or laborers. As chapter 8 shows, their annual income per unit of land was actually greater than that of their landlords. Some tenants became affluent. Therefore, registered landownership indicated only one aspect of the socioeconomic condition of immigrant households.

While recognizing the complex property regime in Shuangcheng, the book highlights the importance of registered landownership in social formation. In this sense, it also sheds light on the different roles wealth and income played in the socioeconomic life in rural China.36 Despite the financial well-being generated by income from the land, eventually, it was a combination of the political and economic rights associated with land entitlements that defined social conflicts and the formation of social groups. Not only did registered landownership indicate the elite status of metropolitan and rural bannermen, but the land entitlements also became a source of power, by which the metropolitan and rural bannermen could increase their landholdings. Those who successfully maintained their official landownership continued to be the upper class even after the regime changed from the Qing to the Republic of China.

Agency and Structural Inequality

The process of social formation in Shuangcheng vividly illustrates the interplay of agency and structure. Here agency refers to the capacity of human beings to develop strategies and to shape their circumstances (Sewell 1992; Emirbayer and Mische 1998). This capacity includes both people deriving a perception of social relations from the existing schema and developing strategies to work with or work around that structure. Works by sociologists demonstrate that all humans not only have agency but also the capacity to exercise agency in everyday life (Goffman 1959, 1982). The agency of Shuangcheng bannermen could first be seen in the development of village communities and households as independent social units to organize everyday life. Due to the highly institutionalized nature of the Eight Banners and the background of the settlement as a social-engineering project, the state planned villages and households to be administrative units. Yet, after settlement, these administrative units soon developed into autonomous entities. Villages became independent communities, organizing land clearings and supervising land transfers outside the state’s purview. Households, with rights equivalent to landownership, enjoyed the autonomy to make decisions regarding production, consumption, and the management of their lands.

The style of local governance in Shuangcheng, as elsewhere, gave immigrants considerable room to exercise their agency. It has been well established that in late imperial China the state intentionally kept the local government lean. Therefore, the organization of everyday life became largely the responsibility of various social organizations, such as village communities and families. While local elites made efforts to disseminate state ideology and laws to the local communities, customs played a major role in directing civil affairs, such as land transfer, land rental, marriage, and adoption (Scogin 1994; Huang 2001; Cohen 2004). Moreover, Melissa Macauley (2001, 332) points out that compared with that of the Ottoman Empire, the Qing government’s effort to implement laws and policies before the twentieth century represents “an instance of legal simplification without the state-building dimension.” Under the logic of limited administration, the officials had no intention of actually changing local practices but merely kept the state vision of local society in paper to facilitate tasks of local governance, such as taxation. Interestingly, although the Qing meticulously carried out the Shuangcheng settlement and built a state-dominated system, in governing civil affairs officials still followed the principle of indirect rule and kept state policies in paper. As chapter 5 shows, despite the state’s prohibition, local officials accommodated land sales among bannermen, so long as they followed state policy and reported these cases as official land transfers. Therefore, despite the state-dominated system, village communities, households, and individuals in Shuangcheng enjoyed the same level of agency as residents in rural communities elsewhere in nineteenth-century China.

Shuangcheng immigrants lived in multiple structures. In addition to the state-dominated system and the unequal assignment of land entitlements, the general economic conditions and family demography also constituted the structural settings. These structures served as the schema for immigrants to develop strategies to cope up or change social relations. At the same time, these structures both provided resources for and constrained the agency of immigrants. Moreover, these structures affected the agency of immigrant groups differently, depending on their position in the social hierarchy and other characteristics.37 When making important decisions regarding the fortune of their family, immigrants had to consider these constraints and opportunities and make use of available resources in pursuing their interests. In turn, these decisions affected family demography and socioeconomic status, thereby creating variations in the pattern of land distribution. Thus, the interplay between the agency of immigrants and the above structural settings underlay the process of social formation in Shuangcheng. Eventually, the subsequent economic and demographic differentiation, social mobility, and social reproduction shaped the social hierarchy.

The economic conditions in Shuangcheng—the low labor to land ratio and the low level of economic development—affected the wealth status of the Shuangcheng immigrants. As a newly settled area, Shuangcheng had abundant land and few people. Although the frontier setting created hardships for the settlers, it still offered most immigrants opportunities to become wealthy. Not only did the state allocate a large amount of land to metropolitan and rural bannermen, but the immigrants who moved there without a government order and therefore had no land entitlement had the leverage to acquire use rights of land. Moreover, in the initial years of settlement, the frontier setting also affected the formation of social organizations and social groups. Most of the immigrants moving into the area lacked the support of extended families and had to fight the wildness and clear land on their own. Only families that have enough laborers to farm the land could survive. This was true for both official and unofficial immigrants. Immigrant households organized to clear land, rented their land out, or hired laborers. Along with this process of securing laborers, immigrants also built village communities and formed other social networks, which then shaped the social organizations and hierarchy. Finally, the low level of development indicates that capital had very limited influence on land distribution in Shangcheng.

Family demography, because it affected production and consumption, had significant impact on the wealth status of immigrant households. Birth, marriage, and death are the major events in the life cycle of a household. The outcome of these events—fertility, timing of marriage and remarriage, and mortality—determines the household size, measured by the number of members and its gender and age composition. The household size and gender-age composition indicate the numbers of its available laborers and of the non-working-age members in need of economic support (Cain 1978). Since labor is an important factor in agricultural production, both conditions significantly affected the ability of a rural family to manage its land. A household with few laborers and a lot of non-working-age members had a shortage of labor in production and pressure in consumption. Without strategies for surmounting these difficulties, family members had to exchange land or other property for money, and potentially leading to downward wealth mobility. Existing studies on the demographic behaviors in preindustrial Europe and Asia have revealed a strong correlation between household socioeconomic status and the demographic success of families; families with higher socioeconomic status were likely to have more surviving children, better chances of entering into marriages, and lower risk of dying during a period of economic hardship (Harrell 1985; Campbell, Lee, and Bengtsson 2004; Clark and Hamilton 2006; Shiue 2008; Tsuya et al. 2010; Bengtsson 2014). This is also true in Shuangcheng (Chen, Lee, and Campbell 2010; Chen, Campbell, and Lee 2014). This correlation can be explained as a two-way effect: better economic conditions gave family members the resources to achieve demographic success; and, at the same time, the demographic circumstances of a household also influenced the direction of its wealth mobility.

In addition to its effects in production and consumption, the impact of family demography on the intergenerational transmission of wealth deserves special attention. Existing studies on wealth distribution in Western countries have demonstrated that intergenerational transmission of wealth, or inheritance, is a major cause of wealth inequality (Oulton 1976; Menchik 1979; Piketty 2000, 2014; Arrondel and Grange 2006; Harbury and Hitchins 2012). Since wealthier families have more assets to pass down to the next generation, inheritance results in the reproduction of the status of the family. At the same time, different principles of inheritance practice also affected wealth mobility. For example, in preindustrial Europe, wealth was transmitted through the right of primogeniture, through which a family passed down all of its property to the eldest son. The practice of primogeniture helped to maintain a concentrated pattern of wealth distribution. In societies in which families divide their properties among all the children, children from larger families tend to have much less wealth than their parents (Menchik 1979). In late imperial China, partible inheritance was the norm, and families divided their properties equally among all the sons. If families always had two or more sons in the next generation and the members did not acquire more land through other paths, the practice of partible inheritance resulted in the dissipation of family wealth in a few generations’ time (Myers 1970, 159–66; Lavely and Wong 1992).

In Shuangcheng, family demography interacted with the state’s land-allocation policies and inheritance practices to influence the wealth mobility of individual households. First of all, having at least one surviving male heir was crucial if immigrant households were to continue the family line and maintain the family’s wealth. Families without a male heir would eventually become extinct. Once families successfully secured heirs, they could choose how to pass down their property. Under the Eight Banners, the state adopted the practice of primogeniture inheritance to regulate and register household landholding. Some families followed the state regulations, but others continued practicing partible inheritance (Chen 2009, 222–61).38 The coexistence of primogeniture and partible inheritance rules gave immigrant households room to exercise agency and make choices based on the specific family situation. Often times, the choice reflected the power dynamics within the family. Thus, variations in inheritance practices and the number of heirs also resulted in disparities in wealth status.

Finally, one of the most important components of social formation in Shuangcheng was that immigrants actively used the structure of the state—the government institution and the land-allocation policies—to pursue their interests. The state-dominated system made state power conveniently available to immigrants. By differentiating land entitlements, the state created discrepancies in the agency of different immigrant groups. For the privileged metropolitan and rural bannermen, the state served as an important resource that gave them an advantage in the competition for wealth with other immigrant groups. As the book shows, metropolitan and rural bannermen actively maintained their boundaries and blocked other groups’ access to land entitlements, through mechanisms that sociologists label “rent seeking” and “opportunity hoarding” (Tilly 1998, 74–116; Wang 2008, 18–21). Meanwhile, the groups that were excluded from land allocation faced tremendous constraints in exercising their agency. As chapter 3 shows, even though some floating bannermen and civilian commoners privately cleared land on their own, they lost the land to metropolitan and rural bannermen.

The interaction between the agency of immigrants and the state structure became even more complicated, because the different interests of state representatives provided immigrants multiple channels through which to appeal to state authority. Although the state had a strong presence in Shuangcheng, the central, provincial, and local governments often had different interests in governance. Thus, state power in local society was presented in various forms, which offered communities and individuals abundant choices. Aware of the multitude of state representatives, immigrants tactfully chose their allies, forming coalitions with some state authorities while resisting others. The different interests of state representatives not only offered the privileged groups more room to exercise agency, but also made state a source of power for the disadvantaged groups. For example, although the central government prohibited civilian commoners from settling and owning land in Shuangcheng, by acquiring the personal sponsorship of local officials, some civilian commoners were able to work as contractors to organize land clearing and thereby occupy a large amount of land. Thus, state power was so convenient to use that it became prevailing in local society.

Eventually, both social reproduction on the basis of population categories and social mobility at the household level helped to reinforce the state-designated hierarchy. In the long term, the interaction between general economic conditions, demography, and state policies gave rise to demographic and economic differentials at the household level. Capable households had opportunities to maintain their socioeconomic status or achieve upward social mobility, while incapable families were located at the bottom of the hierarchy. However, the majority of the members of the privileged groups were able to maintain their wealth status. The boundaries between population categories, especially, were fortified when the state served as a major resource for immigrants to exercise agency. As immigrants sought to use state power to maintain or promote their status, the state structure deepened their knowledge about the social relations. Ultimately, immigrants developed distinct identities based on the state-designated population categories.

Data Sources and Organization

Shuangcheng has rich and comprehensive archival sources. Unlike the archival sources of most local societies during this period, which mainly center on local administration and legal cases, the archival sources available for Shuangcheng include edicts, court memorials, official correspondence, legal cases, and banner household and land registers. These voluminous documents provide information not only on local administration and socioeconomic life but also on the policies and the rationales of the emperor and the officials in the central, provincial, and local governments. The edicts and memorials preserved in the First Historical Archives in Beijing offer details about the motivation of the central government and the processes of Shuangcheng settlement from 1812 onward. The archives of the local banner and civilian governments in the Qing and the Republican periods provide continuous documentation of local life, beginning in 1852.39

The comprehensive archival sources are suitable for analyses using a mixture of qualitative and quantitative methods. Records from 338 volumes of household registers and twenty-three volumes of land registers for the period 1866–1913 constitute a uniquely rich dataset: the China Multi-Generational Panel Dataset-Shuangcheng (CMGPD-SC). With 1,346,826 annual observations on 107,890 individuals, the dataset covers all the official immigrants—metropolitan and rural bannermen—and the registered unofficial banner immigrants and their descendants living in Shangcheng’s 120 banner villages (Wang et al. 2016). Moreover, with 37,187 plot inventories for 8,182 individual landowners, the dataset also includes farmland registered in Shuangcheng from 1866 to 1906. This dataset makes possible not only analyzing the overall pattern of land distribution but also tracing the upward and downward mobility of individual families throughout the time period.

These land and population data surpass the data used by existing studies both in scale, covering an entire county, and in depth, providing longitudinal individual- and household-level information on local socioeconomic conditions. Existing studies of land distribution in historical China are based on either aggregate county-level data or household-level data from a few villages.40 The aggregate data omit details of the individual landowners; data collected from a few villages are likely to provide only piecemeal glimpses of a more complex agrarian society. In imperial China, the county was the lowest level of bureaucratic structure. Individual counties not only had a certain unity but also were large enough to include a variety of social groups with different socioeconomic statuses. Thus, taking a county as the unit of analysis generates a more comprehensive view of social categorization and wealth inequality.

The administrative documents and legal cases available in the traditional archives make it possible to explore in detail the socioeconomic processes that generated the patterns of land distribution. These documents complement the dataset by providing rich information on the interactions and tensions among officials and different immigrant groups and on the mentality of the local people. In other words, the book fleshes out the statistical patterns of social categorization and land distribution with vivid pictures of local practices of land accumulation and their interplay with government policy. Gazetteers, survey reports published by the Japanese Southern Manchuria Railway Company, and genealogies and oral histories collected during fieldwork supplement these sources to provide an account of Shuangcheng’s history.

To present a complex picture of social formation under state domination across a long span of time, the narrative is organized thematically while maintaining a sense of the chronological unfolding of events. Chapters are grouped into two themes—“State-Building” (Part I) and “Social Development and Stratification” (Part II)—to address the two processes that formed this immigrant society. An analysis of how the development of each theme affected the distribution of land drives the narrative of each part.

Part I, which consists of chapters 2 to 4, provides an overview of the processes by which the state built institutions and consolidated its power between the 1810s and 1850s. Chapter 2 examines the implementation of Shuangcheng settlement to highlight the state’s efforts to clear the existing boundaries among the immigrants: the state eliminated existing social organizations among the settlers by scattering households from the same place of origin or same descent group among different villages. This boundary-clearing procedure provided a foundation on which the state could build a two-tier hierarchy, with metropolitan bannermen at the top and rural bannermen on the bottom. Chapter 3 analyzes the population categories recorded in the state household registers and the unregistered population to explore the ways the state built new boundaries among immigrants. It shows that, by assigning differential land entitlements to the registration categories, the state created a new social hierarchy. Chapter 4 examines the issue of local governance in Shuangcheng to illuminate two processes of consolidation in this newly settled frontier society: the consolidation of the society per se, marked by immigrants reaching an agreement on a local identity, and the consolidation of the control of local government by the imperial court. Although the state built this society from the top down, it took the central government thirty years to embed the local government in the bureaucratic system.

Part II, which consists of chapters 5 to 8, explores the social construction of categorical boundaries from the 1840s to the 1910s. Chapters 5 and 6 examine the development of village communities and family, respectively, and explore immigrants’ activities of wealth accumulation surrounding these two organizations. These two chapters show that, despite the state policies of land allocation and population registration, savvy immigrants strategically coordinated local practices with state policies to create their own upward and downward wealth mobility. By actively using the state as a source of power in land accumulation, immigrant groups reinforced the state-built boundaries. Chapter 7 analyzes the patterns of land distribution between 1870 and 1906 to explain why the state-mandated social hierarchy endured in Shuangcheng. It shows that land distribution among bannermen with land entitlements exhibited a pattern of stratification without concentration. This pattern of land distribution sustained a stable landowner class consisting of metropolitan and rural bannermen. Finally, chapter 8 shows that, by the end of the Qing, the unequal assignment of land entitlements had transformed the population categories into groups with distinctive identities. These identities continued to serve as the basis of social grouping and to define local conflicts in the early period of the Republic of China.

Notes

1. The term “Manchuria” used to be controversial because it is associated with the colonial interests of Russia and Japan in this region in the nineteenth and twentieth centuries. However, Mark Elliot (2000) reveals that beginning in the seventeenth century, the term “Manchuria” gradually developed into a typonym referring to the area in Northeast Asia that the Manchu rulers of the Qing claimed as their place of origin, into which they invested a unique identity. Therefore, in this book, I use “Manchuria” as a historical term to refer to the region that in the twentieth century has become Northeast China.

2. About the itinerary of the first group of metropolitan bannermen, see Jiang Jixian’s memorial on 1824.1.28 (JJCLFZZ 03-3387-38), and Rongzhao and Qiying’s memorial on 1824.2.22 (JJCLFZZ 03-2846-34). On many occasions, Shuangcheng is also called Shuangchengbao or Shuangchengpu.

3. The name list of the first group of metropolitan banner immigrants archived in the Ningguta vice commander-in-chief’s office in 1824 (NGTFDTYMDA, 52: 289).

4. The total number of households is summarized from China Multi-Generational Panel Data: Shuangcheng (CMGPD-SC) and the report of the Shuangcheng banner government regarding the baojia system in 1866 (SCPZGYMDA 160: 636-1: 354–60).

5. Compared to the 1860s, the number of registered households and size of the population in 1910 significantly increased as a result of immigration and the expansion of administrative area and government registration.

6. Wang (2005) and Von Glahn (2012) pay attention to the historical precedents of hukou household registration in PRC and trace the history of household registration in China back to its origins in the sixth century BC. Both studies consider the hukou system a continuation of the civil registration practices in imperial China. While this point is well taken, the hukou system shares more features with the banner registration in Shuangcheng than it does with civil registration in the past.

7. Since the 1970s, numerous studies have examined this trend, including the two edited volumes by the Center for East Asian Studies at the University of Chicago (Jones 1979; Tsou 1981b); Kuhn (1980); Rowe (1984, 1992); Rankin (1986); and Esherick and Rankin (1990).

8. For example, Davies et al. (2011) estimate that in 2000, 71 percent of the global wealth was concentrated in the hands of 10 percent of adults, and that about half of the adult population had nearly nothing.

9. “Liaodong” refers to the geographical area east of the Liao River in today’s Liaoning province. In the early seventeenth century, Nurhachi and Hong Taiji established the Manchu State in this area.

10. In the Qing bureaucracy, at almost every level of government, the court stipulated that some posts could only be filled by Manchu and Mongol bannermen. Moreover, the state also arranged special examinations for Manchu and Mongol bannermen (Rhoads 2000; Elliott 2001).

11. The three ethnicities—Manchu, Mongol, and Han—refer to institutionalized population categories. The ethnic composition of bannermen was actually more complicated than this. There were also Xibe, Korean, and some other small ethnic groups.

12. A metropolitan bannerman with the lowest soldier’s rank received two taels of silver as monthly salary; whereas a garrison bannerman of the same rank only received one tael of silver (Ding 2003, 220). Bannermen in the garrisons received both silver and grain salaries, with the exception of those in Shengjing, who did not receive a grain salary. Instead, they worked on state-allocated land to supplement their livelihoods.

13. Before the nineteenth century, bannermen in the garrisons were not allowed to take exams in the provinces where the garrisons were located. To take exams, they had to travel to Beijing, and this inconvenience prevented many from entering the government (Ding 2003, 219–21).

14. For example, in 1771, the court identified more than six thousand widowers, widows, and orphans in the metropolitan banner population who could barely make a living. Some of them had even become paupers (Liu 2008, 719).

15. See the articles collected in HCJSWB, book 18, juan 35, “baqi shengji,” 1–45.

16. Ding (1985) and Wei Y. (2010) studied in detail the 1742 relocation of metropolitan bannermen to Lalin.

17. Shuangcheng is about 75 kilometers from Acheng, where the capital of the Jin dynasty was located.

18. The Willow Palisade (liutiaobian) was a barrier the Qing government erected in 1681 to prevent Han Chinese from moving farther into Manchuria and encroaching on banner land. It was constructed of two parallel one-meter-high earthen levees, spaced three-and-a-half meters apart, and crowned with a fence of willow fronds (Edmonds 1979).

19. The only exception was in what is present day Liaoning, where the government of the Fengtian prefect administered all the civilian affairs.

20. The state had established a civilian local government in Jilin in the 1730s to govern a handful of settlements of Han-Chinese immigrants who had entered Jilin surreptitiously and were registered by the government (Lee 1970; Fan 2007, 222–23).

21. Under the equal-field system, the state owned the land and allocated a plot to each adult man and woman. By farming their plots, households paid taxes and provided labor services to the state (Han 1984). On military farms during the Ming dynasty, the state also owned the land and assigned one plot to each household (Wang 1965).

22. The population figure in 1850 is from Ho (1959, 283, appendix 2). The population figure in 1907 is from Fan (2007, 70). Liang (1980) provides similar results.

23. See the memorials by Ming’an on 1878.9.9 and in 1881 (QDJLDASLXB, 58–62, 7).

24. See Ming’an’s memorial on 1882.3.22, which is enclosed in the correspondence between the Office of the General of Jilin and the Shuangcheng banner government (SCPXLYMDA, no. 14936).

25. See the twenty-eight rules made by the General Bureau of Tax Counting and Land Opening (qingfu fanghuang zongju) on 1902.7.15 (SCPZGYMDA, 268: 1195: 1–18). Although the decision was made in 1902, it took the government three years to prepare for the transition. Also see the order from the Office of the General of Jilin to the banner government in 1906.2 (SCPZGYMDA, 275: 1237: 35–39)

26. See the order of the provincial government on 1906.3.27 (SCPZGYMDA, 276: 1238: 225–26).

27. Shuangcheng still occupies a leading position in grain production today. In 2005, the grain yield of Shuangcheng ranked tenth among all counties in China (GJTJJ 2005).

28. Some scholars have considered the high agricultural productivity and the commercialization of agricultural products in Manchuria as signs of economic development and advancement in agricultural production (Yi 1990, 163–65). Recent studies of Chinese economic history emphasize the growing output per capita and structural change in the economy—such as specialization of agricultural production and investment in capital and technology—as the indicators of economic development (Huang 1985, 1990; Myers 1991). In this sense, the agricultural economy in northern Manchuria in the early twentieth century did not develop (Zhao 1972; Reardon-Anderson 2005, 196–98).

29. Upon its establishment in 1882, the civilian government acquired the rights to manage the banner land. But in 1885, the banner government was able to get these rights back (Ren 2013). Thus the civilian government only managed the land in Shuangcheng for a very brief period.

>30. The continuation of the banner administration in Shuangcheng was provided for in the abdication agreement between the Qing and the Republic of China. The emperor’s abdication was made contingent upon the Republican government’s agreement to maintain the Eight Banner system to administer the affairs of bannermen (Tong 1994).

31. In his study of the hukou system of contemporary China, Fei-ling Wang (2005, 4–9) synthesizes a theory of institutional exclusion as the framework. I borrow this concept to show the parallels between the Eight Banners in Shuangcheng and the hukou system.

32. This incomplete landownership in China has given rise to scholarly debate regarding the nature of landownership in early China over whether the land was state-owned or privately owned. The articles collected in Lishi yanjiu bianjibu (1957) best represent the various positions in this debate. Yang Guozhen (2009) reconciles this debate by pointing out the incomplete form of landownership that persisted throughout the imperial time.

33. Before the eighth century, state landownership was the dominant form, although the state allowed private land transactions very early on. The state did not collect taxes on land, but levied a poll tax and required that commoners provide labor services to the state, which it justified by considering these as repayment to the state for having authorized land use rights. In 780, a tax reform by the Tang state introduced the principle of tax collection based on household land and property; this reform marked the state’s recognition of private landownership. This transition continued in the following dynasties and was finally completed in the mid-eighteenth century, when the Qing state began to collect taxes exclusively on land.

34. There is a rich body of scholarship on the two-tiered form of landownership in traditional China. The earliest studies date back to the 1930s. Some recent scholarship includes Chao and Chen (1982); Huang (1985; 2001, 99–118); and Yang (2009, 70–105). For a complete review of the scholarship, see Cao (2012).

35. Because this practice was widespread in China, regional variations existed.

36. In recent years, economic historians worldwide have paid attention to differences between wealth and income, especially the implications for social inequality. In general, scholars point out that wealth inequality is more persistent than income inequality because of the inheritance practices that transmit wealth from one generation to the next. See Lindert (2000); Keister (2000); and Piketty (2000, 2014).

37. Sewell (1992, 20–21) discusses the different agency of social actors based on social position and the social system under which they live.

38. Primogeniture in property inheritance was practiced not only by bannermen in Shuangcheng but also by those working on the manor lands in the surrounding areas of Beijing (Qiu 2014, 258–60).

39. See Wang et al. (2016, appendix B) for a more detailed discussion of the archival sources.

40. Chao and Chen (1982); Chao (1986, 2003); Qin and Su (1996); Hu (2012) have used government land registers to study land distribution. However, because of the issues of preservation, the land registers used by previous studies are scattered and only available for a few villages. Another source for studying land distribution is survey data collected in the early twentieth century. These data provide more contextual information, but they also are conducted at the village level. See Myers (1970) and Huang (1985).