HOUSEHOLDS AT THE MARGINS

Long before the Song, Chinese states had created monopolies as a way to tax non-agrarian goods like salt and tea. Like the decision to allow land to circulate on the private market, this policy emerged in part from tax shortfalls. In the late Tang, these monopolies—especially on salt—made up a major component in state finance.⁵ By the Song, tea households (chahu), smelter households (yehu), and saltern households (zaohu) all supplied the state directly with their respective products, albeit in highly varied and regionally dependent ways.⁶ Activist ministers like Wang Anshi and Cai Jing expanded the tea and salt monopolies even further. In some regions, these monopolies could be key to the extension of state power to non-agrarian populations.⁷ In addition to taxing these specialized households, the Song also imposed labor service on its commoner population, including a regular corvée to cut firewood, and more erratic levies to log large timber.⁸ In addition to their recognized importance in state finances, these household levies were a way for the Song to derive both control and revenue from non-agrarian environments.

The spread of household registration to non-agrarian peoples took on an entirely new character under nomadic rulers. Innovations began under the Khitan-ruled Liao dynasty, contemporaries of the Northern Song. In order to tax the steppe and forest peoples of the north as well as the agrarian population of the south, the Khitan created a dual administration, imposing a head tax on the northern groups and land taxes on the sedentary farmers.⁹ This dual system was highly influential. The Jurchen Jin dynasty adopted it when it conquered the Liao in the early 1100s, and the Mongols adopted it when they conquered the Jin, in part under the tutelage of a Khitan noble named Yelü Chucai.¹⁰ In these early stages, the dual system focused on incorporating sedentary farmers into nomadic states. Yet as they integrated North China into their empire, the Mongols went far beyond their Liao and Jin predecessors. Within the sedentary population, they oversaw a proliferation of increasingly specific household categories, including separate classifications for artisans (jiang), Chinese military households (junhu), and a wide range of other smaller professional groups including Confucian scholars (ru), physicians (yi), musicians (yueren), and diviners (yinyang).¹¹ The Mongols also retained the smelter and saltern households clustered around mines and salt marshes.¹² Tax extraction remained capricious until well into the reign of Kublai Khan (1260–94).¹³ Nonetheless, the basic outlines of this complex household system were in place by the time the Mongols conquered South China from the Song in 1279.

As they incorporated the former Song territories, the Mongol household system shifted again, this time to incorporate the distinctive non-agrarian groups of South China. Reorganization of the former Song territories began with the imposition of existing categories, starting around the time of the first provisional census in 1290.¹⁴ But it also involved the creation of new household groupings to incorporate hunting, fishing, and mining groups. By 1294, an island jurisdiction near Ningbo created new categories for tea households, boat households (chuanhu) taxed in shark skins, and hunting households (buhu) taxed in fox pelts.¹⁵ By the early 1300s, one county near Nanjing (Jinling) had more than eight hundred gold-panning households (taojin hu), including the ancestors of the Ming founder Zhu Yuanzhang.¹⁶ These specialized tax categories appear quite similar to the better-documented system used by the Qing for taxing hunters, pearl divers, and mushroom pickers in the north, a system that probably developed from the Mongol legacy.¹⁷

By the 1300s, some sources began to organize the new household groups into superordinate categories like North Chinese (Hanren) and Southerners (Nanren), categories that many historians have viewed as a racial hierarchy with Mongols at the top and their sedentary subjects at the bottom.¹⁸ But while the Mongols may have had both implicit and explicit bias against their Chinese subjects, this was only loosely systematized.¹⁹ Throughout the late thirteenth and early fourteenth centuries, household categories were principally organized for tax purposes, and in an erratic and highly localized fashion.²⁰ Far from a uniform or ideological imposition of racial hierarchies, they emerged through a process of trying to tax disparate, complex, and mobile populations. Like the Song’s tea and saltern households—and the hunting groups later organized by the Qing—the Yuan’s hunting and fishing families represented an attempt to incorporate new peoples and new environments into the state’s fiscal regime.

VILLAGES AND THEIR DISCONTENTS

Mongol rule in China declined starting in the 1350s, before giving way to the radical vision of Zhu Yuanzhang and his Ming dynasty in 1368. Perhaps due to his personal history at the margins, and certainly in reaction to the excessive extractions of the Mongols, Zhu attempted to create a system of self-sufficient villages. When the Ming compiled a census and land surveys, the goal was not to maximize revenue; they were intermediate steps toward the reorganization of the population into administrative villages (lijia) starting in 1381 and culminating with the yellow registers a decade later.²¹ This village system represented the centerpiece of Zhu’s policies, used for both organizing revenue and social engineering. Nominally, each village was created as a group of 110 households that oversaw tax collection, labor service, policing, and dispute resolution. The ten wealthiest households took turns serving as village head (lizhang), each taking on responsibilities for ensuring the village’s tax payments for one year in the ten-year rotation. The hundred ordinary households were split into ten “tithing” groups (jia) and performed the more menial duties, also on a decennial basis.²² Villages were further grouped into a spatial hierarchy of wards (tu), townships (du), and cantons (xiang) within each county.²³ Officials used these village hierarchies to make the government self-sufficient. Each office set quotas for the goods needed to maintain itself, including a wide range of items like fuel, paper, and wax for government offices; arrows and uniforms for military garrisons; and even game for official banquets.²⁴ Officials then divided these quotas among subordinate jurisdictions: between the counties in each prefecture and province, and between the townships and villages within each county. While based on earlier systems of mutual surveillance, the early Ming villages represented a new high-water mark of the penetration of governmental oversight beyond the limits of the formal state.²⁵ In areas that had only been marginally integrated into the Song and Yuan, village administration also marked the beginnings of a widespread documentary culture and fixed social units in relation to the state instead of purely by family or tribe.²⁶

Yet while Zhu Yuanzhang reintegrated many of the Yuan’s professional household categories into uniform administrative villages, he retained several important distinctions between status groups, including the main division between commoners (min), artisans (jiang), and military (jun), as well as the more locally specific categories for saltern and tea households.²⁷ Zhu even extended and amplified some of the marginal household categories that counted groups engaged in woodland and wetland economies. In 1382, the year following village registration, Zhu also required boat people to register as fishing households (yuhu) at river mooring stations (hebo suo) throughout the empire.²⁸ In addition to subjecting them to mutual surveillance, this registration held fishing households responsible for annual payments of marine goods. While based in “boat” or “fishing households” of the Yuan, the mooring stations centralized taxation of a peripatetic population and extended it to regions of the southeast coast largely untouched by earlier states.²⁹ In other areas, the Ming retained hunting households (liehu or buhu) that had been established in the Yuan, gradually expanding the program in the fifteenth century.³⁰ The Ming even created a separate category for three thousand households near Nanjing who were specifically required to cut reed fuel for the capital.³¹ Elsewhere, the Ming worked to collect the major products of each local environment: game, hides, and feathers from Fujian; tung oil and lacquer from Jiangxi; timber and bamboo from Zhejiang; and, most importantly, fuel.³² While Zhu Yuanzhang imagined an empire of uniform villages, this vision gave way to a more pragmatic administration that aimed to incorporate and tax non-agrarian peoples and non-agrarian landscapes as well. In the end, this was a less radical departure from the Mongols than it appeared.

The village system also brought a new slate of fiscal problems, including many related to the goods quotas, which were too inflexible to respond to changes in local environments or governmental needs. While village standardization allowed for easier budgeting, the household system was also rife with regional irregularities. Because taxes were based on official demand, levies were not distributed uniformly or according to local productivity. Border regions, transit corridors, and the hinterlands of the capital were taxed especially highly to meet the needs of nearby government offices. Paradoxically, these policies also opened the door to a radical departure from the ideals behind them, as the greater state penetration into local economies enabled a massive expansion of state levies.

In 1398, the Ming founder died and was succeeded by a grandson, whose reign had barely begun before he was deposed by his uncle, who seized the throne to rule as the Yongle emperor. While Yongle governed with his father’s autocratic style, he showed no commitment to the principles of self-sufficiency behind the tax quotas and village system. Instead, he oversaw a massive expansion of the physical infrastructure of the state—building his personal estate at Beijing into a massive new capital, dredging the Grand Canal to supply Beijing, and launching expeditions to the Indian Ocean, the steppe, and Vietnam. The Yongle reign effectively marked the Ming’s second founding. It left two long-term legacies: dual capitals—Nanjing on the Yangzi and Beijing in the north—and the social and environmental consequences of a massive expansion of the command economy.

Leaving the specifics of building ships and palaces to later chapters, it is worth emphasizing the sheer scale of labor service requisitions to supply them. Between 1406 and 1420, perhaps a million laborers were conscripted throughout the empire to construct the imperial palaces in Beijing.³³ By very rough estimate, more than a million large trees were cut from the frontiers to supply this construction, requiring another labor force of a million or more loggers.³⁴ Between 1411 and 1415, another 165,000 laborers were conscripted to dredge the Grand Canal and build embankments.³⁵ The dikes were constructed from wooden fascines, which probably required a comparable force of corvée laborers to cut timber and bamboo. While these laborers were drawn from all parts of the empire, the burden fell especially hard on artisan households and on the regions neighboring the projects themselves. To pay for his large projects, Yongle printed large quantities of paper money. By 1425, government-issued paper currency circulated as low as 2 percent of its face value and was effectively abandoned.³⁶ Collectively, these efforts strained both forced labor drafts and cash economies to their breaking points.

When the Yongle emperor died in 1424, the empire must have heaved a collective sigh of exhaustion and relief. By then, the Beijing palaces and Grand Canal were largely completed. The court launched one final fleet under Zheng He’s command in 1433 before canceling the missions entirely. The culmination of these large-scale projects greatly reduced the demand for labor, yet it is clear that imperial policy also shifted away from such outsize demands on labor and material. Under pressure to decrease the massive and irregular corvée burdens imposed by Yongle, his successors the Hongxi and Xuande emperors sought to return to policies of self-sufficiency. They canceled many projects outright. In 1425, Hongxi issued an edict that “wherever the government had placed restrictions [jin] on mountain workshops, gardens, forests, lakes, wetlands, kilns and foundries, fruit trees and beehives, all [was] to be returned to the common people.”³⁷ Further edicts under the Xuande emperor clarified and broadened this rule.³⁸ The state also faced the unintended consequences of Yongle-era monetary expansionism. Following the collapse of the paper money supply, rich households hoarded silver and copper cash, plunging markets into currency famines and depression.

The excessive levies of the Yongle era also led to widespread tax evasion and even emigration from the most heavily taxed areas. To avoid reporting for corvée, some households fled their registration entirely, either absconding to the frontier or becoming subservient to larger households. Others falsified their registration status, hiding wealth and workers, and even changing household category to avoid the more onerous forms of labor service. By the mid-1400s, the official census had little correspondence with the actual population and entire villages were filled with ghost households. This only increased the burden on those families whose registration remained current.³⁹ Due to the collapse of the currency and retrenchment from large projects, the trend toward tax evasion occurred at a time when the bureaucracy was both overextended and resource-poor. For more than half a century, the household census and land registers contained little more than empty figures, often copied from the previous decennial surveys.⁴⁰ Population and landholding figures are almost entirely missing for the years between 1421 and 1491, and sometimes even later.⁴¹

SILVER ACCOUNTING

In the mid-fifteenth century, officials began to innovate new ways to function within the constraints imposed by their predecessors without resorting to the same extractive tendencies that led to the near collapse of the early Ming system. Through local experimentation, officials in the interior south gradually began to resolve the worst problems with the corvée system. In several Jiangxi counties in the 1430s, Ke Xian introduced an “equalized corvée method” (junyao fa) that merged corvée under four main headings called “four levies” (sichai). The four headings included administrative village duties (lijia), which were mostly related to tax collection; “equalized corvée” (junyao), which grouped together most of the miscellaneous goods levies; and more self-evident categories for postal service (yizhuan) and militia (minbing). In the 1440s and 1450s, the next generation of officials brought this reform to higher-level jurisdictions: Han Yong and Xia Shi to the rest of Jiangxi, and Zhu Ying to Fujian, Guangdong, and Shaanxi.⁴² Much like the mid-Yuan tax reforms that consolidated land tax categories, Ke’s equalized corvée made little change to the levies themselves; but by classifying them into standard categories, he made it easier to redistribute the burden among households and villages. This provided temporary relief for the communities most damaged by earlier extractions, but it also failed to resolve the fundamental inequities that led to tax flight.

The second major shift in household levies came on the heels of the equalized corvée reforms, when an influx of silver allowed officials to convert household impositions to cash payments. As officials rebalanced corvée assessments, they also began to substitute silver surcharges for some of the in-kind levies of goods and labor. This began in the 1450s, when the local magistrate Han Yong converted some county-level levies for ritual goods into payments called “public expense silver” (gongfei yin).⁴³ Over the next eighty years, other officials in Jiangxi and western Fujian expanded this conversion to encompass most household levies. Conveniently, the equalized corvée reforms had already compiled these levies into neat and consistent categories, making substitutions easier; in some places they were even called “equalized silver” (junping yin).⁴⁴ Yet whether assessed directly or commuted to silver, the quotas became uneven over time, leading inevitably to tax evasion.⁴⁵ As long as levies were assessed on the household, families continued to flee or falsify their status to avoid extractions.

By the mid-1500s, local officials again reported widespread tax flight across the regions of the south most responsible for levies of forest products. In southern Jiangxi, the magistrate Hai Rui reported, “[Whereas] prior to 1551, [this county] had forty-four villages, today it has only thirty-four … [and] there are many half villages, or even villages with only one, two, or three parts in ten.”⁴⁶ In a nearby county, the magistrate Qian Qi wrote, “One village often takes on many villages’ levies, [and] one household often takes on several households’ corvée. Requisitions are excessive, and goods and labor are insufficient.”⁴⁷ Xu Jie, another local official, summed up the situation in Jiangxi, where he said, “[People] are not troubled by the land tax, they are troubled by corvée.”⁴⁸

In the mid-1500s, officials finally resolved the paradox of corvée: instead of applying fixed quotas to mobile and changing populations, they imposed them on immovable assets. The first generation of reforms, started by Ge Gaiyi in 1522, followed the previous strategy of commuting levies to silver and redistributing them. But instead of redistributing this silver quota by village, he divided each county’s corvée quota by its total tax assessment and imposed it as a surcharge on each picul of tax grain. He called this “village equalization” (lijia junping).⁴⁹ Around this time, the leading scholar Gui E submitted a similar proposal to the Board of Revenue, intending to carry it out empire-wide, but received no response.⁵⁰ In the 1550s and 1560s, a second generation of reformers took up the idea again, dividing the total corvée duty in each county by its total acreage, ignoring the problematic census, and creating a single line-item tax. Replacing “single line reform” (yitiao bianfa 一條變法) with a more poetic homophone, they called this the “single whip method” (yitiaobian fa 一條鞭法). For two decades, province-level officials including Wang Zongmu, Cai Kejian, and Zhou Rudou attempted to convert and redistribute corvée throughout Jiangxi. In the face of opposition from princes with estates in the province, the measure repeatedly failed. Finally, in 1572, a third generation of officials led by Liu Guangji promulgated the policy province-wide.⁵¹ Meanwhile, Wang Zongmu and Hai Rui, both of whom had experience with the single whip at low-level posts in Jiangxi, spread the new accounting method to Shandong and the Southern Metropolitan Region, while Pang Shangpeng implemented similar policies in Zhejiang.⁵² Finally, in 1580, Grand Secretary Zhang Juzheng promulgated the single whip empire-wide.⁵³

The shift to silver accounting in the late fifteenth and the sixteenth centuries came from the grass roots, as local officials learned hard truths about the changing fiscal landscape. They realized that the land tax was harder to avoid than labor levies, if only because the land was fixed and workers could move. People continually gave birth, died, and moved, rendering population data quickly obsolete, but land remained in place. This meant that once land had been registered, officials needed only to update information on its ownership. Proprietors were also generally inclined to keep the state abreast of land transactions: buyers had incentives to register property to have an official record of ownership in case of a title contest, and sellers had incentives to change the registration to decrease their tax liabilities. Placing the silver surcharge on acreage therefore made it far harder to evade than the earlier levies based on household composition. Yet despite its advantages over earlier reforms, the single whip method was essentially an accounting trick that redistributed the existing tax quota. To reform Jiangxi’s taxes in 1570, Liu Guangji did not send agents into the countryside to survey landholdings or count households; he locked seven leading officials and a tax expert in the examination hall to perform calculations.⁵⁴

Despite the lack of administrative outreach, the single whip reforms totally transformed the Ming state’s position in the economy, replacing direct levies of goods and labor with a silver budget. In fifteenth-century Chizhou, villages rotated responsibility to send 104 woodcutters to cut fuel for county, prefecture, and Nanjing offices; in the sixteenth century, the prefecture hired 104 workers, each paid twelve taels of silver out of the land tax surcharge.⁵⁵ Many levies of woodland products were also replaced with goods bought with tax silver. Prior to the single whip reforms, Jiangxi collected just over two thousand catties of tung oil (around one thousand kilograms), most of which was sent directly to the warehouse that supplied court workshops; following the reforms, tung-producing areas split a tax surcharge of sixty silver taels, which they forwarded to court to purchase oil.⁵⁶ Almost all woodland levies likewise shifted to line items in a silver budget.

The conversion of thousands of discrete revenue items into their cash equivalents made little overall difference in state budgets. Using the best available figures from seventeenth-century Jiangxi, non-agrarian, non-textile goods accounted for only 2 percent of the value of the autumn tax and essentially none of the summer tax.⁵⁷ But the effect on laborers was profound, effectively transforming a forced labor economy into a cash economy. In heavily wooded areas, forest products were often the only goods to sell to earn the cash needed for tax payments. The 1632 gazetteer from Kaihua County, in western Zhejiang, noted the importance of timber exports in meeting this new expense: “In Kaihua County there is little farmland and the people plant fir for their livelihood. They log every thirty to forty years; this is called ‘clearing the forest.’ In this county fir is the best local product, followed by ginger and lacquer, and then charcoal. The profits from ginger, lacquer, and charcoal are only one-fifth those of timber. If you ask the elders, they will say that the profits from fir are not less than ten thousand [taels] per year. For this reason few households flee the land tax.”⁵⁸ Kaihua was just one of dozens of counties where local livelihoods relied overwhelmingly on forest products and that now needed to sell them to raise cash. Even in Liaoning, in the far northeast, households previously responsible for cutting fuel and fir timber for the state now owed silver payments instead.⁵⁹

As forest communities were exposed to the market, they found that each product brought its own commercial strengths and weaknesses. Timber, especially fir, was the most valuable, but it required risky multi-decade investments. Other products offered shorter harvest cycles but lower profits. These included bamboo, both structural poles and edible shoots; firewood and charcoal; dyestuffs like indigo; fiber crops like hemp, ramie, mulberry bark, and palm fronds; and drugs and spices like ginger. Other goods could be harvested repeatedly once trees matured. These included resins like lacquer and pine tar (songzhi); oleiferous fruits like tung, tree tallow (wujiu youzhi), and camellia seed oil (chayin you); and a wide range of edible fruits and nuts. Like Kaihua, most communities diversified their options by planting a variety of forest crops. There were clear regional specializations: tallow trees (wujiu) near the mouth of the Yangzi, tung in Yuanzhou, indigo in Ji’an, drugs in Linjiang, paper mulberry in Guangxin and central Hunan, tea in Zhejiang and northern Fujian, and citrus and lychees in Quanzhou. Some forest production fed substantial industrial development, including pulp for Yanshan’s papermakers and fuel for Jingdezhen’s porcelain kilns. The runaway growth of these commercial-industrial towns (zhen) in the sixteenth and seventeenth centuries stymied attempts to fit them into the normative hierarchy of administrative cities.⁶⁰ But fir was king, probably accounting for more than half of forest acreage across broad swaths of South China. Throughout this forest belt, the state removed itself from controlling production or transport of most of these goods. Instead, the market was made by two major diasporas working in concert with the forest owners: timber merchants, who were principally from Huizhou, at the center of the Jiangnan forest belt, and forest workers, who were overwhelmingly from the Wuyi Mountains at the corner of Fujian, Jiangxi, and Guangdong.

TRANSPLANTS AND MIGRANTS

Of the two groups of sojourners involved in spreading commercial forestry, Huizhou merchants had the longer documented history in the trade. Traveling abroad from their mountainous prefecture in the middle of Jiangnan, Huizhou merchants dominated markets throughout South China by the mid- to late Ming.⁶¹ The rise of this remote and mountainous region to commercial preeminence was a direct outgrowth of its role in wood markets. In the twelfth century, Huizhou was at the epicenter of the revolution in tree planting. Huizhou also had a direct river route to the Song capital at Hangzhou and received special treatment under the tariff regulations of the twelfth century.⁶² This gave Huizhou a key advantage in the nascent market in commercially planted timber. As more timber entered the cash economy, gradually in the fourteenth and fifteenth centuries, and in a great rush in the sixteenth and seventeenth centuries, Huizhou merchants dominated the sale of other regions’ timber as well. The Kaihua gazetteer quoted above cites the special role played by Huizhou merchants in this market. While noting that timber enabled Kaihua landowners to meet their tax burdens, it added the caveat that “they must let men from Huizhou clear the surplus [timber] and transport these goods to Jiangnan without obstruction.”⁶³ Similar anecdotes confirmed the position of Huizhou merchants throughout the Yangzi River basin and down the southeastern coast. Through their market advantage established in the Southern Song, Huizhou traders developed the connections, the capital, and the expertise to emerge as the preeminent timber wholesalers in the Ming.

Huizhou’s importance in wood markets was further cemented by the growth of nearby Jingdezhen as a porcelain production center. A market town just south of Huizhou, Jingdezhen developed as its kilns were tapped to supply porcelain to the Yuan court in the thirteenth century and to the Ming court in the fifteenth century.⁶⁴ In the sixteenth and seventeenth centuries, the state removed most monopolistic strictures on porcelain production, and Jingdezhen became the preeminent maker of blue-and-white ware, not just for the court, but for the world market.⁶⁵ Throughout the growth of Jingdezhen into a major industrial center, Huizhou was its main fuel supplier, again via direct water routes, and Huizhou merchants were involved in shipping ceramics out of Jingdezhen.⁶⁶ Huizhou moneylenders also became Jingdezhen’s main source of working capital to bridge potters from production to sale.⁶⁷ Once again, their early position in the market allowed them to dominate.

From their advantageous position in Hangzhou’s officially regulated timber markets, Huizhou merchants expanded horizontally to become the most important middlemen in the timber trade. Building on their positional advantage in Jingdezhen’s fuel markets, Huizhou merchants expanded vertically to dominate trade and finance in the world’s greatest porcelain industry. By the late Ming, they were by far the most important timber traders across the south—not only in neighboring regions like Jingdezhen and Kaihua, but throughout the Yangzi River networks; in Fujian, Guangdong, and Guangxi; and along the Grand Canal (see map 3.1).⁶⁸ As they traveled across the south, buying and selling timber, fuel, and a plethora of other goods, Huizhou merchants were also a key vector for the transmission of expertise on forest management, financing, and market conditions.

As Huizhou merchants took pole position in the trade in wood products, tree planting became especially associated with another diaspora: peoples from the Wuyi Mountains in Fujian. Until the sixteenth century, this region was populated by a heterogeneous and shifting group of tax refugees from the lowlands and non-Sinitic upland peoples.⁶⁹ These groups interfaced with the Chinese state indirectly, in part by submitting tribute in tea, timber, or mine or animal products. This changed with the twilight of direct levies in the sixteenth century, a policy shift that made these marginal highlanders choose a side: either fuller integration into the growing commerce in forest products or retreat from participation in lowland society. A sixteenth-century account notes this bifurcation:

In Fujian there are wandering people that grow tea, mainly of the three surnames Pan, Lan, and Lü. They used to share a single ancestor but have since divided.⁷⁰ Those who do not enter their names into household registration wander the cliffs and abandoned lands, living there by farming and hunting to feed themselves, but not paying land tax or corvée. They wear their hair bound into buns and go barefoot. Each is subject to a chief, who they call elders, and who wear kerchiefs and long robes. [Others] travel abroad to various places, never staying for long: from Tingzhou they go to various prefectures in Jiangxi, where they produce fir to export via the mouth of the Ben River [a tributary of the Yangzi]; to Huizhou, where they produce firs to export via the mouth of the Rao River [a tributary of Poyang Lake]; and to Zhangzhou, where they produce fir that is sent to eastern Zhejiang by sea.⁷¹

This passage neatly summarizes how the fissiparous mountain societies of southwestern Fujian split into two discrete groups. The first group—those who did not register their households or pay taxes—retained their identity as non-state peoples. They came to be known as She, perhaps a term for the shifting cultivation they practiced. The second group traveled abroad from their Wuyi homeland, entering lowland society as forest laborers. They came to be known as Hakka (kejia or kehu), often translated as “guest families” but perhaps better understood as “sojourners.”⁷² While the sixteenth-century emergence of the Hakka was driven by multiple factors, the shifts in forest corvée are the missing piece of this puzzle. Previous scholarship has focused on the role of Hakka migrants as miners—especially in Guangdong—and in growing cash crops like hemp, ramie, and tobacco.⁷³ But as this anecdote shows, they were also heavily involved in the forest economy, planting timber on plantations from coastal Fujian to Huizhou and Jiangxi (map 3.1). The pull of growing markets for forest products allowed the Hakka a place at the fringes of the Chinese state as forest and mining specialists. But unlike in previous periods, their role in the forest economy was demarcated through private arrangements rather than officially designated household categories.

VILLAGERS AND SOJOURNERS

For hundreds of years, China’s household registration system was predicated on keeping families in place. The state wanted people to stay in their villages so they could be surveilled, and, more importantly, so they could be taxed. By fits and starts, the Song, Yuan, and early Ming states extended the household system from the main tax base of farmers to tea growers, salt producers, miners, fisherfolk, hunters, and dozens of other non-agrarian peoples; in the process, they assigned these peoples to specialized households in an attempt to fix them in administrative space. Conversely, when Chinese states wanted people to be mobile, they placed them in groups that allowed—or enforced—mobility, especially craft (jiang) and military (jun) households. Arguably, this predilection toward enforced localization persists in modern China’s household (hukou) system. Yet the assumption of fixed residence was a poor model of behavior, even for farmers, who fled and falsified their household registration to avoid excessive corvée. It was even less accurate for the shifting vocations of fisherfolk, placer miners, hunters, swidden cultivators, loggers, and the other peripatetic peoples of the non-agrarian fringe.

To make the thankless job of meeting quotas still more difficult, tax collectors had to deal with changes in the land as well as the movement of people. Landowners gradually converted much of the diverse natural woodland of South China into artificial forest plantations. They greatly increased production of a handful of tree species—especially pine, fir, and bamboo—at the cost of substantially reduced habitats for a much wider array of woodland plants and animals. A small number of forest owners curtailed the rights of a much larger number of others, who lost their freedom to hunt, gather, pasture animals, cultivate crops, and cut wood in what had long been open-access areas. Like the assumptions of fixed residence, static tax quotas also worked poorly in a changing fiscal landscape. Even grain and fiber-crop yields could fluctuate with the weather and decline as soils were depleted. The premise of meeting fixed targets of wild goods like furs and mushrooms was even less sustainable, as Jonathan Schlesinger shows in his study of Qing levies in Manchuria and Mongolia.⁷⁴

As officials struggled to reallocate quotas and rebalance the corvée system, the commercial economy offered another solution. For centuries, rural entrepreneurs had invested in planting timber trees and other forest crops. Starting in the late 1400s, these commercial forest economies were buoyed by an influx of silver. By fits and starts, county magistrates began to take advantage of the growing money supply by converting their local levies from the command economy to the cash economy. Not coincidentally, these reforms emerged from the hotbeds of commercial silviculture in the middle south, especially Jiangxi and Fujian. Eventually, generations of gradual policy change came together in the revolutionary single whip reform, an accounting method that reduced the plethora of labor and wild goods extractions into a single surcharge assessed on fixed, cultivated land. In the 1560s and 1570s, provincial administrators promulgated the single whip across their jurisdictions; in 1580, Zhang Juzheng took the method empire-wide. Now, instead of taxing wax and timber directly, officials bought them out of a silver budget, and instead of providing fuel or game to the state as a condition of their registration, many households now looked for similar work on the market.

In addition to centralizing accounting, the single whip reforms meant that the state intervened far less in local relationships between land and labor. As these reforms did away with the direct pressures of labor service, they also created a new imperative to earn cash to pay taxes. The shifts in forest markets and goods levies created both new pressures and new opportunities for households on the wooded fringes of agrarian society. Some, like the Huizhou merchants, leaned hard into the silver economy, registering their property and their commerce with the state. Some, like the She, stayed on the margins as shifting cultivators and non-state peoples. And some, like the Hakka, found themselves in a rather unhappy middle ground. The end of corvée did not spell an end to coercion; it simply left relations between land and labor to be worked out between individual households.⁷⁵ If anything, the end of official labor drafts led to a proliferation of new forms of private, contractual subordination, including shifts in bond servitude, or subordinate relations to others’ households, and changes in tenancy, or subordinate relations to others’ landholdings.⁷⁶ In chapter 4, I explore the implications of this contractual market for the management of commercial forest plantations.