From The Journal of Asian Studies 61, no. 2, May 2002
IN THE GREAT DIVERGENCE, Kenneth Pomeranz (2000) proposes a radical revision of our understanding of the pattern of economic evolution in the eastern and western ends of Eurasia over the course of the early modern and modern periods, roughly late Ming and Qing. A recent restatement of the standard or traditional view can be found in the macro-economic historian Angus Maddison’s account of world economic development in the very long run, The World Economy: A Millennial Perspective (2001), which sums up his argument in Chinese Economic Performance in the Long Run (1998). According to Maddison, “Western Europe overtook China . . . in per capita performance in the fourteenth century. Thereafter China [was] . . . more or less stagnant in per capita terms until the second half of the twentieth century” (2001, 44). In contrast, Pomeranz insists that if the comparative focus is placed, as only makes sense, not on Europe or China as a whole—both of which contained the most disparate regions at vastly different levels of economic development—but on the most advanced, or core, areas within each, it can be seen that, by as late as 1800, there was little to choose between them, in terms of the character of the economy, the nature of growth, or its results (2000, 7–8).1 In both the most advanced regions of China, specifically the Yangzi delta, and the most advanced regions of Europe, specifically England, the institutional frameworks and demographic patterns were equally favorable to growth, the trajectories of economic evolution were similar, the level of economic development in terms of the standard macroeconomic indicators (growth of capital stock, productivity, GDP per capita, etc.) were basically indistinguishable, and the standards of living were at a roughly common level. It was only during the first half of the nineteenth century that England (and then Europe) embarked upon a distinctive path of growth that led to a definitive divergence from the trajectory of the Yangzi delta. It was able to do so, however, not as a consequence of any advantage possessed by its domestic economy, but rather as a result of its unique form of mercantile state and merchant companies that made possible its access to the land, raw materials, and above all slave labor, of its American colonies.
In what follows, we shall begin by specifying Pomeranz’s thesis and spelling out in highly schematic form the main lines of our critique. We shall then develop our argument by contrasting the economic evolution of the Yangzi delta and England in two successive phases, the early modern period and that of the classical industrial revolution, and counter-posing the outcomes in terms of level of development and standard of living.
Pomeranz’s thesis is complex. At the most general level, he sees Smithian growth taking place as a consequence of the deepening of the division of labor, by means of the growth of exchange driven by population increase. In this conception, the rise of dense population tended to make for economic development by making for the growth of markets and thereby ever greater specialization—so long as institutional or demographic barriers to growth were absent, or had been transcended. As Pomeranz puts it, “[p]opulation density is not the sole determinant of Smith’s ‘extent of the market’. . . . But for elaborate specialization to be developed in many areas of economic activity—food, clothing production, building, transport, and exchange itself, there is ultimately no substitute for having many people within affordable physical or cultural distance” (2000, 26). The causal chain, for Pomeranz, thus runs from demographic growth to the increase of demand to the increase in supply via specialization and the growth of the division of labor—making for a long-term trend of moderately increasing output per unit of labor input (2000, 98–99, 226, 212–13, 215). In Pomeranz’s view, demographic expansion, accompanied by the rise of exchange, thus brought both similar, equally impressive processes of economic evolution in the Yangzi delta and England and similar, equally impressive outcomes in terms of welfare. It did so, first, because the Yangzi delta encountered no greater institutional barriers to growth than did England. Its land, commodity, and labor markets were as free, while the nature of its commercial property and the rules surrounding it no less supportive of economic growth (2000, 17–18, 70–85). It did so, secondly, because the Yangzi delta demographic regime was no more subject to Malthusian processes that ran counter to the requirements of capital accumulation than was its English counterpart. Chinese families controlled their fertility (more or less) to the same degree as their English counterparts and thereby prevented to an equal degree their surpluses being swallowed up by too many children, directly or via subdivision on inheritance (2000, 40–41). It did so, finally, because England possessed no decisive advantage in technology over China—or, perhaps more precisely put, because advantages that England possessed in terms of industrial technology were counter-balanced by Chinese superiority in terms of agricultural technology (2000, 47–48). As Pomeranz puts it, “it is easy to imagine [England’s] marked technological backwardness in the largest sector of eighteenth-century economies [agriculture] having a significance as great as whatever advantages it had in other sectors” (2000, 45).
The outcome was that, in the mid- to late eighteenth century, neither England’s level of development nor the level of welfare of its population was above the Yangzi delta’s; indeed, it is likely the opposite was the case. English capital accumulation was no greater than China’s, and any relative difference in the cost of capital did not necessarily go to England’s favor (2000, 31–36, 42–43, 170–72, 178–79). The level of English productivity was no higher than was the Yangzi delta’s, any disparity in England’s favor in terms of manufacturing productivity probably outweighed by the Yangzi delta’s advantage in agricultural productivity (2000, 45, 216–18). As a result, levels of nourishment and well-being more generally were similar as definitively evidenced in the rough equality of life expectancy in the two places (2000, 36–40).
On the other hand, Pomeranz follows Wrigley (1988), himself inspired by the classical economists Smith, Malthus, and Ricardo, in seeing Smithian growth as tending to come up against an ultimately impassable ceiling resulting from the limited supply of land. In what Wrigley terms the “organic economy,” which basically persisted until the industrial revolution, the land provided not only all of the food for the population, but also all of the basic raw materials, by way of plants and animals. In time, demographic growth was thus bound to lead to rising and conflicting demands on the land for food and raw materials. This would impel, argues Pomeranz, a resort to ever more labor-intensive methods to compensate for land scarcity, and, sooner or later, Malthusian-cum-ecological crises. It was only the harnessing of inorganic, i.e., mineral, sources of power that made possible the avoidance of the stationary state described by the classical economists. Coal, steam power, and ultimately the production of chemical fertilizers for agriculture insured the transition to the inorganic economy that was, for Pomeranz, the defining achievement of the industrial revolution.
Pomeranz sees the barrier set by the limited supply of land as not just theoretical but a looming reality in the late eighteenth and nineteenth centuries. All across Eurasia, including both the Yangzi delta and England, economies were coming up against shortages of food, fuel, building materials, and fibers, and these shortages were forcing them to turn increasingly to labor-using techniques to prevent a decline in living standards (2000, 12, 216–42).2 Pomeranz explicitly rejects Sugihara Kaoru’s contrasting characterizations of development in the East as labor intensive and in the West as capital intensive. “There are many signs,” he says, “that substantial regions in Europe were headed down a more labor-intensive path . . . [and] we will find evidence in aspects of agriculture and proto-industry throughout Europe (including England)” (2000, 13). All else being equal, between 1750 and 1850, both the Yangzi delta and the English economy were careering toward what Pomeranz calls the “protoindustrial cul-de-sac” (2000, 206–7, 241).
Nor could the forms of internal expansion pursued by the Yangzi delta and England/Europe, which amounted in both cases to the simple extension outward of the area of settlement, produce peripheral economies that could aid decisively in transcending the land constraint that was increasingly fettering development in each region. This was because neither the Yangzi delta’s periphery, particularly the middle Yangzi region of Hunan-Hubei and Anhui, nor England’s within Europe, particularly Eastern Europe and Russia, was able to offer sufficient markets for their respective core’s exports of manufactures, with the result that neither was able to supply, in return, the food and other raw material supplies that both cores needed. Not only did producers in both internal peripheries fail to demand the core’s industrial goods in sufficient quantities, at least in the Chinese case, they ultimately began to produce such goods themselves (2000, 22, 249–51, 287–88).
What ultimately opened the path for England’s great divergence from the Yangzi delta was the distinctive form of English and European mercantile expansion, which had no counterpart in the East. This made possible the establishment of the unique raw material and food-producing peripheries using coerced slave labor that proved indispensable in enabling England, and later Europe, to transcend the tendency to Malthusian-cum-ecological crisis, while the Yangzi delta was unable to do so. As Pomeranz puts it, “the projection of interstate rivalries overseas . . . [along with] joint stock companies and licensed monopolies turned out to have unique advantages for the pursuit of armed long-distance trade and the creation of export-oriented colonies.” These “relieved the strain on Europe’s supply of what was truly scarce: land and energy” (2000, 19–20). By contrast, in the Yangzi delta, “smoothly functioning regional markets and interdependencies conflicted with the growth of empire-wide markets, especially after about 1780; this made it harder for one or two leading regions to keep growing and to avoid having to adopt even more labor-intensive strategies for conserving land and land-intensive products. Thus, freedom and growth in the peripheries without dramatic technological change led the country as a whole toward an economic cul-de-sac” (2000, 22).
Mercantile states and merchant companies, which did distinguish themselves from anything that could be found in China, were thus able to make use of ample coercive capacities to impose a distinctive form of political economy in the American colonies that enabled the core economies to meet their critical foodstuff and raw material requirements. Specifically, populations of highly specialized slave producers, who (for the most part) confined their output to a single monocrop and therefore required imports of both food and manufactures to secure their subsistence, produced raw materials, initially on command, that the core needed in order to avoid coming up against the land constraint. As Pomeranz puts it, the ecological relief provisioned by the New World “was predicated not merely on the natural bounty of the New World, but also on ways in which the slave trade and other features of European colonial systems created a new kind of periphery, which enabled Europe to exchange an ever growing volume of manufactured exports for an ever growing volume of landintensive products” (2000, 20, emphasis in original). New World sugar and grain for food, cotton for fibers for manufacturing, and timber for heating enabled England, then Europe, to forego the massive acreages and additional labor that would otherwise have had to be allocated to produce them. England “was able to escape the protoindustrial cul-de-sac . . . in large part because the exploitation of the New World made it unnecessary to mobilize the huge numbers of additional workers who would have been needed to use [its] own land in much more intensive and ecologically sustainable ways” (2000, 264).
As Pomeranz concludes, “rather than looking at other advanced economies in the sixteenth through eighteenth centuries as cases of ‘Europe [Britain] manque´,’ it probably makes more sense to look at western Europe [Britain] in this period as a none-too-unusual economy; it became a fortunate freak only when unexpected and significant discontinuities in the late eighteenth and especially nineteenth century enabled it to break through the fundamental constraints of energy use and resource availability that had previously limited everyone’s horizon. And while the new energy came largely from a surge in the extraction and use of English coal . . . Europe’s [Britain’s] ability to take advantage of a new world of mineral-derived energy . . . required flows of various New World resources” (207).
Critique and Alternative
The burden of our critique is that, from the dawn of the early modern period, the institutional framework or system of social-property relations that structured the English economy was radically different from that of its medieval period (1100–1400) and, more to the point, from that of the Yangzi delta during the Qing (1644–1912). As a consequence, the main economic agents in England and the Yangzi delta faced sharply differing constraints and opportunities and therefore found that it made sense to adopt very different economic strategies, or rules for reproduction. The aggregate outcome was that England’s path of economic evolution diverged decisively from that of the Yangzi delta over the course of the early modern period (1500–1750), as it also did from that of most of the rest of Europe at the same time. It was this already existing decisive divergence, not England’s capacity to gain access to the American colonies and their land-saving staple crops, that explains why the two economies distanced themselves ever increasingly from the middle of the eighteenth century, during the period of the classical industrial revolution (1750–1850). Our thesis, stated in the most schematic manner possible, is thus two-fold. In the Yangzi delta, the main economic agents possessed direct non-market access to the means of their reproduction. They were therefore shielded from the requirement to allocate their resources in the most productive manner in response to competition. As a result, they were enabled to allocate their resources in ways that, while individually sensible, ran counter to the aggregate requirements of economic development, with the consequence that the region experienced a Malthusian pattern of economic evolution that ultimately issued, in the eighteenth and nineteenth centuries, in demographic-cum-ecological crisis. In England, in contrast to the Yangzi delta, the main economic agents had lost the capacity to secure their economic reproduction either through extra-economic coercion of the direct producers or their possession of the full means of subsistence. They were therefore both free and compelled by competition to allocate their resources so as to maximize their rate of return (the gains from trade). The region experienced, as a result, a Smithian pattern of economic evolution, or self-sustaining growth, that brought it, in the eighteenth and nineteenth centuries, to the edge neither of demographic nor ecological crisis, but to the industrial revolution.3
The Economic Evolution of the Yangzi Delta and England in the Early Modern Period, 1500–1750
Framework of Social-Property Relations and the Trajectory of Population and Property
Pomeranz sees no reason to view either the institutional framework that obtained in the Yangzi delta or that which prevailed in Britain as more propitious for economic development, because both provided for well-specified and secure property rights, as well as processes of family formation that Pomeranz believes were compatible with the requirements of capital accumulation. The fact remains that the systems of socialproperty relations of each region, though well specified and secure, endowed the main economic agents with rights to very different kinds and amounts of resources and entailed very different ways to gain access to these resources. These agents thus confronted very different economic possibilities and limits, and as a consequence their decisions, taken in aggregate, made for strikingly divergent economy-wide patterns.
In the Yangzi delta, by virtue of the reigning system of social-property relations, both of the main economic agents—peasants and landlords—possessed direct, nonmarket access to the means of their economic reproduction. As a result, they were under no compulsion to buy necessary inputs on the market, could therefore avoid dependence upon the market, and were thus freed from the necessity to enter into competitive production to survive (which is in no way to imply that they wished to avoid involvement in the market). Thus shielded from the pressure of competition, they were enabled to allocate their resources so as to pursue certain goals that were in their own interest but that were nonetheless non-economic in the strict sense of maximizing the gains from trade. Contra Pomeranz, the aggregate outcome was indeed a pattern of family formation, of allocation of property, and of demographic expansion in the Yangzi delta that ran counter to the needs of capital accumulation and Smithian growth more generally. In England, by contrast, economic agents of the sort found in the Yangzi delta—both possessing peasants and lordly takers of rent by extra-economic means—though dominant during the medieval period, had been largely eliminated. Thus, the main economic agents throughout the economy— especially tenant farmers—although in possession of their means of production (tools, animals, and so on), were separated from their full means of economic reproduction, specifically the land. They were therefore dependent on the market for key inputs (specifically the land that they leased), obliged for that reason to produce competitively in order to survive, and thus compelled to adopt a profit-maximizing approach to their microdemographic choices about marriage and child-bearing and their microeconomic decisions concerning the allocation of their resources, especially the land. The aggregate outcome was a pattern of family formation and utilization of land and capital, constrained by competition in production, which provided a powerful foundation for Smithian growth.
Beginning in the late Ming and accelerating over the course of the seventeenth century, peasants in the Yangzi delta secured, most often through purchases, various customary forms of possession and effective permanent tenancy, particularly in the paddy zone.4 They also gained fixed rents, exemptions from rent procurements against second crops, and a variety of favorable terms that reduced the weight of rents. Many of these gains were secured as a consequence of the mid-seventeenth-century population collapse, which strengthened tenants’ bargaining position vis a` vis landlords. They were subsequently consolidated through political struggles in the form of rent resistance movements. As a consequence, some peasants secured full ownership in their land (particularly in northern Zhejiang) while others secured favorable terms as possessors of “topsoil rights” or secure tenants, who could only be expelled for failure to pay their rent. In any case, rents could not be varied, and this, coupled with security in land, amounted to peasants’ effective property (though not, in the Chinese system, legal ownership) in the land (see Bernhardt 1992, 21–27, 228; Huang 2001, 99–118; Liu 1997, 261–79; 1980; Shih 1992, 136–41, 161–64; Walker 1999, 77–79; Rawski 1972, 19–24; Zhou and Xie 1986, 241–52, 271–75, 301–19; Yang 1988, 91–110; Li 1993, 96–113).
As the reverse side of the process by which peasants consolidated their hold on the land, landlords found themselves not only fully divorced from production, but also deprived of the ability to vary rents in accord with the supply and demand for land and thus of the capacity to remove tenants who could not pay what the market would bear.5 By the eighteenth century, they had largely removed themselves from the countryside and taken up residence in the towns (Bernhardt 1992, 17, 19, 21; Shih 1992, 148–55). At the same time, however, these landlords had consolidated their legal ownership of the land and thereby their ability to take what was an essentially politically set level of rent, which, in the Yangzi delta, amounted to 40 to 50 percent of the summer harvest—rice in the paddy zone and cotton in the cotton belt (Bernhardt 1992, 21). Since the burden of taxes had been shifted onto property owners, and rentier landlords paid their taxes out of their rent receipts, the Qing state upheld and enforced the right of landlords to garner rents from their tenants (Huang 1990, 42; 1996, 81; Bernhardt 1992, 13, 30; Jing 1994, 66–67; Mazumdar 1998, 214–16). In the Yangzi delta, the Chinese elite was thus able to extract roughly 30 to 40 percent of the annual agricultural output (Huang 1990, 103).
Since the amount of landlords’ rent was politically fixed and thus entirely independent of peasants’ output, landlords had no reason to concern themselves with, let alone to invest in, agricultural production (Bernhardt 1992, 21, 27). They were therefore free to spend their income on conspicuous consumption, on investment in pawnshops and trade, on the purchase of more land, and, above all, on preparing their children for the civil service examination and/or the purchase of official degrees/offices: official position was still the premier route to wealth and power (e.g., Beattie 1979; Chang 1955; Huang 1985; Elman 1992). Landlords’ high level of extraction was thus a major rentier element at the heart of the delta’s political economy that Pomeranz ignores. It had the effect of channeling massive resources that would otherwise have gone to amplify expenditures by peasants on their means of production and means of consumption into unproductive expenditures.6 This siphoning of surplus restricted a good deal further what would have been, in any case, a highly cramped pattern of peasant-led agricultural and industrial development.
By providing them direct, non-market access to most if not all of their means of subsistence, peasants’ property constituted a powerful shield to competition and thereby enabled them to prioritize critical goals that would not have been feasible to pursue had they been dependent upon the market for their inputs. Peasants thus involved themselves in the market to secure the gains of trade, but did so only to the extent that their so doing did not render them subject to the competitive constraint and thus compelled to eschew highly desired objectives. In order to secure social insurance against illness and old age, as well as to insure the continuity of the patriline necessary for ancestral worship, peasants typically tried to assure that sons who would care for them and perform the appropriate rites survived to adulthood (Fei 1939, 72– 79; Wolf and Huang 1980, 62–65, 164; Wakefield 1998, 198; Shiga 1978, 135– 36). Given the high rates of infant and child mortality, they could not therefore, as a rule, avoid having many children, while also making sure all their sons married early and in turn produced sons of their own who could, if the need arose, also be counted on to support them (Wolf 2001, 136–37, 146–49, 151; 1985, 185; Wolf and Huang 1980, 133–38; Bray 1997, 332–34, 336–43; Liu 1995, 126–27). In turn, they subdivided their holdings, in land, tools, and liquid wealth, in order to provide those sons with the wherewithal for their own maintenance, early marriage, and family formation. The outcome was large numbers of children per woman, rapid population growth (up to a point) and, consequently, a trend to ever smaller peasant plots and the de-concentration of peasant wealth more generally—a pattern that would not have been supportable had peasants been required to form their families and allocate their resources in accord with the requirements of competition and profit maximization (e.g., Wakefield 1998, 185–91, 198, 202–3, 208–9; Shiga 1967, 1978, 113, 117, 135–36; Jing 1994, 54–59; Fei 1939, 65–69; Huang 1996, 25–28, 60–61; Bernhardt 1999, 49–52, 196–97).
Wolf (2001), citing his own work and that of others, estimates that the total fertility rate in late imperial China was at least 7.4, with the number still higher among some Zhejiang women (136–37).7 Population was therefore fast growing, at least through the firsthalf of the eighteenth century (see also Telford 1995, 48–49). In the small Yangzi delta, it tripled between 1400 and 1800, although most of this increase had taken place by the early to mid-eighteenth century. As a consequence of this demographic expansion, plus peasants’ partitioning on inheritance, the average size of household plots (assuming five people per family) fell from 2.9 acres to 1 acre in the small Yangzi delta (Huang 1990, 341–42, table B1).8 Meanwhile, in the big delta, population grew from twenty million in 1620 and 1690,9 to about thirty million in 1750, and thirty-six million in 1850, and the average size of family holdings fell correspondingly—from 1.875 acres in 1620 and 1690 to 1.25 acres in 1750 and 1 acre in 1850 (see Tables 1 and 2 below). Capital at peasants’ disposal was also increasingly dispersed, leaving them with ever diminishing ratios of means of production to labor (capital-labor ratios), although data on this are hard to come by.10 By 1800, population had more or less stabilized and holdings remained at about the same level until the Taiping rebellion and the resulting mid-nineteenth-century mortality crisis allowed for a brief period of land aggregation. But by the early twentieth century, the traditional processes of subdivision had commenced, making again for ever smaller holdings.
The key to England’s economic evolution in the early modern period and beyond was the consolidation of a system of social-property relations in the countryside that broke decisively from that which had prevailed during most of the medieval epoch.
Throughout the medieval period, agriculture had been largely in the hands of economic agents who, like their counterparts in the early modern Yangzi delta, by and large held direct, non-market access to the land, tools, and labor power that they required for their reproduction. Peasants possessed their plots, defending their rights (especially to inherit) by manorial custom and their peasant communities. Lords, for their part, generally took a rent by extra-economic coercion—although as population grew and the number of peasants with insufficient land to provide subsistence expanded, they also turned to farming their demesnes using wage labor or leasing them to peasant tenants (Postan 1966). Largely freed from the competitive constraint, medieval English peasants, like those in the Yangzi delta, tended to subdivide holdings, as well as, it seems, to marry early and to have many children. This was in the service again of assuring the survival of sons who would support them in case of illness or old age. The outcome was high fertility, which made for rapid population growth and ever tinier plots. Between around 1100 and 1300, English population appears to have about tripled (although the numbers remain the subject of controversy) (Hallam 1988, 536–37; Smith 1991, 49; Wrigley 2000, 123–24 n.19). As a consequence, by the late thirteenth century, the average size of peasant plots had shrunk to fourteen acres, and the median size was significantly lower (Allen 1992, 62–64, tables 4.1–4.3). Ten acres was about the minimum required for a family’s subsistence (Titow 1969, 89). During the late fifteenth and sixteenth centuries, in the wake of the late medieval agrarian crisis that was set off by the Black Death (1348–49), the system of socialproperty relations in the countryside was radically transformed through a complex series of developments that can only be crudely encapsulated here (see Brenner  1985, especially 291–99). Following the severe population decline of the late fourteenth century, peasants succeeded, by way of resistance and flight, in destroying the prevailing system of lordly exaction by extra-economic compulsion. Nevertheless, from the fifteenth century onwards, having failed to reinstate serfdom, lords did succeed in asserting their absolute property rights to the greater part of the land. They consolidated their hold on what were, in terms of western European norms of the time, unusually large demesnes. They expanded, moreover, those already large demesnes by appropriating peasant customary land which had been left vacant in the demographic downturn. They acceded, finally, to land held by customary tenants who lacked both the right to pass on their holdings on inheritance and the right to invariable fines on the transfer of their holdings. The emergent class of commercial landlords, unable, as the feudal lords had been, to take their rents by extra-economic coercion, were obliged to depend on rents determined by supply and demand—i.e., what the market would bear. The emergent class of direct producers, now largely separated from their means of subsistence (the land), though still possessing the means of production (tools and the like), were correspondingly obliged to maintain themselves through taking up commercial leases on a competitive land market. Compelled therefore to produce competitively to survive economically, these tenant farmers had to adopt an approach to their economic production that diverged sharply from that of England’s medieval peasantry, as well as from their counterparts in the Yangzi delta during the Qing era.
Subject to the competitive constraint, commercial farmers could not find it in their interest to subdivide holdings, as this would have made for increasingly noneconomic units of production.11 Nor were large families in their economic interest, since these entailed the material support for a number of years of children who could not add to family income as much as they cost to maintain. Because, moreover, the commercial farmers of early modern England, unlike the medieval peasant producers who preceded them, did not subdivide their holdings, they could not directly endow their children with plots that could serve as the basis for family formation. Sons had, therefore, to accumulate on their own the material basis to marry and to support a family—a process that inevitably took time. As a result, marriage occurred substantially later and was less universal than it had been during the medieval epoch or in the Yangzi delta of the Qing era. At the same time, because large farms operated by wealthier tenants tended to beat out smaller ones operated by poorer ones in the competitive struggle (see below), there was a powerful long-term tendency to the build-up of holdings and ultimately a strong turn to wage labor. The outcome, on the one hand, was smaller families (or, more precisely, fewer children per woman), which made for slower population growth (though this may have been partly compensated by falling mortality). On the other hand, one saw the end of subdivision and the rise of ever larger farms. Pomeranz is therefore mi