State-building as the Cause of Institutional Change and Fixation:

The Chinese Case


Kent Deng (e-mail:



I. Introduction

Given the popularity of New Institutionalism and New Institutional Economics in economic history, it remains rather unclear why and how a particular type of institutions took shape and evolved in the past which eventually led to what is called the ‘great divergence’ (citing Pomeranz, 2000).

The explanation given by Douglass North is a Bosurapian one. He argues that that it was population pressure that generated institutional changes in history (see D.C. North, 1981, Structure and Change in Economic History, ch. 7 and pp. 109-10; also E. Boserup, 1965, The Conditions of Agricultural Growth: the Economies of Agrarian Change under Population Pressure). However, there is no systematic proof although it has been regarded more or less as an axiom.

This paper looks into Chinese experience to test this axiom as China’s population growth makes the country a good case to test the Northian notion of the origin of institutional changes. Indeed, from the sheer size and momentum of population growth in China, it should have had the most advanced institutions in the world to facilitate growth in both the short run and long run.

But empirically, China’s institutions which were associated with state-building and which was characterised by private property ownership developed before the population pressure was able to build up. Most intriguingly, China’s key institutions kept going with only minimal amendments/alternation when its population multiplied. Hence, the Northian prime mover for institutional changes had little effect on Chinese history.

One may adopt the old view of oriental despotism and argue that the Chinese institutional fixation was achieved through coercion. Coercion may work in the short run, but not in the long run: two millennia in the Chinese case. So, the sustainability of the Chinese institutions had to be voluntary among the vast majority in society.

In the end, the paper will argue that from the Chinese experience population was neutral in the making of institutions. So, the ultimate cause of institutional change will have to be found in other areas such as national sovereignty and security, political philosophy, political and economic bargaining power between strata, and the peculiar resource endowments for the economy.

II. State-building in China, 770 b.c.–221 b.c.

We have to trace back this far to know why and how China had a very different path compared with the rest of Eurasia.

1. New state-building triggered by a military revolution

In history, there were two main determinants for state-building or state-formation: internal order and external defence. To reduce threat to social stability is the ultimate reason for state-building and state-rebuilding. The Chinese experience also shows that the nature and multitude of the threat determined the type of state: a pluralist competitive kingdom system or a unified empire.

During the early stage of Chinese civilisation, a pluralist competitive kingdom system was well established and entrenched for half a millennium from 770 b.c. to 221 b.c. following the collapse of a feudal like system under the Zhuo (c. 1030–771 b.c.).

During this period of multiple states, political units were small and decentralised across China. And, by 476 b.c., internal stability and social order seem to have been the main concerns most of the time among these units. This was the period when Confucius lived who never saw building a pan-China empire was either necessary or feasible. So, his agenda included good government, social justice and some degree of equality. Indeed, for that matter, Confucius never saw the need for state-rebuilding geared towards a simple government for all the citizens. What he viewed as the norm was a decentralised Zhou feudalism.

An equilibrium of multiple states was maintained before it finally collapsed in 476 b.c. which was followed by a period of 254 years of wars. Before 475 b.c., confrontation and competition among states were not uncommon. But, the main tendency was to fight for hegemony and sphere of influence on East Asian mainland, not to unify the politically fragmented territories (Wang et al. 1986: ch. 3).

The break out of the civil wars after 476 b.c. seemed to have a lot to do with the development of military technology in four areas (1) the use of iron for lighter and tougher weapons to replace bronze (Wei et al. 1993: 10–11), (2) the emergence of calvary with unprecedented speed and manoeuvrability (Wang et al. 1986: 37), (3) the spread of new military tactics embodied in Sunzi’s art of war (Sunzi Bingfa, Master Sun’s Art of War) of the fifth century b.c. (Tian et al. 1990: 79–90), and (4) the emergence of a new army made largely of non-professional, drafted soldiers under a conscription system, presumably as a result of the increased casualty rate inflicted on the military aristocracy following the military revolution. From then on, army discipline rather than individual soldier’s bravery became the main issue (Sun Wu. n.d., Sunzi Bingfa, Master Sun’s Art of War. Reprint, 1995. Beijing: Beijing Yanshan Press, pp. 20, 53, 83, 86, 137); and logistics (mainly food and fodder) instead of weapons became the main concern (ibid., pp. 28-30, 81).

This was a full scale military revolution which eventually tipped the political balance among the numerous units (as many as 160, see Wang et al. 1986: 89), as Sun Wu famously put it (Sun Wu. n.d., Sunzi Bingfa, Master Sun’s Art of War. Reprint, 1995. Beijing: Beijing Yanshan Press, p. 19):

War is such a state affair, critical in terms of life or death, survival or demise, that it cannot be overlooked.

Meanwhile, China became a huge theatre of civil wars which lasted the record time. That is until 221 b.c.


2. State-building induced new social structure and new economic institutions: the free peasantry and private land ownership

This process of military revolution was at the same time a process of state-building as a new type of state began to take shape, develop and replace the old type. The new state was centrally controlled in line with a military command system, manned by salary-paid bureaucrats holding uninhabitable offices. This suited well with rapid territorial expansion without the constraint associated with personal links under feudalism. Constant wars also changed the composition of the military forces. Instead of using professional soldiers, large numbers of new recruits were in high demand from all walks of life. Among many trades, the agricultural sector seemed to fill the bill the best by supplying labour-toughened and relatively disciplined and simple-minded males. The sector also provided the state food and revenue, crucial for sustaining military operations. In addition, settlement and re-settlement of peasants in newly capture areas were a cheap and effective way to control the new territory. To reward the peasantry in an exchange for its support of the wars, a new institution emerged which was marked by private property rights with the focus on private land-ownership. The age-old Zhou chessboard-field system (jingtianzhi), a communal land-ownership, was replaced. After that, China’s land-ownership as a whole was not changed until the early 1950s when communists decided to copy collectivisation from Stalin.

By the fourth century b.c. all the main players in China were going up for it in a frenzy in the fear that they may have been defeated and annexed.

So, the result of the new state-building was a package of new army, new administration and new economic institutions. They were anything but feudal. Given that the new state was free from the kind of constraints traditional associated with decentralised feudalism, it had a strong tendency to expand across a vast territory: the geographic barrier such as desert, high mountains and seas were the only limits. Because of that, this state-building was at the same time empire-building. The Qin Kingdom successfully combined the two in one and hence ended the lasting Warring State Period (475–221 b.c.).

In this context, early China was far less static and far less homogeneous than one might think; and consequently adopting this choice was far less straightforward and far less smooth than one might imagine.

At this point, it is worth mentioning that the chariot war affair was associated with slavery in the Western Roman Empire but with ‘feudalism’ in Zhuo China. And, the cavalry warfare was linked to feudalism in post-Roman Europe but to predominant private landholding in post-475 b.c. China. More important, the establishment of extensive private land-ownership was associated with capitalist development in Western Europe and Japan while it was the backbone of the ‘traditional’, non-capitalist development in China. In a comparative sense, China is a case of ‘precocity’. Or, the globally, the world is dichotomous and China is hence a different ‘species’ altogether.


3. The Qin success in state-building

Among the main players, the Qin became most successful and eventually unified China in 221 b.c.

The Qin Kingdom was considered as a marginal kingdom which was largely left out of the thriving commercial activities in North China Plain region. Qin was benefited, mainly passively, by incoming traders of the other states. Seeing the vacuum of the influence of money in Qin politics, Lü Buwei, a successful merchant of the Wei origin (now Henan), took the advantage to invest in the Qin crown prince which yielded him handsome returns: Lü was later appointed Prime Minister (xiangguo) and practically controlled the politics of the kingdom during his heyday (Sima 91 b.c.: ‘Biography of Lü Buwei’).

But commercial backwardness did not handicap Qin’s rise as a champion among the competing powers. This was especially true after Shang Yang (c. 390–338 b.c.), a Legalist by training, was employed in 356 b.c. as a policy adviser by Duke Xiao of the Qin. According to Shang Yang, ‘Once attracted by farming, men become simple-minded and thus reliable to fight the war’ (Shang Y. 338 b.c.: ch. ‘Agriculture and War’). He specified that to build a strong political unit the ideal proportion of the peasants should occupy 90 per cent of the total population (ibid.) and made this his target in his reform. The first reform, master-minded by Shang Yang, took place in 359 b.c. with radical measures aiming at building a landholding, tax-paying and soldiers-conscripting peasantry with two key features: (1) ‘promoting farming and confining trade’ (zhongnong yishang) through discriminative taxes and (2) dissolving the extended family types and promoting the nuclear family as the taxable unit. Indeed, the gross families embracing inclusively several generations occurred in Han times and was in the long run not a norm. It was practiced exclusively among the super rich. So, Shang Yang’s reform did not create any new family pattern. Rather, it streamlined the existing pattern under the Zhou (Xie W. 1990: 270–1, 276–8). Moreover, the fact that to increase employment in agriculture became Shang Yang’s target means that the farmers had constituted a much smaller proportion in the general population by the time of reform.

A year later, Shang Yang launched his second reform (1) to replace the age-old chessboard-field system with tax-liable household-cum-farms so that the farmers had more incentives to produce more and better, (2) to privatise state land and to permit trade in land properties, (3) to attract farmers from other kingdoms to settle in Qin by offering tax advantages and private ownership (called laimin, literally ‘attracting immigrants’) in order to maximise the recruitment of the native Qin males for the army, (4) to establish the system of prefectures and counties (junxianzhi) under a single commanding centre to allow the Qin state apparatus to govern territories of a great many times of its core area, (5) to establish a universal census register system for taxation and army recruitment purposes, and (6) to uniformise weights and measures to lower the transaction costs in tax collection (Shang Y. 338 b.c.; Zheng L. 1989).

The main consequences of the Qin reform in the agricultural sector were at least three (1) the reinforced property rights (2) the abandonment of fallow and (3) the spread of the plough. They were closely entwined.

According to the unearthed ‘Qin Code written on Bamboo Slips’ (the Yunmen Qinlü), household properties were carefully defined and protected. For example, in the clause for ‘household internal crime’ (jiazui), the law exempts the charge against the son of criminal offences such as injuring the father’s employees and stealing the father’s livestock only if the son lives with his father. In the clause for "stealing from the employer’s parents", the law exempts the charge against the servant who steals from the employer’s parents only if the parents do not live with the employer (see Xie W. 1990: 269).

By definition, the abandonment of fallow went hand in hand with private land-ownership. In other words, land held in common encouraged fallow whereas private land-ownership led to crop rotation. The ultimate reason for that lies in the degree of land scarcity. Once fallow ceased to exist, the maintenance and improvement of land fertility depended greatly on human inputs whereby the plough, which had been the tool for reclamation and sowing, became an indispensable tool for the maintenance and improvement of land fertility. Not surprisingly, by the early Han Period, fallow had been regarded as backward, inefficient and even barbaric. Its domain retreated to lands in China’s periphery. The plough was promoted under the Imperial order and crop rotation-based intensive farming encouraged (see Deng 1993a: chs 3–4).

After the reforms, obtaining superiority in food supply and military power, the Qin was ready to build an empire on the East Asia mainland. That the Qin knocked its rivals over one after another like dominoes showed just how powerful Shang Yang’s new institution was. The military superiority was clearly shown in the life-sized terracotta infantry and cavalry unearthed from Emperor Qin Shihuang’s tomb. The ‘army’ was equipped with advanced weaponry of that time (bronze-cast and chrome-treated dagger axes, spears, swords and a crossbow mechanism, as well as a horse bridle; and possibly the armour for the infantrymen) have been discovered . It includes nearly 6,000 foot-soldiers and 2,000 cavalrymen with chariots and horses. This indeed reflects the emperor’s military accomplishments (Guisso et al. 1989: 56–69). The whole process was carefully planned and executed. From the available data, the impact of this early reform was profound: during the Ming–Qing Period, as much as 80–90 per cent of the total employment in China was still provided by the agricultural sector (Feuerwerker 1984: 299, 302, 312–13; Chao 1986: ch. 3).

The establishment of the Qin state apparatus in turn reflects (1) the completion of a structural change (2) the beginning of the actual dominance of the agricultural sector in the economy. Also, the combination of landholding peasantry and physiocratic bureaucracy offered society little chance to turn back to a less regulated economy as in the Shang and late Zhou periods unless some sort of revolution occurred. Even alien conquerors were unable to change the inertia of the interlocking trinary system.

Apart from property rights, those kingdoms which did not practise physiocracy also paid a heavy price. Later, the Qin rulers found themselves caught in a similar situation when Emperor Qin Shihuang, the big-headed victor, drove 60 per cent of the eligible male adults (mainly landholding peasants) to the very limit by (1) treating large numbers of farmers, 1.2 million of them, as forced labourers in building the Great Wall (changcheng), National Highways (chidao), Imperial Palaces (efanggong), the Underground Mausoleum and by (2) sending 800,000 troops, consisting of peasant soldiers to fight the Huns (xiongnu) in the North and Viets (baiyue) in the South. In 209 b.c. his non-physiocratic policy triggered a nation-wide armed rebellion of the seemingly docile and meek peasants (Shi W. 1986: 18–19). This rebellion consequently ended the glorious 635-year-long history of Qin (841–206 b.c.). Only then, Chinese history realised what a monster it created–the mechanism to bring the system back on track.

Both the Qin state and the peasantry benefited from the agriculture-based power and triumph. Firstly, the Qin’s victory confirmed the winning physiocracy–agriculture formula which provided the Qin with an abundant supply of revenue and well-fed and well-disciplined peasant soldiers, crucial to the Qin cause of unification. Secondly, as the Qin ruling class gained huge territorial possessions, the peasantry was rewarded immediately after Qin’s historic victory under a nation-wide scheme of 216 b.c. called shi qianshou zishitian, literally ‘allowing the commoner-farmers to claim and own land’ (Sima 91 b.c.: ch. ‘Biography of Emperor Qin Shihuang’). The private individual land-ownership expanded and prevailed as the Qin system was replicated in all the newly-captured areas that once belonged to the defeated kingdoms.

Here four points can be made (1) farming alone, no matter how productive it appeared, was not a sufficient condition for the ancient Chinese to become stuck to it; (2) agricultural favouritism was not a sheer technology-based choice for production but rather an institution which combined economic, political interests/incentives and natural endowments; (3) the entrenchment of agricultural dominance was not a God-given practice but the result of an institutional choice through long-lasting violent conflicts among different competing groups; (4) military power played a crucial role in the shaping of the macro, cross-regional economy on East Asian mainland.

The Qin case demonstrates that the military power and conquest were not solely responsible for the making of the Chinese ‘pan-agrarian’ economy. Nor does it mean that a military group had a free will to change the economy. Military power was a necessary but not sufficient condition to convert much of the empire to a farming zone. Other conditions included endowments like geographic locations, climatic patterns and soil types played a part at the very least.

What the military power did for the Qin was to lower the opportunity costs incurred from other choices through coercion. Such military power-based choice did not always work, the best examples in Chinese history being the Northern and Southern Dynasties (420–581 a.d.) and the Mongol Yuan (1271–1368 a.d.). The former period was marked by a time of warfare and schism during which China was broken into pieces by nomadic invaders from the north and west (Elvin 1973: 44). The Xianbei, one of the ‘Five Barbarian Tribes’ (wuhu) conquered North China and established the Northern Wei in 386 A.D. In the process, agriculture in the North was in ruins: production was disrupted, farmers killed, and cultivated land abandoned. As the expansion of the Northern Wei ceased, and with it the spoils of war, which had been an important income source for the ruling class, two-thirds of the Xianbei population became idlers. The combination of the decline of farming and exhaustion of the spoils of war resulted in a severe economic crisis. Emperor Xiaowen (r. 471–499 a.d.) had to launch a reform in 485 a.d. by rebuilding agriculture through re-establishing the landholding and land-owning peasantry (Tang Z. 1956; Elvin 1973: 47–51). During the Mongol 40-year long campaign (ended in 1279 a.d.) against China, a suggestion was made to the Mongolian Taizong Emperor that (Song L. 1371 vol. 153: no. 146): ‘Since the Chinese are useless to our cause, they should be killed off so that their land can be converted to grazing land.’ Such a plan was carried out in the occupied regions. During and after the campaign, millions of Chinese were killed or captured to be slaves or serfs (quding); vast agrarian areas were enclosed as grazing land; horses belonging to the Chinese were confiscated, and horse-raising was forbidden in Chinese communities; autumn tillage for the second crop was forbidden in the North so that the land could be used for grazing after the summer harvest (Wang Q. 1586; Perkins 1969: 23–4, 197–9; ZNK 1984 vol. 2: 51–3; Zheng X. et al. 1984: 242–4, 254–5). This policy of transforming the Chinese agrarian economy was stopped only because of the loss of revenue as argued by Yelü Chucai, a high-ranking courtier of Persian origin (Wright and Twitchett 1962: 19–20, 189–216).

So, military power-based choice was highly conditional: the military ruler only had a limited number of choices. This is particularly so when the economy was not at crossroads but on a well-set developmental path simply because a crossroads economy had less resistance than that on a path due to the differences in opportunity costs associated with ‘path dependency’ (see Mokyr 1990: ch. 7; Rosenberg 1994: pt. 1). In a crossroads economy, choices bore more or less equal weight due to similar opportunity costs for those choice. In a well-established economy, the opportunity cost for resuming the same pattern was often the lowest. So, it was rational to go for the old path unless the old pattern faced a major crisis and thus lowered the opportunity costs for a change to take place.


III. From hegemony to empire: how state-building grew into empire-building

1. External pressure and its ‘endogenisation’

During the lasting civil wars from 475 b.c. to 221 b.c., there was one major factor laid latent: external invasion and conquest from nomads in the Steppes. In effect, even during the civil wars when the Chinese fight against each other, they had to keep one eye on the Steppes. Instead of building Walls against each other, all the warring Chinese kingdoms (Yan, Zhao, Qi, Wei, Qin, Zheng, Han, Chu and Wu) formed a collective defence system with their Walls facing the Tungus, Huns, Yuezhis and Qiangs. Here it is only too evident that the common enemy came from the Steppes during the internally turbulent era of the Warring States Period (see Figure 1).

Figure 1. The Great Walls during the Warring States and the Qin, 475–207 b.c.

Source: Based on information form Yu J. 1986: 97; ZDC 1990; Zhao X. et al. 1991: 69–84, 111–13;

Hou W. 1992: 784, 1332–3.

Note: 1. The Yellow River (with its modern course). 2. The Great Walls at c. 400 b.c. 3. The unified Great Wall under the Qin, c. 207 b.c. Kingdoms during at c. 400 b.c.: A–Yan; B–Zhao; C–Qi; D–Wei; E–Qin; F–Zheng; G–Han; H–Chu and J–Wu. 4. Main events of nomadic invasions with dates.

Not surprisingly, the political unification under the Qin was only a half way to complete the process of empire-building. After 221 b.c., the state priority was refined. Now, the main concern for the new state was no longer how to unify China but how to shield off nomads and safeguard China’s internal order and stability. This was an urgent matter as through increasing contacts the nomads were by now well aware of both the material wealth and the defence weakness of China. In other words, it is not accident that in the period from 475 b.c. to 221 b.c., the pursuit of hegemony was replaced by the ambition to annex other political units to gain the advantage of size for more resources to get the upper hand in a changed game. Indeed, a case of economies of scale. So, steadily, the number of political units was reduced from three digits to two and a single in the convergent process of political unification.

At the point of China’s unification, the conversion of state-building to empire-building proved to be not only feasible but also necessary if China wanted to keep the fruit of its early state-building. The military priority was shifted from internal offence to external defence, but executed by the same army and supported by the same peasantry under the same bureaucracy.

With the battle line redrawn, the Chinese strategy was to substitute a permanent defence wall for the poor quality of their soldiers. Or, simply to substitute a high quality capital-intensive construction for a low quality labour force (here being an army). Therefore, under the first emperor Qin Shihuang (r. 221–210 b.c.), all the previous walls were integrated into one Great Wall. And, China became the only ‘walled empire’ in human history (again, see Figure 1). Indeed, the construction of the Great Wall marked the major turning point at which China’s state-building was replaced by empire-building. However, the whole process did not end until the sixth century a.d. marked by the construction of the Grand Canal which served as the artery for logistics to sustain frontier garrisons and bureaucrats in the north. Between the Northern Song (960–1127 a.d.) and the Qing (1644–1911 a.d.) dynasties, an average of 401,000 tons of rice were transported each year by the state under the scheme of the ‘annual shipping of rice to the north’ (caoyun) (Liang F. 1980: 293–4, 303, 375, 394, 396.).

The establishment of the Great Wall defence nevertheless showed that the internal and external order was regulated and maintained. The end result of this empire-building was peace in China in the long run.


2. Changed function and role of the single state in an empire: national defence

In the order of the importance, the main functions of the single state in China was (1) national defence, and (2) internal order. The aim was to maintain peace. This marked the difference in nature between the Chinese state and a modern developmental state, or even an out-looking expansionist state (the Mongol type) and mercantilistic state (the European type) in world history.

Such a priority was well justified as the threat from the nomads continued after 221 b.c. and ‘wall-crossing’ invasions were rather frequent (see Figure 2). Indeed, once the enemies managed to cross the Great Wall, there was not much the Chinese could do. Most telling, China was conquered in its totality twice by the wall-crossing enemies: the Yuan Mongols (r. 1271–1368) and the Qing Manchus (r. 1644–1911). Thus, the empire-building itself was well justified. From these two disastrous alien conquests, there can be no doubt that China began to suffer diminishing returns from its Great Wall defence in the later part of the thirteen century after some one and half millennia of practice.

Figure 2. Major Events of Wall-Crossing Invasions, 206 b.c.–1644 a.d.

Source: Based on information form Yu J. 1986: 98; ZDC 1990; Zhao X. et al. 1991: 111–18, 244–8; Hou W. 1992: 312, 452–5, 580, 805–6, 1120–2, 1332–3.

Note: 1. The Yellow River (with its modern course); 2. The Han Great-Wall line, c. 100 b.c. 3. The Ming Great Walls, c. 1560 a.d. 4. Main events of nomadic invasions from north and west with dates (in a.d. unless indicated). Three capital cities of various periods: Chang-an (202 b.c.–190 a.d., 304–420, 535–907), Kaifeng (960–1126, 1214–35) and Beijing (1414–1911). The main nomad groups in the north are listed according to the chronological order from the Huns to the Manchus.


In this context, the main role of the Chinese state is revealed: national defence. So, the Chinese empire showed a characteristic of tension externally but relaxation internally. Within the Great Wall, China’s internal policing was sluggish. The Chinese government administration stopped at the counties (with a total number of some 1,000 during the Ming–Qing Period). Each such county had only a handful salaried officials to serve an average of some 399,000 residents (as at 1833, see Liang 1980: 10). Inevitably, local autonomy was warranted. Law and order at the grass roots level of rural towns and villages, was in the private hands of local elites known as the ‘gentry’. The best the Chinese state managed to achieve was to supervise a neighbourhood watch network (called baojia during the Ming), a system which was developed from the archaic model of communal responsibility known as the Zhou chessboard-field system. It suited China’s overwhelmingly private economy reasonably well. Only after 1949 was this autonomy ruthlessly broken up under the dictatorship of Mao Zedong whose party tentacles penetrated all villages and factories alike. But, that Maoist state was a completely different animal which is of outright totalitarianism. It is thus reasonable to state that China was able to accommodate a totalitarian state but decided not to adopt it due to deliberate choices during the entire premodern era.

In comparison, China’s troops in garrison, some times as many as one million men, along the Great Wall defence line were much better organised. As in the Ming case, the Chinese had a five-layer system which consisted of (1) strategic points (zhen), (2) main branches (lu), (3) passes (guan), (4) castles (cheng) and (5) beacon towers (tai) (for the complexity of the Great Wall defence system, see Zhao X. et al. 1991: 111–36, 244–79). According to History of the Ming Dynasty, the distribution of the troops among the 11 strategic points along the walled defence line is as follows (based on Zhang T. 1735: ch. ‘Military’):

Location Province Troops

Beizhen Liaoning 99,875

Qianxi Hebei 107,813

Changping Hebei 19,039

Baoding Hebei 34,697

Xuanhua Hebei 151,452

Datong Shanxi 135,778

Taiyuan Shanxi 57,611

Yulin Shaanxi 80,196

Yinchuan Ningxia 71,693

Guyuan Ningxia 126,919

Zhangye Gansu 91,571

Total 6 976,644

On average, 134 soldiers were allocated to each kilometre of the Wall with a density of one soldier per 7.5 metres. In this respect, the Great Wall defence line was not only capital intensive but also labour intensive.

This was not all. There was the second Ming echelon with 400,000–500,000 combatants under 72 divisions (called jingjun, literally ‘the Army for the Capital’) stationed in the capital city Beijing, less than 100 kilometres away from the Juyongguan Pass along the Great Wall (Zhao X. et al. 1987: 411–14, 417). This meant the Great Wall-related troops under the Ming numbered 1.4–1.5 million. Given that the total number of the Ming armed forces was between 1.8 and 2.8 million men (Zhao X. et al. 1987: 404), the standing army for the Great Walls thus occupied 50–83 per cent of China’s total.

It is worth noting that the Ming was not an unusual period of heavy border defence. It has been estimated that under Emperors Qinshihuang (r. 221–210 b.c.) and Wudi (r. 140–88 b.c.), the Chinese border troops (bianbing) maintained 800,000–900,000 (Huang and Chen 1997: 91). Given that (1) normally the northern-border troops occupied 30 per cent of China’s total armed forces (ibid.), (2) a quarter of the population was probably qualified to serve the army, and (3) in 2 a.d., China had a peak of population growth at around 60 million (Liang F. 1980: 4); the participation rate in China’s defence would be at least 15 per cent of the total population. In some periods, the Great Wall defence system alone involved 19 per cent of China’s total population (Chao 1986: 50). Considering the size and the long border line of the Empire, the heavy concentration of military presence along the northern and northwestern frontiers shows the equally heavy pressure from the nomads.

The rationale for such a set-up was almost certainly to minimise the cost of the state by prioritising national defence. As a result, bureaucrats occupied only a tiny minority of China's total population (see Table 1). Thus, the Chinese state is well qualified as a minimalistic state, cheap and small: After all, the state only controlled a low percentage of GDP, normally not more than 10 percent (Feuerwerker 1984; 322; Will 1990). In my observation, it was less than one percent (see Table 1). The only way to achieve this was to ‘autonomise’ (or privatise) law and order at grass roots level. The popularisation of Confucian values and code of conduct was highly appropriate in terms of reducing both social and private costs of such autonomy.


Table 1. Population of Scholar-officials in the Major Dynasties

Dynasty No. of officials Population % in total population


Western Han 132,805 59,594,978 (2 B.C.) 0.22

Eastern Han 152,986 56,486,856 (157 A.D.) 0.27

Sui 195,937 46,019,956 (609 A.D.) 0.42

Tang 368,668 52,919,309 (755 A.D.) 0.70

Song 24,000 -

Yuan 16,425 59,848,964 (1291 A.D.) 0.03

Ming 24,683 59,873,305 (1381 A.D.) 0.04

80,000 61,852,810 (1474 A.D.) 0.13

Average 124,438 56,656,596 0.26

Source: Jin and Liu (1984: 26).

The changed function and role of the single state in a vast empire in China did not undercut the early-invented social structure and economic institutions of the free peasantry and private land ownership. Rather, the empire was carried on the back of the landholding peasantry.


IV. State-building/empire-building and institutional fixation

Other roles which were derived from the main role of national defence were all focused on protection and promotion of the source of tax revenue and the pool of soldiers and occupational forces (in the form of settlers in the frontier regions, civilian or paramilitary). As explained earlier, the main reliable source of taxation and army recruits happened to come from the agricultural sector and permanent frontier setters happened to be farmers in the process of state-building under the Qin. Naturally, the state's favour tilted towards agriculture which cost the least for the same returns.

1. Physiocracy, ‘agrocracy’ and private property right in agriculture

This agricultural favouritism should not be viewed as sheer accident. Rather it was a result of the interplay between the demand for and supply of resources during state-building. There was a great deal of government inputs in the rise of the agricultural sector in line with the principle of feeding the goose for more eggs. Among many things, the Qin state invested heavily in the new institution of private property rights in order to create new incentives for the peasantry to produce more and better. Both the state and peasantry were benefited from this interplay. The former had reliable revenue and the latter private properties and some economic freedom. This was a Pareto optimum.

Once the mutually beneficial relationship between the state and agriculture became sustainable, and once the employment opportunity and provision from agriculture became sustainable within the economy as the agricultural GDP share overweighed other sectors and the farming population dominated China’s total, China was led to a unique growth path of agricultural dominance in both the economy and state politics. Other policy choices and development choices were not longer cost-free (in terms of either accounting or opportunities costs). More so with the continuous returns made from the agricultural sector.

The state protection and promotion of agriculture was crystallised in Chinese physiocracy, a package of ideology and policy measures geared towards agriculture. In turn, the Chinese bureaucracy became what can be called an ‘agrarian bureaucratic state’ (Goldstone 1991: 41), or simply ‘agrocracy’.

i. Determinatives for physiocracy

Why did the Chinese state have to be physiocratic? Why not mercantile like some European states in the sixteenth and seventeenth centuries? These two questions strike modern observers. It is often believed that physiocracy was a creation of the government or the upper class. In a sense it was, given that policies had to be made and implemented by some organisation in society. However, since such policies were so consistent in spite of the periodic change of governments, some having been not even Chinese in origin (such as those of the Tartar Jin and Manchu Qing) it seems that there were some ‘invisible hands’ in control of the process. In other words, a physiocratic state in China did not occur by accident. Rather, it was a result of a carefully-calculated purpose and careful calculation. The purpose was to protect economic interests and promote a type of economic life. From the analyses of agricultural dominance and the power of the Chinese peasantry, agricultural fundamentalism in China was not a free choice by governments, but a ‘must’ if they wanted to rule the country as seen from physiocracy and the principle of the People as the Foundation (minben).

The invisible hands were primarily the social structure, and secondarily Confucian belief and empire building. In contrast, by the seventeenth and eighteenth centuries, Europe had already developed an advanced financial system including the creation of government bonds and annuities investment that were guaranteed as safe as land. Governments in Europe greatly extended their role in capital markets and were also responsible for the appearance of mortgage markets (Jones 1988: 137). No parallel existed in China. It can thus be suggested that a non-agricultural alternative, which was discovered by European societies, diminished the need to protect agriculture.

Then, the question becomes why did the Chinese fail to find a similar alternative? The answer is Chinese society did not provide a strong incentive to seek any alternatives, as demonstrated (1) in the unique replication of small landholding and land-owning households, and (2) in the fact that the landholding and land-owning peasantry was a main determinant of physiocracy.

From the peasant point of view, physiocracy and minben ideology provided the public with a yardstick to judge government socio-economic policies and to ensure that those policies did not become empty talk. If the majority of the population became worse off, it was the government that was to blame. Bad administration would give enough reason for the peasants to rebel to change the government. The Chinese state could not afford to ignore the interests of the majority that were constituted by the peasants.

From the Chinese elite point of view, there was a correlation between agriculture and peace. They were fully aware of the danger of massive peasant armed rebellions and of the fact that a crisis was often preceded by misconduct on the part of the ruling group. They believed that the central power must be careful in handling the peasants and agriculture because only wise rulers could occupy the throne for long, a lesson drawn by imperial historians from the downfall of the glorious Qin Dynasty (Lee 1969: 148). A golden rule was to ensure the masses were properly fed, as described in Huainanzi (Duke of Huinan) as follows (Liu A. 179–122 b.c.): "Food is the foundation of people’s life; people are the foundation of the country; the country is the foundation of the monarch."

From the Chinese state point of view, physiocracy was in fact the ramification of the agricultural and free peasantry’s dominance in Chinese lives. This is because (1) the peasantry was the stratum of agricultural performers and thus embodied the interest in agriculture; (2) peasants were the overwhelming majority in China’s population; and (3) in addition, the Confucian education–civil service system provided a path for peasant sons to become bureaucrats, which was also encouraged by law. This further safeguarded and reinforced the peasant interest: peasant sons in power were more likely to protect agriculture just for the sake of their own farming families. (4) Agriculture was, as mentioned earlier, the main source of government revenue and for manpower for public works and defence forces. Agriculture was thus central to social stability in China: ultimately, it determined the ebb and flow of the latter. Throughout Chinese history, whenever and for whatever reasons agriculture was disturbed and weakened, social disequilibrium would occur and the country would either be defeated by barbarians through the weakening of the Chinese standing army, which relied on the agricultural sector for soldiers and the food to feed them, or would collapse internally following peasant armed rebellions due to peasant discontent (for disturbances of agriculture and fall of dynasties, see Wang 1936). On the other hand, the more prosperous the agriculture, the stronger the country, with greater tax revenue and a more generous supply of food and fighting men; and the stronger the country, the less trouble the barbarians and the peasants would cause. It is easy to understand why agriculture became an issue on the ancient Chinese political agenda and was so closely watched over by the central authorities (Chen 1911: 50, 55; chs 21, 26, 30–1). That it became a matter of life and death for any dynasty in dealing with agriculture and the peasantry means the Chinese state had virtually no choice but to be physiocratic. Obviously, peasantry and agriculture were the cause and the physiographic state was the effect.

At this point, many hypotheses about China need to be re-evaluated. For example, according to mainstream opinion, the Chinese state was perceived to be hostile towards commercial activities and as favouring agriculture, which consequently caused China’s inability to become industrialised (Perkins 1967: 478). This view seems plausible at first glance. However, when we trace further to ask why the Chinese state had to be so anti-merchant (which is by definition different from anti-market), the ultimate reason can only be established in the Chinese ‘mindset’. Such argument ignores the basis of public choice: a society does not live on mindset but economic activities, inputs and outputs, investment and returns. Mindset does not belong to any of these economic categories.

ii. Origin and development of physiocracy

The idea of protecting agriculture was crystalised in nongben, meaning physiocracy or agricultural fundamentalism (see Broadbent 1978: 104; Zhao K. 1984; Guo W. 1988: 6; Yan S. 1988). This physiocratic ideology originated in no later than the Spring and Autumn Period (770–476 b.c.), and was shared by nine major schools of thought of that time: Taoism, the Yinyang school, the Legalists, the School of Logicians, Mohism, the Political Strategists, the Eclectics, the Agriculturists and Confucianism. All of them claimed either directly or indirectly that agricultural prosperity was the prerequisite for making a political unit powerful (see Chen 1911: ch. 11).

Undoubtedly, Confucianism played an important role in philosophising physiocratic ideas and indoctrinating agriculture as the most respectable economic activity in China (Chao 1986: 106). For example, in Guoyu, one of the Confucian classics, the idea of physiocracy is expressed as follows:

The greatest business of the people is agriculture. From agriculture, the millet which is used for the sacrifice to God is produced; the density of population grows; the expense of the business is supplied; social harmony and peace arise; the multiplication of wealth begins; and the characters of honesty, great-mindedness, integrity and solidity become a general habit of the people.

(quoted in Chen 1911: 381)

Confucianism also indoctrinated the Chinese social hierarchy in which the peasantry was recognised as higher than other working classes and the merchant class was put down at the bottom of the class ranking. Arguably, Confucianism was at best a theoretical source of physiocracy to reflect the Chinese socio-economic reality, since it did not create farming nor the landholding and land-owning peasantry.

The ideology had several aspects. First, agriculture was regarded as benye, or the ‘Principal Occupation’. Priority was given to the protection of the agricultural sector of the economy, while other sectors which competed with agriculture for labour, capital and other resources were confined so that agricultural dominance was not undermined (Twitchett 1968; Lau 1984: 21). Second, farming as an occupation received great respect and farmers were accorded considerable dignity: they ranked in society above other commoners such as artisans (gong) and merchants (shang), as observed in as early as the seventh century b.c. by Guan Zhong (?–645 b.c.) (see also Fairbank 1965: 17–27; Mokyr 1990: 230–1). Third, to encourage and protect agriculture as well as to assist the peasantry were considered the dominant economic policy by the governing institutions (Lee 1969: pts 1–2).

Surrounding this ‘policy core’, there was often a physiocratic package. (1) Policy measures employed included control of grain prices by establishing the floor price level and buying in surplus grain in good years, exemplified in the policy of Li Kui (c. 455–395 b.c.); (2) the protection of private land-ownership, as in the deeds of Shang Yang (c. 390–338 b.c.); (3) assistance with irrigation projects, under such policies as those of Sang Hongyang (152–80 b.c.); (4) demonstration of new and better techniques, as in the policy of Emperor Wudi (140–88 b.c.); (5) famine relief, as in the undertakings of Emperor Zhaodi (86–74 b.c.); (6) periodic land allotment, as proposed in the agrarianism theory and corresponding policy of Li Anshi (443–493 a.d.). Moreover, tillage (litian) was one of the criteria for awarding titles to citizens and for exempting farmers from taxation (for the Warring States Period see Shang Y. 338 b.c.: ch. ‘Jinling’ and Guan Z. n.d.: ch. ‘Shanquanshu’; for the Qing Dynasty see Cheng Q. 1865 vol. 9: 635). Furthermore, members of ‘marginal classes’ of non-landholding individuals and non-scholars such as artisans and merchants were persuaded to buy land to join the mainstream, as practised in post-Song times. Often, sons of these marginal classes become landholders and/or scholars rather than to follow in their fathers’ footsteps, thanks to the class mobility mechanism in society. Last but not the least, it was not unheard of for the emperors, ‘the Sons of Heaven’, to ‘practise farming’ on the imperial farms in order to identify himself with the peasantry and to set up an ultimate example for the younger generation to follow (Ji X. 1955: 54; Lee 1969: 67, 159; Merson 1989: 12–13). In the context of its origin, it is not surprising that this physiocratic ideology dominated government economic policy-making until Qing times (1644–1911 a.d.) (Broadbent 1978: 104; Zhao K. 1984; Yan S. 1988). After being long absent in medieval Europe, a physiocratic state occurred only briefly in the eighteenth century when this Oriental ideology was introduced to French politics through Quesnay (Waverick 1946; Schumpeter 1954). Therefore, it is largely irrelevant to talk about a physiocratic Europe.

Among all these measures, in terms of supporting the small landholding and land-owning farms, the most important were the following three (1) a low tax rate, (2) grain price control and (3) low interest loans which were designed to help small farmers to sustain their production cycle. The tax rate, for example, is estimated as 10 per cent of the household’s total output from Han to Tang times (see Appendix 7). From the Song to Qing on average the government tax revenue counted for merely 7.1–9.3 per cent of the total agricultural GDP or 5–7 per cent of China’s total GDP (Feuerwerker 1984: 299–300, 302). Given that the Chinese agricultural sector was able to produce about a quarter of its products as surplus (ibid.), the amount which the Chinese state tapped was only one-third of that surplus.

One may argue that this figure was a formal tax rate and that informal taxes and duties should also be included. There is little doubt that not only informal taxes and duties existed–and they sometimes got out of control and destroyed the economy–but things went wrong with formal taxes as well. Policy fluctuations were common. Excessive rent-seeking haunted the Chinese Empire. However, heavy tax on agriculture was recognised by both the Confucian elite and ordinary peasants as non-physiocratic, unorthodox and abnormal. Although the picture was not so black-and-white, in most cases, physiocrasy prevailed in the end, as evidently shown from the aftermath of each major peasant rebellion throughout Chinese history.

However, physiocratic dominance in the Chinese state politics did not mean that there was no room for non-agricultural activities. Quite the contrary, physiocracy in China allowed the market to exist and sometimes to prosper. For instance, Sima Qian (145–86 b.c.), the best known historian throughout premodern Chinese history, believed that trade was indispensable in society. So much so, in his The Book of History (91 b.c.), he included a section on market activities (‘Shihuo Zhi’) and merchant biographies, which became a standard for writing dynastic histories. Of course, the market was not ‘free’, it was regulated and controlled to protect the agricultural sector. But, no matter how tight the state monopoly was on some domestic and most international trade appeared to be, there was always a place for trade and trade expansion as long as farming was not sacrificed. It is indeed quite another matter as to whether the professional merchants felt comfortable with such control. In this context, the stated status of the merchant class reflected the real place of the class in society. Scholars often blame the Chinese state for ‘putting the merchant class down’. Oddly enough, few scholars have blamed the Chinese state for honouring and collaborating with the peasantry too much as if the peasantry naturally deserved it. The same attitude should have been consistently applied to all classes.


2. Consequence and impact of physiocracy and ‘agrocracy’

i. Belief of ‘people as the foundation’ to recognise the rights of the peasantry

Physiocracy in premodern China was supported and entwined with the ideology of minben, min meaning ‘the amorphous, indeterminate mass of peasants’ (Hall and Ames 1987: 139), and ben ‘basis’ or ‘foundation’. Therefore, minben can be translated as ‘People as the Foundation’ or ‘the Peasantry as the Foundation’. Here, it is not an issue of passion toward a particular occupation, rather it is a concern of resource and returns (hence the ‘goose’ for eggs).

Interestingly, in spite of the philosophical and political differences between Confucianism and Taoism, scholars from both sides tried to find a way to reach social harmony and took a strong stand against the tyrannical policies of Legalism. The idea which both Confucianism and Taoism came up with was this minben. After the political and socio-economic disaster caused by the Qin Legalists (Jia Y. c. 200–168 b.c.), minben became well entrenched in China’s political thinking (see Chen 1911: 77–9), which can be taken as a sign of the social and political consciousness of the Chinese literati (Liu Z. 1987). As a result, in modern China, even the ruthless ruler Mao Zedong had to disguise himself as a benevolent sage by labelling his régime as one of ‘serving the people’ (weirenmin fuwu).

In Lao–Zhuang philosophy, or early Taoism, minben was embraced in its laissez-faire framework known as wuwei erzhi (literally "order and equilibrium will be achieved without ruler’s intervention"). Taoism had its turn to serve the state to guide policy-making in the early period of the Han Dynasty, because of the total bankruptcy of the heavy-handed Legalist approach under the previous Qin régime. The Legalists’ rejection of the minben principle made the Qin pay a heavy price in that it was the shortest-lived dynasty in Chinese history (see Fan W. 1964–5 vol. 2: 50; Shao Q. 1985: 6; Sun C. 1986: 618). Confucianism was later favoured because the Taoist potential for passivity and the laissez-faire attitude became exhausted (see Ban G. 82 a.d. vol. 56; Jian B. 1983: 485–93). Confucianism, by comparison, was more active, disciplined and utilitarian. The idea of minben was also more explicitly manifested by the Confucian School (see Shao Q. 1985). According to orthodox Confucians, humans are the most precious beings between Heaven and Earth (Zhang Z. 1987). Mencius explicitly states that (1) the people are of supreme importance; (2) rulers have to win the people’s support and (3) an effective way to win people’s support is to ensure the people against starvation in bad years (for the English translation, see Lau 1984: 291, 145, 21). Mencius further held that (Mencius Warring State Period: ch. ‘Lianghui Wang’), ‘only he who is able to protect people is entitled to become the ruler (baomin erwang).’

The legitimacy of government, according to minben, was indeed conditional: government had to ensure that the masses were employed, fed and thus relatively content. Otherwise, the emperor would lose Heaven’s favour and a rebellion could be justified as an antidote to disequilibrium. In world history, there were probably more mass armed rebellions of peasants against the corrupt central authorities in China than anywhere else: for instance, from 1644 to 1721, across China there were 176 armed rebellions, an average of more than two each year (Xie G. 1956). Sometimes these rebellions succeeded in overthrowing the target dynasties, including the Qin, Han (both Western and Eastern), Sui, Yuan and Ming (Jin and Liu 1984: 115–19).

The unique phenomenon of rebelling to change dynasty was another face of the same coin of the Chinese landholding free peasantry-based society. In comparison, frequent massive armed rebellions did not occur under a system such as the medieval feudalism of Europe or Japan, in which the majority of the people were restrained by a decentralised social structure, often characterised by personal legal and economic bondage. In such societies, a census register system was neither necessary nor possible, because of the principle, "my subject’s subjects are not mine". Nor was it necessary in medieval Europe for mass armed rebellions to be organised to change unbearable social conditions: there were many other bloodless means of breaking out of personal bondage, for instance by escaping from individual masters or overlords. It can thus be safely assumed that in Europe more peasants or serfs preferred becoming free bourgeois to joining in violent rebellions, simply because the potential cost of the latter was much higher than the former. Furthermore, in both Europe and Japan the decentralised political system, the personal bondage which existed within the small, separate political units, as well as the economic differences between these, handicapped any attempt at organised manoeuvres by the ruled in rural areas. Further, since the free citizens of the political alliance of independent city states in medieval Europe were not the social majority, they could not play the same role as their Chinese counterparts. In the case of the free Germanic and Slavic communes, there was no need for the members to rebel, since the underdevelopment of private ownership hindered the alienation of the political power from the public interest (Engels 1942). Therefore, ideas like ‘People as the Foundation’ was not very relevant in Feudal Europe and Japan.

Naturally, in order to gain and maintain this legitimacy, the ruler had to protect and encourage agriculture as the means to satisfy the peasantry. This was the link between minben and physiocracy. On the whole, physiocracy and the principle of People as the Foundation were simple justifications and confirmations of the importance of agriculture and the peasantry in premodern Chinese society. Of course, they also reinforced the agricultural dominance and the rights of the peasantry.

The best example of the nongben–minben paradigm in action can be seen in the policy of Emperor Taizong (627–49 a.d.) of the Tang Dynasty. Drawing a lesson from the previous tyrant Sui (581–618 a.d.) who was destroyed by rebels, Emperor Taizong resumed the old dogma of minben and openly recognised that (1) it is the people that maintain the monarch (min yang jun); (2) it is the people that select and allow the monarch to rule (min ze jun); and (3) it is conditional for people to follow and obey the monarch (min guiyu jun). His policy was designed to establish harmony between the people and the ruler by practising abstinence among the ruling class and laissez-faire towards the ordinary people: tax and corvée were reduced or exempted, agriculture was protected and promoted (Liu and Zhang 1991: 70–3). As a result, the Tang Dynasty was prosperous and long lasting.

From the above analysis it is easy to understand just how misleading the so-called Oriental Despotism is when applied to the Chinese case. The Chinese state, including the monarch, Imperial Court and the bureaucracy, was not God and could not have a free hand to do whatever it wished to, because the power of the Chinese state needed to be mandated in the name of tian (or t’ien, meaning Heaven or the deity which has the transcendent power over society) by millions of ordinary free peasants. As Hall and Ames point out (1987: 145), "in the Confucian tradition, there is a frequent association between t’ien conventionally rendered ‘Heaven’, and the masses (min)", "T’ien has compassion on the masses (min): whatever they desire, t’ien is sure to effect" and, "T’ien sees as the masses (min) see; t’ien hears as they hear".

Therefore, the Chinese political system should not be interpreted as totalitarianism or absolutism because the ruling clique never obtained such power even when it tried–the illusion of state’s total control in premodern China being rooted in the sophisticated bureaucratic machine of the Chinese Empire (Wittfogel 1957); nor should it be interpreted as a form of upper class democracy because the Chinese system lacked political equality. At its best, the Chinese established and maintained a system equivalent to what is often called ‘soft-authoritarianism’.

ii. Reinforcement of agricultural dominance in favour of the peasantry

No doubt, physiocracy reinforced agricultural dominance in the economy. ‘Coaxing and enticing people into agriculture’ (quannong) was a standard economic policy guide right from the beginning of the Empire’s history. Accordingly, tillage (litian) became one of the criteria for awarding titles to citizens and for exempting farmers from taxation in the Qin Kingdom of the Warring States Period (see Shang Y. 338 b.c.: ch. ‘Jinling’; Guan Z. Warring States: ch. ‘Shanquanshu’). Later, the virtues of ‘Piety and Tillage’ (xiaodi litian) were established by Emperor Huidi (r. 194–188 b.c.) of the Han Dynasty as the criteria for selecting model citizens to enjoy corvee exemption (Ban G. 82 a.d.: ch. ‘Biography of Emperor Huidi’), and from then on it also remained among the criteria for the selection of bureaucrats until the Tang (Xu S. 1838: 17, 27). In the Yuan Dynasty, an elderly and skilful farmer was appointed Head of Farming Group (shezhang) to take charge of tillage and local security of fifty households (Zheng Z. 1938: 202). Under Qing law, successful farmers were to be appointed to the official position of Eighth Grade (bapin) to assist agricultural affairs (Cheng Q. 1865 vol. 9: 635; Lee 1969: 117). Li Yinggui, an expert in growing double cropping rice, even became closely allied to Emperor Kangxi (Kong X. 1983). Monetary reward was available, too. In 1828, a magistrate of Suzhou Prefecture honoured four farmers who successfully put the method of pit cultivation into practice with an official banquet and a prize in silver bullion (Chen Z. 1958: 369).

iii. Check on the growth of professional merchants

An immediate consequence of the physiocratic policies was discrimination against the activities of the merchant class. Although it is speculative to pin-point the Chinese physiocratic state and its policies as the prohibitive factor responsible for the lack of capitalist development in traditional China, the policies certainly made such a development more difficult than otherwise. This is particularly true in terms of the lack of government support for pro-merchant institutions: commercial laws and pro-market property rights, as Jones points out that in China the ‘political structure did not establish a legal basis for sufficient new economic activity outside agriculture’ (1990: 20).

The rationale for the anti-merchant attitude and policies was multifold. Firstly, technically, it was easier for government to collect tax and recruit young soldiers from settled farming households than from mobile merchants, a fact that was repeatedly pointed out by officials throughout Chinese history. In other words, without modern technology, the monitoring costs were much higher in dealing with merchants than in dealing with land-bond farmers. Such a lower cost in financial administration itself provide the government with enough incentives to protect the agricultural sector. In the end, it became a self-realising cycle: because it was cheaper to tax the agricultural sector, the sector was protected and nurtured for taxation purposes; and a prosperous agricultural sector made taxation burden lighter in relative terms. So, when observations are made, as in Murphey’s phrase (1954: 358): ‘the imperial revenue was at most periods largely from the land tax and from the government trade monopolies’, modern scholars tend to neglect the historical dynamics of this self-realisation. Moreover, technically, it was more feasible to impose state control on a weaker group such as the merchants than a strong group like the peasantry. Once again, it was a matter of the social cost and social benefit.

Secondly, in a typically Chinese view of ‘zero–sum gain’ towards resource allocation, merchants were regarded as the main competitor for resources against the peasants and the state. This ‘zero–sum’ view was natural in that a premodern society did not have a device to measure economic gains nor a clear vision of the process of value-adding outside agriculture (as theorised by classical economics in the analyses of Smithian ‘trade–accounting cost’ and Ricardian ‘trade–opportunity cost’), both crucial in the development of capitalism. Accordingly, enrichment of merchants was viewed as having been achieved at the expense of peasant welfare and government revenue (Fu and Wang 1982: 358; Han L. 1986: 421). This was true not only in China but also in France as seen from the influential School of Physiocracy headed by François Quesnay (1694–1774) who excluded completely commerce in his input–output model. Since the Chinese had freedom to choose their own occupations between the agricultural and non-agricultural sectors (Chao 1986: 2, 3, 5), a potential danger existed that more profitable sectors of the economy would attract too much labour and capital from agriculture. Typically, as an ancient (Han Dynasty) saying goes, ‘to become rich from poverty, working in agriculture is not as effective as crafts industry while crafts industry is not as effective as commerce (yongpin qiufu, nong buru gong, gong buru shang)’ (Fu and Wang 1982: 358; Han L. 1986: 421). Hence, the agricultural sector was protected by the authorities by means of blocking any heavy flow from agriculture and guiding other sectors such as industry and trade in directions which were of benefit, or at least not harmful, to agriculture. That a constant effort was required to check the over-growth of the merchant class was only part of the ‘social engineering’.

Thirdly, interests of the merchant class were subject to violation and erosion from the state in the form of heavy discriminative tax burden and difficulty to enter officialdom. Merchant property rights were not as well protected as peasants’. All these aimed at creating disincentives among the merchant class to control their influence and growth.

Finally, empirically, the masses in premodern China, who consisted of millions of small landholding and land-owning farmers, had a strong sense of equality, an ideal of Great Harmony (datong). They could tolerate to a great extent a rich state but not rich individuals. This was the reason the government trade monopoly and marketeering seldom upset the peasantry.

iv. Bias in technological development toward agriculture

The Chinese agrocracy, together with the agricultural cult, helps to explain why technological development became so highly selective and so clearly agriculture biased. Although the Chinese proved to be highly creative in a wide range of fields, a great many Chinese inventions and innovations withered before reaching their full potential, as revealed by Needham’s work (1954–90). In contrast, agriculture was one of the few areas which received constant scholarly attention throughout long-term Chinese history (Deng 1993a). This was largely due to the systematic promotion and support of agriculture with the physiocratic state functioning as a filter and amplifier in the manipulation of the developmental path for science and technology. This manipulation was possible because there existed a degree of bureaucratic conformity; and different ideas were often seen as unsafe and therefore there were incentives to discourage innovations (Fairbank 1965: 114). Within the ‘safe’ areas of study, useful results were counted as merits. To take technical books as an example, agricultural books (nongshu) were written and published consistently, even during the chaotic periods by different levels of the governments with a wide range of intellectuals being involved. As a result, agriculture and its related fields (such as water control and irrigation for production, transportation and communication for internal migration and food distribution) became exceptions when Chinese science and technology on the whole slowed down and stagnated in the later dynasties (Elvin 1973: 298).

Of course, the Chinese monarch often appeared apathetic towards technological progress. This is not surprising. The monarch was largely a symbol of the Empire, and the ‘dirty work’, including administrative routine and contingent treatment and the encouragement of technological process, were done by the bureaucrats at the front and thus often became hidden. A great deal of work was done at local and individual levels without any involvement of the central authorities (Deng 1993a). So, the reason for the monarch to be lazy was precisely because of the hard-working Confucian bureaucracy. In the realm of agriculture, China’s technological development was achieved in private hands or from the joint efforts of the bureaucracy and individual farmers. In that sense, Mokyr’s viewpoint that in premodern China technological development was a dependent variable to the state is questionable (Mokyr 1990: ch. 9). The difference between China and Europe was not whether and how the state was involved. Rather, the difference was that in China the development was a dependent variable to the a peculiar pro-agricultural socio-economic structure and thus biased towards farming interests. The Chinese state was heavily involved as one of the parties in this biased technological development. There is no evidence that the Chinese state monopolised the entire process of technological development over any length of time (see Deng 1993a and 1997). So, the importance of the government’s role should not be overplayed. Typically, Mokyr maintains that (1990: 238) ‘By the fifteenth century, the role of the imperial government in both invention and innovation was far less remarkable than it had been in medieval times, and no other entity in China was in a position to replace the state in promoting technological progress. There were no substitutes for the state in China. In Europe, precisely because technological change was private in nature and took place in a decentralized, politically competitive setting, it could be sustained in the long run.’

v. The making and supply of agrocrats to man the empire

Chinese physiocracy and agrocracy had their personae in Confucian officials who personally took initiatives to fulfil and materialise the physiocratic belief and policies. The best example were those who were dubbed ‘model officials’. According to Ershisi Shi, the Chinese dynastic histories, these model officials had three major deeds: (1) they substantially improved people's economic life, (2) they guided and civilised people, and (3) they were good judges in legal disputes between individuals. Table 2 contains some model officials of the same period (Han Dynasty) who took the initiative in organising and instructing agricultural production in the regions where they were in charge, including constructing irrigation networks and introducing new methods of tillage (such as cattle-draught ploughing) and new species of plants. Most of them were of middle rank: magistrates of prefectures. Some of these officials became so popular that peasants built temples to show respect for their deeds in promoting local agriculture (for two cases see Wang B. 1987).

Table 2. Han Model Officials who Improved Agriculture


Name Source

1. Ci Chong Fan Y. 445 A.D.: 'Biography of Model Officials'

2. Cui Shi Fan Y. 445 A.D.: 'Biography of Cui Shi'

3. Cui Yuan Fan Y. 445 A.D.: 'Biography of Cui Shi'

4. Du Shi Fan Y. 445 A.D.: vol. 31

5. Fan Ye Fan Y. 445 'Biography of Fan Ye'

6. He Chan Fan Y. 445 A.D.: 'Biography of He Chang'

7. Ji Zi Fan Y. 445 A.D.: 'History of the Eastern Barbarians'

8. Liu Kuan Fan Y. 445 A.D.: 'Biography of Liu Kuan'

9. Ni Kuan Ban G. c. 82 A.D.: vol. 58

10. Ren Yang Fan Y. 445 A.D.: 'Biography of Model Officials'

11. Wang Chong Li D. 466 A.D.: vol. 11

12. Wang Jing Fan Y. 445 A.D.: 'Biography of Wang Jing'

13. Wen Weng Chang Q. 354 A.D.: vol. 3

14. Zhang Dao Li D. 466 A.D.: vol. 5

15. Zhang Kan Fan Y. 445 A.D.: 'Biography of Zhang Kan'

16. Zhao Xinchen Ban G. c. 82 A.D.: 'Biography of Zhao Xinchen'



Table 3 shows the deeds from another angle, in which the same crop (rice) cultivation is considered.

Table 3. Rice Cultivation and Officials' Activities


Period Name Rank Action Code


Zhou Ximen Bao County Irrigation project 1

Li Bing Province Irrigation project 1

Han Zhao Guo Minister New method 2

Three Kingdoms Liu Fu Province Irrigation project 3

Zheng Hun County Irrigation project 4

N.+ S. Ren Yang County Teaching ploughing 5

Sui Ji Ye Province Irrigation project 6

Tang Zhang Jiuling Province New rice variety 7

Li Hantong Province New rice variety 8

Yu Di Province New rice variety 9

Song Zhenzong Emperor New rice variety 10

Zhu Xi Province Seed importing 11

Su Shi Prefecture New tool 12

He Chenju Province New rice variety 13

Ming Xu Guangqi Minister New rice variety 14

Guo Mian County Irrigation project 15

Qing Kangxi Emperor New rice variety 16


Note: a. (1) Sima 91 b.c.: ch. 'Book on Rivers and Channels'; (2) Ban G. 82 a.d.: ch. 'Economy'; (3) Chen S. c. 280 a.d.: ch. 'Biography of Chen Shuo' and Wei S. c. 554 a.d.: ch. 'Biography of Liu Fu'; (4) Chen S. c. 280 a.d.: ch. 'Biography of Zheng Hun'; (5) Li D. 466 a.d.; (6) Wei Z. 656 a.d.: ch. 'Economy'; (7) Liu X. 945 a.d.: ch. 'Biography of Emperor Xuanzong'; (8) Ouyang 1060 a.d.: ch. 'Biography of Guo Yuanzhen'; (9) ibid: ch. 'Biography of Yu Di'; (10) Tuo T. 1345 a.d.: ch. 'Biography of Emperor Zhenzong'; (11) Zhu X. 1182 a.d., also see Chen Z. 1958: 69–70; (12) Wang W. n.d. vol.6: 2051–2; (13) Tuo T. 1345 a.d.: ch. 'Economy'; (14) Xu G. 1639; (15) Zhang T. 1735 a.d.: ch. 'Record of Rivers and Channels'; (16) See Chapters 2 and 3 of this monograph.

b. N. + S. – Northern and Southern Dynasties.

In effect, the involvement in agriculture among Confucian officials was not just motivated by performing good deeds for promotion. It was much more personal than that due to the institution called ‘salary from land’ or ‘salary land’ (zhitian ). Such land constituted a considerable proportion of state-owned land. For example, between 1068 and 1076 A.D. of the Northern Song Dynasty, the total acreage of salary land was 2,349,000 mu (136, 242 ha), 37 percent of all state-owned land (Liang F. 1980: 290, 292). Salary land was cultivated by tenants, and the rent was the main source of income for officials, which meant that officials actually functioned almost as landlords of state land. The land distribution is illustrated in Table 4.

Table 4. Acreage of Salary Land for the Highest and Lowest Ranks


Period For the highest rank For the lowest rank


1. Northern Wei 1,500 mu (75.0 ha) 600 mu (30.0 ha)

2. Sui 500 mu (38.0 ha) 100 mu (7.6 ha)

3. Tang 1,200 mu (64.8 ha) 200 mu (10.8 ha)

4. Song 2,000 mu (116.0 ha) 200 mu (11.6 ha)

5. Jin 3,000 mu (174.0 ha) 200 mu (11.6 ha)

6. Yuan 1,600 mu (112.0 ha) 100 mu (7.0 ha)

7. Ming 1,600 mu (112.0 ha) 100 mu (7.0 ha)

Source: CBW 1978:151 and Liang F. 1980: 292, 472–5, 488–92

Note: Figures are converted to the modern measure.

Given that officials who run the Chinese empire only occupied a tiny fraction of China’s total population (see Table 1), and that the empire was a farming one, it was inevitable that a large proportion of these officials were agrocrats.


V. Irrelevance of population and land fluctuations to Institutions

What is rather surprising is that the private land ownership-centred economic institutions was to a great extent immune of changes in population and man-land ratios. This is the acid test of the Northian notion of the origin of institutional changes.


1. Elastic supply of land in the long term

Firstly, during the imperial period (221 b.c.-1911 a.d.), overall speaking, the expansion of China’s territory was the precursor of population expansion. This simply means that the land supply in the empire was rather elastic (see Figure 1). The obvious implication here is that the population pressure was not uncontrollably great over the long run for China.


Figure 1. China’s Territorial expansion in absolute terms

Source: Based on `Maps of Chinese History', National Geographic Magazine, July 1991, n.p.

Note: China as in: 1. 207 B.C., 2. 24 A.D., 3. 907 A.D., 4. 1644 A.D., 5. 1911 A.D.


Secondly, the elasticity in land supply was also taken the form of periodically improved man-land ratio following China’s notorious dynastic cycle (see Table 5).


Table 5. Dynastic cycle with Wars and Losses of Population in Chinese History

Population Population Decrease
Period Year before wars* after wars* (%)


1. Qin 221 b.c. 5,000,000

2. Han§ 157 a.d. 10,677,960


Three Kingdoms 263 1,473,423 86.2

(7,672,881) (86.4)

3. Sui§ 609 8,907,546


Tang (1) 618-626 2,000,000 77.5

(2) 650 3,800,000 57.3

4. Tang§ 854 4,955,151

N. Song (1) 976 3,090,504 37.6

(2) 996 4,574,257 7.7

5. S. Song and Jin§ 1193-1195 19,526,273


Yuan (1) 1291 13,430,322 31.2

(59,848,964) (21.6)

(2) 1330 13,400,699 31.4

6. Yuan§ 1330 13,400,699

Ming (1) 1381 10,654,362 20.5


(2) 1403 11,415,829 14.8


7. Ming§ 1626 9,835,426


Qing (1) 1655 -

(14,033,900) (72.8)

(2) 1680 -

(17,094,637) (66.9)


Source: Based on Liang F. 1980:4-11. For the Qin, estimation is based on (1) there were a million people constantly on corvée (Zhou Bodi. 1981. Zhongguo Caizhengshi (Financial History of China), Shanghai: Shanghai People’s Press.: 76; see also Sima Qian. 91 b.c. Shi Ji [The Book of History]. Reprinted, 1982. Beijing: Zhonghua Books: vol. 6 ‘Biography of Emperor Qin Shihuang’), and (2) on average one male adult per family was drafted for the corvée duty, and (3) there were on average five members in a family.

Note: The numbers without parentheses are numbers of households, while those in parentheses are numbers of people. Comparison of the data (1) and (2) from different periods indicates recovery speed.

N. – Northern; S. – Southern. * Registered population; § the starting points of comparison.


Thirdly, the elasticity of land supply also took the form of intensified use of land with an increased yield for each crop (see Figure 2) and an increased number of crops from the same land area over the year (known as ‘multiple cropping index’). In the post-Song period, it was more or less the norm for southern farmers in the Yangzi River region and beyond to grow two crops a year with the combination of one wet rice and one dry crop. Under multiple cropping, the same land area provides the same kind of utility as that of an increased acreage under single cropping. Premodern China was known for its continued intensification of land use. Hence, the elasticity in land supply was maintained most of time.

Figure 2. Yield Increase in China in the Long Run

Source: Yu Yiefei. 1980. 'Zhongguo Lidai Liangshi Pingjun Muchanliang Kaolue' ('An Investigation of the Dynastic Average Grain Output per Mu in China'). Chongqing Shiyuan Xuebao (Bulletin of Chongqing Teachers' College) 3:8-21.

Note: ß interval, from the Northern Wei (386–534) to the Tang (619–907) when state-guaranteed life time lease holding of farming land was practised.


2. Population growth and population pressure in the long run

From Figure 3, the two variables which determine population pressure – land territory and population size – changed over time in China. But the there was a tendency for the man-land ratio to decline. This does not necessarily mean, however, that the supply of land was not elastic.


Figure 3. Population Pressure in China

Source: Based on Liang Fangzhong. 1980. Zhongguo Lidai Use of uninitialized value in concatenation (.) or string at E:\listplex\SYSTEM\SCRIPTS\filearea.cgi line 455, line 426. Huko Tiandi Tianfu Tongji (Dynastic Data of China's Households, Cultivated Land and Land Taxation). Shanghai: Shanghai People's Press: 4–11.

Note: ß interval, from the Northern Wei (386–534) to the Tang (619–907) when state-guaranteed life time lease holding of farming land was practised.


3. Irrelevance to Institutions

Several points can be made under this heading. First, empirically, the establishment of the private land ownership and the making of the landholding free peasantry predated the formation of Chinese empire and the rise of Chinese population (see Table 5). In particular, the empire at 221 b.c. is estimated to have had at most a total population of one million (see Table 5). Yet, private land ownership was it was already well entrenched in China. This institution sustained until the early twentieth century when China experienced unprecedented population explosion (see Figure 3).

Second, the period from the Northern Wei (386–534) to the Tang (619–907), marked by ß in Figures 2 and 3, was the period when the state nationalised farming land and then leased it out to farmers on the permanent basis under the institution called the Land Equalisation System (juntianzhi). It was a disguised private ownership, aiming at guaranteed land supply to farmers. During this very period, China’s man-land ratio was record high with abundant supply of land and the zeal for technical improvement was low.


VI. Final remarks

The state-building, rather than population growth, was the determinant for institutional changes in premodern China. This finding has serious implications. Most important of all, it suggests that the rationale of institutional change may well be Weberian which has little to do with the normal economic variables such as labour, capital, land and so forth.

Secondly, over time, institutions may not change even economic conditions are changed. The fixation of institutions may look like and feel like path dependency. But it is the cause of the ‘path’, not the result of the it.

Thirdly, from the view point of institutions, the great divergence between western Europe and East Asia began very early on, certainly not after the Renaissance, but before Christ.

Fourthly, if institution is a largely an independent variable, it becomes hopeful for a society to be reformed and re-moulded regardless of its indigenous conditions and constraints. So, globalisation is not only possible but feasible.



References (to be included)

Wang Shusheng, Wang Tengda, Fang Renxi, Tian Zhaolin, Shi Shengbi, Bai Taichang, Chen Xijin, He Shaohe, Shao Qing, Zhao Gusong, Zhao Xiukun, Guo Rugui, Tao Wenhuan and Cao Yuexuan. 1986. Zhongguo Junshishi Fujuan Shang (A Military History of China), Vol. 2. Beijing: PLA Press.

Tian Zhaolin, Tao Wenhuan, Chen Xijin, Zhao Xiukun, Shi Shengbi, Chen Yangping, Deng Zezong, Yao Zhihong and Yang Qi. 1990. Zhongguo Junshishi Fujuan Shang (A Military History of China), Vol. 5. Beijing: PLA Press.

Wei Zhenfu, Tian Zhaolin, Yi Xiuliang and Huang Yibing. 1993. Zhongguo Junshishi Fujuan Shang (A Military History of China), Vol. 1. Beijing: PLA Press.