The Deer Was Never Consulted

Four hundred households, then little more than a hundred. I kept returning to those numbers while reading Ch'ien Mu's Traditional Government in Imperial China. They come from a Tang official's account of a district where families had fled, leaving the remaining households to shoulder a tax burden that refused to shrink with the population. A tax quota is a remarkably durable thing. People disappear, fields go uncultivated, and the figure in the register remains.

Ch'ien does not overlook the cruelty of this arrangement. He gives it serious attention, which is one reason I enjoyed the book so much. Still, I found myself lingering over the missing households rather longer than the institutional history required.

The book began as five lectures delivered in Taipei in 1952, later published in Chinese as Zhongguo Lidai Zhengzhi Deshi (中國歷代政治得失). The English translation by Chun-tu Hsueh and George O. Totten appeared in 1982. In a remarkably small compass, Ch'ien traces the government of five dynasties, Han, Tang, Song, Ming and Qing, through their administrative offices, recruitment systems, finances and military arrangements.

I have considerable admiration for the undertaking. To explain two thousand years of political institutions is difficult enough; to make their gradual transformations intelligible is another matter. An office retains its name while its powers migrate elsewhere. A reform quietly becomes the problem for which the next reform is prescribed. The emperor and his ministers contest authority, officials acquire new routes into government, and arrangements intended to be temporary develop the sort of longevity usually associated with bad habits. Ch'ien knows where the changes began, and he understands that institutions acquire their character from the people who use them.

He also refuses an easy verdict on the imperial past. Calling two thousand years of government despotism may have the advantage of brevity, but it saves the historian from rather too much work. Ch'ien wants to know why an institution was established, what alternatives were available, and why reasonable people might have considered it useful. Even when he finds an arrangement defective, he tries to understand the circumstances that produced it.

His circumstances deserve consideration too. Ch'ien left mainland China for Hong Kong in 1949. He delivered these lectures in Taiwan and prepared the manuscript while recovering from an injury in Taichung. In the preface to his earlier Guoshi Dagang (國史大綱, Outline of National History), he had asked readers to approach their country's past with wenqing yu jingyi (溫情與敬意), warmth and respect. There is something particularly affecting about that appeal from a historian whose own world had been so thoroughly unsettled. I cannot read it without feeling that he deserves the same generosity he asks us to extend to the past.

And, the book kept making me wonder about the people who appear in its accounts of government chiefly as the objects of administration.

The Open Door

Near the end, Ch'ien reflects on several tendencies he considers troubling. Among them is China's increasingly levelled society (平鋪的社會). The great hereditary families had declined, political careers had opened to qualified scholars from less privileged backgrounds, and society had become more equal in certain respects. Ch'ien's worry is that this flattening also left society dispersed and poorly organized. He invokes Gu Yanwu's (顧炎武) concern that, without strong local organization, society would have little to rely upon when central authority failed. The lure of office, meanwhile, drew talented men into government and helped produce an increasingly swollen bureaucracy.

I rather enjoyed that complaint about the bureaucracy. A civilization that rewards its brightest young people by encouraging them all to become officials should perhaps not be astonished when it acquires a large number of officials.

Ch'ien also recognizes the achievement behind the problem. As he explains in his preface, the recruitment and examination systems opened a channel between government and society. Men without hereditary claims to power could enter public service, sometimes rising from modest circumstances to positions of considerable authority. Their children would have to earn the same distinction for themselves, or settle for being the children of someone who once held office.

The imperial examinations offered possibilities unknown under a rigidly hereditary order. But education took money and time, and the candidates were drawn from a limited part of society. Even among those eligible, success was rare. Tu Fu (杜甫) sat the examinations and failed. Li Bo (李白) never sat them, commonly attributed to his family's merchant background. Between them, they did reasonably well in poetry. A narrow opening can transform the prospects of those who pass through it while leaving the position of everyone else largely unchanged.

More interesting to me is what happened once a successful candidate crossed the threshold.

A scholar might grow up among farming families, pass his examinations and eventually govern a county resembling the one he had left. He could remember the anxieties of a bad harvest. Perhaps he would be unusually sympathetic to people struggling to pay their taxes; perhaps he would spend his career trying to protect them. There were officials who did exactly that. Yet his livelihood, reputation and prospects now depended upon a hierarchy above him. The people below could petition, complain and occasionally secure his punishment. They had no regular, institutionalized right to appoint or dismiss him.

The examination system selected managers. It did not make the population shareholders.

This is where my habits intrude upon the reading. I keep wanting to know who appoints whom, who pays, who bears the loss when a decision goes badly, and who has an effective way of saying that a manager has ceased to serve the purpose for which he was appointed. It is an irritating habit when applied to a book about imperial China, since the institutions were never designed to satisfy the expectations of a modern shareholder meeting. Nevertheless, the questions persist.

The people were never a single body with a single interest. A wealthy landowner and his tenant might both want peace and a functioning irrigation system, but their agreement could end rather quickly when the cost had to be divided. Merchants, laborers and farming households could have competing claims upon the same magistrate. A conscientious official had to judge among them. His learning and character might help him make that judgement wisely. They could not give the people concerned a right to determine whose interests he should represent.

Ch'ien pays considerable attention to officials who remonstrated with emperors and criticized abuses of power. Those institutions mattered, and some men exercised their duties with remarkable courage. What unsettles me is the distance between an official who speaks for the people and one who must answer to them.

The distance appears again in the historical record.

Ch'ien makes a distinction in his preface between lishi yijian (歷史意見), the views of people living under an institution, and shidai yijian (時代意見), the judgements of later generations. He asks us to take contemporary opinions seriously instead of dismissing the past for failing to conform to our expectations. I agree, and I like the intellectual discipline of it.

But who managed to leave an opinion for us to consider?

Ch'ien points to the memorials of distinguished ministers, and rightly so. They often contain careful observations, direct criticism and a considerable amount of practical knowledge. Yet the educated and governing classes also possessed an overwhelming advantage in producing and preserving the material from which historians reconstruct political opinion. The archive inherited their habits of attention. The people most qualified to argue about government were often the people whom government had qualified to argue.

I do not know what a particular farmer thought about a particular reform unless some trace of it survives. It would be dishonest to supply him with a speech. Still, I cannot accept that the absence of a speech tells us very much about his opinion, much less his capacity to form one.

Managing the Deer Park

Ch'ien himself offers a wonderful example of the difference between the visible hierarchy and the machinery underneath it.

In his discussion of the Ming and Qing, he turns to the xuli (胥吏), the clerks and runners who handled much of the everyday business of government. The formal distinction between the educated official and the lower-status clerk had hardened over time. An official could arrive with considerable rank and little practical knowledge of the office he had been appointed to run. The clerks knew the records, the procedures, the delays that could be arranged, and the small turns of language upon which an administrative decision might depend.

Ch'ien quotes a late-Qing list of the business that passed through their hands: appointments, disciplinary matters, finance, ceremonial affairs, criminal cases, litigation and public works. It is worth lingering over that list. We are accustomed to thinking of bureaucratic paperwork as a kind of residue left behind by the important work of governing. Here the paperwork was the governing. An appointment could be hastened or delayed; a case could be made to appear more or less serious; the administration of a tax depended on the man who knew which register to open and what the entries meant. A magistrate had the authority to issue orders. The clerk knew how an order became an event.

I find this fascinating, partly because it is such a familiar organizational problem. The person with formal authority is rarely the person with all the information, and may be the last person in the building to discover it. Bradly Reed's work on Qing clerks and runners gives this administrative world further depth, while Tuan-Hwee Sng's(孫傳煒) economic analysis explores the difficulty of supervising agents across an enormous empire.

There is a principal-agent problem here. The ruler depends upon officials whose actions he cannot fully observe; officials rely upon subordinates with knowledge of their own. The emperor may wish to preserve revenue and public order over the long term, while a temporary officeholder has his next appointment to consider. A clerk may have reasons to cultivate procedures that make himself indispensable. Nobody has to be a monster for the arrangement to produce unpleasant results, although it certainly provides opportunities for monsters.

Imagine being appointed keeper of a royal deer park. The estate is someone else's. You could spend years improving its habitat, only to be transferred before the herd recovers. Or you could increase this year's hunting yield, produce a splendid report and leave the next keeper to explain the declining population. The owner might be furious when he discovered what had happened, but discovery is the part of the problem that his keeper is paid to manage.

The deer could lose more meat than ever reached the emperor's table.

What I find harder to put aside is that better supervision would address only part of this difficulty. An obedient official might be better for the emperor than a corrupt one, and often better for the population too. Yet obedience alone would not determine whose interests ought to govern the official's decisions. The state could make its agents more obedient. It could not, merely by improving their obedience, make the people their principals.

Now back to the households.

The Treasury's Claim

Ch'ien's account of the Tang tax reforms is where I spent the most time, perhaps because the arithmetic becomes so painfully straightforward. The old zuyongdiao (租庸調) system rested upon land allocation, household registers and obligations in grain, textiles and labour. It required the state to know a remarkable amount about its subjects. Ch'ien describes registers prepared in multiple copies and revised at different intervals, with records of households and able-bodied men kept for years so that officials could compare successive returns.

Think of the administrative effort. Births, deaths, changes in household circumstances, land entering and leaving cultivation. Every alteration had to be recorded, copied, transmitted and checked, and the task was performed across a country of considerable size without the conveniences of modern communications. Ch'ien points out how easily such records deteriorated when the officials responsible became careless, and how much depended upon sustained attention to unglamorous work.

It is tempting to find that quaint. I cannot. Anyone who has ever tried to reconcile records maintained by different offices will recognize the problem, even if the modern offices have computers and considerably fewer excuses.

As the older arrangements broke down, the Two-Tax Law, liangshui fa (兩稅法), introduced under Yang Yan(楊炎) in 780, offered a simpler system. It allowed taxation to follow property and wealth more closely and collected payments twice a year. Ch'ien acknowledges its advantages, including the freedom it offered households that had moved from one locality to another.

The difficulty lay partly in how local tax quotas were established. Ch'ien describes a movement from liangru weichu (量入為出), determining expenditure according to available revenue, towards liangchu weiru (量出為入), requiring revenue to meet established fiscal demands. Quotas based on previous collections could survive changes in the population that had once paid them.

He cites Lu Zhi(陸贄), the Tang official, reporting a township that had fallen from four hundred households to little more than a hundred, and another district that had lost roughly two-thirds of its households. Those still present inherited the burden of those who had fled.

I find the first example difficult to read as merely a defect in taxation. Four hundred households become a hundred or so. The departure of so many families must have had causes and consequences well beyond the figures in the register, although the record does not tell us enough to reconstruct their lives. We can see what the shrinking denominator does to the tax burden. We know much less about the decisions that took the missing families away.

The arithmetic is cruelly simple. If ten households are responsible for a fixed quota and five flee, the remaining five must pay twice as much per household, assuming the burden is divided evenly. That gives some of them a reason to leave too. The district can enter a fiscal decline in which the effort to collect the original revenue makes the eventual collection even less likely.

Then there is the question of money.

Ch'ien quotes a later observation by Li Ao(李翱) about the changing price of silk. At the beginning of the reform, a tax liability of ten thousand cash could be met with two and a half bolts of silk if each bolt fetched four thousand. Decades later, with silk selling at eight hundred cash a bolt, the same nominal sum would require twelve bolts.

Nearly five times the cloth for the same figure on the tax account.

I confess that numbers like these hold my attention longer than the names of most government offices. The nominal obligation has not changed, so the accounts may look reassuringly consistent. A household producing or selling silk would experience something quite different. It would have to surrender much more of what it had made, or sell more of its produce, merely to discharge the liability that appeared unchanged in the government's books. The account has preserved its figure. The household has had to absorb the difference.

And Ch'ien sees it. That matters to me. He is attentive to the cost of reforms that appear administratively sensible, and he does not ask the reader to admire a convenient collection system simply because it was convenient to those collecting the money.

Still, once I started thinking about the accounts in this way, I became less satisfied with the usual balance of institutional gains and losses. A reform can improve the treasury, simplify administration or strengthen central control, and those achievements may benefit the population. The distribution of the gains deserves its own accounting.

The Proprietor and His Subjects

Mancur Olson's distinction between roving and stationary bandits offers one explanation for why even a hereditary ruler might have reason to protect his subjects. A ruler expecting to pass his territory to his descendants has an interest in maintaining productive farms, orderly markets and a stable population. His interest can overlap with the welfare of the people. The arrangement is not without its domestic complications. Too many sons could produce a succession struggle; too few could leave the dynasty searching anxiously for an heir. The emperor's family arrangements had a way of becoming everyone else's problem.

Huang Zongxi (黃宗羲), in Mingyi Daifang Lu (明夷待訪錄), asks a question Olson's calculation cannot answer. He challenges the ruler's treatment of the world as private property and considers how authority might be restrained. His proposal to give schools a role in scrutinizing government is striking, although the responsibility still falls chiefly upon scholars. Even a formidable critic of royal ownership looked primarily to educated men to challenge those who ruled.

Olson helps explain why a proprietor might take care of his property. Huang asks what entitles him to regard the world as his property in the first place.

I would be unfair to Ch'ien if I suggested that his book overlooked ordinary people's welfare. It plainly does not, and the discussion of Tang taxation alone should dispose of that accusation. My regret is more particular. He has taught us to see how institutions develop, why their designers thought them reasonable, and how their consequences changed over time. I find myself wanting to apply another measure as well: how much authority the governed possessed to influence the purposes of government.

Peace mattered. So did irrigation, relief from arbitrary demands and the ability to carry goods to market without being robbed along the way. These were real gains, sometimes enormous ones, and a government that delivered them deserved credit. A farmer whose harvest was saved by an irrigation project had no reason to disdain the benefit merely because he had not voted for the magistrate who organised it.

But when the interests of the landowner and the tenant diverged, or when the treasury's requirements exceeded what the households could bear, who was entitled to decide what good government meant? The population might be well served by a conscientious ruler, or badly served by a careless one. Its fortunes remained heavily dependent upon the qualities and priorities of people it had little regular power to appoint or dismiss.

Ch'ien's title asks us to weigh the gains and losses of successive political institutions. I came away wanting to ask whose gains, and whose losses. The historical sympathy he asks of his readers should surely have room for the people whose opinions so seldom reached the record.

In Sima Qian's (司馬遷) Shiji (史記, Records of the Grand Historian), Kuai Tong (蒯通) describes the fall of Qin through a memorable image: Qin lost its deer, and the world competed to catch it. The deer stood for the imperial throne. But the throne came with the land, the grain, the labour, and the people over whom it conferred authority.

Ch'ien tells us how the hunters organized themselves, divided their responsibilities, recruited their assistants and considered the proper management of their prize. He also explains why their arrangements so often failed.

I finished the book wondering about the deer. It had been sustaining the hunters for a very long time. Nobody seems to have thought its opinion particularly relevant.

Sources

Ch'ien Mu, Traditional Government in Imperial China: A Critical Analysis. Translated by Chun-tu Hsueh and George O. Totten. The Chinese University of Hong Kong Press, 1982. Chinese original, Zhongguo Lidai Zhengzhi Deshi (中國歷代政治得失), particularly the Preface, Tang lecture, Ming lecture and General Conclusion.

Ch'ien Mu, Guoshi Dagang (國史大綱), Outline of National History, Preface, on historical understanding and wenqing yu jingyi (溫情與敬意).

Huang Zongxi, Mingyi Daifang Lu (明夷待訪錄), particularly Yuanjun (原君, On the Ruler), Yuanchen (原臣, On the Minister) and Xuexiao (學校, On Schools).

Mancur Olson, “Dictatorship, Democracy, and Development,” American Political Science Review 87, no. 3 (1993): 567–576.

Tuan-Hwee Sng, “Size and Dynastic Decline: The Principal-Agent Problem in Late Imperial China, 1700–1850,” Explorations in Economic History 54 (2014): 107–127.

Bradly W. Reed, Talons and Teeth: County Clerks and Runners in the Qing Dynasty. Stanford University Press, 2000.

Sima Qian, Shiji “Huaiyin Hou Liezhuan” (史記·淮陰侯列傳), containing Kuai Tong's account of Qin losing its deer.