The Scale of Regional Financial Administration in the Later Part of the Tang Dynasty

As a preliminary step toward clarifying the actual circumstances of regional financial administration in the later part of the Tang Dynasty, let us first survey the scale of regional financial administration in this period. In the financial administration of the later part of the Tang Dynasty, the portion of income coming from revenue sources derived from the twice-a-year tax and monopolies was a fixed amount; in the state as a whole, it was about 36 million guanshi (30 million guan in cash and commodities, and 6 million shi in grain). Of this, one-third, about 12 million guanshi (9.5 million guan, 2 million shi), constituted the shanggong revenue sent to the central government's Ministry of Revenue (Duzhi 度支); the remaining two-thirds, about 24 million guanshi (20.5 million guan, 4 million shi), comprised the regional financial administration, retained by the circuit and prefecture, respectively, as songshi (liushi) and liuzhou cash and commodities.Footnote ⁸

The individual circuits and prefectures were rated according to their populations and their location conditions, and the financial administration of each circuit or prefecture had its fixed amount determined in accord with this evaluation. Aside from the case of Suzhou that has been introduced by Hino (see Table 1), only the most fragmentary information is available to offer concrete examples of the scale of these financial administrations.Footnote ⁹ Here, I will take the Zhexi circuit (which includes Suzhou), and the Jiannan/Xichuan circuit as examples and survey the special characteristics of the scale of circuit and prefecture financial administration.

According to the Yuanhe junxian tuzhi 元和郡縣圖志 (Illustrated gazetteer of Yuanhe prefectures and counties), juan 25, the Zhexi (Zhejiangxi) circuit was the leading regional administrative center in Jiangnan, comprising six prefectures (the capital Runzhou, as well as Changzhou, Suzhou, Hangzhou, Huzhou, and Muzhou) and 37 counties, with 313,772 households and 57,932 qing of arable land. There are no surviving historical documents which tell us the amount of the twice-a-year tax for the whole of the Zhexi circuit and the scale of its financial administration. However, according to an imperial decree dated the renshen 壬申 day of the second month of Yuanhe 2 (March 26, 807):

From this it is clear that with the cash portion of the twice-a-year tax alone, the circuit forwarded tax revenues to the central government in excess of 340,000 guan. Also, according to the biography of Li Deyu 李徳裕 in the Jiu Tang shu 舊唐書, the cash portion of tax revenue retained by the circuit (liushi), which formed the bulk of circuit finances, was 500,000 guan, and that when this was combined with other revenues, the total expenses of the circuit were roughly 630,000 guan.Footnote ¹¹ It is unclear how much tax revenue was in the form of grain, but it seems likely that the money and grain combined were close to one million guanshi.

The prefectures under the jurisdiction of the Zhexi circuit include many large ones. Among them, Suzhou was a pivotal site for the financial administration of the Tang Dynasty state, as the following description indicates:

As can be seen from Table 1, which shows “Tax revenue and distribution for Suzhou,” the combined revenue in cash from the twice-a-year tax and monopolies in late-Tang Suzhou was approximately 590,000 guan, which was divided into three parts: in excess of 300,000 guan that was sent to the central government (shanggong); more than 100,000 guan that was sent to the provincial governor (songshi), and more than 170,000 guan that was retained as cash for the prefecture (liuzhou). The amount of revenue that was grain is unclear, but we can see that in Suzhou, the revenue was roughly equally divided between the portion transmitted to the central government and the portion that made up regional financial administration. There are no historical documents that survive from other prefectures, but it is known that the tax revenue in cash for Hangzhou was 500,000 guan; from this we can tell that its scale was on par with Suzhou's.Footnote ¹³

In the case of Jiannan Xichuan circuit, the circumstances of the circuit as a whole are comparatively well known. The Xichuan circuit comprised 26 prefectures, beginning with Chengdu-fu, and 112 counties.Footnote ¹⁴ The amount of the twice-a-year tax for the entire circuit was 2,013,645 guan and the amount of cash and rice transmitted to the central government as changfu 常賦 “regular payment” (i.e., shanggong) was in excess of 700,000 guanshi. Out of the total amount submitted to the central authorities, the cash portion of the twice-a-year tax came to more than 560,000 guan and the grain portion to 140,000 shi.Footnote ¹⁵ Speaking only of the cash portion of the twice-a-year tax, we can see that about one-fourth went toward making up the central Ministry of Revenue (duzhi) finances as shanggong, and three-fourths made up regional finances (liuzhou and liushi). This is close to the national-level distribution structure that I explained earlier, but it is certainly not the case that these portions were mechanically imposed on a national scale.

When looked at in this way, it is clear that the ratio of division of the twice-a-year tax between the regional and the central government was not uniform and seems to have been determined through the specific relations between each region and the central authorities. In the case of Zhexi circuit and Suzhou, the ratio of tax revenue forwarded to the central government as shanggong was probably high because this was an area that comprised the major revenue source for the central government's financial administration; as the above quotation notes, “at present, the state's finances for the most part come from Jiangnan.” What I have presented above gives us an overview, albeit an insufficient one, of the scale of regional financial administration for which the twice-a-year tax and monopolies served as revenue sources. This was not the full extent of regional financial administration in the later part of the Tang Dynasty, however, for it was managed with an even broader range of content. Below, I would like to take up this concrete content.

Income and Expenditures in Prefectural Financial Administration

On the basis of multiple historical documents, I have compiled Table 2, which shows “A chart of regional (prefectural) financial administration in the late Tang Dynasty.” Below I would like to pursue my investigation on the basis of this chart.

The income for prefectural financial administration comprised three parts: 1) the liuzhou portion of tax revenue that remained in the prefecture; 2) the “money for public use” (gongyongqian 公用錢); and 3) the ever-normal and charity granaries. Previous research has concentrated on elucidating the nature of partitioning in the liuzhou portion. Moreover, with regard to the nature of expenditures, aside from a few brief mentions by Hino Kaisaburō, there has been virtually no consideration of this topic. When it comes to the portion made up of public moneys, it has not been elucidated at all.Footnote ¹⁶ Below, for each of these three areas, I would like to investigate its special characteristics as well as what role it played in the prefecture's financial administration.

First of all, the liuzhou portion constituted the general financial administration of the prefecture; it was a fixed amount made up of the allotment from the twice-a-year tax and liquor-monopoly money (quanjiuqian 權酒錢). This portion was in general managed under the decision-making authority of the prefect (cishi 刺史) of the prefecture. Therefore, the central government did not directly take a hand in the management of financial affairs. The following passage from “Amnesty to commemorate the presentation of an honorific title” (Shang zunhao shewen 上尊號赦文), dated the 23rd day of the 4th month of Huichang 2 (842), is a good example of this:

However, the twice-a-year tax that served as the local government's revenue source was, in systematic terms, unified under the authority of the central Ministry of Revenue (Duzhi 度支), and was generally speaking directed along a chain of command that ran either from the Ministry of Revenue through the Surveillance Commissioners (guanchashi) or Military Commissioners (jiedushi) to the prefecture or in some cases by imperial command to the circuit or prefecture. There were also extraordinary cases in which these funds were reclaimed as part of central expenses.Footnote ¹⁸ Although it is true that the significance of regional financial administration differed depending on the political power relations between the central government and the various regions, the liuzhou and liushi portions of tax revenue were essentially a part of the central government's financial administration. They should be thought of as its property that was consumed in the localities.

Let us examine the structure of expenses for the liuzhou portion. A historical document that provides a clear example of this is the following imperial decree dated the first month of the first year of Huichang (841):

According to this, we can see that there were categories of expenditure through the circuit and prefectural financial administration for: 1) salaries for local bureaucrats; 2) fees for use of inns and stations (communication and transportation expenses); 3) post station labor (communication expenses); and 4) miscellaneous moneys (various other expenses).Footnote ²⁰ Though it is not seen here, with the military reforms of Yuanhe 14 (819), soldiers came to be stationed in each prefecture; henceforth, military expenses came to constitute the core of prefectural finances. A good example of this is the one I included above in Table 1: that the liuzhou section in Suzhou from the Wudi ji is labeled “Retained in Suzhou for food, clothing, military and other expenses.”Footnote ²¹ In this way the basic pattern was that expenses associated with military activity, including soldiers’ salaries and military expenses (equipment, city walls, repairs of moats, etc.) came to constitute the main portion of expenses for the liuzhou portion,Footnote ²² on top of which were added the salaries of the local officials, communication and transportation expenses.

Out of these expenses, salaries for soldiers and bureaucrats constituted the largest part of expenditures.Footnote ²³ There were a fixed number of soldiers and bureaucrats, and with the exception of the fluidity of military action expenses, the expenditures for the liuzhou portion came in large part to have fixed amounts. A good example of this can be seen in an imperial order dated the eighteenth day of the eleventh month of Dahe 3 (829), which reads:

This demonstrates that the liuzhou portion had the same structure of expenses as the financial administration directly under the jurisdiction of the central government's Ministry of Revenue.Footnote ²⁵ We can confirm that from the point of view of expenditures, the liuzhou section was a part of the central financial administration.

Let us turn to the “money for public use” (gongyongqian 公用錢) section. It was this section, in fact, that constituted the distinctly regional income and expenses. This section comprised the following: 1) the “remaining surplus” (huican xianyu 廻殘羨餘); 2) various kinds of capital; and 3) other. It was called “money for public use” and was clearly distinguished in terms of income and expenses from the liuzhou section that formed the general financial administration. Let us look at these components in sequence.

The “remaining surplus” (huican xianyu) is the property left over from the liuzhou tax revenue section (which had been apportioned as a fixed amount); such surplus might occur simply because the liuzhou revenue had not been completely spent, or it might come about as the result of economizing financial expedients (fangyuan 方圓). In other words, it was a reapportioned share of the twice-a-year tax that had been apportioned as liuzhou.Footnote ²⁶ Because of this, it could be used much more flexibly in regional finances than the liuzhou section. For example, the following Bureau of Review memorial from the ninth month of Dahe 4 (830) is one example:

In this way, the “remaining surplus” turns out to have had a diverse range of characteristics; it formed one component of regional financial administration, but at the same time it could also be used for extra levies (zaquelü 雜榷率) to the twice-a-year tax or as a source of revenue for submission to the emperor. Bai Juyi turned the “remaining surplus” into the topic of a poem, criticizing it as follows:

From this poem, it is clear that the category of “surplus items” was by no means merely the leftovers from the general financial administration proper; rather, it could also include commodities that had been forcibly and improperly seized from the peasantry.

Next, the category of “various types of cash capital” (zhuse benqian 諸色本錢) refers to the cash (capital) of both central and regional government offices that was nominally designated “profit-making cash” (shiliqian 食利錢) among other names. This cash was entrusted to money-lenders (zhuoqianhu 捉錢戶) and petty officials (guandian 官典 or xuli 胥吏) who managed it. They used this cash as principal for trade, buying and selling on the market, and to lend out to private individuals. The profits from commerce and interest (at a monthly interest rate of 4–5 percent) were recovered and applied to expenses. For example, a Census Department (Hubu 戶部) memorial from the sixth month of the first year of Huichang (841) reads as follows:

This memorial from the Census Department is a petition raised mainly among the central government officers to make an additional request concerning the newly distributed profit-making cash capital of 20,000 guanwen. Evidence that regional officials are included in this is clear from the fact that the memorial contains the phrase “thus it should be sufficient to sustain the various internal and external (i.e., central and regional) offices for a long time.” In the case of regional officials, the revenue source for this cash that would be used as capital was basically from the redistributed portion of the liushi or liuzhou cash and commodities or from “remaining surplus.”Footnote ³⁰ From Yuanhe 10 (815) onward, however, there was also distribution from the central government's Census Bureau (hubu cao 戶部曹).Footnote ³¹ As I will explain later, the income from the management of these various types of cash capital can be traced back to the Han Dynasty,Footnote ³² but it is something that directly continued the system of the Sui dynasty and is the portion among the regional financial administration of the later part of the Tang with the oldest origins. However, it can be said as well that insofar as the revenue source came from commerce and profit, it was also the aspect where a parasitic quality was most intense.

As for other sources of revenue, there was income too from the commercial tax that was implemented in Yangzhou as well as from the tax-hemp (shuima 税麻) of the same prefecture. There were also circumstances in which bureaucrats spared some of their stipends and applied it to money for public use.Footnote ³³

Looking at the income in the money for public use section as a whole, we can make a basic division between the redistributed portion from liuzhou and liushi tax revenue on the one hand, and an income section that was parasitic with regard to commerce and the circulation of currency on the other; the latter included commercial taxes, high-interest lending, and so forth. From these features, it is inconceivable that there was a fixed amount of income every year. The liuzhou section had as its revenue source the fixed-amount redistribution of the proper tax, that is the twice-a-year tax, and if this constituted the general financial administration, then we can understand the portion of “money for public use” as constituting a special section of the financial administration that was fluid.

Let is consider the expenses corresponding to the “money for public use” section. With regard to the “money for public use” that had its revenue source in “remaining surplus” cash and goods, a certain framework was determined as follows in accord with the proposal made in Dahe 4 (830) by the Bureau of Review:

From this it is clear that the expenses were designated as: 1) construction and repair of government offices and furnishings; 2) entertainment expenses associated with the exchange of public and private envoys and bureaucrats; 3) regional police operations; 4) supplementation of unpaid tax duties; and 5) stockpiling for use in natural disaster. Of the five items enumerated above, stockpiling for use in natural disasters was in particular encouraged as part of the fundamental expenses, and in situations where the “money for public use” was left over, it was encouraged that grain for stockpiling be purchased.Footnote ³⁵ In Dahe 4 and earlier, it seems that there were various kinds of expenditures in addition to the five categories enumerated above, but with this year came the imposition of a certain fixed framework. With regard to this point, I would like to take up the matter again in the next section, below.

In the case of “money for public use” that had its revenue source in profit obtained from various kinds of capital cash, the expenses included the following: 1) construction and repair of government offices and supplies; 2) Entertainment expenses associated with the exchange of public and private envoys and bureaucrats; 3) meals for lingli 令吏 and other low-ranking officers (guanxieliao 官廨料 and gongxieliao 公廨料); 4) administrative operational expenses such as paper and brushes. In some circumstances, it seems that this money was also applied to compensation for officials.Footnote ³⁶ It is unclear how money for public use that derived from other revenue sources was expended, but it is thought that it would have been similar to the above structure of expenditure.

On the whole, the two groups of expense structures outlined above were concerned with the typical official duties of the prefect. A memorial from the Director of the Bureau of Evaluations (kaogong langzhong 考功郎中) in the seventh month of Taizhong 6 (852) reads as follows:

According to this, the typical official duties of prefects can be summarized into five categories: 1) collection of levies and taxes and making the collection equitable; 2) judicial decisions; 3) maintenance and preservation of households and arable land; 4) maintenance and preservation of official buildings; 5) maintenance and preservation of communication and transportation operations. Aside from affairs concerning judicial decisions (number 2), the other official duties were fiscally supported by the expenses from the two sections of liuzhou tax revenue and money for public use. Yet, whereas the liuzhou tax revenue section was of a piece with the Ministry of Revenue's financial administration, when it came to the section of “money for public use,” the scope of the prefect's individual independent judgment was relatively broad, although this section still had a certain connection with the Secretariat-Chancellery (zhongshu menxia 中書門下) as well as the Census Department and Bureau of Review.Footnote ³⁸ In this way, the section of “money for public use” was structured around various kinds of expenses that corresponded to administrative necessities for the prefecture in excess of the typical duties of the prefect. The expenses for which the liuzhou tax revenue section was used had become a fixed amount and permitted virtually no freedom of movement. The “money for public use” not only made it possible to provide some side support for these expenses, but it is also fair to say that this section constituted an independent regional financial administration.

Let us turn to the ever-normal and charity granaries, which constituted the third branch of prefectural financial administration. Funakoshi Taiji has researched ever-normal and charity granaries of the latter half of the Tang dynasty. According to his findings, the system of ever-normal and charity granaries was something newly established from Yuanhe 1 (806) onward. It was a system in which individual regional prefectures stored up grain in preparation for disaster and it had not only the function of providing relief but also of stabilizing commodity prices. When it was first established in Yuanhe 1, two-tenths of the grain portion of the twice-a-year tax that individual prefectures collected was spared and made the source of grain for storage. Subsequently, with a proposal from the Census Department in Kaicheng 1 (836), the source came to be an additional rice levy in the amount of one sheng of rice for each mou of land, whether public or private, but this was halted in Huichang 6 (846). The mission of the ever-normal and charity granaries to serve as sources of relief was fulfilled in the period between the Yuanhe era and Kaicheng 1. Afterward, the ever-normal and charity granaries continued to exist all the way to the end of the Tang, but they degenerated to the point of becoming a system for gathering tax grain that was only charity granary grain in name.Footnote ³⁹ The ever-normal and charity granaries were managed at the prefectural level by the Administrative Supervisor (lushi canjun 録事參軍) and the central Census Department had general oversight. As is clear from their revenue sources as well as their administrative lineage, the ever-normal and charity granaries occupied a place connected to the Census Department and were fundamentally a form of financial administration with strong central aspects.Footnote ⁴⁰ Therefore, even in cases of diversion to regional expenses, this was carried out on the basis of central directives.Footnote ⁴¹

Together with the use of “money for public use” to stockpile grain in the event of natural disaster, the ever-normal and charity granaries formed a two-part insurance expense. In order to make use of the grains that had been stored up in the ever-normal and charity granaries, the decisive authority of an imperial edict was required. This meant that although the system was able to display impressive power when it came to long-term responses, it was ineffective in response to natural disasters that necessitated urgent measures. To use the grain that had been stored away against natural disasters through the expenditure of “money for public use,” by contrast, a decision made at the prefectural level was sufficient, meaning that an ad hoc response to an emergent situation was possible.Footnote ⁴² Had it fulfilled its original purpose, the implementation of ever-normal and charity granaries and the use of “money for public use” to store grain away in the event of natural disasters would have amounted to an insurance expense that, it is fair to say, would have played a major role in the social reproduction of Tang dynasty peasants. However, there was a second dimension to its character; just like the charity granaries and land tax in the first half of the Tang dynasty, it became an additional tax outside of the tax proper, and often transformed into a means of exploiting the peasantry.Footnote ⁴³

Adjusting the Equilibrium of Prefectural Financial Administration

As described above, regional financial administration in the latter half of the Tang dynasty controlled three branches: 1) the liuzhou portion of tax revenue; 2) the portion of “money for public use”; and 3) the ever-normal and charity granaries. Whereas the liuzhou portion was financial administration that was connected to the central government's Ministry of Revenue and the ever-normal and charity granaries was financial administration that was connected to the Census Department, in the case of the money for public use portion, autonomous management by the prefecture was permitted, albeit with certain restrictions.Footnote ⁴⁴ Now then, in the management of the regional financial administration, what sort of mutual relations existed between these three constitutive branches? Let us focus on the equilibrium between income and expenditure.

It is unclear what sort of ratio existed in the composition of these three branches, but needless to say the mainstay of prefectural finances was the liuzhou portion of tax revenue. The liuzhou portion, both in terms of its income sources and the expenses for which it was used, was nearly of a fixed amount. From the first, it was managed on the premise that there would be an equilibrium between income and expenditure. Nevertheless, because of factors like diminished income as a result of fugitive households (taohu 逃戶) or excessive burdens of expenses for military action within areas under its jurisdiction, the prefectural financial administration often failed to maintain a balance of income and expenses. For example, the following is written in an imperial command dated the yihai day of the twelfth month of Changqing 1 (January 9, 822):

The costs associated with military actions were unexpected, contingent expenses, and the repeated eruption and suppression of bandit uprisings in various areas during the later part of the Tang dynasty imposed a significant burden on regional financial administration. When an imbalance in finances emerged because of these sorts of military expenses or unpaid taxes and levies, the shortfall was initially offset by internal contrivance within the liuzhou portion or through drawing on money for public use. For example, the following passage appears in the imperial decree mentioned above from the first year of Huichang (841):

Here, the narrative centers on the financial administration of circuits, but the fact that it also includes prefectural financial administration is clear from the reference to “miscellaneous monies of that prefecture or province.” The unpaid tax of fugitive households was made up for on the balance sheets by means of makeshift contrivances in the portion of miscellaneous expenses within the liuzhou, liushi and other regional finances. As I have explained above, the money for public use that derived its revenue from remaining surplus cash and goods was used to supplement and make good the unpaid taxes of the general peasantry. There was, in addition, also support from prefectural financial administration that came through diverting grain from the ever-normal and charity granaries. An imperial order from the seventeenth day of the first month of Dazhong 1 (847) is a good example. There, the proposal was made to encourage now-fallow paddy fields in various prefectures to be returned to cultivation and their harvest applied to military provisions needed locally and to grain storage for natural disasters. In addition, it was directed that seeds, foodstuffs, and tools for production be covered by diverting grains from the ever-normal and charity granaries.Footnote ⁴⁷ In short, alongside ensuring sustainability by storing up cash and grains in case of natural disaster, one of the functions of money for public use and grains in the ever-normal and charity granaries was to provide backup for the liuzhou section of tax revenue that constituted the core of regional finance administration, thereby helping to maintain its balance.

However, there appear to have been limits to maintaining balance within regional financial administration in this way. At the time, in order to make actions related to financial affairs run smoothly in the three offices of the central government's Finance Commission (sansi 三司)—the Department of Revenue (duzhi), the Salt and Iron Monopoly Bureau (yantie), and the Census Department (hubu)—property placed on government ledgers and called “cash or goods belonging to the government” (shusheng (xisheng) qianwu 屬省(係省)錢物) was retained in the regional areas. There were occasionally times when a circuit or prefecture could borrow the cash or goods that belonged to the government.Footnote ⁴⁸ For example, an imperial command from the yihai day of the first month of Kaicheng 3 (838) contains the following:

In addition, there is the following in the “Amnesty to commemorate the awarding of an honorific title” from the third day of the first month of Taizhong 2 (848):

This is a case in which deficiencies resulting from regional military expenses were made up for by borrowing from the cash and commodities belonging to the Finance Commission's three offices; yet while such adjustments were made to maintain equilibrium, ultimately the debts became irrecoverable loans and repayment was in the end excused by means of an imperial edict.

As discussed above, regional financial administration at the prefectural and other levels was fundamentally premised on a balanced equilibrium of income and expenditures through the principle of fixed amounts. However, as a result of factors such as excessively burdensome military expenses and stagnant collection of tax payments, it was not uncommon for income and expenditures to become imbalanced. In these situations, steps were taken in order to achieve balance within the regional financial administration by a) supplementing from the miscellaneous expenses portion of the liuzhou section; b) supplementing from the money for public use portion; and c) diverting from the ever-normal and charity granaries. In cases where even these measures proved insufficient, d) balance adjustments were effected by borrowing cash or goods belonging to the government; this was property left regionally but which belonged to the three offices of the Finance Commission: the Department of Revenue, the Salt and Iron Monopoly Bureau, and the Department of Census. Going by the principle of fixed amounts in the financial administration frame, on first glance it seems that even financial administration during the latter part of the Tang dynasty had a modern budget function where plans were made to balance income and expenditures. Yet when we consider the actual circumstances at the level of regional financial administration, we see that the situation was more precarious: a constant struggle to get by from one day to the next.

The Centralization of Regional Financial Administration during the Tang–Song Upheaval

The previous section has provided us with a general overview of the actual circumstances of regional financial administration in the latter half of the Tang Dynasty. The problem that arises is its place in history. Speaking broadly, this means asking what historical stage the regional financial administration described above occupies in terms of Chinese history as a whole. Moreover, the time period that is at issue is known as the Tang–Song transition: one of the most prominent periods of transformation in Chinese history. This question will also lead us to inquiring into the significance of this period of transformation from the perspective of the history of financial administration. Our discussion must begin by returning to the Han Dynasty, when the framework for financial administration history in the ancient state of China took shape.

As I made clear in the first chapter of my book Chūgoku kodai no zaisei to kokka (Financial Administration and State in Ancient China),Footnote ⁶⁴ during the period when the ancient state came into being, the special feature of financial administration history was that the peasants’ surplus was gathered by regional prefectures and counties in various forms such as “field tax” (tianzu 田租) and “tax in commutation” (gengfu 更賦; the three types of tax known as gengqian 更錢, suanqian 算錢, and kouqian 口錢), which were in turn accumulated in various commanderies and counties mainly in the form of “stored grain” (weiji 委積). The financial administration of the central government's Chamberlain of the National Treasury (Dasinong 大司農) was composed of redistribution from this regional accumulation. With a scale of between four and six billion in cash, the National Treasury's finances were comparable to no more than roughly 5 percent of total surplus labor in the Han dynasty. Although the Chamberlain of the National Treasury, which was the central government's office of financial administration, had established some authority over this regional accumulation through the reports on financial matters (which were called “Submitted reports” shangji 上計) that the commanderies and princedoms (jun and guo) submitted at the beginning of the year as well as the reports on current financial holdings they produced four times annually, it had still not completely achieved authority over financial administration. To state my conclusion in advance, the penetration of the central government's authority to direct this regional accumulation and have control over regional financial administration still lay in the future. In other words, the establishment of centralized authority over regional finances would continue to be groped toward gradually, a target sought for constantly through all of the transformations of ancient Chinese dynasties. In this way, it was none other than the transition from the latter half of the Tang dynasty to the beginning of the Northern Song that the process of establishing the centralization of authority over financial administration entered its final stage. In what follows, I will take the Northern Wei financial administration (which served as the original model for Tang dynasty financial administration) as a fixed data point and clarify the historical special circumstances of financial administration in the later part of the Tang dynasty.

As I explained in “The Structure of Financial Administration in the Northern Wei,”Footnote ⁶⁵ the fundamental tax collection zudiao 租調 in the Northern Wei was divided into various categories: “official levies” gongdiao 公調 (or “regular levies”; changdiao 常調), which constituted the central financial administration, and “external expenses” (diaowaizhifei 調外之費), which constituted the regional financial administration, as well as “irregular levies” (zadiao 雜調), which constituted occasional expenses by the central government. In Taihe 8 (484), at the time of the imposition of the stipend system, the following record appears in the Wei shu (History of Wei).

With regard to the financial administration ratio among gongdiao and diaowai expenses and the bureaucrats’ stipends, the Tongdian quotes the “Wei ordinances” (Weiling) as follows:

Excluding bureaucrats’ stipends, the ratio of central financial administration to local financial administration was five to two. In comparison to the Han dynasty, the percentage that the central government's financial administration occupies has become overwhelmingly higher. Not only that, but it is also worth bearing in mind that the very amount of regional financial administration is clearly stipulated here, which bespeaks an advance by the central government into local financial administration: an increase in its authority to direct financial administration.

In addition to the right to direct regional financial administration, what also deserves attention is the fact that, with the system of “three-grades and nine-ranks of tax” imposed during the reign of emperor Xianwen in the latter half of the fifth century, a system had already been imposed upon peasant and other households in which tax grain was directly transported to the grain storehouses of the capital, to the grain storehouses of important prefectures, and to those of the household's own prefecture. This means that all property could, at the order of the state, be transported from the region of its production directly to the region in which it was needed. In other words, this means that a network for the exchange of goods was opened across the entire national territory. This was organized in order to secure the financial administrative needs of the central government, as well as the military requirements of the borderland districts in the northwest and those southeast of the Jianghuai.

Before a nationwide system of commodity circulation was completed, the Northern Wei's fiscal distribution came to a halt due to an uprising near the end of that period that originated in the 523 Revolt of the Six Garrisons. What brought the system to completion was the fiscal management under the statues and ordinances (lüling 律令) system in the Sui and early Tang that had reunified the realm. With that, a system of commodity distribution with national scope was organized with the Ministry of Revenue's Bureau of General Accounts (Duzhisi 度支司) as the command center. All property could now be directly transported from the region of its production to the region where it was required and such transport occurred in three patterns: 1) within the same prefecture; 2) to the capital (sending to the central government); 3) external distribution (sending to other prefectures or to the borderland). The local expenses (when sent to the same prefecture) were limited to the stipends of the local officials and communication costs as well as insurance expenses (charity granary).

The framework of local financial administration and the organization of a national scale fiscal distribution system that began in the Northern Wei and was established in the Sui and early Tang were the forerunners of the late Tang period regulation of the amounts of liuzhou (tax for the prefecture) and liushi (tax for the provincial governor) and the amount of shanggong (tax forwarded to the central government). In terms of the formal ratio between the local areas and the central government, if we compare cases where the content is known, then during the Tianbao period of the Mid-Tang (742–755), the ratio was 4:5, or roughly equally split, between finances for the central government (4) vs. finances for the military and the prefecture itself (5); in the late Tang dynasty, the tax for the central government was half of the tax for local government. In comparison to the case of the Northern Wei, when the ratio of gongdiao (taxes forwarded to the central government) to diaowai (taxes for local expenses) was 5:2, these figures show that the proportion of local financial administration became gradually larger. However, we cannot regard the enlargement of the proportion of local financial administration as simply a matter of the decline of the central government's authority to direct fiscal matters.

The fundamental reason for the enlargement in the proportion of the local financial administration in the late Tang dynasty is due first of all to the establishment of a separation between military personnel and farmers during the Kaiyuan period (713–741), and secondly that the laborers who transported property switched from being peasant adult males (zhengding 正丁) to being soldiers and merchants; as a result of that, the amount of local financial administration that was taken up by expenses associated with the costs of supporting soldiers and military actions on the frontier increased dramatically.Footnote ⁶⁸ What is even more important is that after the An Lushan rebellion, because of the dissolution of the empire, in various areas within the internal territory, fanzhen (areas controlled by regional military governors) grew segregated and it came to be the practice that expenses for supporting soldiers and military activity were held in reserve separately. The central government's authority to direct the fiscal matters of these areas controlled by regional military governors was restricted.

It is true that there were differences in the ratio of financial administration between the central government and localities over the course of the Northern Wei through the latter part of the Tang Dynasty. But if we compare the situation to the fiscal management of the Han dynasty where only the amount of redistribution from local areas to the central government was fixed, the fact that a fixed frame was imposed upon local financial administration indicates that the state's fiscal authority had achieved a certain advance. In this way, premised upon a new social structure in which soldiers and peasants were separated, the central government's fiscal authority again began to penetrate even into the framework of local financial administration and a more centralized form of fiscal management became established; it is this process that marked the shift from the late Tang dynasty to the early Song dynasty. Below I will explain the course of events.

In the late Tang dynasty, local financial administration was composed of three parts: 1) the tax revenue retained by the prefecture, or liuzhou portion; 2) the “money for public use” (gongyongqian) portion; 3) the ever-normal and charity granaries portion. Of these, as the research of Funakoshi Taiji has made clear, the ever-normal and charity granaries portion collapsed at the end of the Tang dynasty, and while there was a plan to revive it during the Five Dynasties, this effort ultimately did not succeed.Footnote ⁶⁹ The problem remains with the liuzhou portion and the “money for public use” portion. Let us first of all examine the “money for public use” portion, which had comparatively more independence.

The establishment of “money for public use” goes back to the Kaihuang era of the Sui Dynasty. In regard to an incident from Kaihuang 8 (588), the Sui shu (History of Sui) narrates as follows:

The lending of publicly owned property by the local government offices to private individuals to obtain profit that could be used for administrative expenses is a practice that can already be confirmed in the latter Han.Footnote ⁷¹ And in the Northern Wei, prior to the imposition of the stipend system in the eighth year of Taihe (484), there were occasions when regional expenses corresponding to stipends were raised by allowing merchants to manage publicly owned property.Footnote ⁷² Managing public property in order to raise funds to meet the cost of expenses is something that only rarely comes to the surface of historical records, but it had a long tradition. The practice of using the “public office money” as principal and applying the profit from trade or from high-interest-rate lending to prefectural financial administration was something that was a successor to this tradition. At the latest, it can be traced back not long before this Kaihuang period (581–600). The origins of the other revenue sources that made up the money for public use are at present unclear, but it seems that cash capital in the form of “public office money” was the original form from the early stages. Public office money and other forms of cash capital used for profit-making underwent various transformations with the arrival of the Tang Dynasty, such as becoming the revenue source for the stipends of provincial officials, but in the latter part of the period it came to be an important revenue source of the money for public use.Footnote ⁷³

The first regulations on profit-making cash capital came in the eleventh month of Yuanhe 9 (814), when the Census Department's transaction tax of five wen was extended and applied to this profit-making cash capital.Footnote ⁷⁴ An imperial command issued in the twelfth month of this year contains the following:

These regulations on the management of profit-making cash capital were imposed on the various offices of the central government, but we can assume that they were also applied to the management of cash capital by the various offices of prefectures and other local government entities. This is because eventually it would lead to the regulations that were, as discussed above, imposed on money for public use that derived from the remaining surplus in Dahe 4 (830). As we have already seen, the proper composition of expenses for money for public use at this time was designated as follows: 1) the construction and repair of government offices and supplies; 2) entertainment expenses associated with the exchange of public and private envoys and bureaucrats; 3) regional police operations; 4) supplementation of unpaid tax duties; and 5) stockpiling for use in natural disaster. We can conclude that this signified an expansion of the content of regulation from 814. The regulations on “money for public use” in Dahe 4 was carried out in response to the instructions of the imperial command from the previous year of Dahe 3 (829). The imperial command gave as a reason for its imposition the fact that it was common for prefects to be brought up on charges of misappropriation of funds because there were not clear regulations to be followed.Footnote ⁷⁶ Conversely, this means that up until that point, the matter had been left to the free discretion of the prefects. Since we do not encounter any historical material concerning regulations on money for public use after this point, it seems that the regulations on money for public use continued through the end of the Tang.

Let us turn to the liuzhou section. The liuzhou section was a fixed amount and the contents were almost completely standardized. However, there were occasionally inordinate expenditures by the prefect and others. So far as I have been able to determine, the earliest regulations on these came with an imperial decree dated the second year of Yuanhe (807):

This order became even more concrete with a memorial from the Bureau of Review dated the sixth month of the first year of Changqing (821):

The regulations imposed by the central government upon the liuzhou portion at this stage were a strengthening of oversight concerning prefectural financial administration by the Bureau of Review. They severely cracked down on excessive expense.

This system, upon reaching the Later Tang in the Five Dynasties period, underwent a fundamental change. It was the consolidation of the three offices of the central government's Finance Commission—the Department of Revenue (duzhi), Salt and Iron Monopoly Bureau (yantie), and Census Department (hubu)—under the tax officer (zuyongshi 租庸使) in the first month of Emperor Zhuangzong's reign period Tongguang 2 (924).Footnote ⁷⁹ The tax officer was a position established after the Later Liang responsible for financial administration. It was the forerunner of the later three offices of the Finance Commission. However, around 924, the tax officer not only oversaw the financial affairs of the three offices of the central government, but also began to intervene in local financial administration in ways fundamentally different than before. The content of this transformation is recorded as follows in an entry for 924 in the Zizhi tongjian (“Comprehensive Mirror in Aid of Governance”):

Contradicting the practices up until that point, the tax officer went over the heads of the military commissioner and the surveillance commissioner to give direct instructions about the requisition of property from the various prefectures. In the imperial command was the phrase “What the tax officer has declared is the recent practice of the bogus court,” from which we can see that from the Later Liang, this sort of direct participation in prefectural financial administration began to spread. In the Later Tang, as the tax officers effected the unification of the financial affairs of the three offices of the central government's Finance Commission, they bypassed the military commissioners and the surveillance commissioners; by unifying under their direct control the prefectural financial administration, they attempted to achieve an even greater concentration of their authority over financial affairs. That the military commissioners and surveillance commissioners strongly resisted this move is clear from the Zizhi tongjian account.

What deserves attention is that not only was there concentration of authority to direct financial administration, but that the tax officer placed “public office money,” which constituted an independent section of prefectural financial administration, under its direct jurisdiction. An imperial command dated the fourth month of Tiancheng 1 (926) eliminated the office of tax officer; the following passage appears in the text:

According to this, the Tax Officer clearly tried to place not only the prefecture's “public office money” but that of the Military Commissioner and the Surveillance Commissioner under his direct control. The unification of the three offices of the central government's Finance Commission under the Tax Officer, the direct participation in the financial administration of the prefecture, as well as direct management over “public office money” that was a step in that direction, were all measures undertaken by the central government financial administration offices with an eye toward establishing a more concentrated administration of the financial affairs of the entire country. However, in part because of the political circumstances of the time, this plan foundered after just over two years.Footnote ⁸² Nevertheless, at this time, the unification of the central government's financial administration under a single high minister in the form of the Supervisor of the Three Fiscal Agencies (pansansi 判三司) was realized, and eventually this led to the establishment of the State Finance Commissioner (sansishi 三司使) in the eighth month of Changqing 1 (930).Footnote ⁸³ The establishment of the Supervisor of the Three Fiscal Agencies and the State Finance Commissioner was a step backward from the Tax Officer, but the establishment of unified management of financial affairs by the central government's Finance Commission came to be a bridgehead for the spread of leadership authority of the State Finance Commissioner into local financial administration.

The Later Tang plan for the tax officer to establish concentrated financial administration was achieved in two stages in the course of the unification of national territory in the Northern Song. The first thing to be seized upon was the liuzhou cash and commodities portion. That happened in the second year of Qiande (964). An account in the Xu zizhi tongjian changbian (Continued Comprehensive Mirror in Aid of Governance Long Edition) reads as follows.

With this was planned a step toward transformation of liuzhou cash and goods into cash and goods belonging to the government (xisheng qianwu 係省錢物), for aside from regular expenses in the prefecture, all would be sent to the capital. In other words, this was a thorough narrowing of the scope of free discretion that the prefectures had enjoyed; the “remaining surplus” (huican xianyu), the portion that up until now had served as the revenue source for the “money for public use” section, would now be transferred to the central government, transformed into cash and goods belonging to the government. As a result, it came to be that only the bare necessary minimum of regular expenses whose use had already been decided, such as the salaries of bureaucrats, was retained in the local area. In this way, by means of the transformation of cash and goods for public use into cash and goods belonging to the government, in Kaibao 6 (973), prefectural financial administration was ultimately all put under the direct control of the central government's financial administration and the concentrated authority over financial administration by the State Finance Commissioner was established.

This account is one that can easily give rise to misunderstanding. The last sentence about “liuzhou cash and goods” records an evaluation of the complete transformation of cash and goods for public use into items belonging to the government; it is not a statement that this in fact happened. It is clear on the basis of the above analysis, that the money for public use portion and the liuzhou portion constituted separate lines of prefectural financial affairs. Therefore, it is not the case that the placement of cash and goods for public use onto government ledgers meant the complete transformation of the liuzhou portion into cash and goods belonging to the government. We should observe here that as a result of the policies instituted after Qiande 2 (964), the money for public use over which the prefect held independent discretion as well as the cash and goods from “remaining surplus” of the liuzhou portion that formed its revenue source were placed onto government ledgers; as a result of that, the financial administration of the prefecture was unified under the direction of financial affairs by the central government. The leftover liuzhou portion was the bare minimum for local expenses, and was for expenses in which there was no flexibility. Thus, at this point in time, the financial administration authority of the prefect was ultimately concentrated in the hands of the central government.

Of the three sections comprising the local financial administration in the latter half of the Tang Dynasty, the ever-normal and charity granaries were re-established in the first year of the Yuanhe period (806). The liuzhou section was established along with the imposition of the twice-a-year tax law in the first year of Jianzhong (780) and its regulations began in Yuanhe 2 (807). The money for the public use section was established around the Kaihuang era of the Sui Dynasty (581–600), but along with the implementation of “profit-making cash capital” that had as its revenue source the Census Department’s transaction tax of five wen in Yuanhe 9 (814), there came to be a certain framework that was imposed upon its use. Looked at in this way, we can see that the penetration of the central government's authority over local financial administration in the late Tang Dynasty began in the Yuanhe period. In this way, the process of concentrating the central government's fiscal authority over local financial administration that began in the Yuanhe period came to a turning point with the hasty attempt of installing a tax officer in the Later Tang period of the Five Dynasties. Then, the unification of central financial administration under the State Finance Commissioner was first achieved, and ultimately in the Northern Song, the central control of the liuzhou section, i.e., its transformation to cash and goods belonging to the government was begun; this process was then completed with the transfer of money for public use to government ledgers in Kaibao 6 (973). What were the origins of this?

On what grounds can we explain the process of concentration of the central government's authority to direct financial administration? The matter concerns the state as a whole and clearly is the result of a complex mix of political and economic factors. However, what most deserves our attention here are the military factors. Among the financial administration problems that existed from the latter half of the Tang Dynasty through the Northern Song, the most significant was the need to make expenditures for the cost of the perpetual support of soldiers, who had reached a million in number after the establishment of the division of soldiers and peasants. According to “A scheme presented to the army” (Gongjuntu 供軍圖) submitted in the first month of Kaicheng 2 (837) by Deputy Minister of Revenue Serving General Accounts Wang Yanwei, out of the roughly one million soldiers, excluding the 600,000 soldiers who were compensated by liuzhou and liushi cash and goods, the expenses for supporting the remaining 400,000 was the responsibility of the central government's Department of Revenue.Footnote ⁸⁷ During this time, the authority to direct military affairs across the entire country was scattered, as the following quotation shows:

In spite of this, 40 percent of the fees for maintaining these soldiers came from the central expenses. It is clear that if the authority to direct military matters lacks the backing of financial administration, then it is essentially weak. We can see that at least in a latent sense, from the fundamentals of financial administration, the concentration of military authority to the central government had already begun. In this way, from the Five Dynasties into the beginning of the Song, the authority to direct military affairs that had been scattered came to be unified in the form of the imperial armyFootnote ⁸⁹ and the central government's control over the costs of boarding soldiers and other military expenses advanced, so that the authority to direct financial administration came to be concentrated under the central government. Or, rather we should say that the central government's assumption of authority to direct the military was in a mutually influencing relationship with its assumption of authority to direct financial administration: both proceeded as cause and effect. It is military factors that offer the most fundamental explanation of the roots of the process of centralization of the central government’ s authority over financial administration that began in the Yuanhe period.