Post-Mao Central-Local Fiscal Relations: Local debt and local agency 改革开放后的中央地方财政关系:探索地方债务和地方代理 by Karl Yan

CORN March 2015 Edition.

Post-Mao Central-Local Fiscal Relations: Local debt and local agency


Image from


Fiscal decentralization and tax-sharing has had a profound impact on local government in China. As a result of the 1994 Tax-sharing system, local governments, municipalities and counties have become less self-sufficient because their income levels can no longer meet their expenditure levels. Trying to deal with this situation then, local governments have exercised agency and sought outside the budget and non-budgetary means to make ends meet.



Local governments in China nowadays spend far more than they take in as revenue. In many localities, deficit problems are dealt with mainly through non-budgetary measures and outside-the-budget sources. Among the many problems with these approaches is the lack of oversight and supervision over the sources and uses of these funds. As these tools fall outside the parameters of state supervision and management, local governments have gradually developed a sense of collective agency in utilizing these tools to meet their increasing expenditure. How were these deficit problems and local institutions developed, and how were local agencies operationalized are the two fundamental puzzles this analysis seeks to explore.

In an era when the entire Chinese party-state seems driven by GDP growth, local governments find themselves adopting this a priori assumption and as such are working together to promote such growth. Degrees of and decisions to use collective agency are based upon the careful calculations of financial and economic gains. Local agency here refers to the idea that local governments are relatively autonomous actors that can make their own decisions and follow through their decisions—either within or outside of the state structure.


Rise of local states—financial self-sufficiency (1979-1994)

To trace the rise of local governments in China, we need to go back to the year 1978, when the central government started to experiment with fiscal decentralization. As a result, from 1978 to 1993, local governments were given much more freedom to experiment with economic and political reforms, which included the central government devolving fiscal responsibilities to local governments as well as giving up much of its revenue to local governments (放权让利). These reforms were designed to allow local governments to work more effectively in delivering economic growth and public services.[1]

The decision to decentralize was essentially a forced one because of the central government’s inability to sustain a fiscal balance due to large deficits in 1978 and 1979. The hope was that the increase in revenue at the local level would increase the central government’s income as well. This is because if local government could achieve sustained economic growth, the central government’s revenue would also increase through different revenue-transfer schemes negotiated with local governments. The mindset was that if each piece (e.g. at the local level) of the pie became larger, the pie itself (e.g. at the centre) would be enlarged as well.

But the result of fiscal decentralization over time has significantly weakened the central government’s ability to extract revenue from the local level because local authorities now gained the right to keep revenues to themselves—and didn’t want to pass it to the centre afterward. In this respect then, local governments in the pre-1994 era were much more powerful compared to the central government as the former had direct control of their economic policies and were direct tax collectors. Further, they could also keep much of the tax revenue through strategic bargaining with the central government.

According to Wang Shougang, a scholar well-versed in central-local government relations, there were four strategies which local governments used to negotiate with the centre:[2]

  • Negotiating a better arrangement within the revenue-sharing system. This strategy has often been successful because fiscal decentralization strengthened the bargaining position of local governments as they have control of revenue—precisely what the central government is trying to get its hands on through
  • Taking a non-cooperative attitude towards central policies. Local governments generally see extractive fiscal policies from the central government as detrimental to local interests.
  • Withholding as many funds as possible (藏富于企业). Local government do this through a variety of means, such as tax exemptions, tax breaks, lower tax rates, and granting tax reliefs to local firms.
  • Imposing ad-hoc charges on local enterprises. This strategy is used by local governments to compensate for lost revenues to the central government.

These four strategies have severely limited the central government’s ability to extract revenue from the local levels, and thus have greatly prevented the central government from more easily balancing its national budget. Resource-rich municipalities like Shanghai and Tianjin have more leverage and hence more bargaining power and can negotiate a better arrangement in the revenue-sharing system (Strategy 1). Likewise, they are able to keep more resources in their own hands in order to stimulate their own local economies. Municipalities of provincial status such as Shanghai enjoy much more spending power; for example, the city’s budgetary expenditure increased by 300% from 1983 to 1986.[3]

Further, during this period of time, as local governments came to shoulder infrastructure-building and service provision, these local governments also needed to develop comprehensive bureaucratic structures to deal with urban expansion and rising expenditures. From this period then, local governments gained the ability to make use of funds as they saw fit, and as a result became quite autonomous actors from their principal—the central government.[4]

The initial motive to push for fiscal decentralization was to solve the deficit problem that the central government faced in the late 1970s. However, the policy did not help to solve the problem—it actually made it worse. By the early 1990s, apart from the increase in the size of centre’s deficit since 1979 (see Graph 1), the central government has failed in perhaps its most important duties: it can no longer play an important role in the distribution, re-distribution and stabilization of national resources. In fact, local governments proved themselves much more able to deliver both economic growth and welfare services. This directly resulted in the central government’s issuing of billions in government bonds from 1980 to 1990 in attempt to finance its fiscal deficit.[5]




Local Finance since 1994—Results and Consequences for Local Governments

In light of the dire situation mentioned above, the central government introduced the “Tax-sharing system” (分税制) in 1994. The new tax policy was aimed to strengthen the financial capacity of the central government and better macro-manage the entire country’s economy. In the 1994 financial system reform, the central government clearly outlined the responsibilities between central and local governments in terms of expenditures and revenues (Table 1).

Table 1: Central and Local Expenditures[7]

Central Financial Expenditure

Local (Municipal) Financial Expenditure

·         National security·         Armed police·         Diplomacy and foreign aid

·         Management cost of central government and bureaucracy

·         Basic infrastructure cost under the management of the central government

·         Centrally managed enterprises and new product testing fee

·         Geological surveys

·         Agricultural subsidies made by the Central Finance Ministry

·         Repayment of national and foreign debt

·         Centrally administered public security, supervision and inspection spending

·         Centrally administrated cultural spending

·         Centrally administrated education spending

·         Centrally administrated public health spending

·         Centrally administrated scientific spending

·         Local administration

·         Locally administrated public security, supervision and inspection spending

·         Parts of armed police spending

·         Militia

·         Locally managed infrastructure projects

·         Locally managed firms and new product testing spending

·         Agricultural subsidies

·         City maintenance

·         Locally administrated education spending

·         Locally administrated cultural spending

·         Locally administrated public health spending

·         Price difference (subsidies given to producers so that produces can be sold at a lower price, often set by the local government)

·         Other spending


In terms of revenue, new national tax bureaus were created whose tasks included collecting taxes from centrally administered corporations and agencies. The national taxation system (国税系统) is capable of reaching to the lowest level of formal government in China—townships (镇). Such a comprehensive system allows the centre to bypass provinces (省) and municipalities (市) in tax collection.

Taxes that are directly related to economic growth are then shared by the two levels of governments, while taxes that are suitable for local collection are collected and kept by the local government (see tables 2 and 3 for details). This new tax-sharing system was aimed at solving the central government’s inability to extract enough revenue from the local government in order to balance the former’s budget. It would also mean a shift in the balance of power between central and local governments. After the introduction of the new tax system, the central government’s income steadily increased—and local governments’ income steadily decreased. Revenue-sharing proved to be the classic zero-sum game since the money could only either stay or leave local government coffers; if one level of government’s income has increased in proportion to the total income, the other level of government’s total income must decrease in proportion to the total income.

And although total income and expenditure have risen sharply since 1980, GDP growth has also increased since. The annual budgetary revenue has averaged about 17% of total GDP, peaking in 1980 at nearly 26% and at its lowest in 1995 at just over 10%. Likewise, budgetary expenditure has average nearly 19% of total GDP, hitting a high point at 27% in 1980 and a low point in the years 1995 and 1996 at just over 11% (see graph 2). Therefore, when budgetary revenue and expenditure increase, they are always in proportion to the overall GDP.



Looking at local government’s income in 1992 (2503.86 x 100 billion RMB) and comparing it to that of 1994 (2311.6 x 100 billion RMB), local government total revenue in 1994 is even lower than that of 1992. The transition from decentralization (放权让利) to the new tax-sharing system has clearly straightjacketed local governments’ ability to obtain monetary resources to fund its enlarged spending. To be specific, in 1993, local governments’ financial income was 78% of total financial income, leaving 22% to the central government. However, in 2005, local governments’ financial income decreased to 47.7% of total financial income; by contrast, central government’s financial income was 52.3%.

All of this means that the central government has been undeniably successful in extracting revenue under the new tax-sharing system. Although the new tax-sharing system helped the central government to balance its books, local governments found themselves in deficit because of a low self-sufficiency rate (at 60.5% in 2005) caused by a lower revenue “share” in the new tax-sharing system.[8] Simply put, local government revenues no longer kept up with the amount local governments were spending.

The financial situation at the local level has been problematic since the introduction of the tax-sharing system in 1994. As a result of chronic deficits then, if local governments wanted to keep spending the same amount, they had no other choice but to seek other ways to increase revenue and make up for the missing 40%.

By contrast, the central government’s revenue has been gradually increasing since 1994 (see graph 3). Before then, central government revenue only averaged about 30% of total financial revenue. However, after the tax-sharing system was implemented, central government revenue accounted for about 50% of total revenue. Further, though local expenditure remained around 70% of total expenditure, not only did local government income stagnate but actually decreased over time. In the context of expenditure, central government’s expenditure counted for 39.7% of overall expenditure in 1985 and has been steadily declining. It has hovered around 30% since 1989. This means that, although revenue proportion has been increasing, the central government’s level of spending has not increased, as it always remained around 30% in proportion of total expenditure. Therefore, though the tax-sharing system introduced in 1994 solved the central government’s fiscal problems, it created a new problem for local governments.




Local Agency—dealing with local self-insufficiency

Because local government revenues could not meet their expenditures, these governments needed to find new ways to solve the issue of fiscal self-insufficiency. However, institutional measures within the state parameters were unavailable as the 1994 tax-sharing system clearly defined the responsibilities of tax collection for both central and local governments. This meant that local governments could not add new taxes or give tax breaks, which were two traditional strategies used during the 1980s to increase budgetary revenue. Currently, local government still collects many different kinds of taxes, such as personal income tax, land tax, and city maintenance tax (see table 2 for a full list).


Table 2: Central and Local Tax Responsibilities[9]

Central Taxation

Local Taxation Shared Taxation
1.      Duties2.      Consumer tax and value added tax collected by the customs3.      Consumer tax

4.      Centrally administrated corporations’  corporate tax

5.      Central banks, foreign banks and other financial institutions’ Financial tax

6.      Railway bureaus, Central branches of each of the banks, security and insurance companies’ Income tax, operations tax, revenue and maintenance tax

7.      Revenue from centrally administrated corporations

8.      Foreign trade firms’ export tax return

1.      Business tax (excluding those listed under Central Taxation)2.      Locally administrated corporations’ corporate tax (excluding those listed under Central Taxation)3.      Revenues from locally administrated corporations

4.      Personal income tax

5.      Land tax

6.      City maintenance tax

7.      Investment tax

8.      Property tax

9.      Vehicle and ship tax

10.  Stamping tax

11.  Slaughter tax

12.  Agricultural livestock tax

13.  Special agricultural product tax

14.  Farmland usage tax

15.  Contract tax

16.  Estate tax

17.  Land value-added tax

18.  National land usage tax

1.      Value-added tax (central 75%, local 25%)

2.      Resources tax (Sea and Oil belong to the central government, other types of resources belong to local governments)

3.      Stock Exchange tax (central 50%, local 50%)


However, after the initiation of the 1994 Tax-sharing system, local tax bureaus were created to meet local taxation needs. Further, as social services expanded, new bureaucracies were created to take on these tasks—which meant more expenditure, too. Though the tax-sharing system limited local governments’ financial gains, it bolstered local governments’ capacity to provide services and collect taxes. Therefore, local governments sought non-institutional ways to extract more revenue to sustain their increasing expenditures.

By exercising agency then, local governments often discovered two ways of increasing their budgetary revenue: 1) income outside the financial budget management and 2) non-budgetary income.[10] This type of strategic decision and implementation was operationalized under a rationalist assumption—prioritizing economic gains, which would lead to cadre promotion. Operationalization of these methods was only possible because of the newly bolstered local bureaucracies. Further, with the incentive of promotion, in which economic indicators account for almost one-third of the entire cadre evaluation, unsurprisingly then local leaders from across horizontal and vertical units have been more likely to work together to implement such measures to achieve economic growth in the locality.

Going outside the Budget—How Local Governments pump up revenues

In 1994, in Qufu, Shandong province, 3000 “household registrations”[11] (户口) were sold to generate additional revenue—that is, non-budgetary—to help keep pace with local government expenditures. Each household registration was priced at 3000 RMB (about US$480). In 1994, local governments in China earned a whopping 25 billion RMB (US$3.9 billion) by selling hukou alone.[12]

These non-budgetary incomes are important to consider because they have become main sources of revenue for local governments. And it would be impossible for registration transactions to take place because a single transaction requires cooperation from a number of different local bureaus. Moreover, local governments have been adding new fees for locally administered programs. From the years 2000 to 2003, more than 4,000 (and as many as 10,000) newly created local fees were created to raise revenue.

In a new wave of urban development and modernization, local governments have also used land development as a new way of securing extra income for their localities. Rents, the push of the real estate market, and expansion of land development were some of the methods used by local governments to finance expenditures. Local agency can be exercised through the sale of land as local government has the ultimate political authority over land usage and land rights. Currently, in China, rural farmland is collectively owned and urban properties rights are leased to individuals for 70 years. Therefore, because of a monopolization over land rights and land transfer, many local governments have been ruthless in expropriating farmland and urban real estate from peasants and urban-dwellers respectively, and reselling them at higher prices to land developers for either industrial or commercial purposes.

Further, multiple bureaucratic levels and units—both vertical and horizontal—have worked together to expropriate land, sometimes using force. This “strategic implementation in land expropriation” was often exercised under limited supervision from the central government. Indeed, the tax sharing system shows a division of state power that local governance is solely vested upon local governments.

From 2001-2003, local governments’ income from land development was 910 billion RMB (US$145 billion). This represents a sharp increase from a mere 6.7 billion RMB (US$1.07 billion) in 1998. Local governments’ fees on land development (from construction to sale of real estate) were 30%-40% of the total real estate price. Adding on the 20%-40% land usage tax out of the total real estate price, local governments therefore earn 50%-80% from the total real estate price. This meant that, if the original price of a house was at 1000 RMB (US$160) per square metre, from the sale of this real estate local government earns 500-800 RMB (US$80 to US$128). Therefore, land developers have no choice but to raise the original price in order for the developers to profit. The price difference, as well as below-market compensation, were some of the major grievances causing state-society tensions in contemporary China.


Thus, we can see that the institutions and bureaus that local governments were able to make use of were products of decentralization and the 1994 Tax-sharing system. Further, the agency of local governments can also be traced back to the early reform era, when local governments were encouraged to act as entrepreneurs in finding new ways of balancing their books and achieving high economic growth. As local governments started to experience difficulties in maintaining high levels of growth just as their expenditures were increasing, they reverted to many of the institutions that were created as a result of decentralization and tax-sharing. Now, local governments are using these institutions and bureaucratic mechanisms to extract revenue from their own localities.




KarlYangKarl Yan is a PhD student in political science at the University of Toronto. His research focuses on contentious politics and social movements, local [government] agency and authoritarian politics in China. Karl is also working at DreamNovation, a non-profit organization dedicated to improving social equity in Canada and China.

He can be reached at karl.yan [at]



[1] 刘颖,“新中国成立后中央与地方关系的历史变迁研究述评”,《广西社会科学》,2012年第二期,第200卷,第122-123页.

[2] Wang, Shougang, “Central-Local Fiscal Politics in China” in Changing Central-Local Relations in China: Reform and State Capacity, edited by Jia Hao and Lin Zhimin, Boulder: Westview Press, 1994, pp. 95-97.

[3] Ibid, p. 104

[4] In a principal-agent relationship, the principal appoints an individual or group to act on the principal’s behalf. In theory, the agent acts on behalf of the principal and the interests between the two are well-aligned. However, it becomes a principal-agent problem when the agent becomes autonomous and make decisions that affect the principal.

[5] Wang, p. 106.

[6] China’s Surging Economy, pp. 209-210.

[7] 谢旭人,《中国财政改革30年》,北京:中国财经出版社,2008.

[8] Self-sufficiency here refers to the idea that, on budget, localities are able to match expenditures with revenue from the locality alone, without help or revenue transfers from non-local sources.

[9] 谢旭人,《中国财政改革30年》,北京:中国财经出版社,2008.

[10] In Chinese, this is known as “不纳入财政预算管理的预算外收入”以及“没有纳入预算外管理的非预算收入”.

[11] A “house registration” identifies a person as a resident of a certain area. Such registration grants the registered person all associated rights and welfare of the given area. Registration documents usually have the entire family’s information on them and each family is given one document.

[12] 陈曦,“分税制后中央与地方财政关系分析”,《宜春学院学报》,2012年第2期,第32卷,第32页.

Facebooktwittergoogle_plusredditpinterestlinkedinby feather

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *