Information about Asia and the Pacific Asia y el Pacífico

10. Fiscal Policy in China

Eswar Prasad, Steven Dunaway, and Jahangir Aziz
Published Date:
September 2006
  • ShareShare
Information about Asia and the Pacific Asia y el Pacífico
Show Summary Details

Steven Dunaway and Annalisa Fedelino

In recent years, fiscal policy in China has been prudent. Fiscal deficits have been lower than budgeted, because revenue overperformances relative to the budget targets were not fully spent and the stock of government debt has remained low and even declined in 2004 and 2005 (Figure 10.1). Fiscal policy has largely been guided by the government’s medium-term focus on fiscal consolidation aimed at making room for likely future expenditures on contingent liabilities, such as the banking sector’s large nonperforming loans, and a need for higher social spending as the population ages.

Figure 10.1.Fiscal Developments

Sources: China, Ministry of Finance; and authors’ calculations.

Fiscal policy has also been used, albeit less frequently, for short-term macroeconomic management. For example, as concerns about possible overheating and excessive investment in some sectors emerged in early 2003, fiscal policy was used to contain demand. In the future, as the authorities aim to direct growth away from investment, fiscal policy has an important role to play—for example, by reprioritizing spending in favor of programs that would boost household incomes, including spending on social safety nets and education. To make greater use of fiscal policy for demand management, broader coverage of fiscal accounts and a better assessment of the impact of budget operations on current macroeconomic conditions would be helpful, as well as improvements in cash management to better coordinate fiscal and monetary policies.

Overview of the Budget

The budget deficit has declined from a peak of 3.7 percent of GDP in 1999 to 1.3 percent of GDP in 2005. This consolidation was achieved through revenue increases of more than 5 percent of GDP during this period; strong growth in tax collections, in particular income taxes and value-added taxes, was driven by both buoyant economic activity and improvements in tax administration (Table 10.1). At the same time, spending grew by about 3.5 percent of GDP, mostly in the areas targeted by the government’s development policies. For example, capital expenditures expanded by about 3.5 percent of GDP over the period; expenditures on pension and social welfare programs increased by almost 2 percent of GDP, reflecting the transfer of some social expenditure mandates previously assigned to state-owned enterprises back to the government (Table 10.2)1

Table 10.1Government Revenue 1
(In percent of GDP)
Total revenue12.113.113.815.115.916.216.617.3
Tax revenue11.011.912.713.914.614.715.115.8
of which: taxes on income and profits1.
Value-added tax (VAT)on domestic goods4.
VAT and excises on imports0.
Nontax revenue1.
(As a share of total revenue)
Tax revenue90.791.092.092.492.091.290.791.3
of which: taxes on income and profits15.416.419.223.624.621.722.324.4
VAT on domestic goods35.533.133.332.432.233.033.933.7
VAT and excises on imports5.48.710.910.09.812.713.913.4
Nontax revenue9.
Source: China Ministry of Finance.

The coverage of this data includes the central government, provinces, municipalities, and counties.

Source: China Ministry of Finance.

The coverage of this data includes the central government, provinces, municipalities, and counties.

Table 10.2.Government Expenditure 1
(In percent of GDP)
Total expenditure and net lending14.916.817.017.918.918.618.118.5
Current expenditure11.712.512.713.714.514.414.414.9
of which: Administration and defense3.
Culture, education, public health, science2.
Pensions and social welfare relief0.
Capital expenditure2.
Unrecorded expenditure0.
Memo item: primary expenditure14.016.116.317.118.317.917.718.1
(As a share of total expenditure)
Current expenditure78.374.774.676.676.977.279.380.5
of which: Administration and defense23.522.623.424.625.425.926.426.0
Culture, education, public health, science16.916.716.917.918.318.618.519.1
Pensions and social welfare relief1.
Capital expenditure15.918.820.921.221.321.319.619.5
Unrecorded expenditure5.
Memo item: primary expenditure94.296.095.795.997.096.397.497.6
Source: China Ministry of Finance.

The coverage of this data includes the central government, provinces, municipalities, and counties.

Source: China Ministry of Finance.

The coverage of this data includes the central government, provinces, municipalities, and counties.

Fiscal consolidation over the past few years contrasts sharply with the performance in the years immediately following the Asian financial crisis. At that time, the Chinese authorities attempted to boost domestic demand by stimulating the economy; as a result, the budget deficit widened by about 1.3 percentage points in 1998 and by 1 percentage point of GDP in 1999—the largest consecutive deficit increases recorded in the past two decades.

By and large, both revenue and expenditures have deviated by significant amounts from budget targets (Figure 10.2). Deviations were more sizable for expenditures in the earlier years (with a peak of almost 12 percent, relative to the budget, in 1998), when the government focused on stimulating the economy. This trend reversed in later years, when revenue overperformances relative to the budget (reaching a sizable 12 percent of the budget in 2004) were larger.

Figure 10.2.Overshooting in Revenue and Expenditure

(Outturn less budget, as percent of budget)

Sources: China, Ministry of Finance; and authors’ calculations.

Assessing the Stance of Fiscal Policy

The Chinese authorities have described their post–Asian crisis fiscal policy as being “proactive.” In addition to significant and deliberate spending overruns, special bonds (referred to as “construction bonds”) were issued beginning in 1998. The proceeds of these bond issues were largely on-lent to local governments to be spent on capital projects.

In the 2005 budget, the authorities portray a shift in the fiscal policy stance from “proactive” to “neutral.” The targeted budget deficit was lowered in nominal terms to Y 300 billion—about Y 20 billion lower than the medium-term target announced in 2004—and the issuance of construction bonds was reduced to Y 80 billion (0.5 percent of GDP, about onethird the average annual issuance during 1998–2004).

However, the impact of fiscal policy on the economy was not only the result of changes in spending. Indeed, although spending exceeded budgeted amounts in every year since 1998, revenue was also much stronger. Thus, while the authorities were injecting stimulus from the spending side, they were also withdrawing it from the revenue side through higher taxes.

An assessment of the stance of China’s fiscal policy using the conventional analysis of fiscal impulses takes account of the changes in both spending and revenue. The fiscal impulse is a simple indicator of the impact of fiscal policy on aggregate demand. Its main purpose is to provide a measure of fiscal policy that is adjusted for changes induced by cyclical movements in economic activity.2 The measure can be used only as a guide, because a number of uncertainties surround the analysis, such as major structural changes and incomplete data; results are also sensitive to the choice of assumptions, particularly with respect to trend growth.3 Based on estimates of the fiscal impulse, fiscal policy was broadly contractionary during 2004 and 2005 (Figure 10.3). As expected, spending was expansionary in the earlier years, whereas revenue has systematically exerted a contractionary impact.

Figure 10.3.Fiscal Impulse

(In percent of GDP)

Source: Authors’ calculations.

Strengthening the government’s liquidity management is also important for improving macroeconomic policy operations. In China, government spending takes place unevenly during the year (Figure 10.4). Generally, it takes about 8 to 9 months to execute half of annual expenditures, and about 20 percent of recorded expenditures take place in December alone.

This leads to accumulation of fiscal surpluses during most of the year, with a sharp reversal in December. The associated large swings in government deposits significantly complicate the conduct of monetary policy over the course of a year.

Figure 10.4.Implications of Fiscal Policy Execution for Liquidity Management

Sources: China, Ministry of Finance; and authors’ calculations.

China’s New Fiscal Priorities and Their Impact on the Budget Balance

The Chinese authorities intend to shift spending away from capital projects to social programs. These policy intentions are reflected in a number of official statements and documents, including the 2005 budget speech by the minister of finance and the recently approved 11th Five-Year Plan. Because additional spending in health and education may require the construction of related infrastructure, this shift should not be seen as necessarily implying lower capital expenditures and higher current spending but rather as a move away from “construction of physical infrastructure to social infrastructure.”4

Although these policy intentions are sound, spending increases should be weighed against a number of uncertainties and rigidities. The following are some key considerations:

  • The budget classification is inadequate. Current classification of government expenditures—which is based more on functions and administration than on economic concepts—does not allow proper identification of programs or a clear-cut distinction between current and capital expenditures.
  • The budget classification is not consistent across the various levels of government. Differences in expenditure and revenue classifications undermine coordination of budget formulation and implementation and hamper the consolidation of various budgetary items—a necessary requirement for consistency across various government programs. This is more of a problem in the context of social programs because most of this spending is delivered at the local level. For example, 94 percent of total health spending and 97 percent of total education spending were undertaken by local governments in 2004.
  • There is insufficient monitoring and lack of accountability. Although the central government provides funding for most spending via transfers, thereby serving as an instrument to control and influence spending,5 evidence suggests that monitoring mechanisms and accountability are lacking.
  • Political economy constraints are likely to play a role. The model of development implemented so far in China has rewarded local authorities, promoting investment in fixed capital at the expense of other types of spending, including social programs. The planned shift in spending may actually meet with some resistance.
  • Spending is recorded on a cash basis. Operations that would normally be treated as financing items (such as clearance of arrears or amortization of debt) are included in spending when the corresponding cash payments are made. This has biased spending upward when liabilities were paid back (for example, in 2004, when a large amount of arrears on value-added tax refunds to exporters was cleared) and has biased spending downward when the corresponding liabilities were incurred but not recorded. Hence, care should be taken when looking at spending patterns in China because there is not sufficient information to properly classify and record all government outlays.

Ongoing public financial management reforms would help address these weaknesses, and their implementation should buttress any significant increase in overall spending. This would provide assurance that policy intentions are translated into appropriate actions and that value for money is guaranteed. A new budget classification and chart of account should be promptly implemented, and the single treasury account should continue to be extended nationwide. Over time, the experience with pilots on performance-based budgeting of capital projects may also prove useful.

Significant increases in social spending might be achieved without major cuts in other expenditures or a substantial rise in the budget deficit, if recent history is any guide. Spending has risen significantly over the past several years, even when the budget dictated expenditure cuts. Over the period 1998–2005, the Chinese authorities intended to reduce spending relative to GDP more often than they intended to increase it (Table 10.3 and Figure 10.5)show how budgeted spending in a year compared with actual spending in the preceding year). However, regardless of original intentions, spending has consistently exceeded budget targets. At the same time, these higher levels of spending have not led to a systematic deterioration in the fiscal position because revenue overperformance relative to the budget exceeded the increase in expenditure.

Table 10.3.China: Budgeted and Actual Spending(In percent of GDP)
Memorandum items:
Budgeted change0.7–1.0–0.50.2–0.9–1.5–0.1
Actual change1.–0.3–0.50.7
Source: China Ministry of Finance; and authors’ calculations.
Source: China Ministry of Finance; and authors’ calculations.

Figure 10.5.China: Deviations Between Outturn and Budget, 1998–2005

(In percent of budgeted level)

Sources: China, Ministry of Finance; and authors’ calculations.


Steven Dunaway is Deputy Director, IMF’s Asia and Pacific Department, and is also the IMF’s mission chief for China. Annalisa Fedelino is Senior Economist, IMF’s Fiscal Affairs Department.


The current budget classification, which is being upgraded, does not provide an accurate description of expenditures by economic and functional types. Hence, it is difficult to explain past trends in expenditure programs.


A positive fiscal stance indicates that the budget deficit (surplus) is larger (smaller) than it would be, given the cyclical position of the economy; similarly, a negative fiscal stance indicates that the fiscal deficit (surplus) is smaller (larger) than it would be, given the cyclical position of the economy. The fiscal impulse measures the change in the fiscal stance and indicates the direction and amount of fiscal stimulus.


For simplicity, it is assumed that annual trend growth is 8 percent.


This will also reflect the model chosen by the authorities for the provision of the related services. For example, social infrastructure services may be provided by the private sector, as is the case in a number of countries, under concession contracts or public-private partnerships. This option is not without risks to the government, though, because shifting costs to the private sector may mean significant obligations later on. On the positive side, efficiency gains would be expected to stem from private sector participation. These (and other) issues will be explored in a forthcoming selected issue paper on social programs and policy options in China.


Transfers from the center represent almost half of local governments’ total resources on average, but this share varies significantly across provinces. For example, in 2002 this share ranged from about 20 percent in Beijing and Guangdong to about 60 percent in Inner Mongolia and Guizhou, and 85 percent in Tibet.

    Other Resources Citing This Publication