Chapter

III Monetary and Fiscal Policies in Transition

Author(s):
Gyorgy Szapary, Steven Dunaway, David Burton, and Mario Bléjer
Published Date:
March 1991
Share
  • ShareShare
Show Summary Details

Since market-oriented reforms were set in motion in 1978, the nature and role of macroeconomic policy instruments, particularly monetary and fiscal policy, have been transformed pari passu with the changes in the economic system. As discussed in Chapter II, under the rigid central planning in place before the reforms, enterprises had little or no independent decision-making powers with respect to investment, production, prices, or wages, all of which were basically decided by the planning authority. In this setting, the role of monetary policy was essentially to support the implementation of the physical output targets contained in the plan, while preventing an excessive accumulation of liquidity in the household sector. At the same time, the objective of fiscal policy was limited to the administrative allocation of resources by regulating the rate of capital accumulation and maintaining household incomes at a level consistent with the availability of consumer goods.

As the scope of central planning has diminished and agents have been increasingly guided by market signals, macroeconomic management has required adoption of a consistent set of new, indirect, policy instruments. Such instruments have in fact been created, but difficulties in applying them have arisen from a number of factors. First, the pre-existing techniques and institutions of central planning have been allowed to continue functioning alongside the new ones. Second, the decentralization of economic decision making to economic agents has been interpreted as requiring also a considerable degree of decentralization of economic management from the central government to lower levels. The increasingly powerful role played in the economy by local governments at the provincial, city, and county levels, which see their primary role as supporting the development of their own regions, has severely affected the ability of the central government effectively to utilize the new indirect instruments to maintain macro-economic balance. These fundamental problems have exacerbated the technical difficulties inherent in developing instruments of macroeconomic control suited to the new environment, and periods of inflation or external imbalance have resulted.

The following sections examine the evolution of the roles of monetary and fiscal policies as the reforms have proceeded. They also consider whether there are characteristics of the reform process that inhibit the maintenance of control over aggregate demand, and explore the question of how firm macroeconomic control can be maintained in a partially reformed economic system without resort to policies that negate the basic intent of the reforms.

Monetary Policy

In the economic system that emerged during the 1980s, monetary policy came to play an independent and major role in macroeconomic management. With more financial resources left in the hands of enterprises and households, and with greater freedom of choice over how these resources should be used, money was no longer just a counterpart to physical transactions specified in the plan, but was also important as a store of value and as a medium of exchange for coordinating unplanned market transactions. The expansion in the role of the financial system was also reflected in the increase in the proportion of enterprise investment financed by the banking system. At the same time, price and trade liberalization increased the potential for expansionary monetary policy to be reflected in inflation and external imbalances.

Policy Formulation and Implementation

Prior to the reforms, monetary policy was implemented through a credit plan and a cash plan. The credit plan was the financial counterpart of the physical plan and specified the amount of credit needed by enterprises to implement the output targets. Enterprises were not allowed to hold cash and their holdings of financial assets were closely regulated (the dual payments system). Enterprises were required to remit all their surplus funds to the Government, and in turn the budget supplied the funds for investment. Bank credit was used chiefly to provide working capital and was related to the accumulation of inventories by enterprises. The cash plan covered the various factors that influenced the amount of cash in the economy, principally the payment of wages and the purchase of agricultural products. However, the PBC had little control over most of the factors affecting the amount of currency in circulation, and its role was essentially to monitor cash flows and alert the authorities when deviations from the cash plan occurred. Since most prices, foreign trade, and external capital flows were controlled prior to the reforms, an excess supply of money generally had little impact on prices or the balance of payments but was rather reflected in shortages in the goods market. Given the “soft” budget constraints under which enterprises operated—a characteristic of which was an accommodating credit policy37—the system tended to generate excessive liquidity in the economy which was reflected in the rationing of goods and the accumulation of forced savings. On occasion, however, these pressures were allowed to be passed on to higher prices, such as in 1961 when recorded inflation rose to 16 percent.

In response to the evolving role of monetary policy, there have been major changes in the formulation and implementation of monetary policy since 1979.38 Although the annual credit plan continues to be a central element in the formulation of monetary policy,39 banks have been given greater discretion with respect to loans for working capital.40 A cash plan is also still a part of the policy formulation process, but the target for the increase in currency is now only indicative.41 Also, while the formulation of the credit plan remains essentially a “bottom up” exercise that focuses on the credit needs of borrowers, the PBC has recently begun to evaluate the credit plan in the context of a broader financial program. The latter is based on objectives for economic growth and inflation and includes targets for broad money as well as for currency.

Reflecting the coexistence of planning controls and market mechanisms in China, monetary policy is implemented through what may be viewed as a dual control system that relies on both direct credit controls and indirect levers. Under the credit plan, credit ceilings are established for each specialized and universal bank,42 and the banks, in turn, allocate ceilings to their local branches.43 The extent to which credit ceilings have been enforced, however, has varied considerably in recent years. At times they have been little more than indicative targets, and in some years, reflecting weak enforcement, they have been substantially overshot. On other occasions, particularly when monetary policy was being tightened, credit ceilings have been enforced more strictly. On these occasions the authorities have reemphasized controls over credit allocation to sectors and specific enterprises to ensure that credit went to areas deemed to be of priority, so that the brunt of the credit restraint was borne by nonstate-owned industries.44

Efforts were made in the second half of the 1980s to place greater reliance on indirect instruments of monetary control, especially reserve requirements, interest rates, and PBC lending to banks—as opposed to credit ceilings—in controlling the overall amount of credit. The PBC’s ability to control bank credit indirectly, however, has been limited by the ability of local governments to put pressure on its own branches to extend credit to banks to enable them to meet regional credit needs. As a result, growth in PBC credit to banks itself has at times been a major cause of rapid monetary expansion. Steps were taken during 1987–88 to strengthen PBC head office control over the lending by its branches by giving new loans under the plan a fixed maturity and by placing restrictions on the ability of local branches to extend temporary credits. Subsequently, in 1989, a requirement that the appointment of presidents of branches be approved by head office was introduced, and measures were taken to ensure that short-term PBC credits are withdrawn when they fall due.

With regard to interest rates, which are regulated by the PBC, adjustments were infrequent until 1988 despite sizable fluctuations in the inflation rate. The rigidity of interest rates reflected, in part, concern about the impact of increases in lending rates on enterprises whose profitability was restricted by price controls. Since then, interest rates have been adjusted on several occasions, and the interest rate on long-term savings deposits has been linked to inflation.45 In addition, financial institutions have been given limited flexibility to set interest rates higher than those specified by the PBC.

At the present transitional stage of reform, interest rates appear to be more effective in influencing the demand for deposits than in affecting either the demand for, or, in particular, the allocation of credit. Households are natural optimizing units that face rigid budget constraints, making them sensitive to the returns on financial instruments. This responsiveness was illustrated by the effectiveness of increases in deposit rates in halting a run on deposits during the summer of 1988. By contrast, many state enterprises, especially those that are unprofitable, continue not to be fully responsible for their financial performance and are frequently able to obtain subsidies to cover losses. Hence, their demand for credit may not be strongly influenced by the cost of borrowing.46 The impact of changes in lending rates on credit demand by nonstate enterprises, which account for a growing proportion of economic activity and have to operate within relatively hard budget constraints, is likely to be greater than for the state sector. Nevertheless, the allocation of credit remains determined by the credit plan and local interventions rather than by market mechanisms, and nonstate enterprises have borne the brunt of efforts to ration credit supply.

Monetary Developments

Difficulties in maintaining monetary control under the reforms have been reflected in wide fluctuations in the growth rates of money and credit aggregates (Chart 1 and Table 1). In late 1984 and early 1985, for example, the rate of monetary expansion increased dramatically as the in-creased financial autonomy granted to enterprises led to a sharp increase in the demand for credit to finance investment and wage bonuses that the newly established central bank was not yet strong enough to resist.47

Chart 1.Monetary Developments, 1984–89

(Annual percentage change)

Source: Data provided by the Chinese authorities.

Table 1.Monetary Survey, 1985–891
19881989
198519861987MarchJuneSeptemberDecemberMarchJuneSeptemberDecember
(In billions of yuan; end of period)
Net foreign assets20.83.923.032.034.231.330.327.523.024.037.1
Net domestic assets499.1668.3812.0835.0891.3940.1979.7991.81,033.61,075.91,158.7
Loans to enterprises and individuals628.9814.5980.31,003.21,057.51,102.61,141.81,151.41,180.21,223.21,347.0
Net credit to government2–9.35.820.89.610.317.130.524.330.223.724.7
Other items, net3–120.4–152.1–189.1–177.8–176.4–179.6–192.7–183.9–176.7–170.9–212.9
Money plus quasi-money520.0672.2835.0867.0925.5971.31,009.91,019.21,056.61,099.91,195.9
Money374.0474.1568.5575.3623.8669.9695.0668.5672.0676.5744.2
Currency(98.8)(121.8)(145.4)(144.2)(154.2)(184.9)(213.3)(209.7)(207.9)(208.1)(234.2)
Household demand deposits(39.7)(50.6)(70.8)(75.6)(80.8)(86.5)(94.8)(91.5)(93.3)(92.9)(94.6)
Enterprise deposits(203.0)(262.1)(307.3)(315.9)(348.9)(359.2)(347.7)(329.4)(332.0)(333.5)(367.0)
Official institutions(32.6)(39.6)(44.9)(39.6)(39.9)(39.4)(39.3)(37.9)(38.8)(42.0)(48.4)
Quasi-money145.9198.0266.5291.6301.7301.4314.9350.7384.7423.5451.6
Capital construction deposits(23.5)(25.1)(30.9)(33.0)(34.3)(34.5)(31.2)(31.9)(35.1)(38.0)(42.0)
Household term deposits(122.4)(172.9)(235.6)(258.7)(267.4)(267.0)(283.7)(318.8)(349.6)(385.4)(409.6)
(Percent change from the same period of the preceding year)
Net domestic assets22.133.921.522.226.023.820.718.816.014.418.3
Of which: Loans to enterprises and individuals(22.3)(29.5)(20.3)(22.7)(23.8)(211)(16.5)(14.8)(11.6)(10.9)(18.0)
Money and quasi-money17.129.324.225.728.425.521.017.614.213.2184
Of which: Currency(24.7)(23.3)(19.4)(26.3)(35.8)(45.6)(46.6)(45.4)(34.8)(12.5)(9.8)
Memorandum item: Ratio of currency to deposits (percent)23.522.121.120.020.023.526.825.924.523.324.4
(seasonally adjusted)21.820.519.520.221.323.624.826.326.123.422.6
Sources: People’s Bank of China; and staff estimates

As inflation rose in mid-1985, the PBC took steps to tighten policy, particularly through stricter enforcement of credit ceilings, and money and credit growth rates slowed sharply. Subsequently, concerns about a slowdown in economic growth led to an easing of policy that resulted in broad money growth rates close to 30 percent for much of 1987 and the first three quarters of 1988.48,49 Interestingly, the broad money expansion during this period was fueled not so much by PBC credit to banks as by a large drawdown of banks’ excess reserves (Table 2). The utilization of these excess reserves was made possible by a shift from mandatory to indicative ceilings on banks’ loan portfolios in 1987; it was facilitated by the development of an interbank market that allowed banks to economize on working balances.

Table 2.Operations of the People’s Bank of China, 1985–891(In billions of yuan; end of period)
19881989
198519861987MarchJuneSeptemberDecemberMarchJuneSeptemberDecember
Net foreign assets12.43.715.023.923.922.520.922.924.731.132.9
Claims of financial institutions224.9269.4277.4270.3270.3292.0338.8332.9331.6348.5420.7
Other domestic assets–8.79.629.916.816.839.045.846.951.246.448.1
Loans8.613.022.723.323.326.430.630.231.433.034.5
Budgetary borrowing, net2–9.35.820.89.69.617.130.524.330.223.724.7
Other items, net–7.9–9.3–13.6–16.1–16.1–4.5–15.3–7.6–10.4–10.3–11.1
Reserve money228.6282.7322.3311.0311.0353.4405.5402.7407.5426.0501.7
Liabilities to banks96.4120.1127.4121.3121.3122.1145.4147.0152.8167.1208.2
Required deposits(42.0)(56.5)(67.0)(70.8)(70.8)(85.6)(84.1)(87.2)(92.0)(96.8)(104.2)
Other deposits(47.0)(55.8)(52.8)(43.5)(43.5)(27.4)(50.9)(47.8)(49.0)(58.9)(91.2)
Cash in vault(7.3)(7.8)(7.6)(7.0)(7.0)(9.1)(10.4)(11.9)(11.8)(11.4)(12.8)
Liabilities to nonbanks132.2162.6194.8189.7189.7231.3260.1255.7254.7258.9293.6
Currency in circulation(98.8)(121.8)(145.4)(144.2)(144.2)(184.9)(213.3)(209.7)(207.9)(208.1)(234.2)
Deposits(33.4)(40.8)(49.4)(45.5)(45.5)(46.4)(46.9)(46.0)(46.8)(50.8)(59.4)
Total liabilities228.6282.7322.3311.0311.0353.4405.5402.7407.5426.0501.7
Memorandum: Money multiplier32.272.382.592.792.822.752.492.532.592.5823.6
(seasonally adjusted)24.025.127.327.527.627.026.325.025.425.324.9
Ratio of excess reserves to deposits (percent)12.111.08.26.46.03.76.86.36.17.010.1
(seasonally adjusted)411.110.07.56.66.04.16.26.46.07.89.2
Sources: People’s Bank of China; and staff estimates

At a more fundamental level, monetary developments reflected evolving views about the importance of monetary restraint for price stability and also about the appropriate inflation rate for an economy undergoing rapid development. It was only after inflation had climbed sharply in 1988 that a consensus was reached that monetary policy had to be tightened to fight inflation.50 The approach adopted involved the reintroduction of mandatory credit ceilings, as well as the use of indirect instruments including increases in interest rates, an increase in the reserve requirement, and firm control over the extension of PBC credit. Growth of broad money was thereby reduced significantly to only 13 percent by September 1989, after which the tight stance of policies was eased.

Macroeconomic Impact of Monetary Policy and Transmission Mechanism

Developments in monetary aggregates have had a major impact on both output growth in the short run and on the path of inflation during the 1980s. With regard to output, the monetary contractions of 1985–86 and during late 1988 and 1989 were both accompanied by a sharp slow-down in the growth of industrial production (Chart 2). With little flexibility in interest rates for much of the period since 1978, the transmission mechanism has operated in the first instance primarily through the availability of credit to enterprises and its impact on their demand for raw materials and intermediate inputs and also on their ability to increase wages and bonuses.51 The severity of the decline in output growth following the tightening of monetary policy in the past may have reflected the unresponsiveness of the prices set by some state enterprises to market conditions and also a limited sensitivity of wages and bonuses to demand conditions in the state enterprise sector. A perception by enterprises that credit restraint would not be sustained for very long in the face of a slowdown in industrial production may have contributed to inertia in the growth of wages and bonuses. More recently, adjustments to deposit rates have, through their impact on consumer spending, become a second major channel through which monetary policy affects demand.

Chart 2.Money and Industrial Production, 1984–89

(Annual percentage change)

Source: Data provided by the Chinese authorities.

Despite the rigidity of some prices, rapid growth in credit and money appears to have been the primary cause of the inflationary pressures that have emerged on several occasions.52 When monetary growth has increased, such as in 1984–85 and again in 1987, a surge in inflation has followed only a few months later (Chart 3). Similarly, when monetary policy has been tightened, for example in mid-1985, and again in late 1988 and 1989, inflation has slowed shortly thereafter.53 Cost push forces, including adjustments to administered prices and wage pressures, have probably also played a part in the inflationary process, but these appear to have been secondary. More generally, of course, for cost push pressures to cause sustained inflation, they need to be validated by credit expansion.

Chart 3.Money and Prices, 1984–89

(Annual percentage change)

Source: Data provided by the Chinese authorities.

Issues for Policy

Recent experience with monetary policy, especially the large swings in monetary growth rates, points to the need for modifications in the way policy is formulated and implemented.54

Choice of Intermediate Targets

It is widely agreed that a stable, low-inflation environment is necessary for durable economic growth to be achieved. Experience in China and elsewhere suggests that such conditions are needed for economic reforms to be successfully introduced; if the implementation of reforms is perceived to be linked with rising inflation, popular support for them is likely to be weakened. Furthermore, efforts to reduce demand when inflationary pressures have emerged have typically involved a strengthening of planning controls, and such setbacks to reform have tended to undermine confidence in the durability of the reform process. An important question therefore is how to maintain firm control over monetary policy during the transition to a more market-oriented economy. It goes without saying that making the achievement of price stability a major goal of monetary policy is the crucial first step.

Since there are significant lags between changes in the stance of policy and their impact on inflation or nominal income, an intermediate target (or targets) is likely to be helpful as a guide to policy. Ideally, this should be a variable that the authorities can easily influence, that has a close and stable relationship with the ultimate objectives of policy, and that provides an early warning if policy is deviating from its intended track. Money or credit aggregates have often been found to be useful intermediate targets in many countries, although in recent years financial innovation has often disturbed their relationship with nominal income, undermining their usefulness as policy guides. Other possible indicators include the exchange rate and interest rates, both of which have weaknesses as intermediate targets.

The choice of intermediate targets in China is influenced by the institutional arrangements that have evolved under the reforms. Financial markets are still relatively underdeveloped and interest rates remain regulated by the PBC, ruling them out for the time being as intermediate policy targets. Regarding the exchange rate, the rate in the swap centers is responsive to market forces, but is also influenced by variations in changes in the intensity of administrative controls over the purchase of foreign exchange for imports, limiting its usefulness as an indicator of monetary policy. These considerations suggest that a money or credit aggregate (or aggregates) may be potentially the most useful intermediate target for setting policy at the present stage of reform. The empirical work on the demand for money briefly reported in the appendix tentatively suggests that existence of relatively stable relationships for a range of monetary aggregates, strengthening the case for using monetary aggregates as intermediate targets.

This still leaves to be addressed the question of whether the targeted variable should be a money or a credit aggregate. The choice depends in part on the nature of the exchange rate regime.55 As discussed earlier, China operates a dual exchange system and also tends to use import controls to limit access to both markets with a view to influencing the balance of payments. In this setting, developments in net foreign assets may be to some extent independent of monetary policy, so that the choice between money and credit is not clear cut. One consideration in these circumstances would be that targeting a credit aggregate could result in more rapid monetary growth than desired if a tightening of import controls contributed to a larger than expected buildup of international reserves.56

A further issue concerns whether a narrow or broad aggregate should be targeted. Under the system of cash planning inherited from the traditional central planning model, the amount of cash circulating in the economy is viewed as the principal determinant of inflation. This has led the PBC to focus on currency as the main monetary aggregate. The appropriateness of currency or alternative monetary aggregates as intermediate policy targets can be viewed largely as an empirical question, depending on the strength and stability of the relationships between the different monetary variables and the ultimate objectives of policy. The results of the estimation of the demand for currency in the appendix suggest that currency has an unusual pattern of response to changes in the opportunity cost of holding money as captured by expected inflation. The demand for currency at first rises in response to an increase in expected inflation, only declining to less than its initial level in the long run.57 While the nature of this response in principle does not rule out currency as a monetary indicator, in practice the uncertainty surrounding the precise timing of the phases of the response pattern and difficulties in observing changes in expected inflation restrict the usefulness of currency as an intermediate target.58 This suggests that broader aggregates, the demands for which decline on impact in response to an increase in expected inflation, might make more suitable intermediate targets. Targets for broader aggregates would also have the advantage that they would be more helpful in the formulation of the overall credit plan.

Policy Implementation

Although progress has been made toward greater emphasis on indirect instruments for implementing monetary policy, considerable reliance is still placed on credit ceilings, particularly at times of credit tightening. Together with efforts to direct credit to priority uses, credit ceilings have tended to limit competition in the banking system and have contributed to inefficiency in resource allocation, running counter to the long-run objectives of reforms. At the same time, they have created incentives for intermediation outside the financial channels covered by the credit plan. Placing greater reliance on indirect instruments, and especially giving interest rates a larger role in allocating credit, however, will need to be accompanied by a hardening of enterprise budget constraints to make enterprises more responsive to market signals.59 This, in turn, has to be coordinated with price reform to remove remaining price distortions, to ensure an efficient pattern of resource allocation. Thus, improvements in monetary policy implementation must proceed hand in hand with enterprise and price reforms.

Further steps to improve the control of the central bank’s head office over its branches would help to ensure that whatever monetary targets were established would be adhered to and would help in the achievement of a more steady path for monetary policy. One such measure might be the replacement of provincial PBC branches by regional branches that act across provincial boundaries so as to reduce their susceptibility to political pressure. The general strengthening of the power of the central bank would also help to ensure that whenever monetary policy needed to be tightened action would be timely; to date, course corrections have had to await State Council approval and, as a result, have been delayed allowing inflationary pressures to become more entrenched.

Development of the Financial System

Several new banks and a large number of new nonbank financial institutions have been established and the range of financial instruments widened since the start of the reforms. Also, steps have been taken to increase competition, and to improve the efficiency through the development of an interbank market. However, competition within the financial system is still restricted and the financial market remains segmented, especially along regional lines. This segmentation is reflected, as noted in Chapter II, by the high proportion of interbank market transactions accounted for by banks in the same province.

Measures that might be taken to strengthen the financial system include increasing the autonomy of the specialized banks and making them fully responsible for their financial performance; as noted, reducing the reliance on credit ceilings would also help to promote competition. In addition, the development of securities markets, including the extension of the secondary market for treasury bonds to cover all issues, would help to increase the efficiency of the financial system and would represent a step toward allowing market-determined interest rates to emerge. At the same time, there is a need to strengthen the regulation and supervision of financial institutions to ensure that the soundness of the financial system is not undermined by increasing competition and the emergence of new financial institutions.

Price Liberalization, Repressed Inflation, and Monetary Overhang

Repressed inflation is believed to be a common feature of centrally planned economies and the extent to which it is a characteristic of the Chinese economy is a question that has important implications for the conduct of monetary policy and for the implementation of economic reforms. In particular, further price liberalization would result in a surge in open inflation if there was a significant amount of repressed inflation in the economic system. In those circumstances, it would be desirable to absorb the liquidity overhang before prices were liberalized. This section attempts to assess the likelihood that there has been substantial repressed inflation in the Chinese economy.

Analytical Considerations

It is straightforward to understand how repressed inflation or forced saving could arise in an economy in which all prices are set administratively by the state and black markets are effectively suppressed. A simple example would be an economy with one asset, money, and one commodity whose output is given and which is distributed at a fixed price. In this setting, an increase in the money supply would result in an excess supply of money and excess demand for goods—that is, repressed inflation.60 Also, the real money stock would be determined by the nominal money supply and official prices, not by the demand for real money balances.

The structure of China’s economy, however, no longer corresponds closely to that of a stylized centrally planned economy. Prices have been partially liberalized; by the second half of the 1980s perhaps less than half of transactions took place at state set prices, and adjustments to state set prices have become more frequent as the reforms have progressed.61 A particular feature of the price liberalization has been the introduction for many commodities of a two-tier price system under which quota production is sold at state-set prices, while production in excess of the quota is free to be sold at market determined prices (see Chapter II). In addition, the prices of other products and of many services have been completely liberalized, although the prices of some items sold through state outlets remain regulated.

In analyzing the potential for repressed inflation, it may be helpful first to examine the implications of two-track pricing since this is an important feature of China’s partially liberalized system. With pure two-track pricing individuals and enterprises are free to buy goods at the margin at market-related prices. Hence, money holdings cannot be regarded as involuntary and in that sense there can be no repressed inflation. A stricter test for repressed inflation, however, would be whether the overall price level (a weighted average of administered and market prices) is less than the price level that would emerge under a completely free price system. Theoretical analyses of two-track or equivalent systems suggest that the market prices of goods would be at least as high as the prices that would prevail if the price controls on quota output were removed.62 The strict condition for the absence of repressed inflation, however, requires market prices under the two-track system to be sufficiently above the price level under a fully liberalized system to offset the effect on the overall price level of sales at state-fixed prices. Put differently, it requires that agents do not hold a higher level of real money balances under the two-track system than they would in the complete absence of price controls. Of course, the smaller the proportion of output that is sold at market prices, the higher those prices in relation have to be relative to both the state-set price and the price that would emerge if all price controls were removed.

If the two-track system is expected to persist indefinitely in its present form, then there would seem to be no obvious reason why agents would wish to hold a higher level of real balances than under a fully liberalized system—that is, attempt to shift consumption from the present into the future.63 On the other hand, if households and enterprises expect the price system to be liberalized in the near future, this would restrict the extent to which market prices (particularly of durable goods) can be above their level under a liberalized system. That is to say, as market prices rise, agents may attempt to delay purchases until after liberalization. It is possible, therefore, that the overall price level with a two-track pricing system would be lower than the price level that would prevail with a fully liberalized price system. Market prices under the two-track system, however, would still be responsive to changes in the money supply, since the notional price level under a fully liberalized price system would itself depend on monetary policy. A demand for money function of some form would also exist in these circumstances.

As noted earlier, not all commodities are subject to a two-track system; the prices of some items remain fully regulated, while for other products prices have been completely liberalized. The continued distribution of some commodities at state-set prices allows for the possibility of unsatisfied excess demand for those goods at prevailing prices and for the existence of repressed inflation; the latter would be prevented from emerging only if the prices of fully liberalized products were driven up sufficiently to push the overall price level to its full free market level. However, Ellis and Naughton (1990) show that in a situation where only some goods are rationed at fixed prices, the relationship between saving, on the one hand, and rationing or the price of rationed goods on the other is complex, depending on the characteristics of the goods in question and the nature of household preferences. While this is a partial equilibrium analysis, it suggests that rationing for some goods does not automatically imply repressed inflation in the economy as a whole.

Evidence

In discussions of repressed inflation in China, attention is often focused on the decline in the velocity of money during the 1980s (Table 3). One possible interpretation, advanced by Feltenstein and Ha (1988), is that it reflects an increasing amount of repressed inflation as remaining price controls have prevented prices from adjusting fully to increases in the money supply. An alternative explanation with very different policy implications is that the decline in velocity corresponds to a growing monetization of the economy as market mechanisms gained in importance under the reforms.

Table 3.Velocity of Circulation, 1982–891
19821983198419851986198719881989
Broad money2.182.071.941.811.671.501.491.47
Narrow money2.842.732.632.512.412.222.192.28
Currency13.4312.5311.7010.209.769.088.057.36
Sources: Data provided by the Chinese authorities; and staff estimates

As a starting point in considering whether the decline in velocity reflects repressed inflation, it may be helpful to ask whether money holdings are unusually high in China. An international comparison shows that with the decline in velocity during the 1980s, the ratio of broad money to GNP has in fact become relatively large. Based on data for 1987, the ratio of broad money to GNP in China was 67 percent compared with an average of 50 percent for a sample of 94 countries; China ranked eighteenth in the sample.64 Moreover, the ratio for China looks particularly high when the per capita income of the sample countries is taken into account (Chart 4). Nonetheless, this cannot be regarded as convincing evidence of repressed inflation, and may largely reflect the limited range of alternative financial instruments in China.

Chart 4.Broad Money/GNP: An Inter-Country Comparison1

1 Based on data for 1987.

2 In percent.

3 In thousands of U.S. dollars.

Empirical work on the demand for money casts additional light on the importance of repressed inflation. As discussed above, if money balances were held involuntarily, real money balances would be determined by the nominal money supply and official pricing policy. In these circumstances, it should not be possible to estimate a well-determined money demand function. In particular, real money balances would not be responsive to changes in the opportunity cost of holding money. The empirical work reported in the appendix, however, indicates that it is possible to estimate relatively stable money demand functions for various aggregates using quarterly data for the period 1983–88, and that money demand is sensitive to changes in expected inflation.

While this evidence suggests that money holdings are voluntary, as noted above, it does not rule out the possibility that the overall price level is less than the price level that would prevail if prices were fully liberalized. Given this consideration, the high estimated long-run income elasticities could still reflect, in part at least, repressed inflation broadly defined. This would require that the degree of repressed inflation is correlated with real income. Some evidence to support this view is reported by Feltenstein and Ha, although this is based on data for the period 1976–88 during only a part of which prices were substantially liberalized.65 Their approach also seems to involve circular reasoning in that developments in velocity are used to explain the behavior of the demand for money.66

The apparent responsiveness of both inflation and output to changes in the rate of monetary growth is evidence in favor of the view that repressed inflation may not have been substantial in China. Certainly, if a substantial proportion of real money balances were being held involuntarily, monetary policy would not be expected to be effective and—contrary to experience—would have little impact on either output or inflation. It is also of note that in 1989 there was a buildup of unsold stocks of consumer durables following the tightening of monetary policy. This indicates the absence of excess demand for those goods at prevailing state-set prices. However, the possibility of excess demand for high-quality imported goods in restricted supply remains.

Taken as a whole, the available evidence, while far from decisive, suggests that the magnitude of repressed inflation in China may have been limited. This should give some encouragement to efforts to proceed further with price liberalization. At the same time, the remaining possibility of significant repressed inflation would add weight to the importance of maintaining tight control over monetary policy as liberalization proceeds. The sale of state-owned assets, such as through the privatization of housing, could also help to absorb any excess liquidity in the economy.

Fiscal Policy67

As with monetary policy, the reform process has brought about marked changes in the objectives and mechanisms of fiscal policy. Under strict central planning, the limited objective of fiscal policy is to allocate resources administratively by directly regulating the rate of capital accumulation and by maintaining household incomes at a level consistent with the availability of consumer goods.68 Tax policy as such does not fulfill a major function, since the private sector plays only a minor role in the production process and state enterprises are required to remit all their surplus funds to the Government. Thus, the primary instrument of fiscal policy under central planning is government expenditure. Government funds are channeled back to enterprises through the financing of investment and working capital and in the form of subsidies. Therefore, in addition to its overall allocative function, the budget plays an important redistributive role by shifting resources among industries, sectors, and geographic regions. The level of expenditure is essentially determined by the quantitative output and investment targets set out in the plan and, during periods of overheating, cutbacks in expenditure are used to reduce inflationary and balance of payments pressures.

In the new system of economic management emerging in China, the main macroeconomic function of budgetary policies is to provide the Government with an effective policy tool to manage aggregate demand. To achieve this task, it is clear that fiscal instruments should be flexible and, at the same time, provide policymakers with a fair degree of control and with predictable options and results. Flexibility, however, does not imply that policy regulations have to be loose but that the central authorities have to possess the ability to adapt and change the rules, which have to be precise and transparent, in a timely manner in response to evolving circumstances. How well can fiscal policy in China fulfill this role under present circumstances? To answer this question, it is necessary to consider the characteristics of the existing policy instruments, as currently designed and applied, and assess how suitable they are to serve as efficient tools for short-term macro-economic management.

A second important question regarding the capacity of fiscal policy to fulfill its macroeconomic task relates to whether there are built-in elements in the reform process that tend to create fiscal imbalances. To evaluate this aspect, it is necessary to assess the ability of the emerging system to avoid a deterioration of public finances without implementing policies that negate the basic intent of the reforms. To this end, a review of recent fiscal trends and an analysis of the underlying forces behind those trends are needed.

This section provides some answers to these questions and draws a number of general implications from the Chinese experience. Although it focuses mainly on the analysis of the revenue side, where the nature of the reforms is unique and their policy implications more complex, it also addresses some expenditure issues, principally in the context of their consistency with the need to maintain macroeconomic balance.

Budgetary Developments Within the Reform Process

The budgetary developments described here refer to the state budget that consolidates the financial operations of the central and local governments. Local governments are all those below the national level; thus, the budgets of provinces, counties, municipalities, and townships are included.69

Tax Reforms and Trends in Revenue

Reform of the tax system has been an important part of China’s economic reforms. As noted in Chapter II, in order to increase the autonomy and financial independence of enterprises it was necessary to allow them to retain a portion of their profits, and by 1986, almost all enterprise profits were subject to taxation rather than being fully remitted to the Government.

One of the distinctive features of the current system is the lack of uniformity and the much larger degree of discretion it exhibits compared with the tax systems of market economies. China has a 55 percent income tax rate for medium-size, and large state-owned enterprises and a number of different rates for small, collective, and house-hold enterprises.70 However, on top of the regular income tax, an adjustment tax has been applied to compensate for differential profitability due to factors external to the enterprise, such as the structure of administered output prices in relation to the evolution of input costs, particularly raw materials. Since the adjustment tax is designed to serve as an equalizer, it is applied at rates that vary, sometimes significantly, from case to case, depending on the particular circumstances of the enterprise. A further large element of discretion was introduced into the tax system in the mid-1980s with the inauguration of the enterprise contract responsibility system; under this system the entire tax liability of an enterprise is negotiated on a case-by-case basis.71

Another aspect of tax reform was the introduction, in 1984, of four separate indirect taxes to replace the previous consolidated industrial and commercial tax.72 In China, indirect taxes are directly collected from enterprises and, unlike in market economies, they can rarely be shifted forward because of price controls. They are used, in addition to their revenue-generating potential, to compensate for perceived variations in enterprise profitability arising from the incomplete nature of price reforms. Since potential profitability is difficult to evaluate objectively, indirect taxes contain a discretionary component that translates into a wide range of rates.73 In fact, since many prices are still regulated, the applicable rates of the indirect taxes are determined jointly with the price of the taxable product, with one of the objectives of the tax-cum-price mix being the equalization of profits.74 Their cascading nature, however, makes an accurate tailoring of these taxes difficult to achieve.

A salient feature of the evolution of government revenue since the beginning of the reforms has been the continuous decline of total revenue as a ratio of GNP, from 34 percent in 1978 to 25 percent in 1986 and to 21 percent in 1989 (Table 4). The main reason for this trend has been the steady fall in revenue from the enterprise sector (direct taxation plus remittances). Although part of this fall has been the intended and desirable outcome of the decentralization process, the declining trend also reflects, as discussed in the next section, the unintended tax consequences of the reforms.

Table 4.Government Revenue, 1978–89(In percent of GNP)
19781979–811982–841985–8719881989
Total revenue134.430.027.024.820.420.9
Revenue from enterprises20.617.112.58.35.65.2
Of which: Profit remittances(19.1)(16.1)(11.4)(0.4)(0.3)(0.4)
Profit tax(1.5)(1.0)(1.1)(7.9)(5.3)(4.8)
Taxes on:
Income and profits221.517.813.38.35.65.2
Goods and services311.310.610.110.69.19.2
International trade0.80.91.11.81.11.2
Other taxes1.53.23.03.0
Nontax revenue40.80.81.01.31.73.4
Source: China, Ministry of Finance

Compared with other countries, the level of revenue relative to national income was still somewhat higher in China in 1988 than in countries with comparable per capita income, although the difference had narrowed considerably since the late 1970s (Table 5). On the other hand, compared with other CPEs, China’s revenue level was substantially lower. This reflects, in part, the absence of social security contributions to the budget in China, where retirement benefits—as well as other welfare benefits such as education and health care—are still largely the responsibility of the enterprise sector. Despite the changes brought about by the reforms, China’s tax structure still has more in common with the tax structure of CPEs than with that of other countries. Profit taxes still make up a relatively large part of revenue—a characteristic of CPEs where a substantial portion of revenue comes from the socialized enterprise sector. Other characteristics that China shares with CPEs are the very low level—in fact, in the case of China the virtual absence—of individual income taxation and the low level of taxes on foreign trade.75 In addition, China relies relatively more heavily on domestic indirect taxes than any of the other groups of countries. However, since indirect taxes in China often cannot be shifted through price changes because of price controls, these taxes become akin to a profit tax as their incidence falls fully on the enterprise.

Table 5.International Comparison of Revenue1
Country groupPer capita income2 (In U.S. dollars)Revenue as a percent of GNP/GDPIndividual income taxCorporate Profix taxTaxs on goods and servicesTaxes on internation tradeSocial Secuity contributions
(In percent of total revenue)
China
197821534.060.0332.92.4
1988300219.026.2345.65.5
Low-income countries26516.78.39.626.332.62.2
Middle-income countries1,72025.09.717.824.517.86.7
Industrial market economies11,21033.926.67.625.22.124.8
Centrally planned economies50–8041.0–8.060–64313–300.5–12.5
Sources: International Monetary Fund, Government Financial Statistics Yearbook, 1987; World Bank, World Development Report 1988; and Wanless (1985)

Trends in Public Expenditure

The reforms also had a profound impact on public expenditure, which, as a ratio of GNP, fell from 34 percent in 1978 to 27 percent in 1986 and further to 23 percent in 1989 (Table 6). Most of the reduction took place in capital outlays. The decentralization of the decision-making process and the larger profit retention by the enterprises have naturally meant a reduction in the Government’s role in financing investment. Consequently, capital expenditure relative to GNP more than halved over the past decade. Concurrently, a growing proportion of investment by enterprises has been financed by self-owned funds and bank loans. Within current expenditure, the level of defense spending has fallen the most as the Government demobilized about one million persons.76

Table 6.Government Expenditure, 1978–89(In percent of GNP)
19781979–811982–841985–8719881989
Total expenditure34.133.228.526.422.922.7
Current expenditure19.322.021.119.017.118.2
Of which: Defense4.74.53.02.11.51.6
Economic services5.04.23.22.52.12.2
Subsidies3.26.26.85.95.46.2
To enterprises1(1.0)(0.9)(1.3)(2.9)(3.2)(3.9)
Other2(2.2)(5.4)(5.5)(3.0)(2.2)(2.4)
Capital expenditure14.911.27.47.45.84.5
Source: China, Ministry of Finance

Subsidies in the reform period have averaged approximately 6 percent of GNP. There are two types of subsidies provided through the budget in China—consumption and enterprise subsidies. Initially, consumption subsidies rose sharply as the Government provided compensation to urban consumers for price increases arising from the agricultural reforms and the liberalization of agricultural price controls. Subsequently, these subsidies declined as a ratio of GNP, in line with the policy of relying on increases in productivity and wages rather than on subsidies to maintain and enhance living standards. On the other hand, subsidies to cover enterprise losses have increased steadily. Higher enterprise losses have arisen from the existence of soft budget constraints and the rigidity of administered prices.

In China, as in other CPEs, the level of government expenditure is not a very good indicator of the role of the state in the economy, since the Government, as owner of enterprises, has many other means of command over resources, which is not typically the case in market economies. Nevertheless, some perspective might be gained from an international comparison. As shown in Table 7, the level of expenditure in China is comparable with other low-income countries and substantially lower than in middle-income and industrial countries. A major difference vis-a-vis most countries is the virtual absence of social security and welfare payments from the government budget in China—these are the responsibility of enterprises.

Table 7.International Comparison of Expenditure1
Country groupPer capita income2 (In U.S. dollars)Expenditure as a percent of GNP/GDPCapitalCurrentSocial Security and welfareInterest paymentsSubsidies and current trasnfers
(As percent of total expenditure)
China
197821533.843.556.5n.a.9.3
1988300222.925.474.61.125.1
Low-income countries26522.821.778.34.48.414.0
Middle-income countries1,72029.521.079.010.311.516.4
Industrial market economies11,21037.56.593.536.611.735.5
Sources: International Monetary Fund, Government Financial Statistics Yearbook, 1987; World Bank, World Development Report 1988; and Wanless (1985)

Fiscal Policy Stance

Prior to the reforms, balanced budgets were emphasized and the government budget recorded small surpluses almost every year for three decades. In the initial years of the reforms, the budget was characterized by large deficits, averaging over 3 percent of GNP a year (Table 8). Beginning in 1981, the Government shifted the emphasis of macroeconomic policies toward stabilization and controlling budgetary deficits. Reflecting this policy, the deficit, while fluctuating from one year to another, fell to as low as 0.5 percent of GNP in 1985. However, after 1985, the deficit rose again, reaching 2.6 percent of GNP in 1988, as there was a renewed tendency for revenue to fall faster than expenditure as a ratio of GNP.

Table 8.Budget Deficit and its Financing, 1979–88
19791980198119821983198419851986198719881989
(In percent of GNP)
Budget deficit–5.2–3.3–1.3–1.4–1.7–1.6–0.5–1.9–2.0–2.6–1.8
Financing5.23.31.31.41.71.60.51.92.02.61.8
Domestic4.32.90.61.51.51.30.51.41.41.81.0
People’s Bank of China(4.3)(2.9)(–0.5)(0.6)(0.8)(0.7)(–0.2)(1.3)(0.6)(0.8)(–0.5)
Other domestic(1.1)(0.9)(0.7)(0.6)(0.7)(0.1)(0.8)(1.0)(1.5)
Foreign0.90.40.7–0.10.20.30.50.60.80.8
(As percent of total deficit)
Domestic82.584.944.8102.888.582.997.671.770.668.256.1
People’s Bank of China(82.5)(84.9)(–39.7)(40.8)(44.8)(42.9)(–51.2)(67.9)(29.9)(31.8)(–25.9)
Other domestic(84.5)(62.0)(43.8)(40.0)(148.8)(3.7)(40.7)(36.5)(81.9)
Foreign17.515.155.2–2.811.517.12.428.329.431.844.0
Source: China, Ministry of Finance

In the early years of the reforms, more than 80 percent of the deficits were financed by the central bank. In 1981, the Government began issuing bonds and, since then, such bonds have financed an increasing proportion of the deficit. Foreign borrowing has also become a growing source of finance, averaging about 30 percent a year of the total financing of the deficit since 1986. Until 1986, government borrowing from the central bank had not significantly contributed to monetary expansion, but during 1986–88 such borrowing contributed importantly to reserve money growth. Seen against the increase in inflationary pressure during that period, the fiscal deficit and its financing became factors in the inflationary process, bringing the budget into the center of macroeconomic management. This was amplified by enterprises’ recourse to bank credit, including indirect PBC support, for expenditures that previously were covered by transfers from the Government.

Fiscal Impact of Economic Reforms: Analytical Considerations

Against the background of the previous description, one can now analyze the impact that the reforms have had on China’s public finances, and assess, with a particular focus on revenue, the potential policy implications of the deepening of the reform process.

Four aspects of the reform’s fiscal consequences are discussed in this section. In the area of government revenue, the issues that need examination relate to the factors behind the very slow growth of receipts from the enterprise profit tax and the consequent steady decline in total revenue as a share of national income. The reform, by allowing enterprises to retain a significant proportion of their profits and giving them greater freedom of decision over the use of these profits, necessarily reduces the intermediation role of the budget. Therefore, the decline in revenue relative to GNP has been, to a large extent, intended and has been a natural consequence of the reform process. However, beyond this policy objective, there are additional consequences arising from the reform process that are contributing toward a continuous and unintended reduction in the level of government revenue.

The ways in which the reforms have been implemented seem to have led to the emergence of a long-term elasticity of tax revenue significantly below unity and an erosion of the corporate tax base itself (e.g., the taxable profits of the enterprises). Moreover, the way in which another integral part of the reforms—the growing decentralization of fiscal authority to lower levels of government—has been implemented has tended to reduce the revenue that is transferred to the central government, a development that also has potential implications for macroeconomic management. These three aspects of revenue developments are discussed here in detail. In addition, some expenditure issues associated with reforms are also discussed. A reduction in the intermediation role of the budget necessitates a cut in government spending commensurate with the decline in resources transferred to the budget. The concern in this area is that the use of resources through the budget is being reduced at a slower pace than government revenue and that a possible consequence of the deepening of reforms may be an intensification of this trend.

Low Elasticity of Tax Revenue and Role of Contract Responsibility System

An important factor in the reduced income elasticity of revenue is the contract responsibility system that, in most cases, has introduced a degree of regressiveness into the enterprise income tax. As noted, the contract responsibility system for state enterprises was introduced in the mid-1980s to increase efficiency by providing greater decision-making autonomy to enterprise managers. The system, which has been extended to most state-owned enterprises since 1987, involves a contractual agreement between the management of the enterprise and its owner—that is, the central or local government.77 The management contracts typically cover a period ranging from three to five years and stipulate a minimum amount of profit to be realized or tax to be paid to the Government each year;78 profits above the contracted amount—which usually is either the amount of profit realized in the year prior to the signing of the contract or some fixed rate of increase over that amount—are fully retained by the enterprise or taxed at a lower rate. Firms that fail to meet their minimum profit objective must, in principle, pay their minimum tax obligation from self-owned funds. Currently, about 90 percent of medium-size and large state enterprises operate under the contract responsibility system.

From the fiscal point of view, the expectation was that the contract system would expand the tax base in the longer run through improving enterprise performance. However, the intended results are weakened by the way the tax obligations are set in the contracts. Since below-quota profits are subject to a flat tax and above-quota profits are taxed at a lower (or zero) rate, enterprises retain an increasingly larger portion of their total profits as output and profits expand, resulting, in this manner, in an elasticity of less than unity for the enterprise profit tax.79

An actual example could be useful to illustrate the point. A typical contract embodying the type of continuously decreasing average tax rate and the below-unity elasticity implied by the system is provided in the case of the Beijing Broadcast Equipment Industrial Corporation. The 1987 contract stipulated that a minimum of ¥ 10.6 million in profits should be realized, which, regardless of the actual result achieved, will be taxed at the statutory 55 percent enterprise income tax rate. This guarantees to the Government an income tax revenue of ¥ 5.8 million. Any excess of profit over ¥ 10.6 million would be taxed at the rate of 27.5 percent, therefore reducing the average tax rate below the statutory rate.80 The relationship between average, marginal, and statutory rates implied by this characteristic example is depicted in Figure 1. Note that if contracts are strictly enforced, the elasticity of the profit tax actually increases from zero to 0.5 at the quota level, but, although continuing to rise as profit increase above the quota, it remains consistently below unity.81

Figure 1.Tax Rates Within the Contract System

Note: Since at profit levels below the established quota the tax liability is constant (and is equal to the statutory rate times the quota), the marginal tax rate is zero over that range, while the average rate falls as profits rise. Given the implied zero marginal tax rate, the resulting tax elasticity also equals zero. At the quota level, statutory and average rates coincide and, for profits beyond the quota, the marginal rate rises to 27.5 percent. Since the marginal rate then remains constant at the 27.5 percent rate, which is lower than the statutory 55 percent rate on quota profits, the average rate continues to fall, converging asymptomatically toward the marginal level. Given the one-step increase in the marginal rate, the tax elasticity exhibits a jump from zero to 0.5 at the quota level and rises monotonically thereafter, although it always remains below unity since the marginal rate never exceeds the average.

As currently implemented, the contract system not only works toward reducing the tax/GNP ratio—a problem in itself if the level of expenditure is not falling commensurately—but also limits the flexibility in using tax policy for macroeconomic management. Although it could be argued that long-term contracts are beneficial in encouraging efficiency improvements, by fixing contracts for several years, the system has constrained the ability of the central government to introduce new revenue measures in a timely manner, or to alter current ones, in order to meet changing economic circumstances. The loss of flexibility is all the more serious as the decreasing average tax rate embodied in most contracts tends to reduce builtin stabilizers and to introduce a procyclical aggregate demand element into the tax system.

Erosion of Profits

The second reason for the declining trend in revenue from profit tax—and to some extent from indirect taxes—is related to factors that have had a negative effect on the tax base itself, namely, taxable enterprise profits. While aggregate data on enterprise profits are not available, data on profits of state-owned enterprises with independent accounting, which account for over half of industrial output, declined from 13.5 percent of GNP in 1980 to about 7 percent in 1987.82 This trend is confirmed by estimates that show that enterprise savings declined from about 19 percent of GNP during 1979–81 to 15 percent during 1985–87;83 given the generally regressive taxation of profits, the decline in savings points to a fall in profits. Note that, in the context of the contract system, if taxable profits fall as a proportion of GNP, the tax/income ratio will tend to decline at a compounded exponential rate.84

One of the most important factors contributing to the decline in taxable profits is the continuous presence of price controls at the final product level, which has prevented many enterprises from shifting increased costs to prices. These increased costs have arisen from a variety of sources, most importantly from higher input prices (particularly certain types of raw materials whose prices are not tightly controlled), higher depreciation rates arising from a changeover to more realistic accounting practices, increased burden of higher indirect taxes, and a depreciation of the exchange rate.85 In addition, the increased autonomy of enterprises in the use of their financial resources, coupled with an accommodating credit policy, has led to a relatively rapid increase in wages and benefit payments, which also has tended to dampen profits.86 Finally, China has allowed many enterprises to deduct from pre-tax profits not only the payment of interest on loans but also the repayment of loan principal. This policy has been intended to be a transitory arrangement to put on an even footing enterprises that have to rely on bank loans and those that continue to benefit from budgetary grants to finance investment. The practice has had the effect of reducing the taxable income of enterprises, as reliance on loans to finance investment has increased under the reforms and a growing number of loans have become due.

Regional Decentralization

In a country of China’s size, regional decentralization is dictated by efficiency considerations and the Chinese authorities have viewed the regional decentralization of fiscal powers as an integral part of the reform process. However, the direction in which such decentralization is moving affects the buoyancy of tax revenue and tends to erode the amount of revenue transferred to the central government. In China, local governments, mostly provincial and city governments, are responsible for the collection of virtually all major taxes. The revenue thus collected is shared upward with the next level of government.87 The sharing arrangements are not uniform; they are subject to negotiation and may vary from one case to another. Over the years, the revenue-sharing arrangements have undergone many changes, but, since the inception of the reforms in the late 1970s, the trend has been toward granting local governments a greater degree of fiscal authority and allowing them to retain more revenue.88

This regional decentralization has tended to depress revenue collection, particularly after the introduction of the contract responsibility system for enterprises. Although tax policy and statutory tax rates are nominally set at the national level by the central government, local governments are responsible for the negotiation of management contracts with the enterprises they own. Within the framework of these contract negotiations, local authorities have the power to set effective tax rates through the establishment of quota profits and the rate of taxation of above-quota profits. Furthermore, given the sharing rules, local authorities have a clear stake in keeping as much financial resources as possible within their territory to finance their own preferred projects and are not keen on sharing revenue with the higher levels of government. This objective is often pursued through generous tax treatment granted within the framework of management contracts. In this way resources remain within the local government’s jurisprudence and local authorities can then tap these resources through “voluntary” contributions from enterprises to local projects—contributions which, of course, are not shared with the central government. Thus, while the effective “tax burden” of the enterprises is not reduced and may, in effect, be significantly higher than implied by its tax obligations, explicit tax revenue, and particularly the budgetary revenue of the central government, is eroding.

To address this problem and to provide incentives for local governments to collect and remit taxes, recent revenue-sharing arrangements, introduced mostly since 1988, are based on contracts that feature a quota arrangement, somewhat akin to the system of management contracts with enterprises. Under this system, local governments contract to remit to the higher level of government a predetermined amount of revenue and retain all or part of the revenue above the quota. Typically, as with enterprises, the quota is calculated as the revenue remitted in the year prior to the signing of the contract plus a fixed increment.89

The new revenue-sharing contracts between the central and local governments further weaken the center’s control over, and flexibility in, the use of fiscal policy for macro-economic purposes. This is so for two main reasons. First, the relatively low level of incremental revenue transfer that these contracts appear to set leaves increased amounts of resources in the hands of local governments compounding, as far as central government revenue is concerned, the revenue-depressing effects of the enterprise contract responsibility system.90 Second, the fixed increment featured in the quota arrangement implies that the revenue transmitted to the central government is not influenced by the underlying economic conditions, which affect only the revenue accruing to the local governments. Since the latter do not see themselves as having overall demand management responsibilities,91 the quota arrangement tends to impart a procyclical bias to the fiscal system, a bias that, again, is compounded by the declining average tax rate on enterprise profits under the contract management system.

Local governments can also expand their revenue base by setting up locally owned enterprises. When local governments are allowed to retain a higher proportion of the revenue they collect, there is added incentive for them to set up enterprises even if, from a global point of view, the creation of such enterprises may represent a misallocation of resources.92 The establishment of enterprises shielded from nationwide competition is possible because local governments maintain power to restrict the inflow of goods from other provinces and they can also preempt the use of inputs produced in their own province.93 This, of course, creates serious distortions in the allocation of resources from a national perspective. Finally, the quota arrangement could exacerbate the differences between poor and rich provinces, putting pressure on the central government to transfer more resources to less advanced provinces even as its own revenue base is eroding.

Expenditure Issues

Although the decline in government expenditure as a proportion of national income has been significant, viewed against the built-in factors that depress the elasticity of government revenue, there is a question whether the pace of spending reduction is consistent with future macroeconomic balance. The decline in the level of government expenditure has been constrained by three factors: (1) the incomplete nature of price reforms, which has resulted in increased enterprise losses financed through budgetary subsidies; (2) a perceived need, particularly in the early years of reform, to increase consumer subsidies as a way to compensate urban dwellers for the price hikes resulting from agricultural reforms and to avoid, in this manner, the strengthening of a constituency against the continuation of the reforms; and (3) the limited success of the Government in transferring investment responsibilities to enterprises in proportion with the increase in their retained earnings.

In addition to addressing these issues, the probable expenditure consequences of further reforms also need to be considered against the forces that are at work toward reducing the buoyancy of tax revenue. Despite the temporary halt in the pace of price reform announced in late 1988, the progressive liberalization of goods markets seems to remain an objective of the reform. Given a generalized downward stickiness of prices, the readjustment of relative prices following any significant price liberalization is likely to result in an increase in the price level, especially if accommodated by monetary policy. Such an increase could affect budgetary expenditure in several ways. First, higher commodity prices will raise government expenditure on goods and services, as well as capital spending, which will be affected by higher input prices. Second, price reform is likely to be accompanied by wage reform, which will increase the Government’s wage bill. Third, the demand for consumer subsidies will probably rise in order to mitigate the impact of the price increases.94 Finally, the price reform may also have an effect on the amount of subsidies to loss-making enterprises, although the balance of this effect cannot be assessed a priori; while the freeing of prices will allow many enterprises affected by price controls to become profitable, higher input prices and wages are likely to aggravate the situation of others, increasing pressures for higher subsidies. All told, while the liberalization of prices is expected to raise efficiency and hence should contribute to a strengthening of public finances in the long run, in the short to medium term it is likely to intensify the upward pressures on expenditure.

In addition to price reforms, further enterprise and welfare reforms are under way. These encompass mainly housing, unemployment compensation, and pensions. Although it is difficult to estimate precisely their budgetary impact, it is safe to predict that they will tend to put additional pressure on government expenditure. Further pressure on expenditure will come from the need to increase capital spending to relieve major bottlenecks in transportation, energy, and agriculture. Against these prospects, streamlining the tax system and the revenue-sharing arrangements to enhance the elasticity of revenue becomes crucial to prevent the emergence of serious fiscal imbalances that could undermine continuation of the reforms.

Policy Options

The fiscal arrangements as they have evolved, especially the contract management system and the revenue sharing arrangements between central and local government, have weakened the effectiveness of fiscal policy as a macroeconomic instrument and resulted in a loss of revenue to the Government. Concerning the latter, the appropriate level of government revenue is difficult to assess in an economy undergoing transformation, but the success of the reforms necessitates the maintenance of macroeconomic stability, which requires a degree of stability between revenue and expenditure. The need to maintain such fiscal discipline sets the broad framework within which the financial consequences of reform must be fitted. The Government needs, therefore, to either divest itself further of spending responsibilities or redress the erosion of revenue elasticity that has arisen from the manner in which the reforms are being implemented. Although there is still scope for the Government to relinquish spending responsibilities, there is increasingly limited room to reduce the level of expenditure further. Ways must therefore be found to enhance the elasticity of government revenue.

It will be difficult for the Chinese authorities to prevent the continuous decline in the revenue/GNP ratio within a system in which a large portion of incomes—that is, individual incomes in both rural and urban areas—is only lightly taxed or not subject to taxation at all, while the effective tax rate on those incomes that are taxed is subject to case-by-case bargaining. One way to reverse this trend would be to increase the tax burden of enterprises across the board. Such a measure, however, is not in the spirit of the reforms, since enterprises already carry a major portion of the country’s tax burden.95 An alternative response, which would be much more consistent with the objectives of the reforms, is to reduce the discretionary elements of tax policy. This could be achieved by subjecting all enterprises to a flat uniform income tax rate and by allowing only after-tax profits to be subject to negotiation and contracting between state enterprises and their owners. Such contracted “tax” liability would be similar to the payment of dividends to shareholders, in this case the Government. Experiments with this approach are presently being conducted in several localities.

As regards the revenue-sharing arrangements between the center and provinces, one problem is that they rely excessively on bargaining which leads to a large degree of nonuniformity of treatment. A more important issue, however, is that the system, by leaving more resources in the hands of local governments, may encourage local authorities to use their power to restrict competition and thus lead to the establishment of noneconomic industries. The obvious solution here lies in reducing trade and entry barriers to increase competition.

A related issue, and one that is often debated in China, is the appropriate level of fiscal balance during the reform process. As mentioned before, prior to the reforms, balanced budgets were emphasized. However, the level of tolerable fiscal imbalance during (and after) the reforms is not necessarily the same as the level prevailing in the pre-reform system. Indeed, at least in the early stages of the reforms, the negative macroeconomic consequences of deficits might be offset by changes in the behavior of households, which are themselves reacting to the change in incentives. Before the reforms, household income grew relatively slowly; income has grown faster under the reforms, however, and the Government has been willing to give greater freedom in its disposition. This has resulted in a voluntary decision on the part of households to save a greater proportion of their income.96 In fact, the sharp rise in the household demand for money allowed the Government to reap the benefits of seigniorage and finance a higher budget deficit than it would otherwise have been able to do without putting additional pressure on aggregate demand.97 However, the appropriateness of the fiscal balance cannot be abstracted from the stance of overall demand management policies. Clearly, if credit policy to-ward enterprises is as accommodative as during 1987–88, the room for financing budget deficits is reduced.

Conclusions

In this paper the evolution of economic reforms in China has been examined and an attempt has been made to identify the problems that have emerged, especially in macroeconomic management, and to suggest ways in which these problems might be overcome. In reviewing the progress with reforms, a cyclical pattern in their implementation clearly emerges. Major steps forward in the process have been followed by periods of some retrenchment or very slow advance, as economic problems, most notably excess demand and inflation, have appeared in the wake of major reforms. When economic order has been restored, reforms have proceeded again.

This pattern can be attributed partly to a lack of progress in developing and utilizing effectively the indirect instruments needed for macroeconomic management as planning controls were relaxed and to a tendency for the authorities to fall back on the more familiar methods of direct administrative control of the economy when faced with excess demand pressures. The increased reliance on administrative and planning mechanisms at such times has entailed significant costs, especially through damage to the credibility of the reform process; stop-go cycles in pursuing reforms create uncertainties and undermine confidence in the durability of the economic reforms that have been implemented. As a result, individuals and enterprises may be reluctant to invest in the structures and institutions necessary for the full benefits from reform to be realized.

The success of the reforms is evidently contingent upon the ability of the Government to maintain a stable macro-economic environment as decision making is decentralized and planning controls are relaxed. This paper has identified a number of aspects of monetary and fiscal policy as they have evolved that have limited their ability to fulfill their new roles in macroeconomic management. Many of these reflect tensions that have arisen between the decentralization of decision making necessary to raise economic efficiency and the need to establish and maintain macroeconomic discipline. The latter requires that macro-economic instruments remain firmly under the control of the central authorities.

In the monetary area, surges in the rate of monetary growth have been the main factor behind the inflationary pressures that have periodically appeared. The uneven course of monetary policy has in part reflected difficulties encountered by the newly formed central bank in establishing effective monetary control mechanisms, especially with respect to lending by its regional branches in the face of pressures from local governments to expand credit, and the timely adjustment of interest rates. The strengthening of indirect monetary instruments is also essential if heavy reliance on centrally imposed credit ceilings, which run counter to the thrust of reforms, is to be avoided. Improvements in the framework in which monetary policy is formulated also appear to be necessary so as to ensure that targets for monetary aggregates consistent with the avoidance of a buildup of inflationary pressures are established. The strengthening of the independence of the central bank would contribute to its ability both to develop indirect monetary instruments and to resist pressures for expansionary policies.

An important issue for monetary policy as reforms continue is the extent of the repressed inflation on the liquidity overhang in the economy. The partial price liberalization during the 1980s, especially the introduction of two-track pricing for many products, has limited the scope for repressed inflation by allowing demand pressures to be reflected in the prices of deregulated goods. Support for this view is provided by the apparent responsiveness of inflation to changes in the rate of monetary growth. Moreover, the existence of reasonably well-determined demand relationships for monetary aggregates suggests that money holdings are largely voluntary. These considerations, however, do not rule out the possibility that the price level is below the level that would prevail if all price controls were completely removed. That this may be the case adds weight to the importance of maintaining firm control over monetary policy as liberalization proceeds.

With respect to fiscal policy, the enterprise contract responsibility system and the revenue sharing arrangements between central and local governments have been valuable reform instruments that have infused dynamism into the economy. But, at the same time, they have led to an erosion of revenues that have also weakened the ability of the central government to use fiscal policy for macro-economic management purposes. The incorporation of quotas set in nominal terms over a multiyear period in both types of contracts has resulted in an erosion of revenues, particularly to the central government.

Problems with the current practices also arise from the high degree of discretion they entail. The contract responsibility system has involved case-by-case negotiation of taxes between the Government and the enterprises. This is likely to lead to taxation rules that vary from one enterprise to another and which may be very costly to change once the long-term contracts are signed. The regional revenue sharing arrangements, by leaving increased resources in the hands of local governments and by increasing their powers to set tax rates effectively, have eroded the command of the center over the mobilization of resources and have reduced its capacity to predict the outcome of policy measures. In short, the predominance of bargaining that leads to discretionary tax policies and the way in which decentralization of fiscal authority has been implemented have reduced the flexibility of the system and weakened the effectiveness of taxation as a macroeconomic policy instrument. The discretion entailed in the bargaining process also detracts from the benefits of the reforms because it makes fiscal policy less predictable and less effective and tends to validate distortions that could ultimately negate the benefits of reforms.

Since there is only limited room to reduce expenditure further, ways must be found to raise the elasticity of government revenue while reducing the room for discretion in the tax system. In the case of the enterprise contract system, both objectives could be achieved by subjecting all enterprises to a flat, uniform tax rate, allowing only the amount of after-tax profits to be subject to contracting and negotiation. A separation of taxes to be levied by local and central governments would help to achieve these goals for regional revenue sharing arrangements. More generally, the need to adopt more transparent fiscal rules, while keeping degrees of freedom to accommodate changing developments seems to be one lesson to learn from the Chinese experience.

The strengthening of the tax system along the lines suggested above and the greater reliance on indirect instruments of monetary control would go some way toward hardening enterprise budget constraints. But to be fully effective, especially in bringing about a more efficient use of resources, they would need to be supported by additional measures to increase the financial responsibility of enterprises, including the curbing of subsidies to loss making enterprises, and by the removal of remaining price distortions through further price reform. Although reforms to the present time have proceeded in a rather piecemeal fashion, a strong case can be made for moving ahead on a broad front with a comprehensive reform program on the grounds that the rigidities and distortions still present in the economy would undermine the effectiveness of partial reform efforts.

On the concept of soft budget constraint, developed by Kornai, see for example Kornai (1980 and 1986).

The role of monetary policy prior to the reforms and during 1978–85 is discussed in De Wulf and Goldsbrough (1986).

The credit for fixed investment included in the credit plan is closely tied to the state plan.

Control over the direction of working capital loans was strengthened in 1989 as part of the austerity program.

The cash and credit plans are assembled principally by the PBC and the State Planning Commission (SPC), and the original versions of both or any modifications during the year have to be approved by the State Council.

Since 1989, quarterly as well as annual ceilings have been established.

Exceptions are the branches of specialized and universal banks in Shanghai and Shenzen, which are given their credit ceilings directly by the PBC.

The direction of the bulk of credit to medium and large state-owned enterprises operating in priority sectors was an explicit objective of credit policy in 1989.

Deposit interest rates for enterprises were substantially lower than those for individuals until September 1988 when the differential was eliminated. Previously, the differential had encouraged intermediation among enterprises outside the banking system.

Furthermore, state enterprises have been allowed to deduct from pretax profits some loan principal payments, and this has reduced the effective cost of borrowing and limited the effectiveness of interest rates in eliminating excess demand for credit.

A rumor that the credit available to enterprises in 1985 would be based on the credit received in 1984 may also have been a contributing factor.

The stance of policy was briefly tightened in late 1987.

Percentage increases here refer to changes over a 12-month period.

As inflation picked up during 1988, the return on deposits became sharply negative in real terms, and there were widespread deposit withdrawals and panic buying of goods, adding to inflationary pressures.

Although enterprises are now able to issue bonds, the issuance of these securities remains small in relation to bank lending.

As would be expected, market prices have been particularly responsive to monetary impulses.

In both instances, an intensification of price controls may also have contributed to the slowdown in inflation.

A key issue for monetary policy is the extent to which there is repressed inflation in the economy, and this is discussed in the following section.

With a fixed exchange rate, changes in the money supply generally can be brought about through changes in international reserves—that is, through balance of payments surpluses and deficits—unless exchange controls are used to regulate reserve movements. Hence, the money supply is not under the control of the central bank, making domestic credit (or net domestic assets) more appropriate as an intermediate target. By contrast, under a freely floating exchange rate the money supply may be fully controlled by the monetary authorities and is suitable as a target variable.

A second consideration is the relative importance of money and credit in the transmission mechanism. While this is difficult to establish empirically because money and credit aggregates are closely correlated, recent experience suggests, as noted earlier, that both play a significant role.

This reaction probably reflects the absence of checking accounts in China and also uncertainties about the availability of some consumer goods. Hence, as part of the process of reducing deposit holdings by increasing purchases of commodities, agents temporarily increase their currency holdings.

Unanticipated deposit withdrawals, for example, would usually need to be accommodated at least in the short term.

The soft budget constraint of enterprises also limit the effectiveness of attempts to direct credit to particular sectors or firms. Enterprises whose credit is cut back are able to run up substantial arrears to other enterprises since there is little scope to enforce payment particularly for inputs allocated under the plan.

If a labor market is included in the model, households would reduce effective labor supply (if free to do so), and hence output would also be reduced. The extent of this effect would depend upon the expected future path of the consumption constraint (see Barro and Grossman (1974)), but in general there would still be repressed inflation in the system.

It is difficult to make a precise estimate of the proportion of transactions at state set prices as the extent to which prices are controlled varies by province.

The countries include all those covered by International Financial Statistics and for which the necessary data were available. GDP was used instead of GNP in cases in which the latter was not available.

Portes and Santorum (1987), using data from a somewhat earlier period, also find that disequilibrium in the goods market needs to be taken explicitly into account. See Burton and Ha (1990) for a review of the literature on this subject.

Feltenstein and Ha estimate a modified partial adjustment model in which the demand for real money balances (and expected inflation) is assumed to be based not on the observed price level, but on the price level that would prevail in the absence of price controls—the so-called virtual price level. The latter, in turn, is hypothesized to be a function of the observed price level and the ratio of currency to retail sales (the inverse of velocity). In the estimated model, the coefficient on the ratio of currency to retail sales was significantly different from zero, suggesting that repressed inflation was a factor affecting the demand for money.

This section is based on Blejer and Szapary (1989).

See Fetherston (1983) for a discussion of fiscal policy and Wanless (1985) for a discussion of taxation in CPEs; both provide a description of selected country cases.

The presentation of budgetary data follows the format of the International Monetary Fund’s Manual on Government Finance Statistics and therefore differs from that in national sources. The main differences involve (1) the treatment of foreign and domestic nonbank borrowing that in the national presentation is treated as revenue, while in the Government Financial Statistics format it is considered as financing; (2) the treatment of certain subsidies that are deducted from revenue in the national presentation rather than added to expenditure; and (3) the inclusion of off-budget construction expenditures that are excluded from the national budget presentation.

Mining enterprises involved in oil, natural gas, coal, and mineral production are subject to a progressive resource tax on profits. Collective enterprises are subject to a special profit tax enacted in 1985. It is levied at rates based on an eight-grade progressive scale, ranging from 10 to 55 percent. The tax is collected by, and the revenue from it fully accrues to, the county-level authorities. Many new collective enterprises are exempted from the tax for an initial period that ranges from two to five years.

A detailed description of China’s tax system is provided in World Bank (1989a). See also Easson and Jinyan (1987) and World Bank (1988a).

The four indirect taxes are the product tax, the value-added tax, the business tax, and the urban maintenance and construction tax. The coverage of some of these taxes has been changed over the years, partly with a view to broadening the tax base.

Note that this practice runs counter to the objective of letting profits serve as a guide to resource allocation.

However, efforts have been made to tax individual incomes, which have increased substantially as a consequence of the reforms. As a result, receipts from personal income taxes, while still very low, increased in importance in 1988.

Simultaneously, a considerable number of military plants were converted to civilian use. Many of these operations, which were typically located in remote areas in the interior of the country (“third front” factories) for security reasons, were encouraged to relocate to the cities and coastal areas. See Naughton (1988a).

The central government owns many of the largest state enterprises, but most of the state enterprises are owned by local governments at the provincial, municipal, and county levels.

On the characteristics and evolution of the enterprise reforms in China, see Naughton (1985 and 1986), Symposium on Enterprise Reforms in China (1987), and World Bank (1989c).

The tax/contract relationship can be formalized to illustrate the type of problems posed by the contract system and the factors affecting the elasticity of revenue. A formal derivation is presented in the appendix.

For other examples, see World Bank (1989a).

Clearly, if the contracts are nonenforceable downward—that is, if profits below the quota are only taxed at the statutory rate on realized profits rather than on the quota amount (a practice that seems to take place in some instances)—the elasticity falls from unity below the quota to 0.5 at the quota level.

Source: Statistical Yearbook of China, various issues. A survey of 17 enterprises in three cities (Beijing, Chengdu, Guangzhou) also shows that in 11 of these enterprises, gross profits declined in relation to the value of output between 1983 and 1987 (information obtained by the authors on visits to these enterprises).

Calculated as a residual after deducting from gross domestic saving, state budget and household savings. Gross domestic saving is calculated as gross domestic investment plus or minus the external current account balance. This estimate of saving, however, includes the extrabudgetary surpluses of local governments and is, therefore, not a fully accurate measurement of enterprise savings. (Source: Statistical Yearbook of China, various issues.)

See the appendix for a demonstration.

The nominal effective exchange rate depreciated by about 45 percent between the third quarter of 1983 and the third quarter of 1988. Furthermore, since 1987, an increasing proportion of the transactions has been carried out at foreign exchange transaction centers where the rate, which is more freely affected by market forces, is more depreciated than the administered rate.

Real unit labor costs in industrial enterprises rose by close to 13 percent during 1983–87 (Statistical Yearbook of China).

Typically, cities share with the provincial government and the latter shares with the central government. However, some cities, for example, Chongqing, Shanghai, and Tianjin, share all or part of their revenue directly with the central government.

On the evolution of fiscal relations between the center and the provinces, see Donnithorne (1983), Ferdinand (1987), and Oksenberg and Thong (1987).

For instance, the contract with Guangdong Province sets the amount of revenue to be transmitted to the central government during 1988–90 as the revenue transmitted in 1987 augmented by 9 percent each year. In some cases, the revenue to be transmitted is fixed in nominal terms for a number of years, as in the case of Shanghai.

Since the national budget consolidates all levels of government, taxes collected by local governments, but not remitted to the center, boost overall budgetary revenue. However, local governments do not tend to maintain budgetary surpluses (among other reasons, to avoid worsening their future bargaining position with the central government) and, therefore, increases in retained revenue, particularly if they arise after the budget has been approved, tend to generate higher local expenditure (which is difficult to control at the central level) and do not contribute to the improvement of the consolidated balance.

The behavior of local governments is motivated primarily by local interests and not necessarily by macroeconomic considerations at the national level. Local governments see themselves as being able to influence the level of economic activity in their province but not the general level of prices in the country or the overall balance of payments.

Shirk (1985) and Naughton (1988c) make the same point.

There is a discussion of the considerable barriers to entry in China in Naughton (1988d).

These two issues should, however, be qualified. The principal purpose of wage reform in an ongoing process of economic reform is to bring market forces to bear on the labor market and on production efficiency. Under these circumstances, a wage reform, though expenditure augmenting for the budget in the short run, enhances productivity and hence is likely to have a positive impact on revenue in the long run. Moreover, a wage reform should, eventually, result in a reduction of consumption subsidies as real wages rise.

In addition, there has been a tendency for a “clawback” of enterprise profits through surcharges on “extra” profits and taxes on extra-budgetary receipts. Such surcharges and ad hoc taxes are, of course, inimical to the spirit of the contract system.

Household savings have grown rapidly since the start of the reforms. Qian (1988) has found that in rural areas, where welfare benefits are lacking and there are greater opportunities to invest, such as in housing, the savings rate is much higher than in urban areas.

See above section on price liberalization, repressed inflation, and monetary overhang.

    Other Resources Citing This Publication