Information about Asia and the Pacific Asia y el Pacífico

7 Infrastructure Provision in a Socialist Market Economy

Ehtisham Ahmad, Vito Tanzi, and Qiang Gao
Published Date:
September 1995
  • ShareShare
Information about Asia and the Pacific Asia y el Pacífico
Show Summary Details
Bert Hofman and Richard Newfarmer* 

There is ample international evidence of the complementarity of infrastructure investment and productive investment. Provision of infrastructure (e.g., roads, railways, power) is therefore a precondition for sustainable growth. Also, lack of infrastructure prevents an integration of the domestic market, causing considerable efficiency losses and uneven regional development. In China, infrastructure development has been lagging behind the phenomenal growth rates of the reform period and is increasingly becoming a bottleneck for further economic development. To illustrate, the volume of freight traffic by rail grew 6.1 percent on average over the period 1980–90, whereas the length of railway lines grew only by an annual 0.6 percent; and freight traffic by road increased at an annual rate of 14.6 percent, whereas road length increased by an annual 1.5 percent over the same period. Besides quantity, the quality of infrastructure seems deficient, not least because of the underfunding of maintenance, but also because of inaccuracies in design.

Perhaps most pressing is the problem of financing China’s future investment in infrastructure. Demand is large, not only because the country is opting for a high-growth scenario, but also because there is a deficit situation already. To meet power demand until the year 2000, for example, an annual investment of $15–20 billion is needed, but in 1992, about $12 billion was actually devoted to investment in this sector. At the same time, the forms of financing available to the command economy of the past are dying out: credit allocation through the banking system is bound to shrink during commercialization of the specialized banks and the increasing independence of the central bank. Budgetary financing has already been greatly reduced under pressure of the low tax buoyancy of the tax system. New finance forms will therefore have to replace these waning sources. This paper argues that adequate financing of infrastructure can only be achieved through a proper organization of infrastructure provision.

China’s move to a socialist market economy offers new ways to improve the provision of infrastructure. To maximize the benefits of the market as an allocation mechanism, China needs to fundamentally reconsider the tasks of government in infrastructure provision. A crucial step is to separate government from enterprises, which is brought nearer by the “Regulations for Transformation of the State-Owned Enterprises.” The government’s remaining role in infrastructure, however, is not yet clearly defined; the assignment of functions over government and enterprises is incomplete; and some government functions have not yet found an institutional home. Furthermore, in the context of intergovernmental fiscal reforms, the assignment of government responsibilities in infrastructure over levels of government needs to be addressed. Finally, the financial sector reforms offer an opportunity to address the infrastructure financing problem.

This paper tries to contribute to the rationalization of China’s system of infrastructure provision. The first section sets out a normative framework for the government’s role in infrastructure provision and discusses how government should organize itself to perform its functions. The second section presents principles of the division of government functions over levels of government. The third derives the implications of the government’s role for the financing of infrastructure. The last section spells out China’s options.

Because the government’s role in enterprises engaged in individual market goods production will diminish—after a period of restructuring that may require some budgetary involvement—this paper will not treat such investment expenditures, even though at present the government is still involved. Nor will it treat capital expenditures necessary for other government tasks, such as education or health care. Space limits prevent the desirable differential treatment of the various sectors that are lumped together as “infrastructure.”

What Should Government Do in Infrastructure?

Provision Versus Production

Government has a role in the provision of infrastructure (roads, telecommunication, railways, airports, power, ports, and waterways) because, if left without government intervention, the market is illequipped to provide it efficiently. Inefficient provision can be due to a number of characteristics of infrastructure: (1) infrastructure usually requires large, lumpy investments, with many coordination problems in implementation and management; (2) some of these goods (electricity, ports) may show characteristics of natural monopolies, where the minimum optimal scale is close to the market size; (3) pricing of services by the market could be prohibitively costly from an economic perspective (uncongested roads); and (4) the market price could be considered too high from an equity perspective.

In a socialist market economy, government intervention does not imply government production.“Infrastructure” is a bundle of functions, many of which can be more efficiently performed by enterprises, either independent state-owned enterprises, township and village enterprises and collectives, or foreign companies. The art of government policy in infrastructure is to design interventions such that maximum use is made of the allocation through the market mechanism.


Government regulation can create conditions for a competitive market. For example, power generation could be treated as a market good, as long as enough competition in supply exists. Effective competition requires the establishment of a national or superregional grid to which competing power suppliers can deliver. Regulations on the conditions of supply, and standardization of regional infrastructure, such as regional rail tracks would greatly enhance competition among nongovernmental suppliers of services. Government would enhance competition also by providing a level playing field for all enterprise—regardless of ownership—by means of a corporate law. Even if competition cannot be ensured, a regulated nongovernmental enterprise often outperforms a monopolistic government agency. Stability and predictability of government regulation is essential for enterprises to engage in long-gestating investments usually involved in infrastructure.


Government regulations may equally prevent emergence of a market: price regulation still plays an important role in infrastructure and prevents services, such as power supply and transport, from being fully marketed. Moreover, inefficient pricing has led to inefficient use of existing capacity, and the installation of an inappropriate production structure, geared toward the use of uneconomically priced infrastructure. Infrastructure is more likely to be economically used when charges for infrastructure services reflect economic costs, and adequate pricing can generate investable surpluses for sectors that now depend on China’s public finances, such as electricity. Thus, in the move toward the socialist market economy, continued price reform in infrastructure services is essential. Adequate pricing would, therefore, help solve both economic and budgetary problems.


Even if nongovernmental entities could competitively supply infrastructure and infrastructure services, it is hard for them to generate the information necessary to plan for future needs. Moreover, for numerous infrastructure projects, government policy decisions concerning location, environmental impact, disappropriation of land, overall development policy, and so on need to be taken into account in future needs. This information and planning is basically a public good and needs to be provided by government, as nongovernmental entities will have no interest in generating all the information themselves. In China, the planning of infrastructure investment is situated in the State Planning Commission, the local planning bureaus, and the line ministries. The line ministries, however, have been deeply involved in day-to-day management of enterprises involved in infrastructure production, a task that should gradually disappear. A refocusing of the line ministries toward planning seems, therefore, desirable. The involvement of the Ministry of Finance in planning for infrastructure has traditionally been minimal, even though the budget finances part of the infrastructure investments through the People’s Construction Bank of China, and, more important, has often been saddled with recurrent costs. Coordination between the Five-Year Development Plan and the (multiyear) budget would contribute to the attainability of the Plan.


For government, the economic returns should be decisive in providing infrastructure not the private returns of enterprises. Government, therefore, needs project appraisal capacity in order to decide on the need for new infrastructure. In China, the capacity to appraise infrastructure projects is to a considerable extent concentrated in the People’s Construction Bank of China, but the stress is on financial appraisal. Economic appraisal seems centered at the State Planning Commission, which can provide subsidized funding and materials for those projects that are economically but not financially viable. The State Investment Corporations, a 1988 innovation in the planning system, have not been able to play the substantial role in appraisal that was envisioned for them. Their potential role as “investment houses” for certain sectors has not been realized, partially because they only had a single finance channel at their disposal. The Ministry of Finance seems not involved in economic analysis of infrastructure projects, even though these projects compete for budgetary resources with alternative government spending. Financial appraisal is done by the financing agents, such as the People’s Construction Bank of China, and is approved by the State Planning Commission as part of the investment approval system. For those projects in which no government finance is required in any form, such approval seems superfluous.

Project Implementation and Service Management

Enterprises are often in a better position to implement infrastructure investment and manage infrastructure services than a government agency. In many countries, construction and management of power generation is done by enterprises. Road construction and maintenance are well-defined functions and can be handled by enterprises, without direct government management. In these functions, the scope for improvement seems the largest, given the often poor implementation record for infrastructure under the planning system and the poor record of maintenance that led to an insufficient level of infrastructure services. Thus, the management of infrastructure services could in many cases be considered for market provision. Managerial autonomy, and full commercial risk for the enterprise, has been shown to be more efficient than more limited forms of enterprise involvement, such as management contracts. Even if the market is found unfit to provide certain infrastructure services (ports, roads), efficiency can be served by delegating responsibility to semiautonomous nonprofit agencies (port authorities, railways), or by devolving responsibilities to lower levels of government. In China, the experience with autonomous port authorities is a promising example.


Supervision of the infrastructure investment project is equally an essential task for a government that desires high-quality infrastructure assets, in line with the contracted features. In many countries, supervision is done by the same government agency in charge of project appraisal, as much of the same information is needed for both functions. If the management of infrastructure services is devolved to enterprises, regular reviews on the services delivered by the enterprises is desirable to ensure the regulated or contracted quality and quantity of services is met. Finally, supervision of the general competitive structure of the market for infrastructure is necessary to ensure the most efficient outcome for the users of the infrastructure services.

In China, supervision of government-funded infrastructure investment projects is the task of the People’s Construction Bank of China. It oversees the implementation of the project and the disbursements of funds. This supervision has not always been effective, and cost overruns and delays have been persistent in infrastructure investment projects. In the market for infrastructure services, such as the power supply, the government would supervise competitive practices and agreed standards of service. Supervision of the infrastructure services seems to be effected by direct management involvement of the line ministries involved. Again, a retrenchment of these ministries to the task of supervising general regulations and quality standards is necessary to increase the independence of enterprises providing infrastructure services. Supervision of competition is usually done by a separate agency (a cartel bureau, antitrust bureau). China seems to lack such a supervisory agency and has until now opted for direct price control through the Price Bureau.


If the price for which a competitive firm can supply infrastructure is seen as too high—for reasons of equity, efficiency, or development—subsidies can be given, instead of shifting production to government. To minimize the budgetary costs, while leaving producer incentives undistorted, auctioning or competitive bidding may offer a good alternative: the supplier that requires the least subsidy for a given infrastructure service at a given—politically acceptable—price would win the bid. In fact, for the construction of infrastructure projects, this is becoming more and more a common practice in China, but the bidding can be extended to the management and service delivery. The distinction between a government agency and a nongovernmental enterprise becomes vague, if the amount of subsidy on which the enterprise depends becomes high; once the subsidy element dominates, the enterprise could be considered as government.

Potential Competition

Even with a government agency implementing projects, or managing services, competition can play a useful role. The threat of being put out of business by a more efficient enterprise would enhance a government agency’s efficiency. Therefore, regular reconsideration of the supplier can greatly enhance efficiency. Such reconsideration could be prescribed in national legislation, but central government could also enhance pressures on local suppliers by making information available on the costs of infrastructure in various localities. A local government that seeks the optimal policy for its constituents would then choose the most efficient supplier.

China’s budgeting techniques are not geared toward regular reconsideration of public expenditures: budget formulation is highly incremental, owing to the base-number method, and auditing concentrates on the correspondence of budget and expenditures, and not on the effectiveness or on the output generated with the expenditures. Regular expenditure reviews and more program-oriented budgeting seem, therefore, desirable to increase the quality of infrastructure service delivery. Such reviews should be restricted, however, to the government’s own expenditures and should not replace the direct control over enterprises.

Central and Local Tasks: Who Does What?

Local Autonomy in China

In China, fiscal decentralization has given substantial autonomy to subnational government on the expenditure side, although the responsibility for infrastructure is nowhere codified. The constitution merely mentions “regional development” and “economic affairs” as overlapping responsibilities of local government and state council. On the revenue side, however, the local government has, de jure, no autonomy over tax rates or bases. As a consequence, and under pressure of the low revenue buoyancy of the budgetary taxes, a wealth of extrabudgetary fees, charges, and levies have evolved to finance standard government functions, among which is the provision of infrastructure. Moreover, although de jure local governments have no access to borrowing, this is easily circumvented by shifting the borrowing obligation to state-owned enterprises, while reaping the revenues directly by extrabudgetary levies on the borrowing firm, or indirectly by obliging the borrowing firm to perform government functions. China’s organization of government still rests on the principle of “dual leadership,” making local planning agencies, finance bureaus, railway bureaus, and so on, accountable both to the central government and to the local people’s congress. Increased local involvement in infrastructure investment is reflected in the share of locally supported capital construction projects, which rose from 46.5 percent in 1985 to 55.8 percent in 1992, a development contrary to budgetary trends, in which the central government now has over 75 percent of total government capital construction.

Rationale for Decentralization

Conceptually, responsibility for the provision of infrastructure should be put at the level that benefits from it. This implies that the national government should be concerned with national infrastructure projects (national railways, waterways, roads) and overall coordination; whereas the provincial government should be concerned with the same projects but on a local scale, and municipal and lower levels of government should be responsible for municipal, township, and village infrastructure.

Benefit areas of infrastructure are rarely clearly defined, however, national roads can be used for local traffic; ports have benefits beyond the municipality of location; and power stations may supply customers anywhere in the country once they are on a national grid. Responsibility for infrastructure with superregional benefits could be assigned to a higher level of government, whose constituency will benefit the whole area, but this need not be done. In many countries, a wealth of interregional cooperation bodies have responsibility for superregional infrastructure. In China, horizontal cooperation between localities is still rare; most of the coordination takes place at a higher level of government.

The rationale for devolution of responsibility is that local governments will provide the most efficient infrastructure, as they are better aware of local needs and are under closer scrutiny of their constituents than is the central government. Accountability can only be operational, if the local government has a substantial amount of autonomy over its own finance, both on the expenditure side and the revenue side: the fiscal adage for decentralization is that whoever decides on expenditures should pay for expenditures. Financial autonomy is not enough, however: the institutions that should perform the government tasks in infrastructure should be able to operate relatively independently from the central government, as far as local infrastructure is concerned.

User Charges

Assigning responsibility over the infrastructure provision to lower governments is not sufficient in China’s context, where substantial local autonomy exists. The incentive for provision should be supplied with the responsibility. A strong incentive for local government to provide infrastructure stems from user charges. If infrastructure supply is no longer a loss-making operation, local governments would be willing to perform the necessary functions to enable nongovernmental production of infrastructure. Economically excludable goods, such as power, ports, airports, and certain roads or bridges, would be able to generate substantial revenue streams. Even if the efficient charge would not total costs of provision, lump-sum charges, for example, for gas or electricity connection, could make public provision a cost-recovering activity.

User charges, if charged by government agencies substantially dependent on general budgetary means, should be in the budget; these funds—except for a share necessary to stimulate collection—are in principle budgetary resources, whose allocation should be decided upon in the annual budget process. In China, many charges and levies by government agencies are kept at the agency as “extrabudgetary funds.” This practice distorts efficient allocation of government resources and jeopardizes government budget control. Bringing such charges within the budget could be a start for the creation of a legalized local tax base.

Local Tax Base

Benefit taxes, taxes that are closely related to the use of infrastructure, are a second incentive for local governments to provide infrastructure. Examples are vehicle or license plate taxes, fuel taxes, and driver’s license charges, which proxy taxing road use directly; boat taxes and boating licenses for waterways; and passenger taxes for airports. Central government restrictions on local charges should be gradually abolished, if user charges are to be an incentive for local provision of infrastructure.

A third incentive for local governments to provide infrastructure is to assign them a tax base that is highly correlated with the provision of infrastructure. In many countries, property taxes are assigned to the municipal government. Property values, and thus property taxes, vary with the supply of complementary goods, including infrastructure. In China, property markets are still underdeveloped, and most property is directly or indirectly owned by the government and rented out far below economic costs. This does not imply, however, that no user taxes can be levied, either on private agents or on enterprises. Increasing mobility of capital and labor would increase the mechanism by which the tax base is increased, and subnational governments would become more responsive to demands for infrastructure. The central government can promote such benevolent tax competition by increasing the mobility of production factors. An active policy to integrate local capital markets and establish labor markets should, therefore, be the central government’s priority, and it should concentrate on the legal structure for local tax bases, including regulations on tax bases, and allowed rates and bases.

Role of Central Government

In a large country like China, most of the infrastructure can be considered as local, and central government involvement need not be large, but could be highly costly. The paramount task for the central government is the integration of the country by national road networks, rail networks, electricity grids, and so on. The tasks of planning, appraising, supervising, and—where applicable—financing such projects should lie with central agents.

National standards and guidelines should guarantee the integration of local infrastructure with national networks. Laws and regulations should set benchmarks for local government taxes and user charges, and budget supervision should ensure sound practices in budgeting for infrastructure, notably, concerning the recurrent cost implications of a government-managed infrastructure service. Central agents can serve as expert advisors for local governments in planning, appraising, and supervising particularly complicated projects. The central government can also stimulate cooperation among regions by serving as facilitator and coordinator of interregional projects.

Finally, the central government should be responsible for the legal framework that allows competition in infrastructure projects and services. Company laws and tax laws that treat each ownership type equally are necessary to level the playing field. Antitrust laws and antitrust supervision are central tasks as well, because they preserve an internal market.

How to Finance Infrastructure Investments

Provided an infrastructure project is economically viable, it will generate a future stream of benefits. Depending on economic feasibility of exclusion, political choices, and choice of organization, this benefit stream is reflected in market prices, user charges, benefit taxation, or budgetary subsidies that reflect the political valuation of infrastructure services. Before the benefits can be substantiated, infrastructure usually requires lumpy investments that need financing. As stated before, the amount of savings is not the bottleneck for infrastructure investment; it is the intermediation. The government’s role of direct resource mobilization is shrinking in China. In line with the overall retrenchment of the government, the budgetary role of investment finance has decreased strongly over the reform period, from almost 23 percent in 1982 to 7.4 percent in 1992 (Table 1).1 Much of the financing has been taken over by domestic loans from banks. Banks obtain most of their funding in the short and medium term and are less suited for the long-term investments required for infrastructure. Moreover, infrastructure finance is still dominated by “policy loans” through the specialized banks, which are incompatible with the desired commercialization of the banks. This makes development of new means of infrastructure finance a pressing problem.

Table 1.Financing Investment: Sources of Funds(In percent)
Total fixed asset investment
State appropriation22.77.4
Domestic loans14.321.0
Foreign investment4.96.2
Self-raised funds47.853.8
Fixed investment by state-owned units
State appropriation35.412.4
Domestic loans14.327.4
Foreign investment6.410.4
Self-raised funds34.840.2

Bonds and Equity

In many market economies, intermediation for infrastructure finance is done by the capital market. Infrastructure that can generate economically a future income stream that covers the investment costs on the project can in principle be managed as a profit-seeking enterprise, either state owned or otherwise. In a perfect capital market, such an enterprise would have independent access to the capital market, and the government need not be involved in financing such an enterprise. Even a partial float of equity in state-owned enterprises that implement infrastructure projects or deliver services can be a powerful tool for the owner to monitor its performance, and for competitors to learn which enterprise is efficient in delivery. In China, however, financial market liberalization is still in its infancy, and bonds and equity still play a minor role, funding about 11 percent of total investment. Especially for enterprises, access is limited, with only 7.8 percent of total investment financed by enterprise securities (Table 2).

Table 2.Securities Market Development(In percent)
Issues in 1992
Treasury bonds31.9
National investment bonds9.9
Enterprise bonds29.5
Financial bonds19.5

The primary task for the government in conjunction with opening infrastructure for profit-seeking enterprises is therefore to actively pursue a deepening of the capital market by issuing regulations and instituting the oversight necessary for such a market. At present, a substantial amount of securities issue and trade takes place within an institutional vacuum, which not only affects the quality of the investments financed, but also puts strains on the government’s macroeconomic control. In many countries, access to capital markets is regulated, or scheduled according to a timetable, in order to not overburden the capital market, and prudency regulation requires minimum standards of reporting and accounting. Care, however, should be taken that such macroregulation does not interfere with microeconomic efficiency of capital allocation. Clarity of property rights is a necessary condition for increased state-owned-enterprise access to (international) capital markets, and incorporation of those enterprises engaged in infrastructure is indispensable. Since state-owned enterprises are to be considered as independent enterprises, the government should bear no responsibility for their debt issued on the capital market.

Government Guarantees

If carefully handled, government guarantees can be instrumental in enhancing capital market access for enterprises investing in infrastructure. Enterprises with little or no reputation on the (international) capital market may need government guarantees to gain access to those markets. Such guarantees are a government obligation, and it is sound practice in other countries to explicitly take account of them in the budgetary process, by including them as a memorandum item. Moreover, since these guarantees will be called in the event of default of the borrower, provision should be made in the budget to capitalize a fund that will be depleted in the case of default. Failure to do so can seriously destabilize government finance in the future. Providing guarantees requires a solid project analysis—both economically and financially—by government, if no irresponsible risks are to be assumed. Only explicit guarantees, however, should be recognized by government.

Development Banks

A variety of countries have used a special institution to finance infrastructure and other high priority projects, and recently the idea of a “policy bank” has been introduced to China, but the plans are still very much in a state of flux. International experience shows that development banks are rarely successful, but those that are, are neither banks nor do they actively conduct policy. Rather, the institute is designed to implement government policy that requires long-term capital, say, a development bank. Policy choices, such as priority sectors, the total amount of credit the development bank may grant and subsidy on interest or principal repayments cannot be decided by such a bank but are agreed upon in the political process. The agency cannot be used for budgetary tasks such as subsidizing loss-making state-owned enterprises. The development bank can instead concentrate on project selection and the supervision of project implementation and service delivery. Vehicles for investment in infrastructure by the agency can be either enterprise-issued bonds or equity, direct loans, or local government bonds or loans. The allocation of funds should be transparent, with economic and financial rates of return as the guiding principles. The development bank could work under commercial criteria, if financially nonviable but economically desirable projects remain in the realm of budgetary finance.

The funding of a development bank should reflect the autonomy perceived for it. If it is to act as a fiscal agent, funding directly from the budget is the most appropriate procedure. The Ministry of Finance would then issue bonds and onlend them to the agency, providing funding for desired interest rate subsidies from the current budget. Alternatively, the development bank can be set at arm’s length from government, in which case funding with equity and bonds would be appropriate, possibly with government guarantees as set out above. The development bank should still receive current funding from the budget for the interest rate subsidies the government sees fit. Finally, the agency could be operated as an enterprise, incorporated, and with its own capital and borrowing, and independent from the budget. Such a corporation would be viable for sectors that will become profitable in the course of reforms, such as the power sector. However, one would hardly consider such a corporation a development bank.

Essential for macroeconomic stability is that the development bank have neither access to domestic commercial bank credit nor credit from the Construction People’s Bank of China. Furthermore, credit creation in the monetary sense should be forbidden, and only fully funded investments should be made. Deposit taking by the development bank would be undesirable. Essential for microeconomic efficiency is that the development bank have the capacity and authority to appraise and supervise infrastructure projects, and enforce repayments of its loans. Thus, of the roles of government as spelled out earlier, a development agent can perform three: appraisal, project implementation, and finance. This description comes close to the present role of the People’s Construction Bank of China; the biggest difference is that the People’s Construction Bank of China takes deposits and has only limited economic appraisal capacity.

Many of the institutions similar to the development bank sought by China have “gone bad” through lax supervision of projects and the absence of provisioning for bad loans. Clear rules on outside portfolio audit, on the publication of accounts and reports, and on loss provisioning will be critical for the long-term viability of such an institution. The governance of the institution would be enhanced by solid rules for the board of directors, financial reporting, and—if applicable—shareholding. Such rules would ideally isolate the agency from direct interference with individual finance decisions, much as a central bank usually functions better when independent of day-to-day politics. A high level of technical expertise of the employees is crucial for success.

Local Government Access to the Capital Market

The major macroeconomic influence of the government stems from the expansionary effect of deficits, as individuals do not fully offset government deficits with private savings. Within a decentralized fiscal system, an additional problem concerning deficit financing arises, when the subnational governments may overissue debt hoping to be bailed out by the central government, if future debt repayments could not be met. However, forbidding subnational deficits altogether leaves possible efficiency gains unexploited, since government infrastructure investments that create long-term benefits and future tax revenue can best be financed by debt issue. Moreover, regulations forbidding local borrowing can sometimes be circumvented, as in China where borrowing for government tasks is sometimes shifted to state-owned enterprises. To maintain macroeconomic stability while allowing at least some local government borrowing, several schemes have proven effective. In most countries of the Organization for Economic Cooperation and Development (OECD), borrowing is restricted to capital investment. Additional regulations usually require interest and amortization to be paid from current income to a certain percentage of current revenue, and so on, and sometimes restrict borrowing to specific purposes (such as energy projects). The aggregate borrowing of subnational governments is often preset by the central government and regularly administered by a special bank for subnational governments.

Local Government and Development Banks

Subnational governments usually have access to development bank lending in countries where such institutions exist, precisely because the local government is heavily involved in those projects that such a bank intends to fund. Development banks could play a useful role in channeling savings to infrastructure investments by enterprises, or by localities. As stated earlier, however, the development banks should be agencies rather than banks. For better access to such a bank, a decon-centrated organization seems advisable in a country as large as China. However, compartmentalization and quota disbursement per province should be avoided to maintain overall efficient allocation of capital as much as possible. Local development agencies already exist in China and take responsibility for industrial zone development in a number of localities, such as in Shanghai. Care should be taken that such agencies do not jeopardize central government macroeconomic control.

Capital Grants

Capital grants from central to local government are a means to fund capital projects of local governments. The advantage is that the central government keeps complete control over the general government’s access to the capital market; the disadvantage is that grants restrict local governments’ accountability of the funds, and they require a mechanism by which such grants are distributed. The latter problem could be dealt with by establishing explicit norms for disbursements from a capital fund, which should take into account rates of return of the projects, and possible regional development goals. Central regulation and standardization of applications for such a capital fund is needed to prevent rent seeking by localities. Local accountability could be enhanced by turning the grants into loans from central to local government, but in the Chinese context this seems to open yet another avenue for negotiation and renegotiation of repayments. Future disbursements, however, could be made conditional on the repayment of previous loans. Cofinancing requirements on local governments may equally be used to enhance local accountability.

Options for China

China faces challenges in the supply, quality, and financing of infrastructure. In moving toward a market economy, the prospects for meeting these challenges are enhanced. However, to make maximum use of the market mechanism, while maintaining crucial functions for government, a clear division of labor between enterprises and government, and central and local government is necessary for successful provision of infrastructure and infrastructure services.

State-owned enterprises involved in infrastructure supply to enhance their efficiency need more independence from government, and implementation of the “Regulations for Transforming the Operating Mechanisms of State-Owned Enterprises” is a high priority for China’s government. Price reform in infrastructure services should make a considerable part of these services accessible for profit-seeking enterprises. Corporatization under a corporate law, adoption of international standards of accounting, and modern governance procedures should enhance the access of state-owned enterprises to the capital market. A competition law should provide equal access to the market for infrastructure services for enterprises of all ownership.

Government should retrench as far as possible from direct involvement in production of infrastructure. Instead, it should concentrate on regulation, standardization, planning, supervision, and—if necessary—subsidization. Institutional homes for the remaining government tasks need to be clarified or developed and strengthened if missing or malfunctioning. Institutions that perform functions that are becoming redundant, such as the enterprise management by line ministries, need a reorientation of their tasks and a reconsideration of their size.

In order to use the market in infrastructure provision, accelerated price reforms should turn many of the sectors involved into commercially viable projects that could be implemented and managed by profit-oriented enterprises with various ownership structures.

Incentives for local government to provide infrastructure should be strengthened. User charges that cover costs would greatly enhance these incentives. A primary tool for this is also to assign local government a tax base that is highly elastic with respect to infrastructure; China’s authorities should consider this in the course of fiscal reform. Central government regulations concerning user charges should be reduced, to allow local governments the legal means to adjust charges to local conditions. Interregional cooperation in infrastructure should be enhanced, in order to relieve central government from some of the financial burden. Central government should concentrate on key projects with national impact. It should further regulate local government activities to the extent necessary for the national integration of local infrastructure.

Government legislation should lay the basis for a deepening of the capital market in an orderly way. Incorporation of state-owned enterprises involved in infrastructure would greatly enhance access to the capital market, and would thereby increase the pool of funds available for infrastructure, by means of equity and bonds. Independent state-owned enterprises imply that they themselves should be responsible for their debts, although explicit, carefully handled government guarantees may play a limited role to overcome reputation problems.

The central government clearly has an important role in mobilizing finance for infrastructure, but that role is considerably different from today’s role. Rather than raising revenues via taxation and then investing directly in infrastructure, a more decentralized fiscal system would allow the government to restrict its role to using its implicit financial and policy framework guarantee to mobilize private savings, both foreign and domestic. This role would require that the government create a mechanism to tap local and international bond and equity markets, and use the mechanism to channel resources to selected projects carried out by provincial and even private companies. Such a mechanism could take one of several forms: (1) a fund that would take minority equity positions and hold debt instruments issued by enterprises supplying infrastructure with the project, and that would be financed through local and foreign bond issues; (2) a development bank that would have direct loans to enterprises and that could receive (if necessary) regular budget capitalization and interest subsidies, but that would rely primarily on issuing its paper into the domestic market; or (3) a “take-out financing agency” that would operate like the fund, but finance only projects that have completed construction, taking out the financing provided by either commercial banks or the government during the construction phase.

As part of the new financing mechanisms, the fund or development bank could review the project appraisal and, by providing finance, would certify the financial and economic viability of the project. The key principles are that the fund or agency (1) would receive finance on market terms or as line-item transfers from the budget, and not be a deposit-taking institution; (2) would make loans on contractual terms only to reformed corporate entities operating in reformed policy regimes; and (3) would offer those terms at market rates except where the government provides interest subsidies specifically designated throughout the budget.

This role is substantially different from those alluded to in the press for the so-called policy bank. The objectives of the latter are to consolidate “policy-based” credits with infrastructure investments as well as those going to loss-making entities elsewhere. Such a policy bank cannot be viable in the market if it inherits a nonperforming portfolio, and must make loans to loss-making or high-risk entities on the basis of judgments made by those outside the policy bank. Finally, it cannot be viable if it subsidizes borrowers by lending at below-market terms and relies excessively on budgetary finance.

If the development bank is (partially) funded from the budget, this should be explicitly budgeted for, and the policy goals of such budgetary funding should be clearly stated. To the extent that the agency is allowed own funding from the capital market, the government should make clear whether this is considered a government obligation. In the latter case, the agency’s funding should be explicit in the budget, and its net borrowing requirement should be considered part of the government deficit.

In order to match the financing needs involved in local infrastructure, central government should consider giving local government controlled access to the capital market. Macroeconomic control and efficiency would gain from moving away from the present situation of indirect, uncontrolled access. A development bank could be a means of providing such access.


The World Bank.


Although recent data are unavailable, the 1986 figures show a much higher budget participation in infrastructure projects. Economywide, state appropriation constituted 14.5 percent of the investment funding, while transport projects were on average financed for 33 percent from the budget.

    Other Resources Citing This Publication