Delegating Which Responsibilities?

Which level of government provides a service is often determined as much by history and politics as by economic efficiency. On efficiency grounds, central authorities would provide services whose benefits spill over beyond the local area, such as defense, postal services, national highways, ports and civil aviation, environmental policies, and so on. Policies affecting macroeconomic stabilization and income distribution are also seen as central responsibilities. Subnational governments typically provide services that mainly benefit their local area, such as construction and maintenance of local roads, law enforcement, city maintenance, and such public utilities as water, sewerage, and various forms of energy distribution. (Alternatively, some of these services can be provided privately or contracted out to private providers)

Because of benefit spillovers, education and health are seen in some countries as national responsibilities. In others, they are provided subnationally—often to a standard set by the central government, and supported by grants that finance some or all of their provision, conditional on the level of the service. In most transition economies, however, practically no conditions have been attached to grants, and few standards have been set for the important responsibilities now increasingly being transferred to the subnational level.

How have the transition economies addressed expenditure assignment? Local governments have heavy spending responsibilities—accounting for some 30 percent of total public expenditure on average. Changing spending responsibilities, privatization, and the continued need for a robust social safety net suggest this share will likely rise further. Fiscal pressures have led many countries to shift important spending responsibilities to the subnational level, even where these are not appropriately "local responsibilities" by the above efficiency criteria.

In almost all transition economies, for example, social welfare spending has been shifted to the local level; major infrastructure projects, pensions and unemployment compensation, and some health and education spending have also been shifted (table 1). While local governments can sometimes administer such programs better (being "closer to the people") they cannot deliver them effectively if underfunded. Where adequate spending is critical to the transition, ensuring it is a national policy objective. The correct design of local revenues and transfers is thus crucial to achieving national social goals. There are also other problems:

•Murky spending responsibilities. In Bulgaria, Hungary and Poland, local governments have major, but vaguely defined, expenditure responsibilities in primary education, local transportation, and environment, as well as in housing and related services. The laws do not distinguish, for example, whether responsibility for primary education implies responsibility for capital spending on school buildings or just for spending on recurrent costs.

•Emerging service differentials. Russia and Ukraine have put the cart before the horse by explicitly assigning revenues to the subnational sector without formally assigning expenditures: subnational spending decisions are being driven by revenue availability instead of the other way around. This approach virtually guarantees the emergence of interregional differences in service provision, depending on whether the region has a rich—or poor—tax base. Indeed, such differentials are increasingly visible in both countries, especially in key social sectors such as health.

•Limited spending discretion. Bulgaria and Romania allow virtually no local expenditure autonomy, but they are not alone. In almost all transition countries, there are central mandates that define spending "norms" for schools and hospitals, require public services to be provided with user charges set below costs, or centrally define wage and pension increases. In Bulgaria, the Ministry of Health mandates expenditures of municipal hospitals down to the inventory of medicine.

•Divesting enterprises' social assets. Enterprise privatization will likely imply major new spending responsibilities for local governments in the transition economies. Enterprises built schools and kindergartens, constructed housing, and sometimes built roads, sewer lines, commercial infrastructure and social facilities. These expenditures (few will be privatized with the enterprise) will likely become the responsibility of local government. Few transition countries have yet estimated the fiscal costs for local governments of this divestiture of social assets and infrastructure.

Charge for Services Whenever Possible

Despite the rhetoric of autonomy in the new local self-government laws, subnational governments enjoy only limited fiscal discretion. Local tax bases as well as rates continue to be determined by the center, even for such minor levies as a dog tax. With few ways to raise revenues, hard-pressed local governments have run arrears, demanded increased transfers from the center, or attempted to borrow, potentially prejudicing stabilization. Some exploit the enterprises they own for revenue purposes, potentially thwarting privatization. Some have even become entrepreneurial (the Moscow city government contributed land as its equity stake in McDonald's).

User charges. Price policy was a distributional tool in the transition economies, although some would argue that it resulted in little equity, purchased at a high price in efficiency losses. Health services and education were free, as were transport, utilities, and many urban services. The "benefit approach" to local public finance suggests that whenever possible (i.e. where a beneficiary can be found), local services should be charged, and paid for, by those receiving the benefits. Not surprisingly, newly elected subnational governments in the transition economies find it difficult to increase user charges. Much greater use needs to be made of them, for financial and efficiency reasons; to avoid large and unpopular distributional shifts, some form of "social pricing" (lifeline pricing, vouchers, etc.), or coordination with other policies (e.g., social assistance, pensions), may be necessary.

Choosing Revenue Sources

Subnational taxes. Where user charges are impractical, services should be financed by local taxes. Local taxes are defined as those that fall on local residents and for which local authorities control the tax rate, define the tax base, and obtain the revenue. "Piggybacking" arrangements (local surcharges on central taxes) are effectively local taxes and can increase local revenues at a low administrative cost. Piggybacking also increases the efficiency of local spending, compared with the transition economies' "tax sharing" arrangements, which simply benefit local governments without them having any say, responsibility, or accountability for the tax levied.

Some taxes are less suited to subnational governments than others. In particular, to minimize distortions, a high degree of uniformity is desirable with respect to the corporate income tax and the VAT. Most taxes currently specifically assigned to subnational governments in transitional countries are nuisance taxes rather than robust own-revenue sources. In Russia, the revenue yield from twenty-one local taxes (including taxes on the resale of used computers, horse racing, the use of logos in advertising, and dog owners) is likely less than 0.5 percent of GDP, or 2 percent of subnational expenditures. In Hungary, where local taxes include the property tax, poll tax, a local business tax, and a bed tax in tourist areas, the situation is only slightly better. In addition, throughout the transition economies, effective local fiscal management is hampered by central mandates on both local tax rates and the tax base.

Property taxes. These taxes have some potential. Current yields are minimal (much of the housing stock is still owned by public enterprises or local governments themselves, and is not easily taxed) and even in OECD countries the property tax is not a major source of state finances. To improve its yield, up-to-date cadastres, and the many centrally mandated exemptions (such as Hungary's ten-year exemption for any dwelling with improvements—a bath, a new roof) will also need to be eliminated. Automobile taxation could also be explored: increased automobile ownership makes this a growing tax base that may also be distributionally and environmentally attractive.

Excise taxes and single-stage retail taxes. These taxes would seem to be good subnational taxes, as they are in market economies. In the transition economies studied, tobacco or alcohol excises are an important revenue source at the national level; however, major administrative reforms would be needed to convert them into subnational taxes. Excises are currently levied on producers who are located in only a few regions (Russia's taxpaying cigarette factories are located in only 21 of its 2,000 rayons). Excise revenues would benefit only these few localities. As for local retail sales taxes common in market economies, implementing them in addition to the VAT, while theoretically feasible (Canada's provinces do this), would be costly, complex, and probably inconsistent with the tax regimes of the European Union.

Tax sharing. In transition economies, tax sharing between central and subnational governments is very common and uniformly takes place on a so-called derivation basis, i.e., revenues flow back to the locality where they were collected. In Hungary, Poland, Russia, and Ukraine—and in all the countries of the FSU—some or all personal income tax (PIT) is shared (table 2). Russia also shares VAT, the corporate tax, and certain natural resource taxes with oblast governments. Tax sharing is a simple mechanism and guarantees some degree of revenue certainty. But it does little to enhance accountability or efficiency. And local governments are vulnerable to changes in the rules of the game. In Hungary, local governments initially received 100 percent of the PIT; as fiscal pressures on the center have increased, they now receive just 30 percent.

Derivation-based sharing means that resources are channeled to higher income areas where the tax bases and therefore revenue collections are largest. It is thus inherently counter-equalizing; the more resources accrue on a derivation basis, the more richer localities will benefit. This may be a problem where regional inequalities are serious, as in Romania or Russia, and where the intergovernmental system lacks other instruments such as transfers (see box) to address such imbalances. In deciding on the weight to be given to equalization, transition economies must balance the tradeoff with growth. In some countries, vast disparities between regions make equalization a political priority. In others, such as Russia, still in the process of nation-building, the need to maintain political unity may be greater than the need for equity: some of Russia's better off oblasts have threatened to opt out of the revenue sharing system if too much is shared with poorer regions.

Surcharges. Transition economies should give serious consideration to surcharges. Such piggybacking is common in many OECD countries, and would promote fiscal accountability and improve local fiscal discretion; subnational governments could also rely on the superior central administrative machinery for tax collection. The PIT lends itself especially well to piggybacking, since arguably, local residents, who would pay the surtax, benefit from local services. Surcharges on either the corporate tax or VAT would be undesirable. Instead of receiving a centrally mandated share of the PIT (as is now the case in most transition countries), subnational governments would surcharge the national PIT to receive "their share." Depending on the level of the local surcharge, the combined rate could be less or more than the previous central rate.

A final note: The decentralization of the socialist state and the ongoing reforms in subnational finance are of considerable importance. With local governments now responsible for key spending on infrastructure, education, health, housing, and the safety net, the strengthening of subnational governments is essential not only to support the evolving private sector and to provide the human and physical infrastructure for growth, but to ensure the social acceptability of the transition process. The intergovernmental systems being put in place will have a major impact on the lives of those undergoing one of the major upheavals of our time —the move from a command to a market economy. It is essential that they be designed to meet this challenge.