Abstract

The government sector is generally defined to encompass the authorities that are engaged in the pursuit of public purposes by providing nonmarket services and effecting income transfers financed by levies on other sectors in the economy. 2/ These functions are performed by all entities of the general government sector, which mainly comprise: (1) the central government; (2) state governments; and (3) local governments. Social security funds and departmental enterprises 3/ are also included in the definition of the general government and are treated as belonging to the level of government at which they operate. Government-owned or controlled financial institutions are classified as public financial institutions, however, rather than as part of general government.

The Role of Government and Fiscal Reform in an Economy in Transition 1/

1. Definition of the government and the public sector

The government sector is generally defined to encompass the authorities that are engaged in the pursuit of public purposes by providing nonmarket services and effecting income transfers financed by levies on other sectors in the economy. 2/ These functions are performed by all entities of the general government sector, which mainly comprise: (1) the central government; (2) state governments; and (3) local governments. Social security funds and departmental enterprises 3/ are also included in the definition of the general government and are treated as belonging to the level of government at which they operate. Government-owned or controlled financial institutions are classified as public financial institutions, however, rather than as part of general government.

In market economies, nonfinancial public enterprises that produce and sell goods and services are excluded from the government sector but included in the definition of the public sector. Such enterprises represent government-owned or public utilities and other nationalized or government established enterprises. In the context of centrally planned economies, these enterprises may represent the dominant part of the productive sector of the economy, intended to play a similar role in the supply of goods and services as private enterprises in a market economy. It may, therefore, not be of significant analytical benefit to use the definition of the public sector in these economies.

The definition of general government outlined above essentially applies to both centrally planned and market economies. A closer scrutiny reveals, however, significant differences between the two systems as regards the role of government, the institutional structure of the government sector, and the instruments employed in the interaction of government with the other sectors of the economy.

In market economies, with the private sector playing a leading role in promoting economic growth and the public sector designed to provide a stable and supportive environment, fiscal policy is focused on the achievement of macroeconomic stability, allocative efficiency, and distributional equity consistent with prevailing social preferences. In centrally planned economies, the government sector assumed to a far larger extent direct responsibility for the control of economic processes.

Such control was exercised in pursuit of a development strategy that involved a large scale mobilization of resources to promote rapid industrialization dominated by a heavy industrial sector. The government further implemented an ambitious program of resource transfers to achieve an egalitarian distribution of income and consumption in society, consistent with the priorities established by the political leadership.

The enhanced role of government in controlling resource allocation was reflected in a pervasive presence of government in the economy through specific institutions, such as central planning agencies, boards for the control of prices, wages, and foreign trade, and branch ministries administering a large state enterprise sector. Institutions that provide legal, regulatory, and administrative services to support the efficient operation of markets were essentially absent. The specific emphasis on the allocative and redistributive responsibilities of government and the priorities underlying the development strategy pursued under socialist central planning led to a large scale intermediation of resources through the budget, raising the share of government expenditures in GDP by 10-15 percent above the typical level in major market economies. These factors also determined the structure of government expenditures and revenues. The composition of expenditures was heavily weighted in favor of subsidies and transfers. The bulk of revenues stemmed from receipts from profits, payroll taxes, and turnover taxes while revenues from individual income taxes remained low by international standards.

The tax structure was marked by a considerable disregard for efficiency. High profit tax receipts were facilitated by an inflation of enterprise residual income through unrealistic amortization rules, very low interest rates paid to enterprises, and, especially, tightly controlled wages. Wages from the socialized sector were divorced from the marginal productivity of labor, as they were supplemented at the household level by extensive social benefits in kind, intermediated through the government sector. This practice drove a significant wedge between factor incomes and disposable incomes and thus between the social and private return on effort, with adverse consequences for economic efficiency.

In centrally planned economies, it was largely the prerogative of the plan to carry out the role played by the price system in market economies. Taxes, and especially those on profits, represented a primary instrument to enforce the plan. As a result, profit taxes were highly differentiated and the tax parameters were frequently adjusted in response to the diverse and changing priorities of the plan. Such taxes were in practice often quite arbitrary and the result of a bargaining process between enterprises and the fiscal authorities, typically effecting a redistribution of resources from efficient to inefficient enterprises.

The scope for income taxation was limited by the low level of socialized sector wages and the consideration that the primary distribution of such labor income was adequate and required no further correction. Schedular income taxes were mainly used to impose prohibitive tax rates on incomes of artists and self-employed professionals and craftsmen in order to bring such incomes closer in line with those from the socialized sector and to contain private sector economic activities.

The above differences between centrally planned and market economies manifest themselves to some degree also in the various stages of transition in which formerly centrally planned economies find themselves in the process of transformation to market economies. Thus, they may have an important bearing on the size of the government sector, the forecasting of budget aggregates, and the design and conduct of fiscal policy in the transition period.

2. Fiscal reforms in Hungary and fiscal developments in 1989

a. Hungarian fiscal reforms

Economic reforms undertaken in Hungary since the early 1950s also encompassed various efforts to modify the role of fiscal policy. During the initial period of traditional central planning the budget was an instrument to enforce plan implementation through massive ad hoc transfers of resources among state enterprises and a large scale redistribution of incomes among households.

The most comprehensive attempt at market-oriented reform in a centrally planned economy was initiated with the introduction of the New Economic Mechanism (NEM) in Hungary in 1968. The NEM was aimed at replacing direct by indirect control over socialized enterprises with limited financial self management. This method was intended to achieve the priorities set out in the national economic plans with greater efficiency. The fiscal reform measures mainly involved the adoption of nominally parametric fiscal instruments to redistribute profits in pursuit of production, trade, income, and price policy objectives as well as a streamlining of the complex system of price subsidies and turnover taxes; more than 2,500 implicit rates were replaced by about 1,000 explicit turnover tax rates.

A significant step was also taken to rationalize the fiscal regime with respect to international trade, in conjunction with a reform of the exchange system. The multitude of exchange rates, which had been highly differentiated through a complex system of price equalization taxes and subsidies, was transformed into a system of unified exchange rates vis-à-vis the nonruble and the ruble areas, respectively. For trade with the nonruble area, the former complex fiscal regime was replaced by an import tariff system with a wide dispersion of rates. Since foreign trade remained subject to considerable direct control, however, the import duties became primarily an instrument of bargaining for trade concessions with nonruble trading partners. With respect to the ruble area, a system of nonparametric “producers’ differential turnover taxes” was applied to bridge the gap between the rather arbitrary prices applied in ruble trade transactions—converted at the unified forint/ruble exchange rate—and Hungarian domestic prices that were linked to a substantial degree to world market prices. 1/

In the course of the 1970s, especially following the sharp rise of petroleum and other raw material prices beginning in 1973, fiscal intervention again became more pervasive and discretionary. Various new production taxes and tax preferences were introduced in an effort to insulate the domestic economy from a full pass through of price developments in the world markets. The complexity of the tax system was exacerbated by a strong differentiation of the fiscal measures among branches and even individual enterprises.

A renewed effort to simplify that tax system was undertaken in conjunction with a comprehensive price reform introduced in 1980. The latter reforms initiated a shift in the burden of taxation from the factors of production to profits, as production taxes were lowered and the 5 percent charge on the state owned capital of enterprises was rescinded, while profit tax rates were raised.

The gains in transparency and predictability of fiscal instruments were eroded again, however, in the following period, by frequent tax adjustments and a confiscatory withdrawal of past retained earnings from enterprises. The latter revenues were allocated through the budget and extra budgetary funds to supporting weak enterprises in crisis industries and in areas of economic decline. The complexity of the system was further enhanced by the introduction of new taxes on the wage bill and the profits of enterprises and an anti-cyclical investment tax in 1985. These and other steps also brought about a reversal of the earlier shift of the burden of taxation from factors of production to profits.

Although a significant step to simplify the structure of turnover taxes was also intended by the authorities in the framework of the 1985 tax reforms, the effort actually implemented fell considerably short of original plans. Thus, the fiscal system in place in the mid-1980s continued to be marked by multiple taxation with multiple aims, violating the general principle of neutrality, transparency, stability, and equity of taxation.

A broad overhaul of the system was initiated subsequently, to be implemented during 1988-89. In a first phase, a value added tax—called a general turnover tax (GTT)—and a personal income tax (PIT) were adopted at the beginning of 1988. The contemporaneous elimination or modification of other taxes, including the municipal contribution on profits, the net worth tax, the investment tax, and wage taxes paid by socialized enterprises, affected about 90 percent of general government revenue, leaving essentially unchanged only taxes on international trade. The GTT replaced the multitude of existing turnover taxes. The tax base of the GTT was defined narrowly, excluding food, basic services, construction materials, personal imports, and other items and the GTT was fully rebated on exports. As a consequence, rates were set high by international standards. They encompassed a standard rate of 25 percent, which yielded 95 percent of GTT revenue, and a reduced rate of 15 percent; a zero rate applied to goods accounting for more than 40 percent of consumer expenditures; a rebate on investment expenditures of socialized enterprises was to be phased in gradually during 1988–92.

The PIT superseded a complex system of schedular income taxes with high statutory marginal rates on incomes from self-employment and intellectual and cultural activities. For the first time, a taxation of wages earned by individuals in the socialized sector and of interest income was established. With the introduction of the PIT, socialized sector wages were grossed-up to preserve the net income of employees. The tax burden on incomes from private economic activities became heavier as the new income tax was applied on aggregate income at the applicable marginal rate, instead of a separate taxation of incomes from different sources under the previous schedular income tax regime and an additional 25 percent entrepreneurial tax was imposed on the profits of private undertakings. The initial PIT rates ranged from 20 percent to 60 percent. About 20 percent of potential taxpayers were expected to fall below the threshold of a general tax-free allowance, equivalent to about two thirds of average earnings in the socialized sector. An additional allowance was granted to employees as a lump sum cost deduction in lieu of itemized expense accounting. Exemptions were applicable to incomes from small scale farming and intellectual activities and most social benefits; interest income on domestic currency financial assets was subjected to a separate flat rate withholding tax of 20 percent; and generous provisions were granted for the deductibility of savings for housing purposes and interest on housing loans. As a result, the PIT was levied on only about one half of household incomes. In 1989, the top marginal tax rate was lowered to 56 percent while brackets were widened and their number reduced to 8; the tax base was also narrowed further.

With the adoption of the PIT, social security contribution rates of employees were changed from a progressive schedule of 3–15 percent to a unified rate of 10 percent and employers’ contribution rates outside the government sector were revised to 43 percent. Moreover, the new contribution rates were applied to the grossed up wages of employees in the socialized sector including the incidence of the PIT. In 1989, employers’ contribution rates in the government sector were also raised to the unified rate of 43 percent (from a previous rate of 10 percent) and social security transactions were moved from the state budget to a separate Social Insurance Fund (SIF). At the same time, a formal unemployment compensation fund was introduced and the subsidization of housing finance was moved from the banking sector to a new housing fund, whose losses are covered by transfers from the state budget. 1/

The second major step toward a more efficient tax system was taken with the rationalization of profit taxation through the adoption of the enterprise profit tax (EPT) at the beginning of 1989. The EPT substituted the previous highly differentiated profit tax as well as taxes levied on incomes of private enterprises. At the same time, several production taxes in metallurgy, the chemical industry, transportation, and public services were also removed. The general EPT rate was initially set at 50 percent, with a temporary surcharge of 4 percent and a reduced rate of 40 percent levied on the first Ft 3 million tranche of taxable profit to favor small businesses. The marginal effective tax rate of the EPT was high by international standards, partly because the tax base was severely narrowed down by several tax preferences maintained for farming, food processing, various services, cultural activities, and especially for joint ventures with foreign partners.

b. Fiscal developments in 1989

The intermediation of resources through the budget in Hungary—at over 60 percent of GDP—has been very high by the standards of major European market economies (see Table 1). In line with the discussion of the specific role of government under socialist central planning in the preceding section, the large share of government in GDP has been chiefly attributable to high transfer payments for subsidies to enterprises and income maintenance programs for households. Together these have accounted for about one third of GDP and one half of government expenditures (Table 2). As the need to restore macroecomomic stability has placed a tight constraint on the allowable budget deficit, a significant effort had to be undertaken by the authorities to mobilize revenue through taxation. These conditions, together with the narrow base of the new taxes, played an important role in the decision to set the initial rates of the GTT and the PIT in 1988 higher than originally contemplated.

Table 1.

Hungary: General Government Accounts

(In percent of GDP)

article image
Sources: IMF, Institute data base and G. Kopits, Fiscal Reform in European Economies in Transition, WP/91/43, April 1991; and Fund staff estimates.
Table 2.

Hungary: General Government Operations

(In percent of GDP)

article image
Source: IMF Institute data base.

Including profit transfers from government-owned financial and nonfinancial enterprises.

Including capital revenue.

Including a small amount of lending less repayments.

Fiscal policy for 1989 was geared to reducing the scope of government intervention, while contributing further to a containment of external imbalances. The authorities launched a medium-term program to curtail transfers to the enterprise and household sectors and planned to reduce government investment and defense outlays, measures that were expected to lower the share of government expenditures in GDP by over 5 percentage points. As the general government overall balance was targeted to record a slight surplus (0.4 percent of GDP) and nontax revenues were expected to weaken with sluggish economic activity, the authorities saw only a moderate scope—amounting to about 3 percent of GDP—for an easing of the tax burden on the economy. Revenue losses were anticipated from a lowering of the PIT rates by an average of 3 percentage points, the scheduled widening of the GTT rebate for investment goods, and a decline in receipts from the differential producers’ turnover tax, mainly attributable to rising ruble import prices. A marked increase was planned, however, in receipts from the taxation of enterprise profits, given the high initial tax rates applied under the newly adopted EPT and the 4 percent surtax added for 1989.

The outturn of general government operations was less satisfactory than planned. Targeted expenditures were overshot—despite a further cut of defense outlays at mid-year—owing in part to the effect of higher than anticipated inflation on spending on goods and services and—through firming interest rates—on interest payments on government debt. A marked shift took place in the structure of central government subsidies, as the cuts in consumer and producer subsidies as well as support to agriculture and CMEA exports were largely offset by a sharp rise in housing loan subsidies, which had been assumed by the budget from the banking system (Table 3).

Table 3.

Hungary: General Government Current Subsidies and Transfers

(In percent of GDP)

article image
Source: IMF Institute data base.

The expenditure overruns in 1989 were not fully matched by stronger than planned revenues. A surge was recorded in nontax revenues, but extra revenues collected by certain budgetary institutions were used for higher capital expenditures, conflicting with the intent of the government to scale down the scope of fiscal intermediation. The collection of tax revenues was markedly lower than planned, due to a significant shortfall of profit tax receipts. EPT revenue was 40 percent lower than expected, partly because profits were squeezed by larger than anticipated wage increases, but more importantly because tax reliefs and refunds granted to state companies and the accumulation of arrears to the government by enterprises in financial distress had been significantly underestimated.

A further shortfall occurred in the producers’ differential turnover tax as the volume of ruble imports fell markedly. The latter shortfall was not mitigated by higher revenue that could have been obtained from an unanticipated surge in private consumption and convertible currency imports in 1989, as buoyant expenditures largely reflected personal imports of consumer durables, which were both exempt from the GTT and subject to a relaxed regime of customs duties. Thus, despite an increase of petroleum excises at mid-year for budgetary considerations, revenue from goods and services also remained below target.

The overall shortfall of tax revenue was tempered, however, by excess receipts from the PIT and social security contributions stemming from sharp overruns of wage increases in socialized enterprises following a liberalization of the wage system and an easing of credit conditions.

Instead of the anticipated small surplus, the overall position of the general government closed with a deficit of over 1 percent of GDP in 1989. In addition, the government had to mobilize resources to finance a continued redemption of official debt to the USSR, equivalent to just over 0.5 percent of GDP in 1989.

In meeting these requirements, the Treasury encountered serious financing difficulties. The placement of government bonds and Treasury bills was hampered by an increasing preference of the public for shares and consumer durables, in light of rising inflationary expectations. In effect, nonbank investors reduced their holdings of government paper during the year. As a result, the government resorted to renewed central bank financing on a significant scale, equivalent to about 3 percent of GDP. This amounted essentially to an onlending of funds borrowed in the international financial markets by the National Bank of Hungary, which has been chiefly responsible for official borrowing from the convertible currency area.

Forecasting Government Transactions

Government operations have a significant impact on the economy, reflecting the size and structure of revenues and expenditures as well as the magnitude and financing of the budget balance. Domestic absorption is affected directly through government expenditures on goods and services and indirectly through the effects of revenues and expenditures on private spending. 1/ Owing to the central role of the budget, fiscal developments have in many instances been the chief cause of major internal and external economic imbalances. Consequently, the assessment and forecasting of government transactions represent key elements in the design of stabilization and adjustment programs.

The following section reviews the general forecasting methods for tax revenues, nontax revenues, government expenditures, and the financing of the budget balance and discusses issues related to the practical application of the pertinent techniques in the case of Hungary and other economies in transition.

1. General methodology of forecasting government transactions

a. Tax revenue

Constraints on data availability may sometimes lead to forecasting tax revenues by an extrapolation of recent trends. Such estimates are not very robust, however, because they assume that revenues will merely reproduce past developments regardless of the evolution of the tax base or the tax system. More reliable forecasts require, therefore, methods that link tax revenues to their bases. In principle, such revenue forecasts should allow for a feedback between taxes and their bases, because changes in tax parameters will affect behaviors that influence the aggregates on which the taxes are levied. Taking into account such linkages requires, however, the construction of a comprehensive econometric model, for which, in may instances, adequate data and other resources may not be available. A more practical approach, called the partial equilibrium approach, is to project revenues on the basis of functions for different types of taxes, assuming given projections of the tax bases.

The development of tax functions requires decisions on the degree of disaggregation and the choice of appropriate proxy tax bases. The breakdown should attempt to distinguish between the major tax categories. Also, a group of taxes should be disaggregated if their bases are quite dissimilar. However, very detailed breakdowns of revenue will increase the likelihood of fluctuations of an increasingly random character. A balance must be struck, therefore, between these conflicting considerations.

After the appropriate revenue classification is selected, consideration needs to be given to the approximation of the tax bases. In principle, each category could be related to the legal base as defined by the tax law. Data on the legal base are, however, often unavailable in a usable form. The revenue classification may also have grouped in a single category taxes subject to different rates or applied to different legal bases. Consequently, an alternative to the legal base is often required. Since major tax categories often relate to the bases that cover large parts of overall economic activity, alternative or proxy bases may be found among the variables comprising the national accounts and the balance of payments. The selection of the most appropriate proxy base will in part depend on whether projections for such variables are generated in other parts of the financial programming exercise. If the latter aggregates are determined in the context of a financial program, a sequential procedure can be employed whereby the tax revenue implications of different base solutions are examined in order to ensure a consistent interaction among macroeconomic variables.

An increase in revenue over a particular period can reflect both automatic and discretionary effects. The former represents the rise in revenue that stems from the growth in the tax base. Discretionary changes in revenue result from modifications of the tax system, including a revision of existing tax rates, a change in the coverage of a tax, or the introduction of a new tax. Analysis and forecasting of revenue developments generally require a distinction between these automatic and discretionary components. A simple method to eliminate the effects of discretionary measures on the growth rate of actual revenues—the proportional adjustment method—is briefly reviewed in Box 1.

The distinction between the automatic and discretionary effects leads to two different concepts for summarizing the response of a tax to developments in the economy. Buoyancy is defined as the ratio of the percentage change in actual tax collections over the data period to the percentage change in GDP. In order to concentrate on just the automatic effects on revenue, however, the alternative concept of elasticity has been developed. Elasticity compares the growth in revenue with the rise in GDP on the assumption that the tax system of a particular year (usually the most recent) had prevailed throughout the period (see Box 2). The interpretation of the size of the elasticity is similar to that for buoyancy with the important difference that it reflects only the automatic response of revenue. Thus, for an observation period in which discretionary changes in the tax system would have yielded an increase in revenue collections, the tax elasticity would be exceeded by the tax buoyancy, which would overstate the automatic response of revenue to a given change in GDP.

Following an adjustment of actual revenue series for the effects of discretionary measures, the major step in the procedure to forecast tax revenues would consist of specifying functional relationships between the adjusted revenue series and the assumed tax base, estimating a set of parameters that characterize the current tax structure, and applying the estimated equations to forecasting revenues for given projections of the tax bases. 1/ If adequate information is not available, more simple alternative methods can be used whose results would be less reliable than those of the above forecasting procedure but still superior to a simple extrapolation of trends. Revenue projections would be prepared by relating tax categories to pertinent macroeconomic variables and making assumptions about factors that might affect the revenue ratios in the forecasting period, including the effects of changes in tax rates, schedules, exemptions, tax administration, and tax avoidance.

Proportional Adjustment Method

Based of information on actual tax revenue (T) and the estimated revenue effect of discretionary measures (D)—the proportional adjustment method allows the derivation of an adjusted revenue (AT) series that approximates the amounts of taxes that would have been paid in prior years if the tax system of the base year (t) had been in force throughout the period. A key assumption is that a given discretionary tax change would have the same proportionate effect on tax receipts in all years. The adjustment procedure can be illustrated by the following formulae:

(1)ATt=Tt
  • since, by definition, no revenue adjustment applies for the base year

(2)ATt1=Tt1[TtTtDt]

Substituting (1) into (2):

(3)ATt1=Tt1[ATtTtDt]

In general, for any period t-n equation (3) can be written as:

(4)ATtn=Ttn[ATtn+1Ttn+1Dtn+1]

Calculation of Buoyancy and Elasticity

Using the notation introduced in Box 1, buoyancy and elasticity are defined by the following formulas;

Buoyancy=TtTt11GDPtGDPt11Elasticity=ATtATt11GDPtGDPt11

If no information is available on the effect of discretionary tax changes, and the number of such changes is small relative to the available number of revenue observations, tax elasticity can be estimated with the help of dummy variables capturing the effect of discretionary changes in a tax equation. For example, for a given discretionary change:

lnTt=a+(b+cDUMt)InGDPt

Where DUMt = 0 prior to the discretionary change; and

  • DUMt = 1 after the discretionary change.

The coefficient b measures tax elasticity, excluding the effect of the discretionary change in revenue collections.

b. Nontax revenue

Nontax revenue usually includes receipts from property income, fees and charges, non-industrial sales, the operating surplus of departmental enterprises, fines, forfeits, and private donations. Systematic relationships may be found for certain components of nontax revenue, but many items often exhibit substantial volatility. Consequently, projections of nontax revenue are often made on the basis of judgement, or simple ad hoc procedures. Sometimes forecasts are prepared for total nontax revenue by applying last year’s ratio of nontax revenue to GDP to the projected value of GDP. However, rather than relying on a mechanistic projection of the aggregate, it might be feasible to allow for expected changes in relevant factors through judgmental adjustments in individual components.

c. Expenditure

Unlike government revenues, there is much less scope for forecasting the level of government expenditures through reliance on causal economic relations. This follows from the essentially political nature of the process through which decisions on public expenditure are made, which means that changes in government expenditure are in large part discretionary. Consideration may, however, be given to endogenous determination of some of the major categories within an economic classification of expenditure. Interest payments will reflect interest rates and the size of the government debt while transfers may be related to the level of economic activity. Wages and salaries are likely to be affected by government incomes policy, as well as prices and pressure in the labor market. Budget estimates are usually available, indicating the authorities’ own expectations for major expenditure categories. Analysis of past relations between estimates and outcomes may indicate errors in such estimates.

d. Financing of the budget balance

Forecasts of financing focus on three primary items: external financing; domestic nonbank borrowing; and domestic borrowing from the banking system.

Regarding external financing projections, the foreign borrowing plans and amortization schedules of finance ministries should be cross-checked with foreign lending projections and amortization schedules of lending agencies. If the level or structure of foreign debt is a problem, the projections should be cast against a broader exercise specifically addressed to the issue of external debt management. Budgetary projections for external finance should be consistent with official capital movements in the balance of payments.

For domestic nonbank borrowing, a forecast is required of the net amount of debt instruments the government will be able to place with domestic financial institutions and other domestic lenders. In this context, consideration should be given to the likely impact on the costs and availability of funds to finance non-government activities. Finally, if appropriate, the required net financing by the banking system is obtained as a residual and the result is checked against the original objectives with respect to monetary expansion.

2. Application to Hungary and other economies in transition

a. Revenue forecasting

The task of projecting government revenues for the reference period in the case of Hungary raises difficulties that are likely to arise also in other economies in transition toward a market economy, for three main reasons:

(1) For most tax categories, functional relationships between revenue outturns and proxy tax bases for the period of central planning may have to be viewed with great prudence. Formal relationships that may be estimated may conceal more a pattern of nonparametric tax regimes or parametric taxes subject to frequent and haphazard changes, than a stable behavioral and institutional environment on which reliable tax functions should be predicated.

(2) The tax regimes applied during the period of central planning were so different from the tax systems that are being introduced in the period of transition that the latter reforms amount to a replacement of virtually all former tax categories by new taxes more typical for market economies. As a consequence, even reasonably stable historical relationships between tax revenues and proxy tax bases that might be obtained for the period of central planning may provide little guidance for the projection of revenues in the transition period.

(3) Even in the possible absence of deliberate fiscal policy measures, structural reforms introduced in other areas of the economy may have significant repercussions on the revenue outturn during the transition period by affecting the revenue base or the parameters defining the structure of the tax system.

(1) Taxes on international trade

A review of the development of the tax system in Hungary suggests that the estimation of a functional relationship between revenues and proxy tax bases for forecasting purposes may be worthwhile only in the case of import duties. Since adequate information is not available on the record and the effects of discretionary adjustments of import tariffs, the functions will be specified for the relationship between actual import duties collected (TM) and the proxy tax base represented by total merchandise imports (c.i.f.) from the nonruble area (NRM). 1/ The parameters of the equations are estimated by the technique of ordinary least squares. Given the limited number of observations, the regression estimates have to be evaluated with caution. The equations are shown in Box 3 with t ratios in parenthesis.

The other major tax that was primarily levied on international trade transactions—the producers’ differential turnover tax—was nonparametric and was mainly a function of the volume and composition of ruble imports and the ruble import prices negotiated in bilateral trade agreements with CMEA partners, the exchange rate of the forint vis-à-vis the ruble, and the domestic prices of the traded products. Given that these factors are subject to discretionary government decisions or the outcome of annual official trade negotiations, it would be difficult to project revenue from the producers’ differential turnover tax without access to adequate background information from official sources. Moreover, revenue from this tax—which amounted to about 4 percent of GDP and 8 percent of total tax revenue in 1989—was highly dependent on the progress of price and trade reforms in Hungary and other CMEA countries; virtually the entire revenue of the tax would be lost as a consequence of a transition to using world market prices in trade among CMEA members.

Regression Equation for Import Duties Period: 1982–89

TMt=7.464+0.169*NRMt(1.92)(10.24)
R¯2=0.94D.W.=1.72
LnTMt=2.920(4.08)+1.169(8.86)  Ln NRMt
R¯2=0.91D.W.=1.59

Where:

  • TM = actual import duties collected.

  • NRM = total merchandise imports (c.i.f.) from the nonruble area, expressed in billions of forint.

(2) Income, payroll, and turnover taxes

In the absence of tax functions and adequate microeconomic information on tax bases and structural parameters of the new taxes, revenue projections can be based on information on tax buoyancies or revenue ratios to proxy tax bases that can be derived from data observed since the inception of the tax reform. However, buoyancy cannot be measured for the EPT—which was introduced only in 1989. Furthermore, buoyancies are difficult to adjust appropriately for the effects of nonrecurrent changes in the structure of the GTT and PIT as well as in the progress of addressing the initial inadequacies of tax administration and of containing tax evasion between the first and the second year of the introduction of the new taxes. Therefore, ratios of revenues to proxy tax bases in the year preceding the reference period are proposed as a guidance for revenue projections

Some available information on the relationship of potential proxy tax bases to aggregates that will be projected in other parts of the financial programming exercise and on parameters describing the structure of the taxes in 1989 is provided in Tables 47. The material is intended to facilitate the task of selecting appropriate proxy tax bases and incorporating assumptions concerning the specification of fiscal policy in the reference period into the tax revenue forecasting exercise.

Table 4.

Hungary: Household Income and the PIT

article image
Sources: Central Statistical Office; Ministry of Finance; and Fund staff estimates.

Fringe benefits, social benefits, and other deductions and allowances.

Subject to a final 20 percent withholding tax.

Table 5.

Hungary: Enterprise Profits and Profit Taxation

article image
Sources: Central Statistical Office; Ministry of Finance; and Fund staff estimates.

Including profit transfers to the budget from government owned financial and nonfinancial enterprises.

Table 6.

Hungary: Household Income and Social Security Contributions

article image
Sources: Central Statistical Office; Ministry of Finance; and Fund staff estimates.

Includes income from self employment.

Excluding contributions from within the government sector.

Table 7.

Hungary: Private Consumption and Turnover Taxes

article image
Sources: Central Statistical Office; Ministry of Finance; and Fund staff estimates.

Exempt from turnover taxes.

Consumer expenditures by residents and nonresidents in Hungary.

The GTT rebate on investment goods was scheduled to be introduced in annual steps of 20 percent during 1988–92.

(3) Other taxes

For the projection of the remaining taxes—property taxes of the local authorities, other taxes on goods and services, and miscellaneous other taxes—it is quite difficult to obtain information on plausible proxy tax bases. It may be appropriate, therefore, to project these tax revenues in the reference year on the basis of their historical ratios to GDP.

(4) Nontax revenue

The usual difficulties in forecasting nontax revenue are exacerbated in the case of Hungary by inaccuracies of the historical data that stem from a poor monitoring of the transactions of central budgetary institutions. Data limitations have interfered with a satisfactory consolidation of these transactions in the central government accounts and may have contributed to the large fluctuations of the series over time. As a result, official projections of nontax revenue have also exhibited considerable forecasting errors.

b. Expenditure forecasting

The difficulties that generally arise in forecasting government outlays are intensified in Hungary and other transition economies by the significant expenditure effects of ongoing systemic reforms. As a consequence, forecasts of government expenditures will be subject to a particularly large degree of uncertainty.

(1) Expenditures on goods and services

Government wages and other operating expenditures are likely to be influenced in a transition economy by institutional reforms aimed at replacing government institutions that had been involved in the direct control of the economy by a government administration that enforces new procedures to underpin the efficient operation of a market economy. Given the numerous indispensable new tasks that government did not address under central planning, while it maintained a highly pervasive but meanwhile redundant presence in the economy and society otherwise, it is difficult to assess a priori the net effect of these reforms on current government outlays. Nominal current expenditures are liable to rise, however, as a consequence of tax, wage, and price reforms. Forecasts have to take into account the effects of wage increases through grossing up when an income tax is introduced and wage controls are relaxed and of price rises that are likely to be associated with a lifting of administrative controls and the adoption of a modern turnover tax system.

(2) Interest payments

Most of the outstanding government debt of Hungary is owed to the central bank or the USSR at fixed nonmarket interest rates. Such rates may be adjusted retroactively, e.g., in the context of banking and financial market reforms. Thus, adequate information on the servicing of outstanding government debt can only be obtained from official sources.

Interest payments on new debt depend on more flexible market oriented interest rates and new borrowing. The iterative process involved in the projection of such interest payments should be noted, as, ceteris paribus, they influence the size of the government deficit which determines the level of new debt that underlies the interest payments.

(3) Subsidies and transfers

In a market economy, subsidy and transfer payments may represent expenditure categories that can plausibly be considered to be endogenously determined, based on established schemes and facilities. It follows from the earlier discussion of the specific role of these expenditures in centrally planned economies, that a marked reduction of the scope of such outlays is a key objective of the restructuring of government and redefinition of the role of the budget during the transition to a market economy. The pace of the moderation of such expenditures depends on the progress of other reforms—notably price reform, the restructuring and privatization of state enterprises, and social security reform. Moveover, as illustrated by Hungary’s experience in 1989, partial setbacks may be encountered as support schemes carried out through quasi-fiscal operations outside the government sector in the period of central planning are brought under the purview of the budget, resulting in a partial temporary relapse of transfer payments. The projection of subsidy and transfer payments is, therefore, contingent on adequate information concerning important elements of the Government’s structural reform program, and highly sensitive to the authorities’ attitude toward the pace and consistent implementation of the reforms.

(4) Capital expenditure

Compared with current expenditures, capital outlays appear to be less predetermined by other important actions associated with the transition to a market economy. A likely decline in capital transfers to a shrinking state enterprise sector may also lend greater flexibility to the projection of capital spending. However, forecasts will have to take account of compelling needs to remedy the neglect of important areas of the infrastructure under socialist central planning. This places limits on the reduction of capital outlays that should not be exceeded lest the restoration of the growth potential of the economy would be jeopardized.

c. Financing of the budget balance

A specific concern regarding forecasts of the financing of the budget balance in a transition economy is related to the typical weakness of capital markets that limits the scope of domestic nonmonetary financing. Forecasters will need to take stock of the financial instruments at the disposal of the authorities in a particular phase of the transition and assess the availability of potential captive financial resources that may be directed by the authorities to investment in government securities.

Exercises and Issues for Discussion

1. Exercises

a. Forecast total general government revenue for 1990 in the format used in Table 8. In preparing the forecast, draw on the following background information:

  • The forecast of nominal GDP at market prices from the output and price projections;

  • The forecast of the value of nonruble imports from the foreign trade/balance of payments projections;

  • The following ratios derived from official macroeconomic forecasts for 1990:

article image
Table 8.

Hungary: General Government Tax Revenue

(In billions of forint)

article image
Source: IMF Institute data base.

b. Calculate revenue buoyancies for the main tax categories.

c. Prepare a forecast of the overall general government balance in 1990 using the format of Table 9. The option is provided to draw on the following information for 1990, recognizing that deviations from projections have regularly occurred in the past and that the figures need to be consistent with the overall macroeconomic scenario.

article image
Table 9.

Hungary: Operations of the General Government

(In billions of forint)

article image
Source: IMF Institute data base.

Including profit transfers from government-owned financial and nonfinancial enterprises.

Including capital revenue.

Including a small amount of lending less repayments.

d. Provide an assessment of the likely resort of the Government to external and domestic nonmonetary financing of the budget balance. The former should be consistent with the projections of capital flows incorporated in the balance of payments workshop. Derive bank financing as a residual. The net repayment of nonconvertible currency debt by the Government in 1990 was officially planned at Ft. 8.6 billion.

2. Issues for discussion

a. Review the likely effect of the following factors on the development of the main components of revenues and expenditures in 1990, commenting on the risks they represent for the realization of the revenue and outlay projections:

  • Fall in real domestic output;

  • Accelerating inflation;

  • Pickup of consumer expenditures;

  • Growth of nonruble imports;

  • Depreciation of the exchange rate;

  • Decline of ruble imports;

  • Price and wage reform in Hungary; and

  • Restructuring and privatization of state enterprises.

b. Discuss the way the tax buoyancies in 1989 are likely to deviate from underlying tax elasticities for the same tax categories and how the buoyancies are likely to change in 1990.

c. Comment on the considerations that determine if the forecast government accounts are consistent with the overall macroeconomic projections.

d. Comment on changes in the structure of revenues and expenditures that appear desirable in the medium term.

1/

This chapter draws on the IBRD Country Economic Memorandum, “Hungary: Reform and Decentralization of the Public Sector,” Annex II, “Fiscal Structure and Developments in Hungary,” forthcoming.

2/

See IMF, Manual on Government Finance Statistics, p. 7.

3/

Departmental enterprises perform specific services for the government sector; they encompass printing and publishing, restaurants in public buildings, and dwellings for government employees.

1/

In addition to imports from the ruble area, the producers’ differential turnover tax was also levied on a number of domestic products, notably energy, whose domestic prices were higher than production costs.

1/

The transfer of housing loan subsidies to the budget removed a disturbing quasi-fiscal function from the Savings Bank, which had represented a formidable obstacle in the way of efforts to complete the reform of the banking system (see Chapter VII).

1/

Government revenues and expenditures also affect economic growth through their influence on the mobilization of savings and the allocation of capital and labor.

1/

In practice, in the framework of the design of Fund supported stabilization and adjustment programs, such tax estimates will generally be required for the preparation of alternative revenue projections, e.g., for an assessment of official forecasts or alternative specifications of the fiscal policy stance. Original revenue projections would usually be elaborated by the fiscal authorities based on detailed information on tax bases and the parameters of individual tax categories.

1/

Expressed in billions of forint (Ft). Nonruble imports in Ft terms are obtained by converting nonruble imports in U.S. dollar terms at the annual average Ft/US$ exchange rate.