Journal Issue

Republic of Poland: Selected Issues

International Monetary Fund
Published Date:
May 2000
  • ShareShare
Show Summary Details

III. Public Expenditure Reform in Poland1

A. Introduction

1. On the road to EU accession, the Polish government is determined to continue the process of fiscal consolidation. At the same time, the authorities are committed to meet their budget targets without an increase in revenues. With grants from the European Union (EU) expected to rise and an anticipated increase in indirect taxes in line with EU countries, this would provide scope for a reduction in direct taxes, as part of a broader strategy to strengthen the role of the private sector. Whether these goals can be met simultaneously, depends ultimately on the government’s ability to control public spending.

2. This chapter discusses the need and scope for public expenditure reform in Poland, in the context of both Poland’s heavy reform agenda and the devolution of expenditure decisions to subnational governments. As a starting point, medium-term projections of total expenditure and its main components are derived in Section B in the absence of discretionary spending measures, taking into account the spending constraints associated with planned structural reforms. These projections serve as a first indication of the magnitude of the spending restraint required to achieve revenue and deficit objectives. Building on these results, Section C explores the scope for expenditure adjustments. Comparisons of the expenditure structure in Poland with other European countries are used to provide tentative indications of areas where adjustments may be warranted and feasible. Section D discusses specific problems of implementing public expenditure reforms in Poland’s decentralized fiscal system.

B. Medium-Term Expenditure Projections and Constraints

3. The first step in the development of a medium-term fiscal strategy is the assessment of future expenditure obligations in the absence of discretionary measures. In the case of Poland, this requires not only projections for “normal” recurrent expenditure obligations, but has to take into account the specific medium-term cost associated with its new NATO membership, prospective EU accession, and implementation of a large structural reform agenda in areas such as coal, steel, and agriculture. The following provides tentative projections of total general government expenditure and its components over the period 2000-05, based on indicated commitments for 2000 and taking into account, where appropriate and feasible, the costs related to planned reforms.

4. Projections of medium-term expenditure developments require assumptions in two areas: (i) the general macroeconomic environment, including real GDP growth, inflation, and the labor market; and (ii) the direct factors of government spending, such as public sector employment and investment decisions. The following analysis focuses on the latter, whereas the macroeconomic environment, for the most part, is assumed to be exogenous.2 Furthermore, the government expenditure projections reflect a “no-policy change scenario”. In this respect, they are based on a number of arbitrary assumptions and should not be interpreted as a reflection of current policy intentions. Instead, they serve as a baseline, or starting point, for possible discretionary policy measures, the scope of which will be discussed in Section C below.

5. The key assumptions underlying the baseline projections are summarized in Box 1 and Table 1. The impact of these assumptions on general government expenditure is illustrated in Table 2, and can be summarized as follows:

Table 1.Poland: Baseline Assumptions for Projections of General Government Expenditure, 2000–2005
Real economy(In percent, unless otherwise indicated)
Real GDP growth5.
Labor productivity growth4.
Employment growth1.
Labor force growth0.
Unemployment rate11.711.310.910.710.510.410.9
GDP deflator (percentage change)
Current spending(Real growth rate, in percent, unless otherwise indicated)
Compensation of employees4.
Real wages4.
Purchases of goods and services 1/8.18.710.711.310.69.99.9
Non-reform related0.
“Low-cost” scenario 2/
Transfers to households2.
Subsidies to enterprises5.
Implicit interest rate on public debt (in percent) 3/
Public debt (in percent of GDP)36.533.631.229.326.724.430.3
Privatization receipts (in billions of zloty)
Capital spending21.57.921.824.85.620.817.1
(In billions of zloty, unless otherwise indicated)
Environmental expenditure 4/8.512.019.029.332.544.924.4
Present value, 1999 U.S. dollars1.
Alternative “low-cost” estimate 2/
Present value, 1999 U.S. dollars0.
Additional infrastructure cost 5/
In billions of U.S. dollars0.,7
Other (non-reform related) investments16.216.818.620.522.524.619.9
In percent of GDP2.

Includes World Bank estimates of operational and maintenance cost of environmental investments, assumed to be implemented gradually over a ten-year period.

Assumes less strict interpretation of EU environmental requirements.

Average rate, derived by dividing total interest payments by the debt stock of the previous period.

Includes investments in the areas of water, air, and waste, based on World Bank staff estimates.

Includes investment in roads and motorways according to Poland’s Transport Investment Plan of Fall 1998.

Includes World Bank estimates of operational and maintenance cost of environmental investments, assumed to be implemented gradually over a ten-year period.

Assumes less strict interpretation of EU environmental requirements.

Average rate, derived by dividing total interest payments by the debt stock of the previous period.

Includes investments in the areas of water, air, and waste, based on World Bank staff estimates.

Includes investment in roads and motorways according to Poland’s Transport Investment Plan of Fall 1998.

Table 2.Poland: General Government Finances, 1999-2005 1/ Baseline Scenario

(In percent of GDP)

Current expenditure41.640.839.939.439.138.638.3-3.3
Compensation of employees8.
Purchases of goods and services9.910.110.410.911.512.112.62.7
Transfers to households18.317.917.016.315.615.014.4-3.9
Unemployment 2/
Subsidies to enterprises1.
Interest payments3.
Capita! expenditure3.54.04,
Memorandum items:
Current spending on education4.
Current spending on health4.
Total spending on defense1.
Other non-reform related consumption8.
Environmental expenditure 3/
Current expenditure0.
Capital expenditure1.
Alternative “low-cost” estimate 3/
Current expenditure0.
Capital expenditure0.
Implicit revenue ratio41.341.740.841.141.541.841.60.3
Nominal GDP (in billions of zloty)611.6693.6777.1860.8947.21,039.21,135.8524.3
Sources: Ministry of Finance and staff estimates and projections.

The presentation is consistent with Table 5 of the 1999 Article IV staff report on Poland. It includes revenue and expenditure of the central government, local authorities, the pension funds (FUS and KRUS), the labor fund, the alimony fund, the fund for rehabilitation, and the environment fund, Other small funds are included in expenditures on a net basis.

Based on a broad measure, including FUS payments for unemployed.

Based on World Bank staff estimates; see footnotes to Table 1.

Sources: Ministry of Finance and staff estimates and projections.

The presentation is consistent with Table 5 of the 1999 Article IV staff report on Poland. It includes revenue and expenditure of the central government, local authorities, the pension funds (FUS and KRUS), the labor fund, the alimony fund, the fund for rehabilitation, and the environment fund, Other small funds are included in expenditures on a net basis.

Based on a broad measure, including FUS payments for unemployed.

Based on World Bank staff estimates; see footnotes to Table 1.

Box 1.Baseline Assumptions for Medium-Term Expenditure Projections

Macroeconomic environment

  • Real GDP is projected to grow by an annual average of 5½ percent during 2000-2005, driven by continuously strong improvements in labor productivity of some 4½ percent, and employment growth of 1 percent a year-slightly below the average rate in 1994-98—assuming that excess labor released in the restructuring process is absorbed by an expanding private sector.

  • The labor force is projected to grow in proportion with the age group of 15-59 years, roughly representing the current “effective” working-age population. As a result of these assumptions, the unemployment rate would fall from some 11¾ percent in 2000 to 10½ in 2005.

  • Inflation, measured by the GDP deflator, is projected to decelerate from 7¾ percent to 3½ percent.

Current expenditure for 2001-2005

  • Projections of the compensation of employees assume that employment in the public sector remains at its 1999 level and real wages grow in line with total labor productivity in the economy.

  • Purchases of goods and services include (1) “reform-related” operational costs associated with large environmental investments required under EU legislation (see below); and (ii) other “non-reform related” expenditures. Regarding the latter, expenditures for education and health are assumed to increase in real terms by 6½ and 8½ percent, respectively, reflecting the need to improve equipment and technology standards. Other purchases of goods and services, with the exception of defense-related outlays, are assumed to grow in line with GDP.

  • Defense-related spending, associated with NATO, is assumed to increase gradually to the planned maximum level of 2.4 percent of GDP by 2004 from an estimated 1¾ percent of GDP in 1999.

  • In projecting transfers to households, the key item, pensions, is linked to World Bank population forecasts of the eligible age group (assuming no change in the effective retirement ages of 59 and 55 years for men and women, respectively), with the incomes of existing pensioners assumed to be constant, in real terms.1 Thus, annual increases in real average pensions result only from new pensioners entering the system, whose income (before retirement) is assumed to grow with labor productivity. Unemployment benefits are assumed to increase in proportion to the number of unemployed, with average benefits (linked to the consumer price index) projected to remain constant in real terms. The latter assumption is also applied to other transfers, projections of which are linked to unemployment developments (20 percent) and growth of the overall population (80 percent).

  • Subsidies to enterprises, including for coal and steel sector restructuring are assumed to grow in line with GDP.

  • Projections of interest payments assume (i) a modest decline in the implicit average interest rate from 7¾ percent in 2000 to 7 percent in 2005, reflecting a declining risk premium associated with lower fiscal deficits and EU accession; and (ii) the use of privatization receipts (projected at some Zl 40 billion over 2000–2002) for deficit financing and debt reduction.

Capital expenditure for 2001–2005

  • Capital expenditure is projected in a two-step approach. First, basic investment needs-after reaching some 2¼ percent of GDP in 2001-are assumed to grow in line with GDP. To these, reform-related capital spending is added, based on government and World Bank staff projections, and covering EU-related environmental investments and additional spending on roads.

1 This is broadly consistent with the indexation rates achieved in recent years.
  • Total government expenditure relative to GDP is projected to fall by 0.3 percentage points between 1999 and 2005, as a rise in capital spending by 2.9 percentage points is more than offset by a decline of 3.3 percentage points in current expenditures.

  • The decline in the current expenditure ratio is driven by a strong reduction in transfers to households relative to GDP, benefiting from a projected fall in unemployment (with assumed effects on unemployment benefits and other transfers) and, more importantly, by a significant reduction in pension expenditures as a share of GDP.3 The latter, equivalent to 2.9 percentage points, reflects both tight indexation rules and a modest increase in the eligible age group. In addition, interest payments are projected to decline by 1.7 percent of GDP, due to a combination of lower interest rates—in response to an improved macroeconomic environment and the assumed accession to the EU—and falling debt ratios, reflecting substantial privatization revenues and a declining deficit. By assumption, subsidies to enterprises including budgetary costs associated with the restructuring of the coal and steel industries and (potentially) the agricultural sector—remain constant, as a share of GDP.4

  • Consumption expenditures relative to GDP are projected to rise by 2.2 percentage points between 1999 and 2005, as a slight fall in the public sector wage bill, in percent of GDP—reflecting the conservative assumptions on government employment (see Box 1)—is more than offset by a sharp increase in purchases of goods and services. The latter reflects, in part, the assumed growth in health- and education-related spending, as well as higher outlays for defense, consistent with the need for a considerable modernization of the military infrastructure under NATO. Nevertheless, due to the declining wage bill, total current spending on education is projected to fall slightly, as a share of GDP, and the rise in health expenditure, equivalent to 0.5 percentage points is modest—well in line with international trends.5

  • The main impact on the purchase of goods and services results, however, from maintenance and operational costs associated with environmental investments required under EU legislation, and aimed at improving water and air quality and waste management. Depending on the strictness with which this legislation is interpreted, annual current expenditures are estimated by World Bank staff to reach US$1.8-US$4.7 billion, once investments are completed. To provide a conservative but realistic assessment of the cost, the baseline projections assume a strict interpretation combined with a back-loaded implementation schedule over a 10-year period, with about half of the remaining investment need completed by 2005. This implies a gradual increase in operational and maintenance cost to 2 percent of GDP by 2005 (see memorandum items in Table 2). Alternatively, the low-cost estimate, based on a more lenient interpretation of the legislation and the same investment profile, would translate into current spending of 0.7 percent of GDP in 2005.

  • Capital spending as a share of GDP is projected to grow by 2.9 percentage points over the projection period, notwithstanding that non-reform related investments are assumed to remain constant as a share of GDP. The overall increase reflects primarily the above-mentioned environmental investments, but also improvements in the road network.6 According to World Bank staff estimates, EU environmental requirements would translate into investment expenditure in the range of US$22. 1-US$42.8 billion, expressed in 1999 prices, depending on the strictness of interpreting existing regulations. With the assumed 10-year implementation period, this would imply capital spending in the range of 2-4 percent of GDP by 2005. Capital spending from additional EU-related investments in roads and motorways are projected at US$0.7 billion annually (0.3 percent of GDP on average), in line with Poland’s Transport Investment Plan of Fall 1998.

6. As illustrated in Table 2, the expenditure projections of the baseline scenario imply that in order for the government to gradually eliminate the public sector deficit by 2004, the revenue ratio would have to rise above 44¾ percent of GDP in 2005, compared with just over 41 percent in 1999. In the case of lower environmental investment requirement, on the other hand, the revenue ratio in 2005 would only be 0.3 percentage points higher than in 1999. Both of these outcomes hinge on the assumption that Poland’s strong economic growth performance continues over the medium term, with average real GDP growth of 5½ percent. Should GDP growth be only one percentage point lower, the revenue ratio would have to climb to almost 47 percent by 2005 (43½ percent with lower investment requirements), in order to meet the deficit objectives (Figure 1). Even with faster economic growth of one percentage point a year, relative to the baseline, revenues in 2005 would still have to reach almost 43 percent of GDP under the high-investment scenario, but could fall below 40 percent of GDP with a more lenient interpretation of EU legislation.

Figure 1.Poland: Revenue Ratios Under Alternative GDP Growth Scenarios, 1999-2005

(In percent of GDP)

Citation: 2000, 60; 10.5089/9781451831887.002.A003

Source: Staff projections.

1/ High (lower) investment requirements result from the assumption of (less) strict interpretation of EU environmental legislation.

2/ Assumes average real GDP growth of 5.5 percent through 2000-2005.

3/ Assumes that real GDP grows by one percentage point less than under the baseline scenario.

4/ Assumes that real GDP grows by one percentage point more than under the baseline scenario.

7. Two main conclusions can be drawn from the discussion above: (i) EU-accession related spending, and specifically the outcome of negotiations on environmental standards, plays a key role in the government’s ability to achieve its medium-term fiscal objectives; and (ii) in most circumstances, spending restraint will be a major element in a successful medium-term strategy.

C. The Scope for Spending Restraint

8. The objective of this section is to identify areas for potential expenditure restraint (relative to the baseline), which would allow the government to achieve its medium-term deficit objectives without the need for higher revenue ratios. With capital spending constrained by reforms linked to Poland’s prospective EU-accession (notwithstanding the uncertainties discussed above), and interest payments determined by the existing debt level, expected privatization revenues, and the overall deficit path, restraint will need to focus on individual components of current primary spending. Before identifying specific measures, a comparison with the expenditure structure of other European countries is used to shed some light on the potential scope for adjustments in Poland’s expenditure composition.

9. In 1998, current government spending in Poland accounted for some 41 percent of GDP and more than 90 percent of total public expenditure, dominated by public consumption and transfers to households, each equivalent to approximately 19 percent of GDP. Interest payments and subsidies, on the other hand, were relatively modest at some 3 percent and 1 percent of GDP, respectively. This structure of public expenditure is not very different from that in most countries in the EU, and both public consumption and transfers as a share of GDP in Poland were either close to or lower than those in the EU, as a whole (which were equivalent to 18 percent and 20½ percent of GDP, respectively, in 1998).7 Thus, if average public spending ratios in the EU were considered a yardstick for Poland, in light of its prospective accession, potential areas for restraint would not be obvious.

10. Such a comparison with spending ratios in the EU as a whole, however, masks the fact that these ratios are dominated by the two largest countries, Germany and France, which also have above-average per capita income levels. A country-by-country comparison shows that, at least within the European Union, there is a clear positive correlation of transfer ratios with per capita income levels (Figure 2, upper panel). This positive correlation with per capita income is not observable for public consumption, as a share of GDP, if all EU members are included in the sample. However, a positive trend does appear, once Luxemburg (with the highest per capita income and the lowest public consumption ratio) is excluded (Figure 2, lower panel). Against this background, Poland’s consumption and transfer ratios are clearly higher than what its income level would suggest.

Figure 2.Poland: Per Capita Income, Government Transfers, and Public Consumption in Poland and Countries of the Eurpoean Union

Source: European Commission, and World Bank Development Indicators.

1/ Not included in creation of trendline.

11. This conclusion is confirmed by a comparison of Poland’s public consumption and transfer ratios with those in other advanced transition economies. While Poland has the second lowest per capita income level among the five transition economies originally considered for early EU accession, its government pays the highest transfers to households, relative to GDP, and its consumption ratio is second only to Estonia’s (Figure 3). From a functional perspective, Poland’s high transfer ratios are mirrored in heavy spending on social security and welfare-a reflection of its generous pension and sickness benefits system.8

Figure 3.Poland: Per Capita Income and Public Spending Components in Selected Transition Economies

Source: World Bank Development Indicators, Government Finance Statistics, and staff estimates.

1/ Data refers to consolidated central government.

12. The comparison of Poland’s expenditure ratios with those of other European countries suggests public consumption and transfers to households as the prime candidates for expenditure adjustments. In addition, there may be further scope to reduce subsidies to enterprises, as restructuring and privatization nears completion. Without prejudice to the government’s policy priorities, the following analysis will concentrate on potential savings in these three areas.

13. As illustrated in Section B above, the overall reduction in current primary spending necessary to meet the government’s deficit objective, is subject to considerable uncertainties, depending crucially on the extent of environmental investment requirements as well as on the desired revenue ratio (including grants) over time. The following discussion assumes that the targeted reduction in current spending as a share of GDP is of the order of 3.5 percentage points between 1999 and 2005. This would be consistent with a broadly stable revenue ratio, even under relatively high environmental investment requirements, and would provide scope for direct tax cuts, as indirect taxes and other revenues (particularly EU grants) are expected to rise.

14. The allocation of savings measures is clearly a political decision to be taken by the Polish government. This section does not attempt to suggest specific actions. Instead, it presents one possible expenditure restraint scenario, based on the following simplified assumptions, summarized in Table 3:

Table 3.Poland: Potential Savings in General Government Expenditures, 2001–2005

(In percent of GDP)

I. Compensation of employees
Baseline: Real wages grow with labor productivity8.
Growth in public sector employment (in percent)
Alternative I: Real wages remain constant in real terns7.
Growth in public sector employment (in percent)
Alternative II: Employment reduction7.
Growth in public sector employment (in percent)-4.3-4.4-4.4-4.4-4.4
Savings (Baseline minus Alternative)
II. Subsidies
Baseline: Constant in percent of GDP1.
Alternative: Fall by 1 percent of GDP in 2003 and beyond1.
Savings (Baseline minus Alternative)
III. Transfers to households (without employment impact)
Baseline: No change in effective retirement age17.016.315.615.014.4
Alternative: Increase in effective retirement age 1/16.915.915.114.213.4
Savings (Baseline minas Alternative)
With employment and growth impact of retirement change1.
Memorandum items:
Employment and growth impact of retirement change
Total expenditure baseline44.044.144.644.244.7
Total expenditure alternative 2/43.342.942.942.041.8
Savings (Baseline minus Alternative)
Additional savings 3/
Source: Staff estimates.

Assumes that each year the effective retirement age increases by half a year. The reduction in the number of pensioners is assumed to lead to an equivalent increase in the number of other transfer recipients.

Assumes that the reduction in pensioners leads to an equivalent increase in employment.

Relative to scenario where reduction in pensions leads to an increase in other transfers.

Source: Staff estimates.

Assumes that each year the effective retirement age increases by half a year. The reduction in the number of pensioners is assumed to lead to an equivalent increase in the number of other transfer recipients.

Assumes that the reduction in pensioners leads to an equivalent increase in employment.

Relative to scenario where reduction in pensions leads to an increase in other transfers.

  • Savings in the compensation of employees are assumed to be achieved by keeping public sector wages constant in real terms. Relative to the baseline, this measure would reduce public spending as a share GDP by 1.5 percentage points by 2005. Alternatively, if real wages increased in line with total labor productivity, public employment would have to be reduced by approximately 4¼ percent annually, in order to achieve the same savings in the total wage bill.

  • Subsidies to enterprises as a ratio of GDP, projected to remain constant in the baseline scenario, are assumed to fall by 1 percentage point in 2003—when the restructuring of the coal and steel sectors is expected to be completed—and to remain at this ratio through 2005. Compared with the baseline, this measure would generate savings relative to GDP of 1 percentage point by 2005.

  • Transfers to households are assumed to be affected by savings in the dominant item, pension payments, accounting for approximately three-quarters of total transfers in 1998. This is assumed to be achieved by an increase in the effective retirement age, relative to the baseline, from its current levels of 59 years for men and 55 years for women.9 While the new pension system, adopted in 1999, is expected to bring about significant savings for the public sector in the longer term (see Box 2 for a summary of the new system), it only applies to workers below the age of 50, and is mandatory only for people younger than 30. Thus, savings over the projection period, relative to the baseline, will have to be achieved through additional measures building on past reforms to curb pension outlays (e.g., tighter eligibility for disability pensions and measures to reduce incentives for early retirement).10 The main effect of such measures would be an increase of the effective retirement age which, for illustrative purposes, is assumed to rise by half a year annually during the 2001-05 period. The savings generated by the assumed reduction in the number of pensioners (about 700,000 people by 2005) depend crucially on whether these are assumed to become unemployed, or alternatively, whether they can be absorbed by the labor market. In the first case, savings relative to GDP would amount to an estimated 1 percentage point, reflecting the much lower replacement income of unemployment benefits (some 30 percent) compared with pensions. In the second case, additional savings of 1.8 percentage point could be generated, resulting not only from the avoided increase in unemployment benefits (some 0.4 percentage points), but also from a higher GDP growth rate.11 In practice, it is likely that a reduction in the number of pensioners would lead to some increase in both employment and unemployment, with total savings relative to GDP projected to be in the range of 1-2.S percentage points.

Box 2.The Pension Reform of 1999 1/

The primary objective of the pension reform adopted in 1999 is to ensure long-term sustainability of the system in the face of foreseeable demographic pressures. In addition, benefits are expected also in the form of higher national savings; greater labor market participation; a smaller gray economy; and deeper asset markets. The following are the main features of the reformed system:

  • The new pension system is offered to non-agricultural workers (with separate provisions for the military, police, and clergy) younger than 50 years, with mandatory participation for workers younger than 30 years in 1999 and new entrants into the labor market.

  • It is a multi-pillar system, involving a direct link between contributions and future pensions benefits, thereby implying stronger incentives to retire later.

  • The first pillar consists of a mandatory, contribution-based pay-as-you go scheme. Besides part of the old-age pensions, it also encompasses all disability and survivors pensions.

  • The second pillar-which is optional for people between 30 and 50 years that were already working in 1999, and mandatory for the younger generation-is a privately-managed funded system. The funds’ portfolio composition is strictly regulated, and activities are monitored by the Pension Fund Supervision Office (UNFE).

  • The third pillar includes all additional, voluntary forms of retirement savings.

  • The government further guarantees a minimum pension equivalent to about 28 percent of the average wage in the enterprise sector in 1999 (with subsequent price indexation) for male retirees aged 65 (60 for women) with a minimum of 25 years (20 years) of service. The pension subsidy would be covered from the state budget.

  • The minimum retirement age is 65 years for men and 60 years for women. The new system does not include provisions for early retirement. However, people who were between 30 and 50 years old in 1999 remain eligible for the provisions under the previous system. There are plans to compensate certain categories of employees for the loss of early retirement privileges through “bridging pensions”, likely to be paid for by employers.

While the gains of the new system will only materialize gradually, cost to the budget in terms of reduced revenues due to the switching to the second pillar are occurring already. However, some savings are expected to be realized over the coming years, as a result of past reforms to the old system (including tighter eligibility for disability pensions, a shift in the indexation base from wages to prices, and a lengthening in the number of working years to be included in the calculation of pension benefits).

1/ For a detailed descriptions of the new system see OECD, 1997-98.

Revenue and Expenditure Ratios After Spending Restraint

(In percent of GPD)

Citation: 2000, 60; 10.5089/9781451831887.002.A003

15. The combined effect of the above measures, is also summarized in Table 3. Given the deficit targets and the projected expenditure ratios in the baseline scenario, total savings generated by 2005 would be equivalent to 3.5 percent of GDP (assuming no impact from pension savings on growth and employment, but rather some increase in unemployment). Under this scenario, the revenue ratio, which falls out as a residual, would be broadly stable. These results illustrate that the government will have to make some tough choices in a highly uncertain environment in order to meet its medium-term fiscal targets. Even if the political will exists, the implementation of reforms may be constrained by the newly decentralized administrative structure.

D. Implementing Expenditure Reforms in a Decentralized System

16. Effective as of the beginning of 1999, the Polish government has implemented a fundamental reform of fiscal administration, which devolved virtually all decision making and financing responsibilities for regional and local public services and investments to the subnational authorities (see Box 3 for a summary of the system’s key features).12 As a result, local governments’ share in total spending of general government is estimated to have risen from about one-fifth in 1998 to almost one-quarter in 1999 (Table 4).13 Reflecting the additional assignments in areas such as roads, education, and public order, the main impact of the devolution affects local governments’ share in capital spending and public consumption, which are estimated to have grown to some 60 percent and 45 percent in 1999, respectively. At the same time, the reform of the health care system, with the establishment of independent health funds within the general government, has further reduced the expenditure share of the central government from more than 40 percent of general government spending in 1998 to an estimated 30 percent in 1999 (see Box 4 for a brief description of the reformed health care system).

Table 4.Poland: Breakdown of Public Spending by Levels of Government, 1998-99 1/

(In percent of general government expenditure)

Central governmentLocal governmentsSocial security funds 2/
Current Expenditure42.728.916.220.841.150.3
Transfers and Subsidies16.518.33.92.779.579.0
Interest payments100.0100.
Capital Expenditure43.
Total Expenditure42.729.419.223.838.046.8
Source: Staff estimates based on Ministry of Finance preliminary data and projections.

Shares exclude intergovernmental transfers.

For coverage see footnote 1 to Table 2.

Source: Staff estimates based on Ministry of Finance preliminary data and projections.

Shares exclude intergovernmental transfers.

For coverage see footnote 1 to Table 2.

Box 3.Structure of the Polish Intergovernmental System

  • The Polish intergovernmental systems consists of four layers of government: the central government, 16 voivodships (provinces), 373 powiats (county or city governments), and 2,425 gminas (local governments).

  • On the expenditure side, the devolution of authority from the central to subnational governments encompasses the provision of a wide range of public services, notably in the areas of education, roads, and health care. The central government maintains responsibility for national defense, income transfers, state roads, state police and the judicial system, as well as the social security system, managed by various funds. Health care has been assigned, as of 1999, to autonomously operating health funds, financed by a dedicated payroll tax and some transfers from the central government to cover payments for people that do not pay income taxes.

  • The voivodships are responsible for general regional policy, the major regional roads, special schools and technical institutes, and larger cultural facilities, and have taken ownership of more specialized regional hospitals; they also supervise the autonomous health finds. Powiats operate at an intermediate level and have authority over secondary education, the municipal police and fire protection, some social welfare services (such as nursing homes), and county-level roads; they also have been transferred ownership of most hospitals. The gminas’ main responsibilities are for primary education, local roads, communal services, basic utilities, various social assistance tasks, and control of local (primary health care) clinics.

  • The revenue sources of the central government include personal income and payroll taxes, corporate income taxes, the VAT, excise and trade duties, and nontax revenues. Subnational governments receive own revenues, including local taxes, fees, and charges, subject to a centrally imposed limit; revenues from shares in the personal and corporate income tax (PIT and CIT); and other central government transfers and subsidies.

  • The bulk of voivodships’ revenues (nearly 80 percent of total projected revenues in 1999) consists of state grants in the form of targeted subsidies earmarked for carrying out specific administrative tasks, and general transfers, including roads and education transfers (which can also be used for other purposes) and equalization grants (to limit differences in per capita tax revenues across voivodships). With very little own revenues (2 percent), the remainder largely stems from a 1.5 percent share of the PIT and a 0.5 percent share of the CIT. Powiats receive a 1 percent share of the PIT (approximately 2 percent of projected revenues in 1999); some own revenue (3 percent); targeted subsidies (some 50 percent); and transfers for roads, education, and equalization (45 percent of revenues). Gminas’ projected revenues for 1999 consist of own revenues (35 percent), including a variety of local taxes (e.g., on real estate and agricultural and forest lands); a 27.6 percent share of the PIT and a 5 percent share of the CIT (some 16 percent of total revenues); and state grants and targeted subsidies (some 25 and 14 percent of revenues, respectively).

  • Subnational government borrowing is limited by law via two rules: (i) each entity’s projected debt service in any given year may not exceed 15 percent of its planned budgetary revenue; and (ii) each entity’s total outstanding debt may not exceed 60 percent of its planned annual revenue. Moreover, subnational governments are not permitted to borrow in foreign currency.

Box 4.The Polish Health Care System1

The reform of the Polish health care system, which became effective at the beginning of 1999, is a major element of the government’s overall policy of decentralization, aimed ultimately at improving the efficiency and quality of services. The main elements of the new system are the following:

  • The organizational structure of the new system consists of 16 autonomous health funds—one for each voivodship, plus an additional branch fund for certain public sector employees-whose responsibility is the financing of health service provided to its members.

  • Financing of health funds is provided through a compulsory insurance system, with contributions equivalent to 7½ percent of taxable income, deducted from personal income taxes withholdings; for households that do not pay personal income taxes (e.g., farmers and unemployed) contributions are paid by the state; from 2000 on, an equalization scheme is expected to redistribute revenues across funds on the basis of enrollment characteristics (e.g., age composition) and fund revenue per enrolled.

  • Each fund is governed by a supervisory board, consisting of representatives appointed by the voivodships; the overall supervision of health funds is the responsibility of Health Insurance Supervisory Office, which is an autonomous agency of the central government.

  • The provision of health care services is primarily undertaken by public hospitals and clinics, owned by local governments.

  • Competition between service providers is to be encouraged, with each health fund negotiating annual contracts with public and private hospitals, clinics, and doctors, on the terms and conditions for payment of health services.

In its first year, the new health care system experienced a number of teething problems: difficulties in monitoring employer transfers of health care contributions to the Social Insurance Fund (ZUS) resulted in a shortfall of revenues to health funds, while uncertainties and flaws in contracts between funds and hospitals created service disincentives and inefficiencies. Many of those problems will disappear within a few years of experience with the new system, and the government is preparing a comprehensive health act, which should provide a legal base for defining missing key details. The introduction of a social insurance system and the split between purchaser and provider functions are important steps in improving the health care system, and further actions should focus on tackling remaining deficiencies in the running of hospitals and insufficient use of primary and preventive care.

1 For a detailed discussion of the main features of the new health care system, see OECD, 1999—2000.

17. Besides administrative advantages, this shift to greater decentralization promises to increase efficiency in the allocation of public services, by allowing a greater responsiveness of spending decisions to the preferences of citizens in different constituencies.14 Notwithstanding the potential efficiency gains, with the devolution of additional spending powers to local authorities, the central government has further reduced its ability to control both the level and composition of overall public spending. However, it maintains the potential to influence local spending decisions indirectly through rules and mechanisms for revenue and borrowing arrangements. The following analysis assesses whether the rules adopted in Poland seem appropriate to ensure an adequate provision of public services without endangering the government’s overall fiscal targets, discussed in the previous sections. The discussion covers arrangements for (i) local governments’ own revenues; (ii) intergovernmental transfers; and (iii) local borrowing; and (iv) highlight other potential risks and areas for improvement.

Own revenues

18. The assignment of own revenues and particularly own taxing powers to local governments is one of the three indirect instruments (together with transfers and borrowing rules) that determine the degree to which the central government can effectively influence subnational expenditure levels.15 As a general rule, larger own revenues tend to increase accountability at the local level but reduce central government control over total public spending (see Box 5 for a brief discussion of the pros and cons).16 The system adopted in Poland entails a relatively strong preference for maintaining central control, as own revenues play a significant role only for the income of gminas, where they account for approximately one-third of total revenues (see Table 5), subject to centrally-imposed limits on both tax rates and fees.17 This approach seems appropriate at the early stages of the administrative reform, to provide the new layers of government (the powiats and voivodships) with a high degree of certainty about their revenues for the coming year. However, once these new entities have gained some experience, there seems to be scope for strengthening their taxing powers (and accountability) without undermining macroeconomic control. For this purpose, the personal income tax, which is currently shared (see below), could be partially assigned to subnational levels (e.g., through a surcharge on a reduced central government rate). Also, single-stage sale and excise taxes, levied on the consumer, are potentially good candidates for downward assignment, possibly with the central government setting ranges for the applicable rates. Finally, as regards the gminas, there is room to improve the real estate tax (currently accounting for some 12 percent of gmina’s revenues) by linking it closer to the actual value of the taxed property.18

Box 5.Pros and Cons of Tax Assignments to Subnational Governments

  • From the viewpoint of promoting fiscal responsibility and accountability, subnational governments should be assigned substantial taxing powers, to establish a close link between the benefits of public services and their price (i.e., taxes).

  • Devolving large taxing powers to lower levels of government, however, constrains the central government’s ability to implement effective stabilization policies—particularly when expenditure decisions have been decentralized already—and, more generally, to influence the overall tax burden in the economy.

  • Distributional concerns can also argue against a significant devolution of taxing powers to local governments, in particular, if there is a large variation in the tax capacity of different regions.

  • The arrangement favored in the literature, and observed in most countries, is a compromise, with some assignment of own sources of revenue to each level of government, complemented by various types of downward revenue-sharing arrangements and other transfers to bridge the gap between revenues and expenditure obligations at the local level.

  • There is broad agreement in the literature that attempts to strengthen the accountability of local governments should focus on those taxes that are (i) less sensitive to income fluctuations (to shelter local governments from cyclical effects and provide the central government with stabilization instruments); (ii) levied on less mobile tax bases (to limit distortionary tax-induced movements of factors of production) and (iii) levied on tax bases that are distributed evenly across regions (to avoid increased regional disparities).

Table 5:Poland: Local Government Revenue Breakdown., 1998–99

(In percent of local governments’ total revenue)

Local Gov.Local Gov.Est.Est.Est.
Own Revenue35.330.
Share in State Revenue24.816.415.91.819.5
Corporate income tax1.
Personal income tax23.
Targeted subsidies14.410.
General transfers25.442.840.649.247.3
Sources: Ministry of Finance and staff estimates.
Sources: Ministry of Finance and staff estimates.

Intergovernmental transfers19

19. With relatively small own revenues, intergovernmental transfers—comprising both revenue-sharing arrangements and grants and accounting for some 70 percent of local governments’ total revenues—are the main vehicle through which the central government can influence subnational spending levels. The design of an appropriate transfer system has to combine three aspects: (i) maintain incentives for local governments to manage expenditures efficiently and responsibly, in order to ensure broad consistency with overall spending targets; (ii) alleviate horizontal imbalances between jurisdictions; and (iii) provide local governments with the residual resources needed to deliver a desired level and quality of public services without a need for excessive borrowing.

20. With regard to establishing appropriate incentives, it is important that the level of transfers in Poland is set centrally and delivered primarily on the basis of objective criteria, rather than in the form of gap filling.20 This is key for imposing expenditure discipline at the local level and maintaining overall expenditure control. In addition, a fair amount of grants (some 80 percent) is provided unconditionally (including road and education transfers, which can be used for alternative purposes). This is important to provide appropriate incentives for an efficient use of resources, as it allows local governments to reallocate saving in one area to other purposes. It also implies that the structure of public expenditure—and specifically the relation between public consumption and capital spending—is determined to a large extent at the local rather than the central level. A potential problem, related to incentives and overall spending control, stems from the revenue—sharing arrangement of personal and corporate income taxes, which may induce procyclical spending behavior-although to a limited extent, as these revenues are not very large as a share of local government income. Nevertheless, if procyclical spending behavior proves a real concern in the future, it could be considered to relate these transfers to a moving average of tax revenues over a number of years, or require subnational governments to build up revenue stabilization funds to even out cyclical fluctuations.

21. Horizontal imbalances—the second aspect of intergovernmental transfers—are addressed by equalization grants. In poviats and voivodships, the grants are based on per capita revenues generated from their shares in the personal and (for voivodships) corporate income tax. The formula applied seeks to compensate each poviat (voivodship), other than the richest, for 85 percent (70 percent) of the difference between the two jurisdictions.21 As relatively small fractions of these entities’ revenues are generated by shared taxes (see above), the equalization effect is limited, particularly for poviats. In the case of gmina’s, equalization is based on a broader revenue measure that also includes own, in addition to shared, taxes, and compensates those gminas generating less than 85 percent of the national average (by 90 percent of the difference between their individual level and the average). By keeping the equalization below 100 percent, the system maintains incentives for local governments to strengthen revenue generation in their own jurisdictions. This, however, comes at the expense of a potentially increasing divergence in the fiscal capacity of different jurisdictions. Whether the current balance between incentive and equity considerations is appropriate, remains to be seen and may require some adjustments in the future.

22. With respect to the adequacy of overall transfer levels (the third and arguably most important of the above-mentioned aspects of an appropriate transfer system), experience so far is inconclusive. On the one hand, preliminary data for the first half of 1999 indicate that local governments have not contracted significant amounts of debt—which may seem to suggest that revenues were adequate to perform their extended expenditure obligations. On the other hand, there is evidence that poviats and voivodships lacked financial resources to continue many of the ongoing investment projects inherited from the central government. The implied shift in the expenditure structure toward higher consumption and lower investment ratios is likely to be reversed, at least partially, as public dissatisfaction (for example, with the standard of roads) grows. It remains, however, an illustration of one of the consequences of fiscal devolution; namely the inability of the central government to determine the structure of overall public spending. Moreover, spending restraint in other areas to offset potentially higher investment spending in the coming years, will be more likely, if local governments face effective budget constraints, and specifically limits on their ability to borrow.

Borrowing arrangements

23. Borrowing arrangements are the third instrument of indirect control over subnational governments’ expenditures, and are key to ensure that revenue constraints become effectively binding over time. Some borrowing is needed to shelter expenditures from cyclical revenue fluctuations and thus avoid procyclical spending behavior. However, especially with the lack of fiscal experience in the newly established poviats and voivodships, some constraints seem prudent to safeguard the overall public sector deficit targets.

24. The subject of local government borrowing is discussed in some detail in Chapter I, and the recommendations developed there also apply here. Specifically, while the rules-based approach adopted in Poland seems generally adequate to contain subnational borrowing without unduly constraining their effective decision-making, the specific limits are set fairly high (see Box 3). Consequently, there may be a case for phasing in the current borrowing limits and combining them with additional incentive-based measures to limit debt accumulation, such as “automatic” reductions in state transfers in case of excessive borrowing and higher capital requirements on bank lending to local (as opposed to central) government units.22

Risks and further areas for improvement

25. The above analysis suggests that the main risks of devolution for the overall fiscal outcome stems not from unduly high allocations of resources to local governments’, or their ability to finance large additional spending from self-generated revenues, but rather from a combination of comprehensive spending obligations and ineffective constraints on subnational borrowing. These risks are exacerbated by weaknesses in the reporting system, under which local governments have to submit their preliminary budgets only by end-March of the current year. To allow the central government scope to react in time to local governments’ spending plans, and to improve the monitoring of their financial situation generally, an advancement of preliminary budget reporting requirements into the pre-budget year—based on the central government’s initial October projections of transfers and subsidies—would seem both feasible and appropriate.

26. In addition, to the risks and problems discussed above, local governments face potentially large contingent liabilities, as a result of the recent health care reform (see Box 4). In the past, health care providers accumulated sizeable arrears, which were taken on by the central government. Under the new health care arrangement, the local governments themselves are ultimately responsible for financial obligations accumulated by their own public health care facilities.23 However, while voivodships are in charge of monitoring the situation in their jurisdiction, data on potential overspending and arrears accumulation of public health care facilities in 1999 are not available. Finally, while the government encourages the privatization of primary health care delivery and a reduction in the existing high density of secondary care general hospitals, local governments may encounter serious difficulties in finding buyers and liquidating institutions with sizeable outstanding obligations. Thus, further reforms to reduce the cost of health care provision and raise the accountability of hospital managements are crucial. As a first step, there is a clear need for improved monitoring.

E. Conclusions

27. With large anticipated expenditure commitments related to structural reforms and EU accession, the Polish government has to make a number of tough choices in order to achieve its medium-term fiscal objectives, namely a balanced position in the general government and avoidance of tax increases to meet this target. The outcome of accession negotiations with the EU—particularly on the adoption of environmental standards on the expenditure side, but also with regard to the provision of grants on the revenue side-will play a critical role for the public finances. In addition, there is a need for spending restraint, with public consumption, transfers to households, and subsidies to enterprises as obvious candidates.

28. The implementation of expenditure reform is complicated by the 1999 administrative reform, which has continued the process of vast devolution of spending authority to subnational governments. This implies that the central government will at best be able to influence overall public spending—through appropriate revenue arrangements and borrowing mechanisms at the subnational level—whereas the structure of public spending, and specifically the allocation of resources to public consumption and investment, respectively, will be determined to a large extent outside the central government. The prime task for the central government, in this context, will be to strike an appropriate balance between the provision of adequate funding to local governments, commensurate with their widened responsibilities, and the implementation of effective constraints on their overall spending levels through incentives and rules limiting their borrowing.


Prepared by Christina Daseking.

This is an obvious simplification, as changes in government expenditure (and revenues) are expected to have an impact on growth, inflation, and employment—and are, in fact, generally motivated by a desire to influence the macroeconomic environment.

The main downside risk to unemployment benefits and other transfers results from the potentially large excess labor in the agricultural sector, that could be transformed into open unemployment in the context of restructuring. However, the projected fiscal implications of this are mitigated by the relatively tight eligibility for unemployment benefits.

Total cost for the coal and steel sector restructuring, in terms of severance packages, closures, and investment requirements, including those borne by the affected companies, have been estimated at Zl 7 billion and Zl 12 billion, respectively, equivalent to some 3 percent of 1999 GDP (see Chapter IV).

Total spending on education in Poland in 1998 was higher as a share of GDP than in the Czech Republic, but lower than in Hungary and Estonia. Health spending relative to GDP was lower in Poland than in any of the three other countries.

Costs associated with the railway restructuring are assumed to be borne by private investors in the course of the planned privatization of part of the railways.

It should be noted that possible differences in coverage and methodology may slightly qualify the results of the country comparison.

For a brief discussion of sickness benefits and recent reforms see Chapter V. Key features of the pension system and its reform are discussed below.

These figures compare with an OECD average of 62 years for men and 61 years for women.

The latter has been achieved by reductions in the replacement income from 76 percent in 1991 to less than 70 percent in 1998—which is still high by international standards. The shift in the indexation base from wages to prices in 1996 will ensure a continuation of this trend.

Assuming that labor productivity remains unchanged, higher employment in this scenario would accelerate GDP growth to an average of 6¼/4 percent over 2000—05, compared with 5¼ percent under the baseline scenario.

The description of the new governmental structure draws on material provided in an internal report on an IMF Government Finance Statistics Mission (see Firestine, 1999).

The terms local and subnational are used here as synonyms, referring to all three layers of subnational government in Poland.

In the economic theory, this idea is supported by the benefit principle, which suggests that a given service should be provided by the level of government that most closely represents the region that benefits from such service. For a further discussion of these issues, see Ahmad, Hewitt, and Ruggiero (1997).

Tax assignment is defined here as the power to determine the base, or at least the rate structure, of a given tax. It does not necessarily imply the collection responsibility, which should take into practical considerations, such as administrative capacity.

For a further discussion see Norregaard (1997).

In the past, these limits have often not been binding, as many gminas preferred to set lower rates.

In the current system, real estate taxes are determined on the basis of the lot and building size, as well as the type of building, irrespective of its market value.

See Ahmad and Craig (1997) for a comprehensive discussion of this topic.

Criteria for road transfers, for example, include road length, traffic density, and accident rates; and education transfers are based on standard subsidies per pupil. Targeted subsidies for duties of state administration (related, for example, to social assistance tasks) are currently derived on the basis of earlier experience in carrying out similar tasks, and are planned to be “standardized” country-wide at some time in the future.

From 2000 on, annual increases in equalization grants to poviats will be based on average additional revenues, rather than additional revenues in the richest poviat.

For a discussion of these proposals see Chapter I.

Operating deficits and payment arrears of these facilities can develop, for example, when contracts negotiated with health funds fail to cover actual costs.

Other Resources Citing This Publication