This Selected Issues paper and Statistical Appendix examines the developments in the intergovernmental fiscal relations for the Republic of Estonia. The paper highlights that intergovernmental relations in Estonia have been marked in recent years by a strong push in the direction of fiscal decentralization. This trend has been part of the broader process of structural change, including privatization and liberalization of markets in Estonia. The paper analyzes the evolution of the financial sector. It also examines European Union accession and the economic policy of Estonia.


This Selected Issues paper and Statistical Appendix examines the developments in the intergovernmental fiscal relations for the Republic of Estonia. The paper highlights that intergovernmental relations in Estonia have been marked in recent years by a strong push in the direction of fiscal decentralization. This trend has been part of the broader process of structural change, including privatization and liberalization of markets in Estonia. The paper analyzes the evolution of the financial sector. It also examines European Union accession and the economic policy of Estonia.

I. Developments in Intergovernmental Fiscal Relations1

1. Intergovernmental relations in Estonia have been marked in recent years by a strong push in the direction of fiscal decentralization. This trend has been part of the broader process of structural change, including privatization and liberalization of markets, in Estonia. While different from these reforms, decentralization is based on a similar thesis, namely that decisions over the production and delivery of goods and services should be allocated to the lowest unit capable of capturing the associated costs and benefits.

2. There is general agreement in Estonia on the virtues of decentralization, but also a growing awareness of its pitfalls.2 Because decentralization has increased the number of actors and budgetary accounts, Estonia faces additional challenges as it sets out to consolidate the impressive gains made in stabilizing the economy. The complexities involved in this task are best summarized in the diverging fiscal performance of the central and local governments during 1993-96; while the central government deficit declined from 1.8 percent to 0.2 percent of GDP, the local government position swung from a surplus of 0.2 percent of GDP to a deficit of 1.1 percent of GDP.3 Intergovernmental relations mainly affect macroeconomic developments through three channels: the assignment and sharing of tax bases and expenditures, the design of transfers to local governments, and the level of local government borrowing. This note examines each of these issues within the Estonian context.

A. Structure of Government

3. The current structure of government in Estonia is the embodiment of the twists and turns in its history over the past 75 years. Between 1920 and 1940, the territory was divided into 11 counties, and by the end of the 1930s, following considerable debate over the organization of local government, a set of reforms reduced the number of municipalities from 365 to 248.4 But the reforms had no time to take hold before the start of the Soviet regime in 1940. A high degree of centralized control over all decisions, including public finances, characterized the next 50 years of government. Under the Soviet system, the little power there was at the local level was concentrated in collective farms.

4. The move toward decentralization began again in late 1989 with the Law on the Foundations of Local Government. As Estonia took the first steps to a democratically-elected government and a market economy, the government structure remained heavily influenced by the Soviet system. There were two levels of local government: the first consisted of 15 counties and 6 republican cities with the status of county, and the second encompassed the towns and municipalities dependent on the county governments or larger towns such as Tallinn.5 In subsequent years, the counties became increasingly important and exercised almost complete control over municipal budgets, though a few of them transferred most powers to their municipalities.

5. A major package of reforms in 1993 introduced a degree of decentralization to the government structure that harked back to the pre-Soviet days of the 1930s; it also reflected the view that the transfer of powers to local governments within a number of counties had already yielded significant gains in efficiency and welfare. The two tiers of local government were effectively eliminated as counties lost their status as autonomous units of local government, becoming instead representatives of the central government, and the republican cities were merged with their surrounding counties. In addition, the municipalities gained control of their budgets with effect from 1994.

6. There are now 254 local governments all with the same status, ranging in size from Tallinn with 420,500 inhabitants to Ruhnu with 63 inhabitants.6 Over two-thirds of these local governments have populations under 3,000.

B. Expenditure Assignments

7. Local governments spent roughly EEK 4.6 billion in 1996, i.e., 22 percent of consolidated government expenditures and 9 percent of GDP (Table 1).7 The lion’s share of local government expenditures, 42 percent of the total, was devoted to education; the “economic sphere,” including road construction and maintenance, and communal services, accounted for 28 percent of local spending; and administration another 12 percent.8 The scale of spending by local governments in Estonia attests to their importance, but it is interesting to note that the expenditures of the Social Insurance Fund (including the Child Support Fund) are higher still. From a cross-country perspective, the share of local government expenditures in Estonia is at the lower end of the range (23-30 percent of general government expenditures) found in such diverse countries as Ireland, the Netherlands, and Hungary, but well below the shares registered throughout Scandinavia (Figure 1).

Table 1.

Estonia: Expenditure by Level of Government, 1993-96

(In millions of kroons)

article image
Sources: Estonian authorities; Fund staff estimates; and Ahmad et al (1997), Table 3.

Average of three years ending in the year shown in parentheses, except in the case of Hungary for which the figure refers to 1995.

Includes state government.

Figure 1.
Figure 1.

Estonia: Expenditure by Government Level in Selected Countries 1/2/

(In percent of GDP)

Citation: IMF Staff Country Reports 1998, 012; 10.5089/9781451812329.002.A001

Sources: Ehtisham Ahmad et al (1997); and Fund staff estimates.1/ Average of three years ending in 1987, except for Norway (1986) and the Netherlands (1988). The data for Hungary and Estonia refer to 1995 and 1996, respectively.2/ Includes social security and welfare which, with the exception of Argentina, is included solely under central government spending.3/ Includes state government in Argentina, Austria, and Brazil.

8. The formal assignment of expenditures under the Law on Local Government Organization (June 1993) indicates that, prima facie, there is no blatant mismatch of economic assignments from the perspective of the standard textbook division of central and local government (Table 2).9 Many services, such as local public transport, road maintenance, garbage collection, that should be provided by local governments on the basis of the “benefit principle” are indeed assigned to local governments.10 Furthermore, typical central government functions such as defense and foreign economic relations are the responsibility of the central government.

Table 2.

Estonia: Expenditure Assignments

article image
Source: World Bank, Estonia: Financing Local Governments, Report No. 14925-EE, December 1995, based on Law on Local Government Organization (June 1993).

The Medical Insurance Fund covers individual medical expenses.

9. Yet a closer examination of the assignment of responsibilities and its functioning reveals at least two issues of concern.11 First, the principle of assigning all components of service delivery to one level of government is not observed in the education sector, which undermines cost management. For example, textbooks and the salaries of primary and secondary school teachers are the responsibility of the central government, while all other expenditures, including maintenance of schools come out of local budgets. Second, according to the Union of Local Governments, up to 167 laws and regulations assign expenditure activities to local governments outside the norms established by the Law on Local Government Organization. In one example, the local government reforms of 1993 gave responsibility for fire and emergency services to the central government, while the Law on Fire Protection states that local governments must provide these services. This might be less of a problem if the additional assignments were backed up by adequate revenue sources and institutional capacity at the local level to implement and monitor programs. In practice, however, there is a risk that such services fall between the cracks and become a function of intergovernmental bargaining over the budgets.

C. Revenue Assignments

10. Local governments have three main sources of revenue: (1) local taxes, fees, and charges; (2) revenue sharing in central government taxes; and (3) transfers from the central government. Under current arrangements, the central government shares only two major taxes, the personal income tax and the land tax, and keeps the revenues from indirect taxes.12 Shared revenues and transfers have been the two key sources of revenue at the local level throughout the transition, but since 1994 when far-reaching local government reforms took effect, there has been an increase in the importance of shared revenues and a corresponding decline in that of transfers.

Local Taxes

11. The Law on Local Taxes (October 1994) allows local governments to levy nine different taxes, notably a poll tax payable by residents between the ages of 18 and 65; a local corporate income tax payable by enterprises and insurance companies registered in the municipality;13 a local sales tax payable by individuals and enterprises engaged in business activities in the municipality; and six other minor taxes on luxury boats, on advertising, on keeping domestic animals, on closing streets for construction or events, on entertainment, and on motor vehicles.

12. The proportion of local tax revenues in total revenues (including transfers) of the municipalities is small, an average of 9 percent during the period 1993-96 (Table 3). Another way to measure this fiscal imbalance is to calculate the share of local government expenditures financed with revenues under the control of local governments. Two calculations of a coefficient of vertical imbalance can be undertaken, one which considers only own local revenues to be under the control of municipalities, and the other which also includes revenues from revenue sharing in this category. A value closer to zero indicates greater fiscal imbalance. These coefficients suggest that the degree of vertical imbalance in Estonia, while still important, has declined considerably (Table 4). According to the first coefficient, the degree of vertical imbalance in Estonia was very high in 1994, but by 1996, had fallen to levels that had prevailed several years earlier in a number of western European countries; when shared taxes are included in revenues controlled by local governments, the coefficient for Estonia is closer still to the figures for those countries.

Table 3.

Estonia: Structure of Local Government Budgets, 1993-96

(In percent)

article image
Sources: Estonian authorities; and Fund staff estimates.
Table 4.

Estonia: Measurement of Vertical Fiscal Imbalance, 1994-96 1/

article image
Sources: World Bank (1995), Table 4; Estonian authorities; and Fund staff estimates.

Coefficient 1 = Share of local government expenditure covered by own revenues; Coefficient 2=Share of local government expenditure covered by both own and shared revenues.

The land tax is included in shared revenue; transfers to local governments and expenditures financed by foreign borrowing through the central government are excluded from local government expenditures.

13. Despite the improvements brought in by the 1994 legislation, local governments in Estonia do not have an adequate level of tax autonomy, i.e. one that would allow them to finance a large share of their expenditure needs through own taxes, thereby raising the efficiency and accountability of the municipalities. Many of the local taxes are no more than “nuisance charges,” and a vehicle registration tax, generally a tax that is equitable and a good source of revenue, has not been widely used outside Tallinn.14 There are other problems with some of the taxes introduced in the 1994 legislation. First, on the basis of principles of tax assignment, it is not generally appropriate to impose a corporate income tax at the local level—municipalities have the opportunity to export taxes (i.e. to tax goods and services provided to the entire county to finance local government services), and apportioning the corporate income of businesses deriving income in more than one region is difficult. Second, the current local sales tax is a turnover tax with adverse cascading effects. Third, the additional local insurance premium tax, which is paid by residents throughout Estonia, is basically a tax for the city of Tallinn because insurance companies are concentrated in the capital.

Shared Taxes

14. Current revenue sharing arrangements give local governments 56 percent of the personal income tax revenues and 100 percent of the national land tax (collected by the National Tax Board). The central government keeps all the revenues from the other principal taxes, i.e. the VAT, excises, and the corporate income tax. There has been considerable change in tax revenue sharing over the years. From 1990 to 1993, the personal income tax revenues were assigned in their entirety to local governments, an allocation which undermined the central government’s control over fiscal aggregates; since then, they have been shared between the central and local governments in a ratio of 48:52 during 1994-95, and in a ratio of 44:56 as of 1996. And initially, both the corporate income tax and the land tax were shared between the two levels of government.15

D. Transfers and Equalization

15. The Law on the Correlation between Municipal and Town Budgets and the State Budget in force from January 1, 1994 sets out the current system of intergovernmental transfers. It calls for the overall level of transfers from the central government to be decided in negotiations between central government representatives and those of local governments, including the Union of Local Governments. However, if the negotiations fail, the central government alone is to decide upon the transfer amount.

16. There are five different types of transfers in Estonia: (1) transfers to achieve equalization support; (2) transfers to compensate local governments for losses in revenue sharing owing to personal income tax exemptions; (3) transfers from the Support Fund designed to cover the gap between actual revenues and prescribed minimum expenditure norms, to support those local governments with unanticipated losses in revenues, and to support local governments in special circumstances; (4) transfers paid out for central government services actually delivered by the local governments; and (5) transfers to support capital investment needs of smaller local governments (the largest five cities are excluded), which are distributed on a discretionary basis.

17. The system of transfers places considerable emphasis on equalization in view of the highly uneven distribution of fiscal resources: the National Tax Board (NTB) collects more than three quarters of all tax revenues in the city of Tallinn, which represents approximately 30 percent of the population.16 In 1995 and 1996, equalization transfers of EEK 759 million (1.8 percent of GDP) and EEK 703 million (1.3 percent of GDP), respectively, were distributed to local governments.

18. The calculations underlying equalization transfers were initially based on the Correlation Law: first, local governments were ranked by the ratio of their qualifying revenues per capita to the national average;17 second, each local government was classified into one of 12 categories according to the ratio of its revenues to the national average; third, a “smoothing” factor was defined for each class ranging from 1.15 for those with ratios below 10 percent of the national average to 0.05 for those local governments with ratios above 110 percent; fourth, individual local government claims to the total funds available for equalization were derived by multiplying the smoothing factor by the population of the respective local government; and fifth, the actual equalization transfer to a local government was obtained by multiplying its relative claim (the ratio of its claim to total claims) by the total fund available for equalization. A new formula for determining equalization transfers was adopted in 1996:

T = (m x ak- an) x 0.9 x ct-1


T = level of equalization transfers to individual municipality

m = coefficient of support for each municipality

ak = average level of qualifying revenues per capita

an = level of qualifying revenues per capita of municipality

ct-1 = population of municipality at beginning of previous year

The coefficient of support is based on agreement between the central government and municipality, and is not expected to change unless the municipality’s circumstances do (e.g. expenditure assignments). This revised methodology is an improvement in two important ways, namely, it is much simpler than the previous one, and it does not allocate resources to Tallinn (which is characterized by a high level of qualifying revenues per capita).

E. Local Government Borrowing

19. Local government borrowing, which consists of own borrowing and of foreign borrowing through the central government (i.e., the use of foreign resources on-lent by the central government), has expanded rapidly in recent years, albeit from a very low base.18 Before 1994, borrowing by local governments at their own initiative was limited to domestic bank credits to bridge revenue shortfalls. In 1994, the stock of commercial bank loans to local governments increased to EEK 108 million, up from EEK 1.5 million in 1993, and thereafter increased by more than 250 percent to EEK 393 million in 1996.19 At the same time, bond issues became an increasingly important channel for local government borrowing. In 1996, bond issues represented 80 percent of own borrowing, though this includes a foreign bond of EEK 480 million (DM 60 million) issued by the city of Tallinn; domestic bond issues were roughly on a par with borrowing from domestic financial institutions (Table 5). The data for 1996 are indicative of the disproportionate share of total local government borrowing undertaken by the city of Tallinn.

Table 5.

Estonia: Local Government Borrowing, 1996

(In millions of kroons)

article image
Sources: Estonian authorities; and Fund staff estimates.

These data are based on a survey conducted by the Ministry of Finance.

Bond of DM 60 million issued by the city of Tallinn in April 1996.


20. Local governments in Estonia have considerable autonomy. Within this framework, the central government has followed a rules-based approach to keeping municipal borrowing in check; one such rule restricts borrowing to investment projects (the so-called “golden rule” adopted in a number of countries). However, the limits on local government borrowing incorporated in the Law on Municipal and Town Budgets (effective January 1, 1994) proved to be inadequate owing mainly to their ambiguous wording. Also, this rules-based approach can end up fostering practices aimed at circumventing the regulations, e.g., the reclassification of expenditures from current to capital, and the use of local government-owned enterprises to borrow for purposes that should be funded through the budget.20

21. To remedy these shortcomings, the government introduced amendments to the law, which were adopted by parliament in May 1997. According to the revised legislation, the stock of outstanding debt of local governments—excluding all loans and transfers from the state budget—cannot exceed 75 percent of revenue projected for the current budget year, and debt service cannot exceed 20 percent of revenue projected for the current year. Furthermore, local governments must forward a copy of the decision to take a loan to the county governor within three days after it becomes effective, and send a copy of the loan contract to the Ministry of Finance within thirty days after the contract has been signed. In addition, in August 1996, the Government of Estonia issued a decree clarifying that the central government would not extend state guarantees—implicit or explicit—against municipal borrowing. By comparison, the other two Baltic countries have instituted strict administrative controls. In Latvia, local governments can borrow only from the Treasury for investment projects deemed worthy, and in Lithuania, local governments can borrow to finance investment expenditures only after receiving central approval (from the State Loan Commission).

F. Conclusions

22. Estonia has made considerable progress in carrying out fiscal decentralization. Just about five years ago, it began the transition to a market economy with a highly centralized system of public finances, local governments acting mainly as administrative units with no independent fiscal responsibility. Since then, its system of intergovernmental relations has promoted institutional settings and processes that allow for the articulation of interests and policy-making based on consensus-building, issues that are raised, for example, in the European Charter of Local Government of 1985.21 There has also been a move to match public services more closely with local demands. The design of transfers from the central to the local governments, with its emphasis on equalization, attempts to place limits on the level of open-ended transfers that could create incentives to attract and self-generate local demand. Moreover, detailed analysis by the World Bank indicates that the transfer mechanism in Estonia has served to reduce the horizontal imbalances of the fiscal system. It lessens the fiscal disparities among local governments with different tax bases and closes the revenue gap of local governments (i.e. the imbalance between expenditure responsibilities and the funding available).

23. The system of intergovernmental relations needs, however, to exploit more fully the benefits of fiscal decentralization. It must also ensure that local governments support the general effort to increase public savings, and that budget constraints are enforced at each level of government. Given the passive role of local government over a fifty year period, this will require in part a concerted move to build up local capacity and a local civil service trained to design, monitor, and implement expenditure programs in its areas of responsibility. Against this backdrop, there are a number of steps that the authorities could take:

  • First, consolidation of local governments and increased cooperation among them in the delivery of public services would improve the efficiency of municipalities. In this regard, the 1998 draft state budget allocates EEK 50 million to get the process started (it will no doubt take a long time to complete): the first ten municipalities to merge into larger units will each receive EEK 3 million, and those that follow will each receive EEK 1 million.22

  • Second, in the case of expenditure activities that are subject to overlapping or competing jurisdictions, the appropriate level of responsibility needs to be clarified. In addition, local capacity to monitor and implement programs under the jurisdiction of municipalities will have to be built up.

  • Third, in order to increase the accountability of local governments and enhance their flexibility to meet the demand for services, a greater share of local expenditures should be financed with local taxes—as distinct from shared taxes which represent a transfer from the central government (albeit out of an earmarked revenue source). One option, already adopted in a number of countries, would be to give each local government the right to apply its own tax rate on the national income tax base.23 A local income tax surcharge would offer local governments a powerful revenue tool and increase the link, at least at the margin, between local public services and taxes payable. This would not exclude the use of supplementary equalization transfers for poorer municipalities. Another option would be to increase municipal user fees.

  • Fourth, the new method for determining equalization transfers, while simpler and better at targeting local governments with per capita revenues lower than the national average, leaves the door open for considerable bargaining between the central government and individual municipalities—the authorities will have to ensure that such negotiations do not undermine the objective of reducing fiscal disparities across regions.

  • Fifth, the borrowing activities of local governments need to be better monitored and the legal restraints on debt obligations vigorously enforced. The authorities have recently taken steps to address this issue. In addition, as part of its prudential responsibilities over the banking system, the Bank of Estonia will require (effective December 1, 1997) that banking groups obtain in advance a letter of no-objection from the Ministry of Finance (which will not represent either an explicit or implicit guarantee) for each additional financial claim acquired on local governments.


  • Ahmad, Ehtisham, Daniel Hewitt, and Edgardo Ruggiero, 1997, “Assigning Expenditure Responsibilities” in Fiscal Federalism in Theory and Practice, ed. by Teresa Ter-Minassian (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Arora, Vivek B., and John Norregaard, 1997, “Intergovernmental Fiscal Relations: The Chinese System in Perspective,IMF Working Paper 97/129 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Bogetić, Željko,Bulgaria” in Fiscal Federalism in Theory and Practice, ed. by Teresa Ter-Minassian (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Norregaard, John, 1997, “Tax Assignment” in Fiscal Federalism in Theory and Practice, ed. by Teresa Ter-Minassian (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Statistical Office of Estonia, 1997, Regional Statistics of Estonia (Tallinn).

  • Ter-Minassian, Teresa, ed., 1997, Fiscal Federalism in Theory and Practice (Washington: International Monetary Fund).

  • Ter-Minassian, Teresa, and John Craig, 1997, “Control of Subnational Government Borrowing” in Fiscal Federalism in Theory and Practice, ed. by Teresa Ter-Minassian (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • World Bank, 1995, Estonia: Financing Local Governments, Report No. 14925-EE (Washington: World Bank).

  • World Bank, 1996, World Development Report: From Plan to Market (New York: Oxford University Press).

  • World Bank, 1997, World Development Report: The State in a Changing World (New York: Oxford University Press).


These papers on selected issues and the statistical tables provide background information to the Staff Report on the 1997 Article IV Consultation and Request for Stand-By Arrangement (EBS/97/212, 11/25/97) for Estonia.


This note draws heavily upon a comprehensive study by the World Bank, Estonia: Financing Local Governments, Report No. 14925-EE, December 1995.


For a summary of the advantages and disadvantages of fiscal decentralization, see World Bank (1997), Chapter 7; and Arora and Norregaard (1997).


These figures are on an unconsolidated basis.


Unless indicated otherwise, the terms local government and municipalities are used interchangeably in this note.


The cities of Tallinn, Tartu, Kohtla-Jarve, Narva, Parnu, and Sillamae.


Consolidated general government expenditures in Estonia, at roughly 40 percent of GDP, are high and may seem to be at odds with a market-based economy and with a relatively low degree of importance attached to the state. However, it is important to keep in mind that the enlarged public sector activities in Estonia are very small compared to other transition economies.


These shares relate to local government expenditures excluding those financed by foreign borrowing through the central government.


For example, see Ahmad et al (1997).


This principle suggests that a given service should be offered by the level of government most closely representing the region that benefits from such service.


The detailed examples given here are based on arrangements through 1995. See World Bank (1995).


Revenues from the tax on water use are also shared with local governments (Law on Environmental Protection, effective 1993).


The local income tax may not exceed 2 percent of taxable income or 0.4 percent of premia for insurance companies.


A national car tax, superseding Tallinn’s local tax, is to be introduced in 1998; all of its revenues will be channeled to the local governments.


The corporate income tax was shared between the central and local governments in a ratio of 65:35 during 1990-92 before reverting back to the central government. The land tax was shared in a ratio of 63:35 during 1990-92, and of 50:50 during 1993-95.


The concentration of fiscal resources in Tallinn is to a large extent due to the concentration of VAT revenues collected from imports, and to a lesser extent, to the collection there of the personal income tax, the corporate income tax, and other indirect taxes.


Qualifying revenues included the land tax and shared revenues from the personal income tax plus transfers for compensation for exemptions; the latter were suspended in 1995. See World Bank (1995).


The data on local government borrowing remain weak. In the only systematic overview of local government borrowing, a survey launched by the Ministry of Finance in April 1996, the data are incomplete in that some municipalities do not respond; are not up-to-date; are not presented consistently (some municipalities include municipal enterprises, others do not); and are not consistent with banking data.


Based on banking data. Domestic bank loans outstanding to local governments amounted to EEK 478 million (3.4 percent of domestic bank credit) at end-June 1997.


In addition, the central government would undertake to pay at least as much in support to the merged local units as they received before the merger. Each municipality that becomes part of a larger unit would also be entitled to EEK 400,000 for merger costs, and EEK 8 per resident for the costs of new local elections.


This system is applied, for example, in Scandinavian countries. See Norregaard (1997).

Republic of Estonia: Selected Issues and Statistical Appendix
Author: International Monetary Fund