Bosnia and Herzegovina: Selected Economic Issues

This paper assesses the extent to which data unreliability could alter the assessment of the macroeconomic challenges ahead. The contributions of the indirect tax authority (ITA) in remedying the flaws are highlighted, and the architectural agenda is discussed. Fiscal sustainability and the government’s bold initiatives to secure it by restructuring the domestic claims have been assessed and key implementation issues in realizing the government’s plans noted. A survey of selected tax policy issues in Bosnia and Herzegovina is also included in the paper.


This paper assesses the extent to which data unreliability could alter the assessment of the macroeconomic challenges ahead. The contributions of the indirect tax authority (ITA) in remedying the flaws are highlighted, and the architectural agenda is discussed. Fiscal sustainability and the government’s bold initiatives to secure it by restructuring the domestic claims have been assessed and key implementation issues in realizing the government’s plans noted. A survey of selected tax policy issues in Bosnia and Herzegovina is also included in the paper.

III. Fiscal Architecture1

A. Introduction

1. General government has an unusually large number of highly autonomous parts. It includes the main layers of government in both Entities (in the RS, Central Government and Municipalities; and in the Federation, Central Government, Cantons, and Municipalities), the State government, and extrabudgetary funds (including the Health Insurance Funds, and Pension Funds) in both Entities. In addition, there are a variety of off-budget foreign donor funded fiscal activities—aid programs (Table 1 and Figure 1). In different ways, each of these units and activities enjoys a high degree of autonomy. This reflects the compromises made at Dayton to secure peace, and the legacy of the fiscal system of Former Yugoslavia.

Table 1.

Bosnia and Herzegovina: General Government, 1999–2004

(In percent of GDP, unless stated otherwise)

article image
Sources: Ministries of Finance; and IMF staff estimates.

Includes transfers to Cantons and extrabudgetary funds.

Pension Fund, Health Fund, and Employment Fund. Also includes the Children’s Fund in the RS.

Includes transfers to extrabudgetary funds.

Data for 1999 and 2000 are incomplete.

Figure 1.
Figure 1.

Bosnia’s Fiscal House: Present Time

Citation: IMF Staff Country Reports 2004, 054; 10.5089/9781451804812.002.A003

2. This architecture greatly complicates fiscal management—so much so that the fiscal system is more often a source of problems than an effective tool to solve them. This note outlines key macroeconomic difficulties—notably conecerning delivery of an appropriate consolidated fiscal balance—that current arrangements give rise to, it notes initiatives under way to correct some of these difficulties, and the issues which remain to be resolved.

B. Four Architectural Failings

3. Current arrangements give rise to macroeconomic concerns in four areas: coordination of policy towards the consolidated fiscal balance; design of indirect tax policy; separation of responsibility for taxes and expenditures; and funding arrangements for the State.

4. First, no institution or body in BiH is responsible for targeting and monitoring the consolidated fiscal balance. In most fiscal systems, this task falls to the central Ministry of Finance. But in BiH, the State Ministry of Finance and Treasury plays a minor role in the overall fiscal system, accounting for less than 4 percent of consolidated government spending. The lack of an institution or arrangement to secure an appropriate consolidated fiscal stance in the otherwise extensive fiscal architecture established by Dayton is particularly remarkable given that a sound consolidated fiscal stance is an essential support for the currency board, and that given the currency board, fiscal policy is the sole remaining discretionary macroeconomic policy instrument.

5. In the absence of such an institution, each independent component of fiscal system has incentives to free-ride on the others, undermining incentives to secure an appropriate fiscal balance. The individual component fiscal institutions—in the Entities and so on—have no incentive to take account of the appropriate aggregate fiscal stance when determining their own balances. In addition to this incentive problem, few of the component parts of government publish data on their activities, which inhibits the ability of “outsiders” to monitor and discipline those developments. The consequent “free-riding” problem undermines essential fiscal support for the currency board.

6. Controls on borrowing are also weak. These are set in the budget laws, with sub central government units and extra budgetary funds often allowed to borrow for investment and short-term purposes, subject to ceilings relative to revenue. But ex ante and ex post central monitoring of these activities is inadequate, it is ambiguous if the ceilings refer to stocks or flows, arrears accumulation is not addressed, and extra budgetary funds are often not covered by the rules. Regulations on the provision of government guarantees are also loose. The weaknesses of these rules compounds the free rider problem by allowing scope for the various component parts of government to act relatively freely.

7. Market discipline—in the form of risk priced into the cost of credit—has proven to be a weak force to address these problems. Certainly, access to official external credit is tightly disciplined by the creditors and is increasingly differentiated between the two Entities. Alongside, access to external and domestic commercial bank credit has, so far, been relatively modest. But access to credit in the form of spending arrears has been large and—by definition—undisciplined. This partly arises from the lack of a system of commercial courts through which payment can be enforced, and it gives rise to part of the challenge of domestic claims on government (See Chapter 4). But even if all creditors were to price country and exchange rate risk into yields for the individual fiscal units, those units would still obtain credit at yields which predominantly reflect the policies of other units. So the component parts of the fiscal system would still lack price incentives to align their individual activities with the necessary aggregate stance. Market discipline on the free rider problem is thus weak, and would still be weak even if tolerance for spending arrears was reduced.

8. So weak is the authorities’ focus on the aggregate fiscal stance that data on consolidated revenue, spending, and the balance are not even collected. The Central Bank has recently initiated work to collect consolidated fiscal data some two years in arrears on a GFS basis. But no institution collects up to date consolidated fiscal data, even though this merely requires adding up the data which the component fiscal institutions produce. This is a particularly acute case of data weaknesses which are discussed in Chapter 2.

9. Second, indirect tax policy has been set and administered at Entity level. Given free flow of goods across the inter-entity borders, this has led to considerable tax arbitrage when policies—and their administration—have not been harmonized. This has greatly complicated preparation of Entity revenue projections and determination of the related spending commitments, and has on occasion caused heavy losses in revenue. Accordingly, this has weakened the necessary overall fiscal support for the currency board. A concerted effort over several years has helped to move indirect tax systems towards harmonization, but until recently, harmonization was maintained solely at the discretion of each of the various component parts of the fiscal system. The Entities also operate different corporate income and personal income taxes, but given the slower pace of arbitrage in activity covered by these taxes, the associated risks to the budgets have been more muted than has been the case for indirect taxes, accordingly giving rise to less macroeconomic risk.

10. Third, tax and expenditure powers are often separated. For example, policy on external tariffs is set by the State, but the revenue accrues to the Entity central governments to fund their spending. In addition, in the RS, sales tax revenue partly accrues to municipalities, and in the Federation, it accrues entirely to the Cantons and municipalities, but sales tax policy in both cases is under the purview of the central Entity governments. The key risk arising from such divorces of decision-making over revenue from that over spending is that policy on revenue takes insufficient account of spending ambitions, putting those spending ambitions and delivery of the appropriate fiscal stance at risk. And this separation does not only refer to tax and spending; external finance raised by the Entities requires authorization by the State.

11. These “separation” risks could theoretically be resolved with effective intergovernmental coordination, but this has proved elusive. Coordination even within coalitions which straddle the different tiers of government has been lacking, perhaps partly reflecting the conflicts which gave rise to these institutional separations in the first place at Dayton. And, in the case of policy on customs tariffs, such coordination between the different tiers of government is further complicated because customs revenue is a much higher share of Federation central government revenue than it is for the RS central government, so that the fiscal implications of any given tariff policy for the two affected Entity governments are very different in each case.

12. Fourth, as a special case of this separation of revenue and spending authority, arrangements to fund the State are ad hoc. Funding for State activities is almost entirely provided directly by the Entities. The related transfers are determined in annual ad hoc negotiations between the Entities and the State, and the monthly transfers are made at the (monthly) discretion of the Entities. This renders the State hostage to the Entities and compromises delivery of an appropriate consolidated fiscal balance for BiH even though the State is small.

13. In addition to these four macroeconomic architectural concerns, a number of further issues arise, straddling macro and micro concerns. There are no transfer arrangements to equalize spending between or within Entities. Partly as a result, pension and budget remuneration rates vary significantly for similar persons between the Entities, between them and the State, and between all these and Brcko District. These inequities aggravate pressures for upward equalization, a force for increased total spending. Lack of coordination of initiatives in defense, judiciary, intelligence services, police, and war veterans have caused a variety of frictions and inefficiencies, and even on public investment projects which affect both Entities—such as road building—coordination is poor. And the proliferation of extra budgetary funds diminishes control and transparency.

14. These characteristics of the fiscal architecture have required an unusually deep engagement by the IMF to address. Understandings in Fund programs on fiscal balances of the component parts of government alongside general embargoes on accrual of new spending arrears have addressed the free rider problem directly, and therefore have established appropriate aggregate target fiscal stances. Understandings on harmonization of indirect tax policy have helped to eliminate destabilizing tax arbitrage. And the Fund has often played a key intermediating role in resolving issues raised by the separations of revenue, spending external borrowing powers and the annual and monthly administrative transfers to the State. The Fund staff have also consolidated fiscal data, ensuring that the evolution of aggregate fiscal developments has been monitored.

15. But the unusually prominent role of the Fund in making the architecture work is symptomatic of the need for profound reform, which is now underway. In addition to standard challenges of operating a fiscal system which arise in any country—unanticipated shocks to the economy or policy and capacity constraints etc—one of the insufficiently noted costs of the multiple failings in the Bosnian fiscal architecture is that the energy required of the authorities simply to operate their system has detracted from the resources they have available to reform it. But reform is nevertheless underway, spearheaded by initiatives in the framework for indirect taxation.

C. Reconstructing the Fiscal Architecture

16. Reforms of the framework for setting indirect tax policy underway make major headway to resolve many of these architectural weaknesses. Most obviously, they resolve finally the problem of harmful tax competition in indirect tax policy and administration. But they also could fill the gap in the Dayton fiscal architecture by establishing an institution with an interest—and possibly eventually a mandate—to set and monitor delivery of appropriate consolidated fiscal stance, though key details on this still need to be determined. And even though, as noted below, these reforms are unlikely to completely resolve the free rider problem, the proposals tackle many of the architectural failings.

17. During 2003, a commission on Indirect Taxation mapped out a new framework closely following earlier IMF recommendations (Box 1)1 Accordingly, the Entity and Brcko customs administrations will be merged into the ITA, all indirect tax legislation will henceforth be State level rather than Entity level legislation, with their parameters determined by the Board of the ITA. This arrangement secures harmonization of administration of all indirect taxes and customs, harmonization of indirect tax policy across the whole country, and, through the voting rules of the Board, a close link is maintained between decisions on indirect tax policy and spending to be funded from the associated revenue.

The Framework Law on Indirect Taxation

The law was approved by all Entity and State legislatures in late 2003. Indirect taxes, as defined in this law, include import duties, excises, the sales tax, the VAT, road fees, and other taxes levied on goods and services. These yield revenue of 17 percent of GDP or close to 40 percent of total government revenue in 2003. It establishes the ITA, an independent State-level organization with its headquarters in Banja Luka (capital of the RS) and four regional offices. The latter will be set-up based on functional organization (e.g., audit, enforcement). The ITA Director was appointed in December 2003 by the State Council of Ministers for a period of five years. All indirect taxes will eventually be collected by the ITA, although there are some exceptions for the transition period until the VAT is introduced (see Chapter 5).

The ITA’s Governing Board (GB) consists of six members (the Entity and State Ministers of Finance, and three experts on indirect taxation, one each to be nominated by the respective governments. The ITA Director and a representative from the CBBH are observers on the Board. The Board meets at least monthly. The Chairman of the GB is to be elected by majority vote from among the Board members for two years. An “Initial Chairman”—who is not a BiH citizen—shall be an additional member of the GB for the first five years and this post has been filled.

The GB is in charge of indirect tax policy. If the GB is unable to reach decisions by consensus, various tie-breaking rules apply.

  • a. Decisions on changing import duties can be taken by simple majority, but this has to include the vote of the State Minister of Finance reflecting the role assigned to the State in Dayton in regard to external tariff policy.

  • b. For other decisions (e.g., on rates or exemptions of the sales tax or VAT), a simple majority of votes is required, including the votes of both Entity Ministers of Finance. This reflects the practice that Entities have designed indirect tax policy and received the associated revenue. It also maintains the link between authority over revenue policy and the spending funded by it at Entity level.

  • c. Decisions on revenue distribution can only be taken by a majority including the votes of all three Ministers of Finance.

The GB will advise all the fiscal authorities on the implications of indirect tax policy for public finances with the objective to prepare fiscally responsible budgets. It can initiate the preparation of new laws, the amendment of existing laws, and the issuance of regulations on indirect tax policy. The GB presents its proposals to the State Council of Ministers, and subsequently to the State legislature, for approval. Prior to this, the GB can give the parliamentary budget committees in the RS, Federation, and Brcko District the opportunity to comment on draft laws or law amendments, if deemed necessary.

All revenue collected by the ITA will accrue to a single account. The funds will be allocated first for external debt service (through the State), then to fund the administrative transfers due from the Entities to the State, then for operating costs of the ITA, and the remainder will be allocated to the Entities and Brcko District according to a formula based on their shares in final consumption. The details of this revenue allocation formula will need to be determined.

18. In addition, the authorities envisage developing the ITA Governing Board into a body to coordinate broader fiscal policies. As the only institution where the three key Ministers of Finance will be formally required to coordinate, it provides a natural venue to develop coordination more generally. The Board could thus double as a National Fiscal Council (NFC). At the outset, such broader coordination will be entirely voluntary to reflect the sensitivities which gave rise to the devolved constitutional arrangements in Dayton. Any development of its role will require consensus.

19. The NFC could address several more of the architectural difficulties noted above. It could provide a venue where the authorities could propose targets for and monitor delivery of the consolidated fiscal balance. Within its current powers, it could also make proposals on the apportionment of the targeted consolidated balance between the various component fiscal units—including the Entities and the State. And it could advise on expenditure reform policies that affect both Entities and the State and guide formation of direct tax and public sector borrowing rules.

20. The NFC’s voluntary status means it can only ameliorate the free rider problem, rather than end it. As indirect tax policy is rendered subject to mutual vetoes, negotiations between the various partners over the exercise of these vetoes could take into account such broader issues, such as the free rider problem—”I will agree to your indirect tax proposal if you agree to a deficit target acceptable to me.” Thereby, the free rider problem could be addressed, even if not completely resolved. Complete resolution of this problem may require formal development of some form of “internal stability pact,” aiming to achieve for Bosnia what the Stability and Growth Pact (SGP) was aimed to achieve for the Euro area. But even leaving aside the tricky issues of enforcement and penalties, the formulation of such a pact will be technically and politically demanding. Given the existence of the State, which covers the whole country, ceilings on borrowing or deficits cannot be defined on a uniform basis relative to GDP for each component of the aggregate fiscal system—the approach implicit in the SGP which defined limits for each member country. So an even more complex system may be required, given the current constitutional arrangements and the reluctance to transform them.

21. In any event, an institutional setup will be required for the NFC. Given the role for it outlined above, its natural membership would be the three Ministers of Finance with the Central Bank as observer. But it could also be the same membership as for the GB of the ITA—the three ministers, plus one expert nominated by each government. It could be supported by a small secretariat, perhaps lodged within the ITA but separated from the ITA’s tax administration functions. And if its role is to deepen, it may need a legal mandate at some point (i.e., delegated decision making power from Entity and State parliaments) as well as well-specified decision-making rules, which may differ by type of issue. All this lies ahead. Note, however, that unlike NFCs elsewhere, which are often aimed to depoliticize fiscal policy by handing it over to experts, the proposal here is motivated by need for more effective fiscal policy coordination, not depoliticization.

22. The ITA and NFC proposals require support from reform in three other areas—within-Entity tax assignments and fiscal practices, the rules on borrowing, and determination of annual funding for the State.

  • In the Federation, the assignment of sales tax revenue to the Cantons and the consequent heavy reliance of its central government on customs revenue concentrates the exposure of both to revenue shocks. A new intra-Federation revenue allocation formula could give each a share of all taxes, though any reforms should anticipate VAT (Chapter 5). Also in the Federation, internal intergovernmental liaison needs strengthening, notably through regular monthly meetings between the Canton and central government Ministers of Finance, and the practice should be formalized.

  • Rules on borrowing by each part of the fiscal system need to be strengthened, notably to ensure that the NFC members have the authority to ensure that their Entities’ consolidated fiscal balance outturns will be consistent with agreements they commit to at the NFC. The practice of accrual of spending arrears has much diminished over the past two years. But the discipline to stop it will require completion of reforms underway to establish commercial courts and the associated enforcement of payment.

  • With the monthly transfers to the State secured by the ITA, a formal system to determine annual administrative transfers to the State is required to replace the ad hoc arrangements now in place. A rule could be based on a percentage share of some revenue streams—perhaps all indirect tax revenue—or on a simple fixed nominal annual increment each year. The former approach would have the drawback that the impact on the State budget of any change in indirect tax policy would be greater than its impact on Entity budgets because all its revenue would come from that source while only part of Entity’s revenue would come from that source. This inequality would cloud decisions on adjustments to indirect tax policy. The latter would avoid these problems, but the number chosen for the automatic annual increase would have an arbitrary element. Either approach would have to allow for adjustments which may be made to the expenditure functions of the state, and for need to encourage the State to secure economies in spending on its activities. A formal rule, on either basis, would allow better medium-term planning by both the Entities and the State and therefore help to secure an appropriate consolidated fiscal stance.

D. Conclusion

23. The fiscal system established at Dayton is undergoing fundamental reform. But key elements of the indirect tax reforms which are spearheading those efforts have yet to be worked out, notably in relation to the tasks and functioning of the proposed NFC. And even if those matters are satisfactorily resolved, key revenue assignments, borrowing rules, and arrangements for the administrative transfers to the State still need to be determined. Critically, the reforms underway retain the spirit of Dayton—pooling sovereignty rather than transferring it, and they therefore work within the compromises which were necessary to secure peace. But they seek to address the worst effects of the Dayton compromises on aggregate fiscal policy. Thus, within those constraints, the fiscal system is becoming a more effective tool with which to solve problems, but there is still some way to go.


Contributed by Peter Doyle and Gunther Taube.


IMF Country Report No 03/204, July 2003, page 15, paragraph 25.

Bosnia and Herzegovina: Selected Economic Issues
Author: International Monetary Fund