Diagnosing and Addressing Serbia’s Structural Fiscal Challenges1
Serbia’s fiscal policies and outcomes have been undermined by the political cycle and relatively weak fiscal institutions. Fund arrangements have helped improve policy implementation but progress has been undone between programs, especially during political transitions. Comprehensive public financial management (PFM) reforms are necessary to strengthen fiscal outcomes.
A. Evolution of Serbia’s Fiscal Challenges
1. Serbia has had a difficult fiscal legacy. The 1990’s were marked by severe macroeconomic imbalances, including a serious output contraction and hyperinflation. High fiscal and quasi-fiscal deficits reflecting chronic government arrears, repressive taxation, and populist public spending initiatives undermined the fiscal position and the economy more broadly. IMF arrangements that started in the early 2000’s were instrumental in bolstering fiscal discipline by scaling back central bank credit to the public sector, advancing structural fiscal reforms to contain the pension and public wage bills, and reorienting resources from current to capital spending. But the fiscal position started to weaken in 2006 (at the height of the pre-crisis boom, when an IMF program expired), exhibiting pro-cyclical features. While the headline deficits did not appear high then, they were considerably higher than the fiscal balances implied by a countercyclical deficit rule that Serbia later adopted in 2010.
General Government Fiscal Balance
Sources: IMF staff estimates.
2. Overall, Serbia’s fiscal performance in the post-1990 period was uneven and relatively disappointing. Periods of improvements in fiscal policies—such as shifting to more countercyclical more rules-based policies (including during the 27-month IMF standby arrangement ended in April 2011)—alternated with setbacks and slippages in fiscal reforms. As a result, Serbia was the only country in the CESEE region that saw a significant increase in the fiscal deficit since the 2008–09 crisis started and the increase in the public debt ratio has correspondingly been among the highest in the region.
CESEE Fiscal Balance
Sources: IMF, World Economic Outlook database.
CESEE: Public Debt
Sources: IMF, World Economic Outlook database.
B. Diagnosing Serbia’s Fiscal Challenges
Assessing political fundamentals of fiscal pressures
3. Both political and institutional factors may have contributed to the fiscal challenges faced during the past decade.
- Political factors. Since the late 1990s elections, Serbia’s has had complex coalition governments. This has arguably been accompanied by political fragmentation (with key sectors split among parties and limited top-down control of economic policies); and reduced accountability for the overall fiscal implications of policy choices.
- Ingrained preference for high public spending. Like other Republics of former Yugoslavia, Serbia inherited a “cradle-to-grave” system of social assistance, as well as a practice of broad-based support to “production.” Despite considerable reforms, targeting of social protection remains insufficient. Likewise, in the aftermath of some unsuccessful privatizations, there was pressure to return those enterprises to state ownership and provide subsidies to keep them operational.
- Relatively weak fiscal institutions. The system of fiscal rules, regulations, and data has been weak and improving very gradually. There have remained gaps and inconsistencies in the data and definitions, which complicate implementation.2 Also, accountability, including in the context of external and internal audit, has remained generally weak. The institutions that are tasked with enforcing accountability (such as the State Audit Institution) have yet to reach their full capacity.
4. As a result, Serbia’s fiscal developments appear to have been partly driven by electoral cycles. Indeed, proximity of elections seems to have been associated with a run-up in fiscal deficits. In some cases, the deficit increases preceded the elections, while in others they were linked to post-election developments. As to the source of deficit increases, it most often reflected spending on public wages and pensions or transfers to local governments.
Fiscal Balance and the Electoral Cycle
Sources: Serbian Ministry of Finance and Economy (MoFE) and IMF staff estimates.
5. Relatively abundant market financing in recent years has reduced incentives for fiscal adjustment. Through 2008, Serbia had relatively low recourse to deficit financing in debt markets, partly because of the implications of the 2006 Paris and London Club restructuring and partly because of the relatively large inflows of privatization proceeds. The 2008 crisis initially tightened the financing constraint for Serbia—mainly because of the drying-up of the potential for privatization proceeds. But the situation changed around mid-2009, with increased market appetite for domestic T-bills. Moreover, Serbia was able to additionally borrow in international markets by issuing a Eurobond in late 2011, and again in late 2012 and early 2013.
Gauging Serbia’s achievements of key fiscal outcomes
6. An important characteristic of any fiscal system is the policymakers’ success in delivering on the key targets. This can be gauged by comparing plans with outcomes—in terms of deficits, revenues, and spending, as well as various sub-categories for different time periods.
7. Over the past few years, fiscal deficits and debt have been significantly above their original targets set three years before. Comparing Serbia’s rolling three-year fiscal plans (adopted in 2007–12), for the general government with outcomes reveals large deviations from plans, which are steadily growing over time. While in the first year of implementation actual deficits exceeded targets by around 2½ percent of GDP on average, in the second and third years this deviation widened to around 5 percent of GDP. The situation with the public debt is similar, and the larger deviations in the ratio (than could be explained by the deficits) reflect additional increases in public debt due to the provision of numerous government guarantees (which are included in the public debt definition in Serbia) to (largely) loss-making public enterprises. They also reflect the effects of periodic bouts of exchange rate depreciation on the stock of the public debt, which is more than 80 percent denominated in foreign currencies.
Projected and Actual Fiscal Balances
Sources: Serbian MoFE and IMF staff estimates.
Projected and Actual Public Debt
Sources: Serbian MoFE and IMF staff estimates.
8. Above-planned increases in the spending ratio appear to be the main reason behind the observed deviations of deficits from plans. The three-year plans targeted a sustained reduction in the spending-to-GDP ratio, which however did not materialize. On the contrary, there was an upward drift in the spending/GDP ratio, resulting in a cumulative increase by more than 5 percentage points between 2007 and 2012. The pattern was not nearly as clear in the case of assessing the revenue/GDP ratio—periodic shortfalls in the ratio alternated with periods of higher-than-targeted revenue/GDP outcomes. While the unexpected increase in the spending ratio could reflect both spending overruns and output shortfalls relative to earlier projections, the role of spending overruns appears dominant, in part because the 2007–12 span covered not only the period of relative weakness, but also captured the end of the boom in 2007–08.
Projected and Actual Public Expenditures
Sources: Serbian MoFE and IMF staff estimates.
9. The upward deficit drift is confirmed through a closer analysis of Serbia’s annual Republican budget process. This exercise complements the above assessment of the medium-term plans, because it is focused on nominal targets (as opposed to percentages of GDP) and regards the more tightly controlled central government level. The budget process starts in the spring of the year preceding the budget year (“t-1”), with the so-called “spring memorandum,” which formulates detailed budget objectives and parameters for the year “t “for the first time. The spring memorandum is followed by: (i) the autumn memorandum that updates the parameters just prior to the final round of budget bargaining; (ii) actual budget law approved by parliament;3 (iii) a supplementary budget in the course of the budget year; and (iv) the outcome estimated after year “t” is over.4 The data for these stages of the annual budget process were examined for 2007–12.
10. Overall, Serbia’s deficits were substantially higher than those targeted at the start of the annual budget process. The increases in the deficits—relative to the spring memorandum—have affected only two of the four stages: (i) that between the spring and the autumn memoranda, which basically reflects the outcomes of initial bargaining between the Ministry of Finance and the spending beneficiaries, as well as updates to the macroeconomic outlook; and (ii) between actual and the supplementary budgets, which also reflects additional bargaining and updates to the macroeconomic outlook, but also—significantly—the pattern of budget execution and new fiscal initiatives. There is no increase in the deficit at two other main stages of the budget process (between the autumn memorandum and the budget law and between the supplementary budget and the actual execution), which essentially reflects the legal requirement not to allow increasing the deficit at those two stages.
Changes in Projected Deficit During the Budget Cycle
Sources: Serbian MoFE and IMF staff estimates.
11. The analysis of the annual deficit drift reveals that it is largely driven by (nominal) expenditure overruns rather than revenue shortfalls. In particular, on the one hand, the revenue projections in the spring memorandum actually have been quite conservative (relative to ex-post outcomes) and tend to be even more conservative in the autumn memoranda. However, during the final rounds of budget negotiations there is a substantial increase in the revenue projections. This seems to be associated with the need to accommodate expenditure pressures under a no-change-in-deficit constraint. Such somewhat unrealistic level of revenues has been maintained in the supplementary budget, and was thus usually (i.e., on average) not achieved. As regards overall nominal spending, there is a strong upward drift at all stages of the budget process, until the final outturn, which is slightly lower than targeted in the supplementary budget. The latter result is a consequence of the legally-binding rule whereby the aggregate spending ceiling as approved in the last supplementary budget has to be met. Thus, the final spending outturn is on average 8½ percent (or almost 3 percent of GDP) higher than targeted at the outset of the budget process. To sum up, the large role of spending overruns in the deficit drift reflects structural rather than cyclical factors (given that the cyclical sensitivity of spending is relatively negligible).
Changes in Projected Revenues During the Budget Cycle
Sources: Serbian MoFE and IMF staff estimates.
Changes in Projected Expenditures During the Budget Cycle
Sources: Serbian MoFE and IMF staff estimates.
12. In-year overruns reflect many causes, but most often new discretionary spending initiatives, both from new legislation and government decisions. These often take the form of new spending legislation passed by parliament that has to be eventually accommodated in the budget law. Also, overruns may reflect a seeming tendency to under-budget certain types of mandatory spending in the bargaining process (for example, social spending). This spending is mandatory, and the under-budgeting has to be fully corrected in a supplementary budget.
13. Analysis of the structure of spending overruns points to a pattern of crowding out productive with unproductive spending during budget execution. Generally, investment spending has been under-executed in budget outturns, while many current spending items have been over-executed. These latter most often regard public wage increases (including bonuses), spending on goods and services, subsidies, social assistance, and net lending.
Identifying fiscal framework gaps
14. Several elements in Serbia’s existing framework are effective and should be preserved. These include: (i) the Budget System Law (BSL), which in many respects reflects (at least on paper) best practices in public financial management, although its implementation is uneven ; (ii) a well-observed requirement that the legal deficit and spending limits (set in the last (typically, supplementary) budget law) be not exceeded; (iii) a requirement to not increase the deficits in the budget law relative to the target set in the autumn memorandum; (iv) several prudent elements related to the framework of public debt, such as including government guarantees in the debt upfront, borrowing rules for local governments, and a strict multi-step process of parliamentary approval of government guarantees.
Nevertheless, significant shortcomings remain in several areas as follows.
- Measurement of the overall fiscal position. The definition and measurement of the general government has a number of weaknesses in Serbia: (i) a complete list of institutions comprising the perimeter of the general government has yet to be precisely defined according to objective criteria (as in many countries, this task has been assigned to the statistics office but progress remains pending); (ii) indirect budget beneficiaries (IBBs), such as schools, universities, or kindergartens—are only partially captured in the government accounts; (iii) the information of government expenditure arrears is not complete and fully reliable at the central government level and limited at the local government level; (iv) there is as yet no system for recording “new” fiscal risks, such as those arising from the relatively recent PPP law and other contingent liabilities (for example, there has recently been an increase in borrowing by public enterprises without a government guarantee) ; (v) the measurement of the fiscal position is done on a cash basis, while a more economically meaningful measurement would be based on accrual methodology; (vi) some of the budget information is not timely (i.e., on IBBs).
- Budget planning. This regards: (i) gradual loss of realism in revenue projections in the course of the budget process; and (ii) overly optimistic expenditure projections at all stages of the budget process, including under-budgeting of some mandatory spending categories, such as social spending and weaknesses in budgeting of investment projects that are financed from project loans; (iii) insufficient capacity for developing fiscal projections for levels of government other than the central level (in particular by cities and municipalities).
- Comprehensiveness of the budget process in other (“non-Republican”) levels of government. Serbia’s budget process is mostly centered on the Republican level, only partially addressing other parts of the general government. While the financial plans of the social security funds are submitted to parliament concurrently with the Republican budget law, they can in principle be subsequently changed outside of the Republican budget process. Also, the financial plan of the Road of Serbia enterprise5 is not prepared at the time of the annual budget law, but some two-three months later after the budget is submitted to parliament. Similarly, the budgets of cities and municipalities are negotiated, presented, and passed sometimes well after the central government budget is adopted by the government.
- In-year budget implementation and controls. Constraints on within-year fiscal drift are weak. On a day-to-day basis, the Treasury has been controlling the pace of expenditure, but mostly with respect to the compliance with the legal framework, which is however continuously changing. While the framework Budget System law formally requires a so-called “pay-as-you-go” rule for new legislative initiatives—whereby any new spending increases or revenue cuts have to be offset by a combination of spending cuts and revenue increases—this rule has not been enforced. While all Serbian laws submitted to parliament are required to include a section on the need for allocating budget funds, the requirement is often addressed in a pro-forma way. Related to that, the budget process effectively permits a drift in both deficit and spending targets.
- Medium- and longer-term orientation of the fiscal framework. The budget laws are annual documents, while the three-year plans do not have a binding nature. The BSL’s pay-as-you go rule requires the assessment and offsetting effects of all initiatives over a three-year period—this latter aspect of the rule is largely neglected. There is no requirement to spell out longer-term implications of selected pieces of fiscal legislation. The absence of a binding medium-term focus in the fiscal framework has permitted large and widening deviations of the fiscal outcomes from the original plans with limited pressure for corrective measures. It has also hampered screening of new fiscal legislation from a longer-term perspective.
- Anchoring of mandatory spending. Mandatory spending accounts for a large share of the budget envelope. For example, spending on public wages and pensions alone amounts to more than 50 percent of general government spending. The lack of anchor for these crucial spending categories—and their high growth in 2006–08—was the main reason for adopting the expenditure indexation sub-rules in the 2010 changes to the BSL. However, these rules, while helping adjustment somewhat, have been undermined by ad-hoc initiatives including bonuses to public employees (and to a lesser extent pensioners), selective increases in public wages, and increases in public employment.
- Public debt reporting, projections, and control. The issues include: (i) different definitions of the public debt in the public debt law and the BSL (the narrower definition of the public debt law is often used); (ii) incomplete coverage of the public debt in all official reports (for example, nonguaranteed debt of local governments, as well as some other items, is excluded); (iii) lack of explicit rules and procedures regarding the calling of government guarantees (the calling of guarantees is not formalized or made public in a systematic way); and (iv) limited reporting and control over contingent liabilities (arising from nonguaranteed debt of public enterprises that can contain an implicit guarantee, public private partnership (PPP) contracts, deposit insurance, arrears, disputed court cases, etc.)..
- Fiscal transparency. A number of issues relate to the presentation of the information to the public that is already available to the policymakers. These include: (i) limited time for public and parliamentary debate for many important pieces of new fiscal legislation (many laws are adopted based on the so-called “urgent procedures”), and (ii) insufficient reporting and information on several important fiscal items (including the final account of the budget and financial statements of some entities comprising the general government, and public enterprises).
- Accountability for budget outcomes. Problems include: (i) relatively vague legal identification of accountability (the BSL envisions that the “government” is collectively responsible for budget implementation); (ii) gaps in internal and external audit procedures (for example, the state audit institution (SAI) is currently building capacity, expanding the coverage of audits (so far incomplete), and building a track record of statements and opinions); (iii) the final account of the central government budget was not discussed, voted, or adopted by parliament for more than a decade; and (iv) lack of sanctions within the set-up of the fiscal federalism (while the budget laws have often included a clause envisioning a possibility of cutting transfers to local governments in case they do not comply with the rules set for those in annual budget laws, the sanction was never (meaningfully) implemented).
- Legal and regulatory rigidities and inconsistencies. Many existing laws, rules, and “government strategies” have not been adopted in a coordinated way, and de-facto may be in actual or potential conflict with each other. Some of these entitle local governments or sectors to a fixed percentage of GDP or of total spending, thus reducing budget flexibility.
C. A Roadmap for Reform
15. The challenges described in the previous section offer a guide to needed reforms. Steps in this direction would also be consistent with fully implementing good practices in fiscal management and transparency.6 These reforms need to be implemented and tailored to Serbia’s conditions. As the challenges are often interrelated, there needs to be a critical mass of steps in the right direction.
16. Progress in selected fiscal framework areas has been made recently, and there is a need to build on it in a more systematic way. In late 2012, reform of the so-called own-source accounts integrated these positions of direct budget beneficiaries into the budget process, which was a significant positive step. Similar progress is being planned with respect to integrating indirect budget beneficiaries in the budget. Some headway has been made—and is further planned—with respect to streamlining the uncontrolled proliferation of para-fiscal fees and charges. The government has also moved to implement a tax administration modernization strategy. All these efforts should continue and preferably intensify.
17. Effectively enforcing the BSL’s pay-as-you-go rule regarding fully offsetting new deficit-increasing initiatives is crucial for achieving fiscal responsibility.7 First, enforcing the rule would send a signal that implementation of existing legal provisions (toward greater fiscal responsibility) is a priority, thereby emphasizing continuity of progress and showing tangible results. Second, striving to continuously enforce the rule will greatly reduce the observed pressures for fiscal deficit drift in the course of the annual budget process, especially at the supplementary budget stage. Third, the requirement to focus on a three-year period for offsetting measures would greatly increase the medium-term orientation of many fiscal decisions.
18. Several practical steps are needed for effectively implementing the pay-as-you-go rule. First, it would require building capacity for a competent and independent analysis of the key fiscal initiatives and the proposed offsetting measures. Such capacity could be created within the Ministry of Finance (whose staff would need to be protected from political influences), and/or could be built within the fiscal council.8 Second, there should be an effectively enforced condition that—without a properly certified analysis of the full package (including offsetting measures), which demonstrate that it would be deficit-neutral—new initiatives cannot be submitted to the government or parliament for consideration. Third, the requirement for the three-year horizon for assessing the neutrality of the package should be strictly enforced, while for changes to the pension a longer-term assessment horizon (say, 10 years) should be considered. The recent experience of Italy suggests that the “pay-as-you-go” rule akin to article 48 of the BSL can be a useful tool in achieving prudent fiscal deficit objectives.
19. There is also a comprehensive agenda of specific structural reforms. These reforms should achieve the following objectives, and lie in the domain of responsibility of the following ministries and agencies.
Better measurement of the fiscal deficits and public debt:
- the Statistics Office would need to (a) produce the full list of entities comprising the general government and (b) start to compile fiscal accounts on an accrual basis;
- Ministry of Finance and Economy (MoFE) should (a) collect full and reliable information on public spending arrears, including at the local government level ; (b) compile a comprehensive database on PPP projects and other contingent liabilities; (c) fully integrate the information on indirect budget beneficiaries (IBBs) in final general government accounts;
- the government would need to propose reconciling the legal definition of the public debt, with a view to aligning it with that of the BSL;
Improved budget planning
- The MoFE should guard against overestimating revenue projections and underestimating spending (in particular mandatory spending);
- The MoFE/Treasury should improve the quality of monthly planning and projections so that problems in budget implementation are revealed at an early stage and proper corrective measures could be adopted quickly;
- Local governments should develop capacity to better project fiscal outcomes;
Greater comprehensiveness of the budget process
- The MoFE would need to (a) integrate the budgets of the social security funds, IBBs, and the Road company (and any other public enterprises that should comprise the general government according to well-specified criteria) into the presentation of the budget law and of targeted outcomes; (b) avoid significant changes in the financial plans of the social security funds and the Road company outside of the broader budget process; (c) make greater effort to assess and coordinate (in terms of exchange of information) the planning of local government budget at the time the budget law is presented to parliament;
- The government would need to (a) alter the rules regarding Social Security Funds (SSFs), in particular helping an efficient reallocation of any extra savings achieved to other levels of government; and (b) effectively integrate the stock of expenditure arrears into the fiscal envelope while adopting plans for resolving the arrears and preventing their future accumulation;
More effective rules regarding the budget’s implementation
- The MoFE should consistently stick to fully observing the budget timetable, as in the preparation of the 2013 budget;
- The government may consider imposing additional rules regarding changing the deficit in the course of the annual budget process (ideally once the pay-as-you-go rule is effectively working), such as limiting the scope for increasing the deficit at the time of the supplementary budget;
Better medium-term budget orientation
- The MoFE could consider (a) introducing nominal binding medium-term spending limits; and (b) improving the realism and quality of rolling three-year medium-term strategies, including by seeking and incorporating feedback from the fiscal council;
Better anchoring and raising the quality of mandatory spending
- The government should consider introducing “claw-back provisions” with respect to the excessive growth in the public sector wage bill and consider automatic corrective measures;
- The MoFE should finalize the register of public employees and propose measures to better control their number;
- The government and the relevant SSFs should undertake specific reforms in the key sectors (particularly pensions, health, education, social assistance) that concern mandatory spending;
Improved quality of investment spending
- The MoFE (or another unit) should assume responsibility for coordinating investment projects and their cost-benefit analysis;
Stronger accountability for budget outcomes
- The State Audit Institution should further build capacity and continue expanding the scope of the audits;
- All ministries and agencies should improve the internal audit function and protect it from political influences;
- The Parliamentary Finance Committee should further build capacity and play a meaningful role in screening the budget and other fiscal initiatives;
- The MoFE should enforce a meaningful, evenhanded, and rule-based system of sanctions for local governments;
- The government should restore personal Minister’s accountability for budget outcomes
Eliminate the scope for inconsistent fiscal legislation
- The government, in cooperation with the MoFE should review, revise, or more clearly interpret legislation and other regulations with a view to lessening the legal rigidities and possible inconsistencies in the spending structure;
Improved fiscal transparency
- The MoFE should (a) report and publish detailed information on spending arrears, including by each budget beneficiary; (b) communicate fiscal developments of predetermined regular intervals, (c) enhance their published reports on the fiscal position and public debt (including by incorporating the IBBs in the deficit position and non-guaranteed debt of local governments as part of general government debt).
- The government should allow full public discussion of the final account of the budget in parliament.
20. It is also desirable that, in due course, the original 2010 fiscal rules framework be credibly reestablished on the basis of healthier foundations. The fiscal deficit rule lays a good foundation for a balanced and countercyclical policy in normal times, while targeting the public debt ceiling of 45 percent of GDP seems to be appropriate for Serbia’s circumstances. Further, the subrules on the key budget spending sub-items such as public wages and pensions would ensure an orderly anchoring once a robust economic recovery takes hold. Finally, the fiscal council has established a useful role in monitoring the fiscal rules, but also in proposing further improvements to the broader fiscal trends and developments, and there is a need to build on this progress continuously. Of course, reestablishment of the elements of the 2010 fiscal rules should recognize that it may take some time—for example, to bring debt back to 45 percent of GDP.
21. Focusing on these reform priorities would favor fiscal responsibility while minimizing adverse effects on growth. The specific reform steps are not cast in stone and would have to be adjusted to the circumstances. But steps in this direction would help achieve desired deficit outcomes while lessening the need for sharp and disruptive adjustment efforts. In particular, transparency-enhancing and other PFM reforms emphasized above would be especially beneficial, since they (i) require relatively small implementation capacity; (ii) do not represent a drag on growth; and (iii) would have positive spill-over effects on the political economy incentives for good fiscal outcomes and on the pace of other fiscal reforms.
International Monetary Fund, 2007, “Manual on Fiscal Transparency,”Fiscal Affairs Department, Washington, D.C. See: http://www.imf.org/external/np/pp/2007/eng/051507m.pdf
Prepared by Bogdan Lissovolik (EUR).
One case in point is the different definition of the public debt in the Law on Public Debt and in the Budget System Law.
While there is technically another stage—that of the government adoption of the budget accompanying its submission to parliament—budgets at that stage are typically not different (in terms of the main budget parameters) from those of the budgets approved by parliament.
For the purpose of this paper, the “final outcome” refers to the outturn as reported by the Ministry of Finance about a month after the budget year is over, as part of regular monthly reporting. This excludes reporting by indirect budget beneficiaries, which is (partly) included into the annual “final account” of the budget.
According to local legislation, Roads of Serbia is a public enterprise, which has however been considered part of the general government in IMF programs.
The rule is contained in article 48 of the BSL, which requires the government to assess financial effects—for the current year and for the following two years—of all new budget initiatives and propose offsetting measures that would neutralize any deficit-increasing effects over this horizon.
For practical purposes, a de-minimis threshold could be set to focus the scarce capacity and expertise on relatively large (macro-relevant) initiatives, while smaller initiatives would undergo a simpler procedure.