## _cr06368

## Source details

**Canonical URL:** [_cr06368](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2006/_cr06368.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2006/_cr06368.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2006/_cr06368.pdf.json)

---

### Alternative current account estimates — overview
- Official CBBH estimates of the current account deficit have fluctuated around 20 percent of GDP, driven by a 50 percent trade deficit and partly offset by large current transfers (mostly workers’ remittances).
- Main statistical shortcomings identified:
  - Inadequate methodology in estimating workers’ remittances.
  - Lack of reporting of international organizations’ employee compensation.
  - Inadequate reporting of donor grants, in particular to local entities.
  - No reporting of cash withdrawals at ATM machines (operations out of banks’ balance sheets).
  - Large past adjustments to trade statistics (underreporting and smuggling); trade statistics improved since 2005 and the introduction of VAT in 2006 increased coverage (doubling registered companies to 32,000 in the first few months of 2006).
- Errors & omissions have been large and positive, around 10 percentage points of GDP per annum, indicating large unrecorded inflows; capital and financial account data suggest the bulk of these inflows are probably current receipts.
- Chapter objective: present alternative current account estimates by (i) using alternative remittance estimates and (ii) deriving the current account from capital and financial account (“below the line”) data; combine approaches and adjust GDP for non-observed activity (NOA).

### Workers’ remittances — evidence and alternative estimates
- Context:
  - Private transfers are about 20 percent of GDP in CBBH balance of payments; three kinds reported: workers’ remittances, other current private transfers, private capital transfers.
  - Focus on workers’ remittances due to alternative data availability.
- Evidence that official remittance data underestimate receipts:
  - CBBH uses 2002 household survey indicating workers’ remittances about KM 2 billion; figure held constant thereafter.
  - Global Development Finance (GDF) reports remittances in Eastern Europe increased by more than 12 percent from 2002 to 2004.
  - Remittances through banks and Western Union increased by almost 50 percent since 2003, while total remittances reported by CBBH remained constant.
  - International experience suggests about 50–60 percent of remittances are in cash even in developed banking systems.
  - Wire transfer costs average 13 percent; only 25 percent of households in Bosnia & Herzegovina have a bank account.
  - WAD (Third Congress) in May 2006 evaluated workers’ remittances at KM 7 billion (about € 3.5 billion); using half of WAD’s number would increase workers’ remittances by 7 percentage points of GDP over official estimates.
  - Private capital transfers estimated by CBBH at 25 percent of remittances; reclassifying in-kind transfers to current transfers would increase workers’ remittances by another 3 percentage points of GDP.
- International and empirical comparisons:
  - Aggregate remittances about 14 percent of GDP (comparable to Albania, Armenia, Slovak Republic, Serbia & Montenegro; lower than Lebanon and Macedonia).
  - Average per-emigrant remittance implied by official data: US$1,275 per year (about US$100 per month), contrasted with US$500 monthly by emigrants in the US and US$150–200 monthly by Russia-based Armenian workers.
  - Empirical literature examples: Mellyn (Philippines) >2x official; Korovilas (Albania) 2–3x official; Roberts (Armenia) 4–5x official.
- Quantitative implications:
  - Private transfers could be as much as 10 percentage points of GDP higher than official estimates—similar to observed errors and omissions.
  - Conservative upward revision: remittances higher by 4½ percentage points of GDP, derived from:
    - Half of capital transfers reclassified as workers’ remittances: about 1.5 percent of GDP.
    - Assuming 60 percent of workers’ remittances are informal and in cash: increases remittances by about 3 percent of GDP.
- Key table excerpts (private transfers, 2003–2005):
  - 2003: Workers' remittances 1,956.9; o/w through banks and Western Union 694.2; Other private transfers 320.3; Private transfers (capital) 492.4; Total 2,769.6; In percent of GDP 20.5; workers' remittances 14.5; Nominal GDP 134,921; Current Account Balance -20.9 percent of GDP.
  - 2004: Workers' remittances 2,042.2; o/w through banks and Western Union 813.2; Other private transfers 387.0; Private transfers (capital) 532.2; Total 2,961.3; In percent of GDP 20.4; workers' remittances 14.1; Nominal GDP 145,341; Current Account Balance -19.3 percent of GDP.
  - 2005: Workers' remittances 2,057.2; o/w through banks and Western Union 1,015.0; Other private transfers 420.7; Private transfers (capital) 492.0; Total 2,969.9; In percent of GDP 19.0; workers' remittances 13.2; Nominal GDP 156,055; Current Account Balance -21.3 percent of GDP.

### Estimating the current account from capital and financial accounts (“below the line”)
- Identity used: CA_t + KA_t = ∆R_t  =>  CA_t = ∆R_t − KA_t  (source formatting presents CA_t = −∆R_t + KA_t).
- Rationale:
  - Capital and financial account data (government net borrowing, bank net borrowing, official reserve movements, FDI, trade credits) generally higher quality than many current account components.
  - Nonbank private sector net borrowing is less reliable but small in absolute terms.
- Adjustments:
  - Capital account should be adjusted for overestimation of capital transfers (in-kind capital transfers shifted to current transfers reduce KA and raise “below the line” CA estimate).
- Result:
  - “Below the line” estimate of the current account deficit in Bosnia & Herzegovina is in the range of 7–13 percent of GDP in recent years.
  - This should be seen as a lower bound because unidentified flows that could finance the current account deficit are not captured.

### Relating the current account deficit to GDP and NOA
- Official GDP is highly underestimated; IMF staff study (IMF Country Report No. 05/198) showed “true” GDP could be 30 to 50 percent higher than official estimates.
- In July 2006, SABH made the first adjustment to official GDP estimates to account for imputed rents, raising the level of GDP by about 10 percent.
- Adjusting GDP upwards by one-third to take these factors into account implies:
  - Current account deficit on official CBBH data is about 14–16 percent of GDP during 2003–05.
  - Alternative estimates derived above are at the 6–14 percent of GDP range.
- Estimating NOA (latent variables method):
  - Isolates “grey economy” factors via energy consumption or money demand equations.
  - Staff extended method using international panel data (variables included currency in circulation, exports per capita, health care expenditure per capita).
  - Results robust across methods: underestimation of GDP ranging from 30 to 50 percent.

### Main conclusion on the “true” current account deficit
- The “true” current account deficit in Bosnia & Herzegovina relative to the size of the economy is probably in the range of 6 to 14 percent of GDP.
- The wide range reflects statistical weaknesses and uncertainties in alternative estimates.

### Public debt sustainability — baseline and adjustment scenarios
- Baseline (no policy change) setup:
  - Extend baseline medium-term DSA to 2050 showing explosive dynamics under unchanged policies.
  - Assumptions include continued strong growth and low inflation for 2006–11, slowdown to annual growth of 4 percent beyond 2011; currency board remains; IDA disbursements fall from ½ percentage point of GDP in 2007 to zero after 2011; multilateral financing declines from 1 percentage point of GDP in 2007 to ½ percentage point in 2011 and zero after 2015; commercial borrowing at LIBOR 5 percent + 200 basis points.
  - Increase in face value of debt by 17 percent of GDP in 2006 associated with settlement of FFCD, war claims, and obligations; additional increases equivalent to 10 percent of GDP in 2007 and 5 percent of GDP in 2008 assumed.
  - Debt servicing costs jump in 2015 when conditional London Club bonds become payable (payable only if GDP per capita reaches US$2,800, as adjusted for German inflation from 1997, for two consecutive years between 2004 and 2017).
  - Privatization receipts assumed at 2 percent of GDP in 2006 and 2007 from sale of RS Telecom.
  - Non-interest current expenditures projected to grow in line with nominal GDP after a hike of ¼ percentage point of GDP in 2007.
  - Capital expenditures decline by 2 percentage points of GDP between 2007 and 2011 and remain at 7½ percent of GDP thereafter.
  - Revenue ratio declines by 2½ percentage points of GDP over 2007–2011 due to lower projected grants (1 percentage point of GDP) and trade taxes (1½ percentage point of GDP); EU Association Agreement tariff reductions could reduce revenue ratio by some 0.4 of a percentage point per year starting in 2007 if schedule similar to FYR Macedonia and Croatia applies.
- Baseline outcomes and risks:
  - Primary deficit increases from 1 percentage point of GDP in 2007 to about 1½ percentage point of GDP in 2011 and further beyond.
  - Debt-to-GDP ratio slightly above 50 percent in the medium-term, then increases rapidly.
  - NPV of debt increases from 30 percent of GDP in 2007 to 35 percent in 2011 and further beyond (discount rate assumed 7 percent).
  - Overall general government deficit increases from 1½ percent of GDP in 2007 to 3 percent of GDP in 2011.
- Adjustment scenario:
  - Same macro assumptions but fiscal policy targets sustained primary surplus of about 1 percent of GDP starting in 2008; resulting overall balance close to zero in 2007–2010 and overall surplus of about ½ percent of GDP thereafter.
  - Outcomes: debt-to-GDP and NPV ratios decline rapidly; debt-to-GDP falling to zero by the end of projection period in main adjustment.
  - Sensitivity:
    - “Low growth” case: annual growth decelerates to 3 percent; under this case NPV and debt-to-GDP ratios barely stabilize in long term and overall general government balance maintained at around 2 percent of GDP as debt servicing costs rise.
    - “High interest” case: risk premium 100 basis points higher (LIBOR+300); adjustment scenario not very sensitive to higher interest assumptions.
- Caveats:
  - If growth lower than assumed, even with 1 percent of GDP primary surpluses the debt-to-GDP ratio may not fall continuously.
  - Long-term projections involve substantial uncertainty.

### Long-term growth prospects (Box 2)
- Medium-term average GDP growth close to 5 over 2007–2011 driven mainly by temporary factors and may be optimistic for long run.
- Cross-country regression-based estimates: long-term GDP growth in range 3–4 percent per annum; per capita potential growth 2½–3½ percent from Levine & Renelt regressions.
- Accounting for institutional development reduces per capita long-term growth to zero in one augmented specification; replacing institutional index with Estonia’s 2005 value increases potential growth.
- With projected population growth of 0.5 percent per year:
  - Results translate into range 3–4 percent (original regressions) and 2½ percent (augmented equations assuming progress in institutional development).
- Predicted long-term per capita GDP growth (percent):
  - LR1: Current level 2.3; With progress 2.3.
  - LR2: Current level 3.2; With progress 3.2.
  - CK: Current level 0.3; With progress 1.9.

### Policy implications and potential sources of fiscal adjustment
- Ensuring long-term public debt sustainability requires more stringent policy than medium-term DSA suggests.
- Primary surpluses of about 1 percentage point of GDP per year are sufficient under baseline growth assumptions to put debt-to-GDP on a gently declining trend.
- Room for savings limited given future expenditure pressures and already high tax burden.
- Fiscal consolidation in last five years relied on nominal expenditure freezes, reduction in reconstruction-related investments, and moderate revenue increases; spending started increasing in 2006 in runup to elections.
- Future spending pressures: settlement war and other claims, property restitution, state-building and EU accession costs, planned infrastructure projects, pension system costs.
- World Bank PEIR identifies potential savings up to 7 percentage points of GDP in wages, other administrative expenses, and transfers to households; savings of this magnitude would:
  - Be more than sufficient to generate required primary surpluses for long-term sustainability.
  - Provide room for meeting additional expenditure needs and possibly reducing the tax burden.

### Expenditure pressures, required adjustment and PEIR savings (quantified)
- Short-term outlook:
  - Revenue gains in 2006 likely partly temporary; primary balance projected to deteriorate to a deficit of 1 percentage point of GDP in 2007 on current trends.
  - State-building agenda generates additional spending increases of ¾ percentage point of GDP beyond 2007.
- Required fiscal adjustment magnitudes:
  - Savings of 3–4 percentage points of GDP—in addition to projected decline in capital expenditures of 2 percentage points—necessary to:
    - Turn projected 2007 primary deficit of 1 percentage point of GDP into a surplus of 1 percentage point of GDP.
    - Offset identified fiscal pressures (¾ percentage point of GDP) and projected decline in revenue ratio (2½ percentage point).
- World Bank PEIR potential savings areas (up to 7 percentage points of GDP):
  - Public wages and employment (judiciary, health sector, other public employees).
  - Education: move to per capita financing, reorient curricula, raise repetition fees, integrate faculties.
  - Subsidies and transfers to non-profits: reduce subsidies to railroads, freeze agriculture subsidies, reduce transfers.
  - Veterans’ benefits, social welfare, child protection: tighten eligibility.
  - Pensions: limit increases to inflation.
  - Pharmaceutical procurement: centralize.
- Quantified spending (percent of GDP) in categories with measures (selected rows):
  - Total spending: 2005: 26.5; 2006: 24.0; 2007: 22.1; 2008: 20.9; 2009: 20.0; 2010: 19.4.
  - Narrow wage bill (excludes defense, police and education): 2005: 5.3; 2006: 4.6; 2007: 4.2; 2008: 3.9; 2009: 3.6; 2010: 3.4.
  - Pensions: 2005: 8.4; 2006: 7.8; 2007: 7.5; 2008: 7.2; 2009: 7.0; 2010: 6.6.
  - Social welfare: 2005: 4.0; 2006: 4.0; 2007: 3.3; 2008: 3.1; 2009: 3.0; 2010: 2.8.
  - Education: 2005: 4.8; 2006: 4.2; 2007: 4.0; 2008: 3.9; 2009: 3.8; 2010: 3.8.
  - Subsidies: 2005: 1.3; 2006: 1.2; 2007: 1.0; 2008: 0.9; 2009: 0.9; 2010: 0.8.
  - Pharmaceutical procurement: 2005: 1.2; 2006: 0.9; 2007: 0.9; 2008: 0.9; 2009: 0.9; 2010: 0.8.
- Implication: identified expenditure reductions exceed what is necessary to achieve primary balance for sustainability, allowing choice and gradual implementation and potential tax reductions.

### Credit flows, external deficit and model-based findings
- Key macro indicators (selected exact numbers preserved):
  - Trade deficit (percent of GDP): 2001: 53.3; 2002: 53.8; 2003: 53.2; 2004: 49.5; 2005: 50.2.
  - Current account deficit (percent of GDP): 2001: 13.3; 2002: 19.1; 2003: 20.9; 2004: 19.3; 2005: 21.3.
  - Credit to the private sector (percent of GDP): 2001: 22.4; 2002: 28.9; 2003: 34.2; 2004: 40.5; 2005: 48.0.
  - Credit to the private sector growth (percent change): 2001: 13.7; 2002: 36.2; 2003: 24.8; 2004: 27.5; 2005: 27.5.
  - Real GDP percentage change: 2001: 3.6; 2002: 5.0; 2003: 4.1; 2004: 5.8; 2005: 5.0.
  - Domestic saving (percent of GDP): 2001: -31.4; 2002: -30.2; 2003: -28.6; 2004: -25.7; 2005: -25.4.
- Observations:
  - Booming credit since 2002 reinforced adverse external trends; bulk of credit flows went to households rather than corporate sector.
  - Exports were only one third of imports in 2005.
  - Between 2001 and 2005 overall fiscal balance strengthened by 5½ percentage points of GDP while credit to the private sector rose by 25¾ percentage points of GDP.
- Theoretical model and empirical estimation:
  - Static general equilibrium approach adapted from liquidity effects models; small open economy with representative household split into consumer, firm, bank, government.
  - Trade balance reduced to function B = B(Φ_F, Φ_C, G, M, E, K, e, P*).
  - Linear regression approximation estimated by GMM on quarterly data Q1 1998–Q2 2004; forecast evaluation Q3 2004–Q3 2005.
- Key GMM coefficient estimates (dependent variable: trade balance):
  - First lag of credit flow to enterprises: -0.21 (t-statistic -2.81, P-value 0.01).
  - First lag of credit flow to households: -1.08 (t-statistic -3.51, P-value 0.00).
  - Second lag of fiscal revenue: 0.39 (t-statistic 1.88, P-value 0.08).
  - Fiscal expenditure: -0.45 (t-statistic -2.17, P-value 0.05).
  - Export price/Import price: 17.41 (t-statistic 6.15, P-value 0.00).
  - R-square: 0.86; Test of overidentifying restrictions (Chi-square, 4 df): 1.68 (P-value 0.79); Number of observations: 23.
- Forecast performance (Q3 2004–Q3 2005):
  - Root mean squared error: 54.68.
  - Mean absolute error: 44.19.
  - Mean absolute percent error: 5.18.
  - Theil inequality coefficient: 0.03.
- Policy implications from model:
  - Composition of credit matters: credit to households has a much larger negative impact on the trade balance than credit to enterprises.
  - Fiscal expenditure has the most immediate impact on the trade balance, followed by credit flows, then fiscal revenue.
  - Counterfactuals for 2006 plausible quarterly paths:
    - Fiscal tightening: one percent of GDP reduction in expenditure over one year generates a 0.45 percent of GDP reduction in the trade deficit.
    - Reduction in flow of credit of one percent of GDP generates a 0.44 percent of GDP reduction in the trade deficit.
    - Conclusion: over one year horizon, fiscal expenditure and credit tightening are roughly equally effective.
  - Implementation constraints:
    - Currency board and open capital account limit monetary tools to restrain credit.
    - Required reserves and prudential measures have limited effectiveness given foreign-bank subsidiaries’ access to parent liquidity.
    - Therefore, fiscal policy (expenditure restraint) is the primary instrument to generate demand restraint.

### Institutional approaches to fiscal coordination — cross-country lessons
- Fragmented fiscal system in Bosnia & Herzegovina; small central State government with weak control over consolidated fiscal policy (origins in 1995 Dayton Peace Accord).
- International community played strong coordinating role historically (OHR “Bonn powers”); NFC established May 2005 as informal body.
- Cross-country mechanisms range from advisory/consultative bodies to legally binding/coercive institutions to hybrids.
- Membership and decision-making:
  - Dominated by politicians; some bodies include non-elected members (Belgium PSBRS, South Africa FFC).
  - Unanimity requirements in many bodies can hinder enforcement; some exceptions permit decisions without unanimity (Australia’s ALC, Argentina’s FCFR, Europe’s SGP).
- Targets and enforcement:
  - Countries target different fiscal variables (consolidated balance, debt levels, sub-national balances, expenditure ceilings).
  - Enforcement varies from explicit penalties (Brazil, Argentina frameworks) to peer pressure and transparency (Switzerland).
  - Credibility of enforcement depends on the enforcing entity’s capacity and willingness to punish non-compliance.
- Monitoring and transparency:
  - Regular, timely, harmonized fiscal reporting critical for coordination.
  - Argentina, Brazil, EU have stronger reporting/monitoring mechanisms; Austria and Belgium have partial or flexible arrangements.
- Role of central bank:
  - Central banks can have indirect but important roles (Belgium governor member of coordinating institution; Australia central bank underwrote and advised on borrowing).
- Cross-country lessons:
  - Institutional choice determined by historical, political, economic context; no one-size-fits-all.
  - Fiscal coordination more important when monetary policy constrained (pegged exchange rate).
  - Institutions facilitate but do not guarantee good policy; political will required.

### Sanctions, enforcement mechanics and credibility
- EU Excessive Deficit Procedure (EDP) provides explicit penalties for deficits above 3 percent of GDP; revised provisions (March 2005) include escape clauses for negative growth or unusual events; in practice no country has been fined.
- Austria: DSP sanctions designed to mimic SGP; few sanctions applied due to flexibility in assessment.
- Brazil: contingency reserve, financial and administrative sanctions (withholding transfers), breach remedies and potential penalties including imprisonment under Fiscal Crimes Law.
- Argentina: sanctions and incentives in framework; federal government exempt from sanctions.
- Australia: national government used transfer leverage historically; legal mechanisms allowed revenue redirection to enforce compliance.
- Credibility: many frameworks allow discretion in application creating enforcement loopholes; frameworks with automatic penalties harder to suspend are preferable.
- Monitoring: timely, harmonized accounting and publication of fiscal indicators essential for credible enforcement.

### Need for effective fiscal coordination in Bosnia & Herzegovina
- As a federal state with substantial powers devolved, effective fiscal coordination is essential; State lacks budgetary resources to assume full responsibility for macro stabilization.
- Currency board implications: maintenance of peg increases importance of consolidated fiscal control; State government may need greater role in setting consolidated deficit target.
- Current NFC weaknesses:
  - NFC lacks legal status; participation of key decision-makers erratic.
  - No formal agreement on harmonized budget calendar or failure-resolution mechanisms.
  - Reliable, timely data on fiscal performance of lower levels lacking.
  - No penalties for non-compliance; public pressure has not been reliable; market discipline limited as country has not borrowed on private markets.
- Minimum recommended measures:
  - Intergovernmental agreement on budget calendar ensuring timely adoption of budgets by all levels.
  - Adoption of reporting system for comprehensive, accurate, timely fiscal data for all government levels and off-budget operations.
  - Consider numerical fiscal rules to guide targets; design enforcement mechanisms with limited discretionary exemptions.
  - Strengthen analytical capacity for budget analysis and empower State to monitor line ministries, other levels and off-budget entities.
- Comparative checklist for Bosnia & Herzegovina (as excerpted):
  - legislative or constitutional underpinnings: N
  - key policy makers represented: Y
  - formal decisions taken: Y
  - formal voting: N
  - formal voting formula/deadlock breaking mechanism: N
  - Penalties specified: N
  - Past Use of Penalties: N
  - Explicit Budget Calendar: N
  - Harmonized Reporting and Accounting Standards: N

_Italic source: Content drawn from the provided excerpt of _cr06368 - References (IMF PDF)._

### References..............................................................................................................

### I.   ALTERNATIVE CURRENT ACCOUNT ESTIMATES

### A. Introduction
- Official CBBH estimates of the current account deficit have fluctuated around 20 percent of GDP, essentially driven by a 50 percent trade deficit and partially compensated by large current transfers (mostly workers’ remittances).
- Main shortcomings of official statistics identified:
  - Inadequate methodology in estimating workers’ remittances.
  - Lack of reporting of international organizations’ employee compensation.
  - Inadequate reporting of donor grants, in particular to local entities.
  - No reporting of cash withdrawals at ATM machines (operations out of banks’ balance sheets).
  - Large past adjustments to trade statistics (underreporting and smuggling), though trade statistics improved since 2005 and the introduction of VAT in 2006 increased coverage (doubling registered companies to 32,000 in the first few months of 2006).
- Errors & omissions have been large and positive, around 10 percentage points of GDP per annum, indicating large unrecorded inflows; financial and capital account data are generally more reliable and suggest the bulk of these inflows are probably current receipts.
- Chapter objective: present a range of plausible alternative current account estimates by (i) using alternative remittance estimates from different methods and (ii) deriving the current account from capital and financial account (“below the line”) data; combine approaches and adjust GDP for non-observed activity (NOA).

### B. Using Alternative Estimates of Workers’ Remittances
- Context and focus:
  - Workers’ remittances are part of private transfers, which are about 20 percent of GDP in the CBBH balance of payments (three kinds reported: workers’ remittances, other current private transfers, private capital transfers).
  - The chapter focuses on workers’ remittances because other categories either have improving coverage or lack alternative data.
- Evidence that official remittance data underestimate receipts:
  - CBBH uses a 2002 household survey indicating workers’ remittances about KM 2 billion; this figure was held constant due to lack of more recent surveys.
  - Global Development Finance (GDF) reports workers’ remittances in Eastern Europe increased by more than 12 percent from 2002 to 2004.
  - Remittances through banks and Western Union increased by almost 50 percent since 2003, while total remittances reported by CBBH remained constant (and declining as percent of GDP), implying an assumed one-for-one shift from informal to formal transfers not consistent with international experience.
  - International experience suggests about 50–60 percent of remittances are in cash even in developed banking systems.
  - Cash transfer drivers: wire transfer costs averaging 13 percent; limited access to banking services (only 25 percent of households in Bosnia & Herzegovina have a bank account); convenience for Bosnians working in neighboring countries.
  - Third Congress of the World Association of Bosnia & Herzegovina Diaspora (WAD) in May 2006 evaluated workers’ remittances at KM 7 billion (about € 3.5 billion), almost three times CBBH reported amount. Using half of WAD’s number would increase workers’ remittances by 7 percentage points of GDP over official estimates.
  - Private capital transfers are estimated by CBBH at 25 percent of remittances, but in-kind transfers (cars, equipment, appliances) are effectively remittances and re-classifying them to current transfers would increase reported workers’ remittances by another 3 percentage points of GDP.
- International comparisons:
  - Aggregate remittances about 14 percent of GDP (comparable to Albania, Armenia, Slovak Republic, Serbia & Montenegro; lower than Lebanon and Macedonia).
  - Average per-emigrant remittance implied by official aggregate data: US$1,275 per year (about US$100 per month), contrasted with US$500 monthly by emigrants in the US and US$150–200 monthly by Russia-based Armenian workers.
- Empirical literature examples of underestimation:
  - Mellyn (Philippines): remittances estimated at more than twice official figures.
  - Korovilas (Albania): remittances 2 to 3 times official estimates.
  - Roberts (Armenia): workers’ remittances 4 to 5 times higher than official figures.
- Quantitative implications:
  - Private transfers could be as much as 10 percentage points of GDP higher than official estimates—similar to observed errors and omissions.
  - A more conservative upward revision: remittances higher by 4½ percentage points of GDP, derived from:
    - Half of capital transfers reclassified as workers’ remittances: about 1.5 percent of GDP.
    - Assuming 60 percent of workers’ remittances are informal and in cash: increases remittances by about 3 percent of GDP.
- Key table excerpts (private current and capital transfers, 2003–2005, in millions of KM and percent of GDP):
  - 2003: Workers' remittances 1,956.9; o/w through banks and Western Union 694.2; Other private transfers 320.3; Private transfers (capital) 492.4; Total 2,769.6; In percent of GDP 20.5; workers' remittances 14.5; Nominal GDP 134,921; Current Account Balance -20.9 percent of GDP.
  - 2004: Workers' remittances 2,042.2; o/w through banks and Western Union 813.2; Other private transfers 387.0; Private transfers (capital) 532.2; Total 2,961.3; In percent of GDP 20.4; workers' remittances 14.1; Nominal GDP 145,341; Current Account Balance -19.3 percent of GDP.
  - 2005: Workers' remittances 2,057.2; o/w through banks and Western Union 1,015.0; Other private transfers 420.7; Private transfers (capital) 492.0; Total 2,969.9; In percent of GDP 19.0; workers' remittances 13.2; Nominal GDP 156,055; Current Account Balance -21.3 percent of GDP.

### C. Estimating the Current Account From Capital and Financial Accounts Data
- Identity used:
  - CA_t + KA_t = ∆R_t  =>  CA_t = ∆R_t − KA_t  (presented as CA_t = −∆R_t + KA_t in the source equation formatting)
- Rationale:
  - Capital and financial account data (government net borrowing, bank net borrowing, official reserve movements, FDI, trade credits) are generally of higher quality than many current account components; nonbank private sector net borrowing is less reliable but small in absolute terms in Bosnia & Herzegovina.
- Adjustments required:
  - Capital account data should be adjusted for overestimation of capital transfers (shift in-kind capital transfers to current transfers reduces KA and raises the “below the line” CA estimate).
- Result:
  - The “below the line” estimate of the current account deficit in Bosnia & Herzegovina is in the range of 7–13 percent of GDP in recent years.
  - This “below the line” estimate reflects identified financial flows and should be seen as a lower bound because it does not capture unidentified flows that could finance the current account deficit.

### D. Relating the Current Account Deficit to the Size of the Economy
- Official GDP is highly underestimated:
  - SABH did not until recently adjust GDP data for the unofficial (grey) economy.
  - IMF staff study (IMF Country Report No. 05/198) showed “true” GDP could be 30 to 50 percent higher than official estimates (Box 1 referenced).

### Main Conclusion
- The chapter’s primary conclusion: the “true” current account deficit in Bosnia & Herzegovina relative to the size of the economy is probably in the range of 6 to 14 percent of GDP.
- The wide range reflects statistical weaknesses in the balance of payments and uncertainties in alternative current account deficit estimates.

_Italic source: Content drawn from the provided excerpt of _cr06368 - References (IMF PDF)._

### 18.      In July 2006, SABH made the first adjustment to the official GDP estimates to

### _cr06368 - 18.      In July 2006, SABH made the first adjustment to the official GDP estimates to

### GDP adjustments and current account implications
- In July 2006, SABH made the first adjustment to the official GDP estimates to account for imputed rents, raising the level of GDP by about 10 percent.
- Adjusting GDP upwards by one-third to take these factors into account implies:
  - The current account deficit on the basis of official CBBH data is about 14–16 percent of GDP during 2003–05.
  - The alternative estimates derived above are at the 6–14 percent of GDP range.
- Other sources may still lead to overestimation of the current account deficit:
  - Official current grants and employee compensation from international organizations could amount to several percentage points of GDP in the late 1990s.
  - These items are not corrected in the paper due to lack of alternative data.
  - Their relative importance has been diminishing over time and is likely to become very small over the medium term.

### Estimating Non-Observed Activity (Box 1)
- Official national accounts in Bosnia & Herzegovina understate activity; evidence includes ratios to GDP, pace of credit growth, and size of the external balance.
- The latent variables method:
  - Isolates “grey economy” factors in an equation explaining energy consumption or money demand.
  - Derives NOA estimates from parameters estimated in energy or money demand equations.
- Staff extended the method using international panel data to estimate the “typical” correlation between latent variables and GDP. Variables included currency in circulation, exports per capita, and health care expenditure per capita.
- Results are robust across methods and point to an underestimation of GDP ranging from 30 to 50 percent.

### Public debt sustainability — Introduction
- Medium-term DSA results can be misleading because existing debt is highly concessional; stabilization of debt ratio requires relatively unambitious primary balances in the medium term.
- Over a longer horizon, growing share of non-concessional borrowing implies rising debt servicing costs and potentially explosive debt dynamics.
- The chapter proceeds in two steps:
  - Extend the baseline medium-term DSA to 2050 showing explosive dynamics under unchanged policies.
  - Present an “adjustment” scenario with sustained primary surpluses of one percentage point of GDP per year; these surpluses are sufficient to keep debt ratios on a gently declining trend under baseline growth assumptions.
- Caveats:
  - If growth is lower than assumed, even with 1 percent of GDP primary surpluses the debt-to-GDP ratio may not fall continuously and may remain broadly stable after an initial decline.
  - Long-term projections involve substantial uncertainty.

### Baseline (no policy change) scenario — key assumptions (Box 1)
- Macroeconomic environment:
  - Continued strong growth and low inflation in line with the staff’s baseline medium-term projections for 2006–11, and a slowdown to the annual growth of 4 percent beyond 2011.
  - The currency board is assumed to stay in place throughout the projection period (or replaced by the euro).
- External financing:
  - IDA disbursements fall from ½ percentage point of GDP in 2007 to zero after 2011.
  - Financing from multilateral institutions on close-to-market terms declines from 1 percentage point of GDP in 2007 to ½ percentage point in 2011 and to zero after 2015.
  - Borrowing on commercial terms assumed at average projected LIBOR at 5 percent + 200 basis points to cover additional financing needs.
- Domestic claims and debt increases:
  - Increase in face value of debt by 17 percent of GDP in 2006 associated with settlement of FFCD, war claims, and general government obligations; average grant element of these settlements is expected to exceed 25 percent.
  - Additional increases in face value of debt equivalent to 10 percent of GDP in 2007 and 5 percent of GDP in 2008 assumed.
- Debt servicing costs:
  - Jump in 2015 when conditional bonds issued as part of agreement with the London Club become payable (payable only if GDP per capita reaches US$2,800, as adjusted for German inflation from 1997, for two consecutive years between 2004 and 2017).
- Privatization receipts assumed at 2 percent of GDP in 2006 and in 2007 from sale of RS Telecom.
- Revenue and expenditure projections:
  - Non-interest current expenditures projected to grow in line with nominal GDP after a hike of ¼ percentage point of GDP in 2007.
  - Capital expenditures decline by 2 percentage points of GDP between 2007 and 2011 and remain at the 2011 level of 7½ percent of GDP thereafter.
  - Revenue ratio declines by 2½ percentage points of GDP over 2007–2011 due to lower projected grants (1 percentage point of GDP) and trade taxes (1½ percentage point of GDP).
  - EU Association Agreement tariff reductions could reduce revenue ratio by some 0.4 of a percentage point per year if a schedule similar to FYR Macedonia and Croatia applies starting in 2007.

### Baseline scenario outcomes and risks
- The baseline generates explosive debt dynamics in the long run:
  - Primary deficit increases from 1 percentage point of GDP in 2007 to about 1½ percentage point of GDP in 2011 and further beyond.
  - Debt-to-GDP ratio is slightly above 50 percent in the medium-term, then increases rapidly thereafter.
  - Net Present Value (NPV) of debt increases from 30 percent of GDP in 2007 to 35 percent in 2011 and further beyond. (All NPV calculations assume a discount rate of 7 percent.)
  - Overall general government deficit increases from 1½ percent of GDP in 2007 to 3 percent of GDP in 2011.

### Adjustment scenario and sensitivity analysis
- Adjustment scenario assumptions:
  - Same macro assumptions as baseline but fiscal policy targets a sustained primary surplus of about 1 percent of GDP starting in 2008.
  - Resulting overall balance close to zero in 2007–2010 and an overall surplus of about ½ percent of GDP thereafter.
- Sensitivity cases:
  - “Low growth” case: gradual deceleration of growth to an annual rate of 3 percent with correspondingly slower revenue growth; expenditures maintained at the same nominal level as main scenario until 2011 and grow in line with lower nominal GDP from 2012 onwards.
  - “High interest” case: risk premium on new borrowing is 100 basis points higher than main scenario (LIBOR+300).
- Adjustment scenario outcomes:
  - Main adjustment scenario: debt-to-GDP and NPV ratios decline rapidly, with debt-to-GDP falling to zero by the end of the projection period.
  - General government balance deteriorates slightly until 2020, returns to balance by about 2030, and moves toward a surplus thereafter.
  - Sensitivity:
    - Adjustment scenario is not very sensitive to higher interest assumptions.
    - Alternative growth assumptions have sizeable impact: in the “low growth” case NPV and debt-to-GDP ratios barely stabilize over the long term, with overall general government balance maintained at around 2 percent of GDP as debt servicing costs rise.

### Long-term growth prospects (Box 2)
- Medium-term average GDP growth close to 5 over 2007–2011 is driven mainly by temporary factors and may be optimistic for long-run projections.
- Cross-country regression-based estimates suggest long-term GDP growth for Bosnia & Herzegovina in the range of 3–4 percent per annum; per capita potential growth 2½–3½ percent from Levine & Renelt regressions.
- Accounting for institutional development reduces per capita long-term growth to zero in one augmented specification; replacing institutional index with Estonia’s 2005 value increases potential growth.
- With projected population growth of 0.5 percent per year, results translate into a range of 3–4 percent (original regressions) and 2½ percent (augmented equations assuming progress in institutional development).
- Predicted long-term per capita GDP growth (in percent) table reported:
  - LR1: Current level of institutional development 2.3; With progress in institutional development 2.3.
  - LR2: Current level 3.2; With progress 3.2.
  - CK: Current level 0.3; With progress 1.9.

### Policy implications and potential sources of adjustment
- Ensuring long-term public debt sustainability requires more stringent policy than medium-term DSA suggests.
- Primary surpluses of about 1 percentage point of GDP per year are sufficient under baseline growth assumptions to place debt-to-GDP on a gently declining trend.
- Room for savings is limited given future expenditure pressures and an already high tax burden.
- Fiscal consolidation achieved in the last five years relied on:
  - Nominal expenditure freezes, particularly on wages.
  - Reduction in reconstruction-related investments.
  - Moderate increases in revenue.
- Spending started increasing in 2006 in the runup to elections and, unless checked, will have a significant full-year impact in 2007.
- Future spending pressures include settlement war and other claims, property restitution, state-building and EU accession costs, planned infrastructure projects, and pension system costs.
- World Bank PEIR identifies potential savings up to 7 percentage points of GDP in wages, other administrative expenses, and transfers to households; savings of this magnitude would:
  - Be more than sufficient to generate the primary surpluses required for long-term public debt sustainability.
  - Provide room for meeting additional expenditure needs and possibly reducing the tax burden.

*Source: IMF staff report content from _cr06368 - 18. In July 2006, SABH made the first adjustment to the official GDP estimates to (PDF chapter/section).*

### 10.      These findings suggest that an even stronger adjustment might be needed to

### _cr06368 - 10.      These findings suggest that an even stronger adjustment might be needed to

### Debt sustainability and adjustment needs
- The “low growth” scenario is considered realistic for Bosnia & Herzegovina in the long run and, combined with high interest rates (not shown), could tip debt dynamics into unsustainability.
- To be on the safe side, an even higher primary surplus than the 1 percent of GDP assumed in the adjustment scenario might be warranted, especially early in the projection period.

### Emerging expenditure pressures and short-term outlook
- Expenditure commitments made during 2006, if fully executed, imply a substantial nominal increase in expenditures on wages and social spending.
- Higher-than-expected revenue in 2006 will cover these increases this year, but revenue gains are partly temporary, so on current trends the primary balance is projected to deteriorate to a deficit of 1 percentage point of GDP in 2007.
- Beyond 2007, the state-building agenda will generate additional spending increases of ¾ percentage point of GDP, mainly from ongoing defense and police reforms and modest strengthening of State institutions before EU accession.
- Large infrastructure investment projects and potential increases in pension outlays may create further expenditure pressures.

### State-building agenda spending profile (Percent of GDP)
- Defense and police spending — total: 2006: 3.4; 2007: 4.3; 2008: 4.4; 2009: 4.1; 2010: 4.3; 2011: 4.1.
- Wage costs:
  - Defense: 2006: 1.3; 2007: 1.4; 2008: 1.2; 2009: 1.1; 2010: 1.2; 2011: 1.1.
  - Police: 2006: 1.8; 2007: 1.8; 2008: 2.0; 2009: 1.9; 2010: 1.9; 2011: 1.8.
- Non-wage costs (defense): 2006: 0.3; 2007: 0.8; 2008: 0.8; 2009: 0.7; 2010: 0.6; 2011: 0.6.
- Institution building: 2006: - ; 2007: 0.3; 2008: 0.4; 2009: 0.4; 2010: 0.5; 2011: 0.5.
- Sources: BIH Ministry of Defense, World Bank, and IMF staff estimates.
- Note on defense cost assumptions: based on Bosnia & Herzegovina Ministry of Defense assumptions, including a temporary wage freeze after unifying defense employees’ wages in Q4 2006; police wage dynamics assumed the same as defense.

### Required fiscal adjustment magnitudes
- To make room for expenditure increases and ensure long-term debt sustainability, savings of 3–4 percentage points of GDP—in addition to the projected decline in capital expenditures of 2 percentage points—are necessary.
- Such an adjustment would:
  - Turn the primary deficit of 1 percentage point of GDP projected in 2007 into a surplus of 1 percentage point of GDP.
  - Offset identified fiscal pressures (¾ percentage point of GDP) and the projected decline in the revenue ratio (2½ percentage point).

### Possible sources of adjustment (World Bank PEIR findings)
- The WB PEIR identifies expenditures savings of up to 7 percentage points of GDP in:
  - Public wages and employment: reducing wage rates in the judiciary; reducing employment in the health sector; reducing wage bill in other categories of public employees by employment cuts and wage freezes.
  - Education: moving to per capita financing of primary and secondary education; reorienting education from technical to general; raising repetition fees; integrating faculties at the university level.
  - Subsidies and transfers to non-profit organizations: reducing subsidies to railroads; freezing agriculture subsidies in nominal terms; reducing transfers to non-profit organizations.
  - Veterans’ benefits, social welfare and child protection: tightening eligibility for veterans and families; reducing allocations for people with non-war-induced disabilities in the FBIH.
  - Pensions: limiting pension increases to inflation.
  - Pharmaceutical procurement: centralizing.
- The PEIR assumes some savings measures would be implemented in 2006; given delay, earliest possible implementation is 2007 and annual estimated savings magnitude would not change significantly with the delay.

### Quantified spending and measures (Percent of GDP)
- Total spending in categories with measures:
  - 2005: 26.5; 2006: 24.0; 2007: 22.1; 2008: 20.9; 2009: 20.0; 2010: 19.4.
- Narrow wage bill (excludes defense, police and education):
  - 2005: 5.3; 2006: 4.6; 2007: 4.2; 2008: 3.9; 2009: 3.6; 2010: 3.4.
- Judiciary: 2005: 0.7; 2006: 0.6; 2007: 0.6; 2008: 0.5; 2009: 0.5; 2010: 0.5.
- Health excl. medical professionals: 2005: 0.3; 2006: 0.2; 2007: 0.2; 2008: 0.2; 2009: 0.1; 2010: 0.1.
- Other categories: 2005: 4.3; 2006: 3.7; 2007: 3.4; 2008: 3.1; 2009: 2.9; 2010: 2.7.
- Education (includes all outlays on education, including wages): 2005: 4.8; 2006: 4.2; 2007: 4.0; 2008: 3.9; 2009: 3.8; 2010: 3.8.
- Subsidies: 2005: 1.3; 2006: 1.2; 2007: 1.0; 2008: 0.9; 2009: 0.9; 2010: 0.8.
- Transfers to non-profits: 2005: 1.8; 2006: 1.6; 2007: 1.4; 2008: 1.3; 2009: 1.2; 2010: 1.1.
- Social welfare (includes social welfare, child protection and veterans benefits): 2005: 4.0; 2006: 4.0; 2007: 3.3; 2008: 3.1; 2009: 3.0; 2010: 2.8.
- Pensions: 2005: 8.4; 2006: 7.8; 2007: 7.5; 2008: 7.2; 2009: 7.0; 2010: 6.6.
- Pharmaceutical procurement: 2005: 1.2; 2006: 0.9; 2007: 0.9; 2008: 0.9; 2009: 0.9; 2010: 0.8.
- Source: World Bank and staff estimates.

### Implications of the expenditure measures
- The identified expenditure reductions exceed what is necessary to achieve a primary balance consistent with long-term debt sustainability, providing flexibility to choose measures or implement them gradually.
- Implementing many of these measures would create room to reduce the heavy tax burden on Bosnia & Herzegovina’s economy while realizing efficiency gains.

---

### Credit flows, fiscal policy, and the external deficit — key observations
- Bosnia & Herzegovina had a large trade deficit—50 percent of GDP in 2005—which has been deteriorating in recent years.
- Booming credit since 2002 reinforced adverse external trends; credit growth driven by entry of foreign-owned banks and a surge in bank deposits after introduction of Euro notes and coin at end-2001.
- Table 1 key macro indicators (selected numbers preserved exactly):
  - Credit to the private sector: 2001: 13.7; 2002: 36.2; 2003: 24.8; 2004: 27.5; 2005: 27.5 (percentage change row).
  - Exports (in Euro) percentage change: 2001: 3.4; 2002: -7.9; 2003: 11.5; 2004: 28.7; 2005: 24.0.
  - Imports (in Euro) percentage change: 2001: 8.3; 2002: 2.5; 2003: 6.0; 2004: 7.6; 2005: 13.6.
  - Retail price index percentage change: 2001: 3.2; 2002: 0.3; 2003: 0.6; 2004: 0.2; 2005: 2.8.
  - Real GDP percentage change: 2001: 3.6; 2002: 5.0; 2003: 4.1; 2004: 5.8; 2005: 5.0.
  - Trade deficit (percent of GDP): 2001: 53.3; 2002: 53.8; 2003: 53.2; 2004: 49.5; 2005: 50.2.
  - Current account deficit (percent of GDP): 2001: 13.3; 2002: 19.1; 2003: 20.9; 2004: 19.3; 2005: 21.3.
  - Credit to the private sector (percent of GDP): 2001: 22.4; 2002: 28.9; 2003: 34.2; 2004: 40.5; 2005: 48.0.
  - General government balance (percent of GDP): 2001: -4.5; 2002: -3.3; 2003: -2.0; 2004: -0.4; 2005: 0.9.
  - Domestic saving (percent of GDP): 2001: -31.4; 2002: -30.2; 2003: -28.6; 2004: -25.7; 2005: -25.4.
- The bulk of credit flows have gone to households rather than the corporate sector; exports were only one third of imports in 2005.
- Between 2001 and 2005 the overall fiscal balance strengthened by 5½ percentage points of GDP while credit to the private sector rose by 25¾ percentage points of GDP.
- GDP growth rates fell from an average 22¾ percent during 1996–2000 to an average of only 4¾ percent during 2001–05; unemployment estimated over 20 percent and poverty rates are high.
- Current account deficit at an estimated 21 percent of GDP in 2005 is well above levels generally associated with sustainability.

### Policy tradeoffs and model-based findings
- Reducing the current account deficit while increasing growth and employment requires deep corporate sector restructuring to raise domestic savings and exports, supported by demand restraint in the short run.
- Demand restraint must come from fiscal tightening, credit restraint, or both.
- The estimated model indicates:
  - Credit to households has a strongly negative impact on the trade balance, well above the impact of credit to enterprises.
  - Fiscal revenue and expenditure both have strong effects on the trade balance.
  - A one percent of GDP contraction in fiscal expenditure has a similar impact on the trade balance as a one percent of GDP reduction in overall credit to the private sector.
- Under the currency board arrangement and open capital account, most traditional monetary policy instruments are unavailable or ineffective in restraining credit.
  - The rate of required reserves—the only monetary policy instrument available under the currency board—has very limited effectiveness given foreign-owned banks with ample liquidity from parents.
- Therefore, the burden of generating demand restraint would have to fall more heavily on fiscal policy, which can be more precisely targeted and controlled.

### Theoretical model framework (overview)
- The chapter develops a static general equilibrium approach adapted from liquidity effects models (Lucas (1990), Fuerst (1992), Christiano and Eichenbaum (1995)) to identify key explanatory variables for the trade balance and to motivate the quarterly empirical regression.
- Economy representation:
  - Small open economy, single representative household split into four agents per period: consumer, firm, bank, and government.
  - Household endowed with fixed amounts of capital (K) and loanable funds (L), and receives foreign aid (A).
  - Two composite tradable commodities: exportable (E) and importable (M); importable produced abroad; exportable produced domestically.
  - All markets competitive; Bosnian agents take foreign currency prices as given.
- Consumer:
  - Utility U(CM, CE); budget constraint C_P*P_M + C_E*P_E = Φ_C + Ψ, where P_M and P_E are importable and exportable prices, Ψ = (1 - Γ)TrK (notation preserved as in source), and Φ_C is bank credit to the consumer (exogenous).
  - Consumer demand functions: C_M = C_M(P_M, P_E, Φ_C, Ψ); C_E = C_E(P_M, P_E, Φ_C, Ψ). Demand decreases in own-price; increases in price of the other commodity, disposable income, and bank credit.
- Firm:
  - Produces exportable using capital and importable inputs F(K_f, M_f); is credit constrained and must borrow F_Φ from the bank (exogenous).
  - Profit maximization and financing constraint lead to input demand functions: M_f = f_M(P_M, P_E, r, Φ_F); K_f = f_K(P_M, P_E, r, Φ_F).
  - Market clearing in capital equates firm demand to available capital K, yielding equilibrium rental rate r = r(P_M, P_E, K, Φ_F).
  - Maximum value function max_F = F_max(P_M, P_E, K, Φ_F) gives total exportable output; export supply equals F_max minus domestic consumption C_E.
  - Export supply increases in P_E, Φ_F, and K; decreases in P_M, Φ_C, and Ψ.

*Source: IMF staff text from _cr06368 - 10.      These findings suggest that an even stronger adjustment might be needed to.*

### 17.      The government obtains funds by taxing the consumer and borrowing from the

### _cr06368 - 17.      The government obtains funds by taxing the consumer and borrowing from the

### Model setup and bank behavior
- Government budget identity (equation (13)): G + rK + Φ = Γ, where Φ represents bank credit to government and is also the fiscal balance.
- Bank lending constraint (equation (14)): G + F + C + R + A + L = − + Φ (as presented in source text), with leftover funds held as reserves (R).
- Bank receives loanable funds and the foreign aid grant from the household and lends following an exogenous credit assessment process; interest rates on bank credit are abstracted from.

### Import and export market clearing; trade balance structure
- Small open economy assumption: Bosnia & Herzegovina is a price taker in exportables and importables markets; quantities determined by optimizing decisions of consumer and firm.
- Combined relations:
  - From combining equations (2) and (13): G + rK = Ψ (equation (15)).
  - Total demand for imports (MΠ) is the sum of consumer and firm demand for the importable commodity.
  - Trade balance in foreign currency (equation (17)) expressed as a function of export and import prices in foreign currency, credit flows and other variables.
- Price-to-exchange-rate relations:
  - *M P = e M P (equation (18))
  - *E P = e E P (equation (19))
  - Where e represents the domestic currency price of a unit of foreign currency.
- Reduced trade balance function (equation (20)): B = B(Φ_F, Φ_C, G, M, E, K, e, P*) as specified in source text.
- Qualitative comparative statics:
  - Credit to households and credit to government have an unambiguous inverse relation with the trade balance.
  - Other explanatory variables (export price, import price, exchange rate, credit to firms, capital stock) have ambiguous net impacts; effects depend on relative sizes and responsiveness of components.

### The estimated model (specification and data)
- Linear regression approximation of equation (20) with seasonal dummies and lag structure yielded equation (21):
  - B_t = a0 + a1 s1 + a2 s2 + a3 s3 + a4 a_t-1 + a5 FR_t + a6 FE_t + a7 PE/M_t + a8 TR_t + ... + ε_t (presented in source text as a full lag structure with seasonal dummies s1, s2, s3; FR = fiscal revenue; FE = fiscal expenditure; PE/M = ratio of export to import prices; TR = trend real GDP).
- Data constraints and proxies:
  - No capital stock or expenditure breakdown of GDP: trend real GDP used as proxy for capital; IMF staff estimates of annual real GDP interpolated using the industrial production index to obtain quarterly real GDP; trend extracted via HP filter.
  - No price indices for Bosnia’s exports or imports: export price index for Europe used as a proxy for Bosnian import price index; import price index for Europe used as proxy for Bosnian export price index.
  - Bank credit data: good quality monetary survey data for bank credit to households and enterprises; credit flows calculated as change in end-period stocks; BIS data on credit by international banks to Bosnian non-bank private sector added to domestic bank credit to enterprises.
  - Credit to government = fiscal deficit (from equation (13)). Quarterly general government revenue and expenditure estimated by interpolating annual data using Entity-level monthly fiscal data; off-budget foreign financed project expenditure interpolated from annual data assuming even distribution through the year.
- Stationarity testing:
  - Ng and Perron (2001) unit root tests applied to all variables using Eviews.
  - Unit root tests reject non-stationarity in all cases (Table 2); levels of all variables used in estimation.

### Estimation method and results
- Sample period: quarterly data from Q1 1998 to Q2 2004. (Forecast evaluation used 2004Q3 to 2005Q3.)
- Estimation method: GMM with instruments including lagged regressors and OP (oil price index); weighting matrix set to be robust to heteroskedasticity and serial correlation.
- Key GMM coefficient estimates (Table 3; dependent variable: trade balance):
  - Constant: -1616.68 (t-statistic -6.25, P-value 0.00)
  - Seasonal dummy for Q1: 29.37 (t-statistic 0.90, P-value 0.39)
  - Seasonal dummy for Q2: -60.90 (t-statistic -3.46, P-value 0.00)
  - Seasonal dummy for Q3: 10.24 (t-statistic 0.37, P-value 0.71)
  - First lag of credit flow to enterprises: -0.21 (t-statistic -2.81, P-value 0.01)
  - First lag of credit flow to households: -1.08 (t-statistic -3.51, P-value 0.00)
  - Second lag of fiscal revenue: 0.39 (t-statistic 1.88, P-value 0.08)
  - Fiscal expenditure: -0.45 (t-statistic -2.17, P-value 0.05)
  - Export price/Import price: 17.41 (t-statistic 6.15, P-value 0.00)
  - First lag of trend real GDP: -0.45 (t-statistic -1.40, P-value 0.19)
- Memorandum and diagnostic statistics:
  - R-square: 0.86
  - Test of overidentifying restrictions (Chi-square test, 4 df): 1.68 (P-value 0.79)
  - Number of observations: 23
- Interpretation:
  - Good overall fit (R-square 0.86).
  - Test of overidentification indicates no evidence of misspecification (p-value 0.79).
  - Most coefficients significant and with expected signs, except trend real GDP and some seasonal dummies.

### Forecast evaluation
- Forecast horizon: Q3 2004 to Q3 2005 (forecast for 2005Q4 excluded due to one-off surge in imports).
- Forecast performance metrics:
  - Root mean squared error: 54.68
  - Mean absolute error: 44.19
  - Mean absolute percent error: 5.18
  - Theil inequality coefficient: 0.03
- Interpretation:
  - Theil inequality coefficient of 0.03 indicates strong forecasting ability (close actual and forecasted trade balances over the 5-quarter horizon).
  - Root mean squared error about Euro 55 million; absolute forecast error on average about 5 percent of the trade balance.

### Policy implications and experiments
- Composition of credit matters:
  - Flow of credit to enterprises has a markedly smaller negative impact on the trade balance than credit to households.
  - Policies targeting restraint of credit to households would be more effective in correcting the external deficit than undifferentiated credit controls.
- Relative speed of policy impact:
  - Fiscal expenditure has the most immediate impact on the trade balance, followed by credit flows, then fiscal revenue.
  - For immediate correction of the trade balance, fiscal expenditure restraint is the preferred option.
- Counterfactual experiments (2006 plausible quarterly paths):
  - A fiscal tightening through a one percent of GDP reduction in expenditure over a one year horizon generates a 0.45 percent of GDP reduction in the trade deficit.
  - A reduction in the flow of credit of one percent of GDP over the same horizon generates a 0.44 percent of GDP reduction in the trade deficit.
  - Conclusion: fiscal expenditure and credit tightening are roughly equally effective over a one year horizon, according to the model.
- Implementation constraints:
  - Currency board and open capital account make it very difficult to target a particular credit growth rate.
  - Typical domestic tools (required reserves, prudential regulation) are often ineffective given a banking system dominated by foreign-bank subsidiaries with access to parent liquidity.
  - Therefore, fiscal policy remains the only instrument that can be precisely calibrated to affect the trade balance.
- Policy recommendation:
  - Near term: focus efforts on fiscal policy (expenditure restraint) to restrain demand and correct the external deficit.
  - Long term: deep structural reforms are required to bring the trade deficit down to sustainable levels.

*Source: Author's calculations and materials in the provided content unit.*

### APPENDIX I: Dataset Used in the Estimation

### APPENDIX I: Dataset Used in the Estimation

### Dataset (raw table)
- Columns as presented: TBPEMTRFRFEOP (Euro million)(Percent)(Euro million)(Euro million)(Euro million)(Euro million)(Euro million)  (Euro, 2000=100)
- Observations (date followed by seven fields as presented):
  - Mar-98-545.698.91,039.77.171.2370.8582.142.9
  - Jun-98-599.096.11,059.36.3-20.1416.6581.739.7
  - Sep-98-656.092.41,078.812.315.3442.0653.138.2
  - Dec-98-717.788.41,098.239.058.7494.4678.432.8
  - Mar-99-533.294.71,117.54.1-57.6399.3682.933.8
  - Jun-99-655.995.61,136.52.523.7542.4745.649.4
  - Sep-99-709.494.31,155.2-10.73.0540.7785.463.5
  - Dec-99-738.892.51,173.714.6-4.6630.2839.774.7
  - Mar-00-475.7102.01,191.810.128.0490.8802.987.9
  - Jun-00-557.8101.41,209.622.236.8568.3830.293.4
  - Sep-00-512.699.51,227.214.318.7540.4821.8107.6
  - Dec-00-661.596.81,244.512.866.9675.8873.1111.2
  - Mar-01-562.2103.21,261.67.3-47.4465.9734.292.0
  - Jun-01-521.7103.11,278.741.7122.8555.3820.699.7
  - Sep-01-693.999.11,295.667.8-88.8663.3853.692.2
  - Dec-01-817.594.01,312.735.361.3716.0895.870.3
  - Mar-02-633.899.21,329.854.860.0542.3701.877.8
  - Jun-02-740.599.91,347.2128.4-10.9616.1793.189.2
  - Sep-02-783.999.41,364.7115.9284.2685.1806.789.2
  - Dec-02-944.894.31,382.588.432.7703.0840.287.4
  - Mar-03-623.1101.21,400.469.3-18.3669.9828.495.1
  - Jun-03-799.0100.61,418.597.211.1743.7919.876.0
  - Sep-03-820.098.71,436.966.884.3812.7933.382.3
  - Dec-03-823.793.21,455.435.961.3830.6966.980.4
  - Mar-04-619.299.21,474.153.4128.0717.5825.583.7
  - Jun-04-868.497.01,492.9126.381.4886.61,011.596.4
  - Sep-04-896.196.11,511.989.884.7842.7909.8108.1
  - Dec-04-926.991.91,531.068.592.0896.51,037.2107.4
  - Mar-05-657.399.61,550.163.980.2793.1898.0114.3
  - Jun-05-942.298.61,569.3124.969.2980.71,070.2131.3
  - Sep-05-980.897.31,588.5110.196.6962.7950.8160.2
  - Dec-05-1,221.094.81,607.7117.6214.0932.61,085.4155.0

### Notes on dataset presentation
- Values are presented in the original compact format from the source, including:
  - Date identifiers (quarterly) and numeric observations concatenated with hyphens.
  - Units and indexing label: (Euro million)(Percent)(Euro million)(Euro million)(Euro million)(Euro million)(Euro million)  (Euro, 2000=100)
- No transformations, aggregations, or imputations have been applied; the dataset is reproduced exactly as provided.

---

### References (selected citations included in the source)
- Bayoumi, T., 1999, “Estimating trade equations from aggregate bilateral data,” IMF Working Paper, No. 99/74 (Washington: International Monetary Fund).
- Brussiere, M., M. Fratzscher, and G. J. Muller, 2004, “Current account dynamics in OECD and EU acceding countries—an intertemporal approach,” Working Paper, No. 311, European Central Bank.
- Christiano, L. J. and M. Eichenbaum, 1995, “Liquidity effects, monetary policy, and the business cycle,” Journal of Money, Credit, and Banking, Vol. 27, pp. 1113–1136.
- Chrystal, K. A. and P. Mizen, 2005, “A dynamic model of money, credit, and consumption: A joint model for the UK household sector,” Journal of Money, Credit, and Banking, Vol. 37, pp. 119–143.
- Fuerst, T. S., 1992, “Liquidity, loanable funds, and real activity,” Journal of Monetary Economics, Vol. 29, pp. 3–24.
- Goldstein, M., and M. Khan, 1985, “Income and price effects in foreign trade,” in R. W. Jones and P. B. Keenen (eds), Handbook of International Economics, Vol II, Elsevier Science Publishers B. V.
- Hansen, B. E., and K. D. West, 2002, “Generalized method of moments and macroeconomics,” Journal of Business and Economic Statistics, Vol. 20, pp. 460–469.
- Jones, J. B., 2003, “The dynamic effects of firm-level borrowing constraints,” Journal of Money, Credit, and Banking, Vol. 35, pp. 743–762.
- Kishan, R. P., and T. P. Opiela, 2000, “Bank size, bank capital, and the bank lending channel,” Journal of Money, Credit, and Banking, Vol. 32, pp. 121–141.
- Li, V. E., 2000, “Household credit and the monetary transmission mechanism,” Journal of Money, Credit, and Banking, Vol. 32, pp. 335–356.
- Letendre, M., 2004, “Semi-parametric predictions of the intertemporal approach to the current account,” Journal of International Economics, Vol. 64, pp. 363–386.
- Lucas, R. E., 1990, “Liquidity and interest rates,” Journal of Economic Theory, Vol. 50, pp. 237–264.
- Mas-Colell, A., M. D. Whinston, and J. R. Green, 1995, “Microeconomic Theory,” Oxford University Press, New York.
- Ng, S. and P. Perron, 2001, “Lag length selection and the construction of unit root tests with good size and power,” Econometrica, Vol. 69, pp. 1519–1554.
- Peek, J. P., E. S. Rosengren, and G. M. B. Tootell, 2003, “Identifying the macroeconomic effect of loan supply shocks,” Journal of Money, Credit, and Banking, Vol. 35, pp. 931–946.
- Rudebusch, G. D., 2005, “Assessing the Lucas critique in monetary policy models,” Journal of Money, Credit, and Banking, Vol. 37, pp. 245–272.
- Stock, J. H., and J. H. Wright, 2000, “GMM with weak identification,” Econometrica, Vol. 68, pp. 1055–1096.
- Stock, J. H., J. H. Wright, and M. Yogo, 2002, “A survey of weak instruments and weak identification in generalized method of moments,” Journal of Business and Economic Statistics, Vol. 20, pp. 518–529.
- Stucka, T., 2004, “The effects of exchange rate change on the trade balance in Croatia,” IMF Working Paper, No. 04/65 (Washington: International Monetary Fund).

---

### Institutional approaches to fiscal coordination (excerpt: chapter opening)
### Key points (paragraphs 1–9)
- The fiscal system in Bosnia & Herzegovina is fragmented and characterized by a small central State government with weak control over consolidated fiscal policy. This situation has its origins in the 1995 Dayton Peace Accord. Political fragmentation has complicated coordination among the State and Entity governments and undermined the authorities’ ability to achieve and preserve macroeconomic stability. The importance of ensuring effective fiscal coordination is heightened by Bosnia & Herzegovina’s euro peg, which effectively removes monetary policy from the tool kit of macroeconomic policies. (Paragraph 1)
- Fiscal coordination among the State and Entity governments was, until recently, achieved through the direct involvement of the Office of the High Representative (OHR), whose so-called ‘Bonn powers’ gave it considerable influence over fiscal policy. The IMF also provided significant guidance, both in its advisory role and through two Fund-supported programs between 1998 and 2004. (Paragraph 2)
- The international community is now playing an increasingly less interventionist role. The establishment in May of 2005 of the National Fiscal Council (NFC) as an informal body of economic policy-makers was intended to help fill the vacuum. The chaos surrounding passage of the 2006 budgets revealed serious weaknesses in the ability of the NFC to coordinate fiscal policy. At the same time, Bosnia & Herzegovina is taking on significant amounts of public debt from the settlement of domestic claims and investment in infrastructure, raising questions about fiscal governance and the need to strengthen the institutional framework for fiscal coordination. (Paragraph 3)
- Many other countries have adopted institutional mechanisms to facilitate fiscal coordination among levels of government. This chapter attempts to identify—among a sample of such countries—useful lessons for Bosnia & Herzegovina in its efforts to establish an efficient institutional setup for fiscal coordination. The chapter is organized as follows: Section B describes factors that affect the choice and efficacy of institutional arrangements (including the relationship between fiscal coordination and a country's exchange rate regime); Section C describes mechanisms used by a small sample of federal states and how they address common challenges; Section D articulates key lessons; Section E discusses application to Bosnia & Herzegovina. (Paragraph 4)
- Fiscal coordination is defined as formal or informal procedures, practices and/or rules that guide fiscal policy decisions of various levels of government intended to secure overall fiscal sustainability and macroeconomic stability. The chapter focuses on fiscal coordination rather than the broader literature on intergovernmental fiscal relations. (Paragraph 5)
- Institutional set-up and administrative procedures are key for fiscal coordination, including the establishment of special coordination bodies in decentralized fiscal systems and the role of political culture in determining appropriateness of particular models. (Paragraph 6)
- There are many institutional “models” and many interconnected political, economic, legal and cultural factors which influence relative effectiveness. It is difficult to establish a clear link between a particular model and fiscal performance; the adoption of an institutional mechanism may reflect a commitment to fiscal prudence rather than be its sole cause. (Paragraph 7)
- The importance of effective fiscal coordination is partly a function of the extent to which monetary policy can be actively used. Where monetary policy is constrained (for example, by a pegged exchange rate), sound fiscal policy and effective fiscal coordination become more important. (Paragraph 8)
- Examples relating fiscal coordination and exchange rate regimes:
  - Argentina and Brazil had pegged exchange rate regimes; Argentina’s currency board collapsed in part due to failure to rein in fiscal profligacy of sub-national governments; weaknesses in consolidated fiscal policy forced Brazil to abandon its crawling peg in 1999.
  - Australia had a pegged exchange rate until 1983; the Australian Loan Council (ALC) required approval of all borrowing by national and sub-national governments; abandonment of the peg in 1983 coincided with a move to a looser, voluntary framework.
  - Switzerland has a fully floating exchange rate and little explicit effort to coordinate fiscal policy; monetary policy can play an active role in macro stabilization.
  - Euro area members are constrained by euro area obligations; discipline is instilled by explicit penalties contained in the Stability and Growth Pact (SGP). (Paragraph 8, bulletized examples)
- The chapter examines institutional mechanisms in Argentina, Australia, Austria, Belgium, Brazil, the European Union, South Africa and Switzerland, assessing legal foundations, involvement of key policy makers, formal powers, comprehensiveness of coverage, and enforcement track record. (Paragraph 9)

*Source: APPENDIX I: Dataset Used in the Estimation (as provided in the supplied content)*

### 10.      Most of the countries reviewed have pursued fiscal coordination through

### _cr06368 - 10.      Most of the countries reviewed have pursued fiscal coordination through

### Approaches and mandates
- Frameworks differ in structure and mandate reflecting historical contexts, political culture and economic conditions.
- Range of mechanisms:
  - Purely advisory/consultative bodies: Financial and Fiscal Commission (FFC) in South Africa; Public Sector Borrowing Requirement Section (PSBRS) of the High Council of Finance (HCF) in Belgium; Conference of Cantonal Ministers of Finance (CCMF) in Switzerland.
  - Legally binding/coercive bodies: Australian Loan Council (ALC) — decisions legally binding on national and sub-national governments (as originally functioning 1927–1983).20
  - Hybrid arrangements: Domestic Stability Pact (DSP) in Austria; Fiscal Responsibility Laws (FRL) in Argentina and Brazil.
- Country-specific notes from Table 1:
  - Argentina: Federal Council for Fiscal Responsibility (FCFR) and the Fiscal Responsibility Law (FRL) — adopted in 2004. The FCFR assesses compliance with the FRL and voluntary Federal-Provincial fiscal pacts and applies sanctions for non-compliance. Nineteen of 24 provinces have signed up to the framework as of June 2006.
  - Australia: Australian Loan Council (ALC) — est. 1927. Permanent supra-national authority mandated to set a ceiling on the amount and distribution of borrowing by national and sub-national governments.
  - Austria: Domestic Stability Pacts (DSP) — 1996, 1999, 2001, 2004 — and a “consultation mechanism”. The DSP sets balance targets for all three levels of government and ensures consistency with national obligations under the SGP.
  - Bosnia & Herzegovina: National Fiscal Council (NFC) — est. 2005. Informal body setting consolidated fiscal balance and its distribution among State and Entity governments.
  - Belgium: Public Sector Borrowing Requirement Section (PBSRS) of the High Council of Finance (HCF) — est. 1989. Technical and advisory body producing ‘advisory opinions’ on borrowing requirements of national and sub-national governments.
  - Brazil: Fiscal Responsibility Law (FRL) and Fiscal Crimes Law (FCL) — adopted 2000. FRL requires balanced primary budgets, defines ceilings for various expenditures, promotes transparency in public accounts and prohibits refinancing of sub-national government debt by the national government; the FCL penalizes government officials who break fiscal parameters.
  - European Union: Council for Economic and Financial Affairs (ECOFIN) and the Stability and Growth Pact (SGP) — 1998. Development of general economic guidelines, coordination of economic policies, surveillance and monitoring of members’ budget policies, and related functions.
  - South Africa: Financial and Fiscal Commission (FFC), Budget Council (BC), and Budget Forum (BF) — est. 1996. FFC constitutionally mandated advisory body providing independent impartial advice on intergovernmental fiscal matters; BC and BF provide fora for discussion of FFC recommendations.
  - Switzerland: Conference of Cantonal Ministers of Finance (CCMF) — est. 1993. Informal forum for cantons to discuss vertical fiscal/financial matters and facilitate common positions.
- Historical/political drivers:
  - ALC established to avoid sub-national competition on international markets for limited financing and was supported by a state-building objective; constitutional amendment ceded power to ALC dominated by national government.20
  - Argentina and Brazil: strong regional/local autonomy (centrifugal forces) constrained willingness to cede sovereignty; both experienced crises in late 1990s/early 2000s linked to failure to coordinate sub-national fiscal policy with national fiscal goals (Argentina: contributed to collapse of currency board; Brazil: devalued exchange rate by 70 percent against the U.S. dollar in 1999). Coordination frameworks evolved differently: Argentina used voluntary participation encouraged via financial incentives; Brazil relied more on fiscal rules enforced by national authorities.
  - Switzerland: strong regional autonomy, but a culture of fiscal prudence at sub-national level avoided crises that compelled Argentina and Brazil to sacrifice sub-national autonomy.
  - South Africa: FFC established as part of post-apartheid state-building; FFC created by 1996 constitution to limit subjective political allocation decisions and provide apolitical advice without formal decision-making power (unlike Australia where institution had constitutional status and decision-making power).

### Membership and decision-making
- Membership composition:
  - Dominated by politicians (Ministers of Finance or Economy or heads of government at national and sub-national levels).
  - Non-elected members included only in PSBRS of Belgium’s HCF and South Africa’s FFC (both advisory bodies). PSBRS membership includes nominees from National Bank of Belgium, the Flemish and Walloon governments, the Brussels Capital Government, and the French Community Government; exclusion: none of its members can be members of the national legislature or community regional councils or elected councils of municipalities with more than 30,000 residents.
  - FFC commissioners appointed by the President or nominated by Executive Councils of the nine provinces and are not permitted to hold office in any political party or organization.
- Independence versus political legitimacy:
  - European Commission (2006) noted a “general perception that independent fiscal institutions have contributed to fiscal discipline” and that such institutions “seem to have a considerable impact on the public debate and the recommendations formulated are generally followed by governments”.24
  - South Africa: concerns that FFC may have become marginal despite strict independence and constitutional mandate; exclusion of government members and politically-associated individuals may have undermined political ownership of FFC advice.25
  - Belgium’s HCF (with explicit links to political constituencies, though excluding public office holders) appears somewhat more effective; suggests need for balance between independence and political legitimacy.26
  - Bosnia & Herzegovina experience indicates direct participation of high-level government members does not guarantee advice will be accepted/acted upon.
- Decision-making rules and voting:
  - Ability to facilitate timely decisions often depends on well-specified voting formula or mechanisms to resolve impasse.
  - Of mechanisms taking formal decisions on fiscal targets or penalties, most require unanimity. Exceptions permitting decisions without unanimity: Australia’s ALC, Argentina’s FCFR, and Europe’s SGP.
  - ECOFIN decisions on SGP compliance mainly by qualified majority (sometimes in consultation with European Parliament).
  - Argentina: consensus not required to impose sanctions, but provincial dominance has effectively prevented enforcement of sub-national fiscal targets.27
  - ALC voting: different rules for different decisions—consolidated borrowing ceiling decision required a simple majority; distribution of loan proceeds required unanimity. National government had two votes plus a “casting vote”; each of six states had one vote. Failure to reach consensus reverted distribution to pre-set formula based on relative net borrowing over previous 5 years.28
  - De Mello (2002b) observation: “in decentralized federations, fiscal consolidation is carried our predominantly by the central government. Because it is more accountable for macroeconomic stability, the central government often has stronger incentives for fiscal discipline and, therefore, bears the brunt of the adjustment costs in periods of fiscal retrenchment”.

### Fiscal targets and determination across levels
- Countries target different fiscal variables (Table 2):
  - Targets include consolidated balance and/or borrowing; consolidated debt level; sub-national balances and/or borrowing; sub-national debt levels; sub-national debt service; and expenditure ceilings.
- Country coverage in Table 2 (matrix presence marked as X in source):
  - Argentina: consolidated balance and/or borrowing (X); sub-national balances and/or borrowing (X); sub-national debt service (X).
  - Australia: consolidated balance and/or borrowing (X); consolidated debt level (X).
  - Austria: consolidated balance and/or borrowing (X); consolidated debt level (X); expenditure ceilings (X).2/
  - Belgium: consolidated balance and/or borrowing (X); consolidated debt level (X); sub-national balances and/or borrowing (X); expenditure ceilings (X).
  - Brazil: consolidated balance and/or borrowing (X); consolidated debt level (X); sub-national balances and/or borrowing (X); sub-national debt levels (X); sub-national debt service (X); expenditure ceilings (X).
  - South Africa: (no X marks reported in supplied excerpt).
  - Switzerland: consolidated balance and/or borrowing (X); consolidated debt level (X); expenditure ceilings (X).1/
- Notes to Table 2:
  - 1/ Excludes most capital spending and IFI-financed current spending.
  - 2/ Targets are set on all municipalities within a state as a group rather than individually.
- Spectrum of approaches to setting targets:
  - Switzerland: makes no systematic attempt to set a consolidated target.29
  - Explicit fiscal rules for all or some levels of government exist in Argentina, Austria, Australia, Brazil, and the EU (negotiated between respective governments).
  - Middle approach: South Africa and Belgium — governments determine their fiscal balances autonomously, but monitored by legislatively- or constitutionally-mandated bodies providing advice on consolidated fiscal stance and intergovernmental fiscal issues.

### Enforcement and continuity
- Enforcement mechanisms vary:
  - Some frameworks apply sanctions for non-compliance (Argentina’s FCFR applies sanctions; Brazil’s FRL and FCL include penalties).
  - In many decision-making bodies unanimity requirements and political dynamics limit enforcement in practice (e.g., Argentina’s provincial dominance hindering enforcement).
- Continuity provisions:
  - Austria’s DSP: negotiated for only a few years at a time; provision ensures that if agreement cannot be reached on a new pact before expiration, the old pact comes back into force, assuring continuity.
- International/integration drivers:
  - European integration and currency union motivated SGP; SGP targets apply only to consolidated general government and responsibility for non-compliance rests with central (national) governments.
  - Diversity of internal coordination among SGP signatories with federal systems: Austria adopted DSPs with numerical parameters and enforcement linked to SGP; Belgium relied mainly on persuasion via HCF.

*Source: _cr06368 - 10.      Most of the countries reviewed have pursued fiscal coordination through (excerpt provided).*

### 24.      A number of mechanisms make use of various types of sanctions to enforce fiscal

### _cr06368 - 24.      A number of mechanisms make use of various types of sanctions to enforce fiscal

### Sanctions frameworks and mechanics
- The Excessive Deficit Procedure (EDP) of the SGP for Euro area countries (e.g., Austria and Belgium) sets explicit penalties for countries exceeding the 3 percent deficit ceiling.
  - Penalties take the form of noninterest-bearing deposits with the European Commission partly linked to the size of the actual deficit.
  - Sanctions can be intensified by an additional deposit not to exceed 0.5 percent of GDP.
  - The deposit is converted into a fine if the excessive deficit has not been corrected after two years.
  - To date, no country has been subject to penalties; ECOFIN Council has suspended or delayed the EDP where judged that a country is making sufficient progress in correcting its excessive deficit.
- Under the revised provisions of March 2005, members can avoid EDP if they experience negative growth (previously annual growth of less than -2 percent) or if a deficit results from an unusual event outside its control or from a severe downturn. The 2005 revision also included “enhanced surveillance” and “peer pressure” to discipline members.
- Many sanction frameworks allow discretion in application, creating potential enforcement loopholes—especially when breaches stem from financial distress and decisions to apply penalties are political.

### Country examples of sanction design and enforcement
- Austria (DSP for sub-national governments)
  - Sanctions levied exclusively on sub-national governments and designed to mimic the European SGP.
  - Non-compliers are to pay 8 percent of the absolute amount of the targeted balance and 15 percent of the shortfall up to a ceiling.
  - Sanctions withheld by the Ministry of Finance and deposited on special account; if adjustments are not made, the deposit is turned into a fine and distributed among complying governments.
  - No sanctions administered to-date due to considerable flexibility in assessment of compliance.
  - The national government instituted penalties for governments that failed to ratify the agreed pact on time; applied once and the penalty was refunded after ratification.
- Brazil
  - Legislative requirement for a contingency reserve in its budget defined in percent of net current revenues.
  - Uses financial and administrative sanctions, e.g., withholding of federal transfers to sub-national governments not effectively collecting own taxes.
  - Breaches of personnel spending ceilings must be eliminated within 8 months or the government is prohibited from engaging in new credit operations.
  - Time frame for corrective action can be increased if the economy contracts by 1 percent or more over the previous four quarters or if a natural catastrophe is declared by the legislature.
  - Fiscal Crimes Law provides sanctions against officials: fines on individuals, loss of job, ineligibility for public office for a maximum of five years, and imprisonment.
- Argentina
  - Framework includes penalties and incentives: compliance rewarded with federal financing at interest rates reflecting extent of provincial consolidation or favorable reschedulings of intergovernmental debt service obligations.
  - If the Executive Committee of the FCRC concludes a violation, the province is given 10 days to explain; FCRC can recommend remedial measures such as publication of the infraction, restrictions on voting power, exclusion from federal tax incentive schemes, limitations on guarantees, denial of new borrowing authority, and limitation on federal discretionary transfers.
  - The federal government is explicitly exempt from sanctions.
- Australia
  - Significant vertical imbalance allows the national government to use discretionary transfers or expenditures to leverage sub-national compliance.
  - Historical precedent: in 1932 the national government, under the Financial Agreements Enforcement Act, redirected certain revenues and bank deposits away from New South Wales to enforce compliance.
- Switzerland
  - No coordinated enforcement framework; federal government adopted rules in 2003 imposing a ceiling on its own expenditures and requiring its budget to be balanced over the cycle.
  - Several cantons have voluntarily adopted fiscal rules, some in cantonal constitutions.
  - Discipline derives from public pressure, competition among cantons, and requirements for referenda on major spending or revenue decisions.

### Credibility and discretion in enforcement
- Credibility of enforcement regimes depends on the ability and willingness of the relevant government to punish non-compliance.
- Many frameworks with formal penalties allow discretion in application (e.g., EU, Argentina, Austria, Brazil), creating potential major loopholes.
  - Austria: imposition of sanctions requires unanimous decision of a four-member ‘conciliation committee’ (two central government and two state/municipal representatives).
  - Brazil: suspension of sanctions requires a decision of parliament.
  - Euro area: EC Directorate can recommend corrective measures, but fines have been practically impossible to implement.
- Absence of explicit penalties does not imply lack of enforcement: informal enforcement via peer pressure and threat of indirect financial repercussions can be effective when the enforcing entity has resources and commitment.
  - Brazil: national government can withhold discretionary transfers to states/municipalities not showing adequate tax collection.
  - Australia: national government has redirected tax revenues or reduced discretionary transfers to non-complying states.

### Monitoring compliance and transparency
- Fiscal coordination requires transparent and regular monitoring, consistent accounting rules, and common indicators of fiscal performance.
- Examples of monitoring arrangements:
  - Argentina’s FRL mandates homogeneous fiscal indicators and requires national and state governments to publish information on budget execution, debt stock, and debt service with no more than a 3-month delay. A commission monitors consistency of budget data among the federal government and provinces. Fiscal reporting requirements are also features of PFO bilateral agreements.
  - Brazil requires timely budget reporting and exchange of information among levels of government; jurisdictions produce a “Fiscal Management Report” three times a year containing indicators on compliance with the FRL. The FRL requires the annual budget law to contain a “Fiscal Targets Annex” assessing compliance with the previous year’s targets. The Ministry of Finance publishes a monthly list of governments that have exceeded their debt limits.
  - EU: national accounting standards set by EUROSTAT to reduce scope for creative accounting under GSP (though scope for creative accounting remains). Austria and Belgium have not made harmonized accounting mandatory for lower levels of government.
  - Belgium’s HCF produces a report every March on compliance with the previous year’s fiscal targets by all levels of government.
  - Austria’s DSP imposes no obligation for individual governments to publish budget balances as part of the DSP, although a fine may be imposed if important budgetary information is withheld from the coordination committee.
  - Switzerland: voluntary efforts to harmonize budget and accounting practices across cantons.
- Australia’s experience since 1983: sub-national governments shifted toward improving fiscal transparency; borrowing in excess of 3 percent of the ALC allocation required a public explanation to the ALC. In 1993 the ALC began publishing a “National Fiscal Outlook” (NFO) report presenting consistent fiscal forecasts; the NFO was discontinued in 1999 as reporting standards improved.

### Role of the central bank
- Central banks can play an important but indirect role in fiscal institutions; they tend not to be direct participants in fiscal coordination discussions.
  - Belgium: central bank governor is a formal member of the coordinating institution.
  - Australia: central bank acted as principal underwriter of government loans and advised on reasonable borrowing amounts; it could refuse to underwrite borrowing deemed macroeconomically inappropriate, creating powerful implicit discipline even if government could float un-underwritten loans.

### Overview of cross-country lessons
- Adoption of a model is determined by historical, political, and economic context; models are not one-size-fits-all and evolve over time.
- Importance of fiscal coordination depends partly on availability of alternative channels for macroeconomic stability (e.g., floating exchange rate and independent central bank reduce reliance on fiscal coordination; pegged exchange rates increase need for fiscal coordination).
- Coordinated fiscal policy is not the same as good fiscal policy; institutions facilitate but do not guarantee good policies.
- No institutional framework can be effective without political will; legislative frameworks and formal rules are no substitute.
- Enforcement responsibility usually rests with national governments; national governments can most easily enforce compliance when they have sufficient resources to leverage sub-national entities.
  - The SGP is the only framework identified that allows penalties on national governments, but sanctions under the SGP have not been applied to any case of non-compliance.
- Discretion in application and enforcement is common across coordination models; frameworks with automatic penalties that require explicit decisions to be suspended are preferable to those that require explicit decisions to impose penalties. The higher the hurdle to suspend sanctions, the less likely suspension is to occur.

*Italic: IMF country report content unit _cr06368, sections 24–35.*

### 36.      As a federal state with significant resources and powers devolved to the sub-

### _cr06368 - 36.      As a federal state with significant resources and powers devolved to the sub-

### Need for effective fiscal coordination
- As a federal state with significant resources and powers devolved to the sub-national level, effective fiscal coordination is essential.
- Unlike in many other decentralized federations (e.g., Australia), the national government in Bosnia & Herzegovina does not itself have the budgetary resources to assume responsibility for fiscal policy in pursuit of macroeconomic stabilization.
- Budgeted State expenditures in 1996 accounted for only 4 percent of GDP compared with almost 50 percent for all other levels of government.

### Currency board implications
- Effective fiscal coordination is particularly important if Bosnia & Herzegovina desires to maintain its currency board since monetary policy is not available as a tool of macroeconomic policy.
- The country examples reviewed signal clear risks in attempting to maintain a currency peg without effective control over consolidated fiscal policy.
- The existence of a pegged exchange rate may point to the need to give the State government a relatively greater say in setting the consolidated deficit target (as in Australia) since it is this level of government that has the most direct responsibility for protecting the national currency.

### Role of institutions in coordination
- At present, the central bank governor has observer status on the NFC.
- More active participation of the Central Bank (CBBH) in the NFC is one way to facilitate fiscal coordination; the CBBH could regularly present its views to the NFC on macroeconomic conditions within which fiscal decisions are to be made.
- At present, this role is filled by Economic Policy and Planning Unit (EPPU) which is accountable to the Council of Ministers. A greater role for the central bank need not supersede that played by the EPPU but could provide a valuable alternative perspective on economic conditions.

### Urgency and institutional weaknesses
- The task of strengthening fiscal coordination is becoming increasingly urgent.
- To date, Bosnia & Herzegovina has relied on pressure from, or outright intervention by, the international community to ensure the appropriateness of fiscal policy.
- The OHR, which until recently had used the powers vested in it as part of the Dayton Accord to ensure responsible fiscal management, has indicated its intention to exit Bosnia & Herzegovina in 2007.
- The current framework for fiscal coordination lacks many elements associated with effective coordination in other countries:
  - The NFC does not yet have legal status and the participation of key decision-makers in the NFC is erratic.
  - There is not yet a formal agreement among the State and Entity governments on a harmonized budget calendar or on how to deal with the failure of policy-makers to come to timely decisions on key elements of fiscal policy.
  - Reliable and timely data on the fiscal performance of lower levels of government is lacking.
  - Enforcement of agreements on fiscal policy among governments is weak: there are no penalties for non-compliance, public pressure has not yet proved a reliable disciplining mechanism, and since Bosnia & Herzegovina has not yet borrowed on private markets, market discipline does not figure prominently.

### Minimum requirements and recommended measures
- At a minimum, effective fiscal coordination will require:
  - Intergovernmental agreement on, and adherence to, a budget calendar that ensures decisions on key fiscal parameters and adoption of annual budgets by all levels of government occur in a timely and predictable manner.
  - Adoption of a reporting system that secures comprehensive, accurate, and timely fiscal data on budgets and on budget execution, for all levels of governments and all off-budget fiscal operations, individually as well as on a consolidated basis.
- Decision-making could be facilitated with adoption of numerical fiscal rules (for example, deficit or debt ceilings or the use of “golden rules”) to help guide the setting of targets.
  - Fiscal rules are neither necessary nor sufficient but could help guide policy-makers as political culture matures and international community involvement diminishes.
  - The “contractual” use of fiscal rules could help address “common pool” challenges associated with the diverse, multi-party character of Bosnia & Herzegovina’s governance.
- Enforcement:
  - Some form of penalty and/or incentive will be necessary to foster fiscal discipline among lower levels of government.
  - The national government does not command sufficient resources to indirectly enforce compliance; enforcement design should limit discretionary use to exceptional circumstances.
- Capacity strengthening:
  - Authorities need to strengthen analytical capabilities for budget analysis and projections over the longer term.
  - The State should be empowered to monitor the operations of line ministries, other levels of government, and off-budget entities.

### Comparative elements (Annex I) — Bosnia & Herzegovina entries
- legislative or constitutional underpinnings: N
- key policy makers represented: Y
- formal decisions taken: Y
- formal voting: N
- formal voting formula/deadlock breaking mechanism: N
- Penalties specified: N
- Past Use of Penalties: N
- Explicit Budget Calendar: N
- Harmonized Reporting and Accounting Standards across levels of government: N

*Source: _cr06368 - 36. As a federal state with significant resources and powers devolved to the sub-*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2006/_cr06368.pdf_