## cr18132 ## Source details **Canonical URL:** [cr18132](https://www.imf.org/-/media/files/publications/cr/2018/cr18132.pdf) ## Other formats - [Markdown version](/-/media/files/publications/cr/2018/cr18132.pdf.md) - [Structured JSON version](/-/media/files/publications/cr/2018/cr18132.pdf.json) --- ### Preface, mission, and purpose - Mission: IMF Fiscal Affairs Department (FAD) visited Tbilisi, Georgia, during September 13 – 25, 2017 to provide technical assistance on enhancing Georgia’s fiscal rules framework. - Mission composition: Torben Hansen (FAD, head); Isabel Rial (FAD); Joao Jalles (FAD); Stephen Farrington (FAD expert); Sami Ylaoutinen (FAD expert). - Key MoF interlocutors and other stakeholders listed in mission participants (Deputy Ministers, heads of MAFD, Budget Department, Tax and Customs Policy, Public Debt and External Financing Department, Treasury Service, Revenue Service; Parliament, National Bank, PBO, SAO). - Purpose: Deliver technical assistance on enhancing Georgia’s fiscal rules framework in response to earlier findings, including the Fiscal Transparency Evaluation (FTE) conducted by FAD in late 2016. - Legal framework in place: Economic Liberty Act (ELA) adopted in 2011, in force 2014, which defines numerical upper limits: - State debt: 60 percent of GDP. - Budget balance (consolidated budget deficit): 3 percent of GDP. - Expenditure rule (ER): expenditures plus the increase in non-financial assets of the consolidated budget to GDP should not exceed 30 percent. - Compliance history: - Debt rule (DR) and budget balance rule (BBR) adhered to since introduction. - Expenditures have exceeded the legislative limit, albeit by a small margin. - FTE findings: gaps in reporting of general government revenue and expenditures relative to GFSM2014; gaps in assessment and reporting on compliance with fiscal rules; recommended review of fiscal rules framework. ### Design of the fiscal rules — principal findings and recommended changes - Design considerations: - Objectives: fiscal sustainability and flexibility to respond to economic shocks. - Context: transitional Georgian economy and intention to gradually move towards the European Union fiscal governance framework. - Rules should be simple and not pro-cyclical. - Assessment of current rules: - Current debt rule (DR) and budget balance rule (BBR) are appropriate and compatible with realistic debt and deficit trajectories. - Structural balance rules would be difficult to implement at this stage. - Expenditure rule (ER) specification as a percentage of GDP has issues, including potential pro-cyclical properties. - Recommendation for ER: - Replace current ER in the ELA with nominal expenditure ceilings set on a rolling basis as part of the MTBF, consistent with the debt and BBRs and medium-term fiscal objectives. - Strengthen three ELA elements: - Corrective mechanisms: - Current corrective mechanism applies to budgetary plans but not to actual outturns. - Introduce ex-post reporting requirements on compliance and require the government to put forward a corrective plan in case of non-compliance. - Escape clause: - Abolish the criterion allowing Parliament to approve a budget not compliant with fiscal rules. - Adopt a clear definition of economic recession (example proposed: two consecutive quarters of negative real year-on-year GDP growth). - Specify who can decide to activate the escape clause and associated reporting requirements. - Independent oversight: - Incorporate clear provisions on independent oversight and verification of government compliance with the fiscal rules. - Recommended timeline (selected design actions): Replace ER with nominal ceilings; introduce ex-post reporting and corrective-plan requirement; abolish parliamentary approval of non-compliant budgets unless escape clause triggered; incorporate independent oversight; incorporate periodic review clause — all proposed for 2018 (MoF proposals). ### Coverage and measurement — gaps, quantified impacts, and recommendations - Main coverage gaps relative to GFSM2014: - General government revenue and expenditures exclude own-source revenue and associated expenditures by Legal Entities of Public Law (LEPLs) and public non-market producers classified as SOEs. - Some transactions between government units and SOEs are not recorded and reported as expenditures in line with GFSM2014. - Debt reporting focuses on central government only; gross debt excludes liabilities of LEPLs and subnational governments. - Quantified implication (based on 2015 data): - Aligning coverage and measurement with GFSM2014 would have resulted in the overall deficit being 0.6 percentage points of GDP higher compared to official data. - LEPLs (institutional and transactional coverage): - Central government controls 496 LEPLs; additional number controlled by subnational governments unspecified. - By end-2015, LEPLs own revenues represented about 12 percent of general government total revenue, and about 17 percent of total expenditures. - Staff estimates that inclusion of LEPLs in general government statistics would: - Increase revenue by 4.5 percent of GDP; - Increase expenditures by 4.3 percent of GDP; and - Decrease the general government deficit by 0.2 percent of GDP. - Authorities estimate LEPLs loans at end-2015 account for 2 million Lari. - Table 3.1 (transactions of LEPLs controlled by central government, 2015–16, percent of GDP): 2015 revenues 9.5; own-revenues 4.5; expenditures 9.3; funded by own-revenues 4.3; overall balance 0.2. (2016 column values provided in source.) - SOEs classification and measurement: - GFSM2014 delineation: classify market vs non-market producers case-by-case; market test often uses 50 percent threshold comparing value of revenue sales and average production costs over at least 3 years. - Staff estimates: 15 percent of the 65 largest SOEs under central government control would not pass the market test in 2015. - Measurement of government support: - Transfers/subsidies to SOEs: recorded as expenditures (above-the-line) and reduce overall deficit. - Loans and capital injections expected to yield financial assets: recorded as financial transactions (below-the-line) and not affect overall deficit. - Georgia practices and issues: - Two main forms referred to as budget lending: on-lending (external) and capital injections. - Table 3.3 (Budget Lending composition, percent of GDP): 2013 total 1.0 (on-lending 0.7; capital injections 0.3), 2014 total 0.9 (0.7/0.3), 2015 total 1.4 (0.6/0.7), 2016 total 1.4 (0.7/0.7). - On-lending debt service payments (Table 3.4, Million Lari and percent): Principal planned vs actual and percent for 2013–2016 (2013: 26.8, 27.4, 102; 2014: 67.6, 55.4, 82; 2015: 78.7, 68.7, 87; 2016: 88.3, 85.7, 97; Av. % 92). Interest and commissions planned vs actual and percent for 2013–2016 (2013: 8.8, 9.6, 109; 2014: 23.2, 17.8, 77; 2015: 26.1, 25.4, 97; 2016: 28.7, 26.2, 91; Av. % 94). - Capital injections often do not meet the “capital injection test” and preliminarily should be recorded as capital transfers (expenditures above-the-line). - Coverage of public debt: - Debt rule coverage narrower: DR limited to central government; budget balance and expenditure rules cover central and subnational governments. - Gross debt compiled on cash basis; does not include accounts payable; compiled in nominal values. - Recommendation: extend institutional coverage of gross debt to general government per GFSM2014 and Guide for Public Sector Debt Statistics (2011), adding liabilities of subnational governments, LEPLs, and reclassified SOEs. - Explicit recommendations (coverage & measurement): - Recommendation 3.1: Include all flows and stocks of LEPLs in fiscal aggregates (December 2017). - Recommendation 3.2: Case-by-case analysis of SOE sector, prioritizing high-risk SOEs for reclassification into general government (December 2018; update every 3 years). - Recommendation 3.3: Include liabilities of LEPLs, subnational governments, and reclassified public non-market entities in gross debt statistics (End 2018). - Recommendation 3.4: Record budget lending to SOEs, including capital injections, in accordance with GFSM2014; record transfers/subsidies above-the-line (December 2017). ### Implementing the fiscal rules — buffers, simulations, MTBF, and reporting enhancements - Safe debt and simulation findings: - Stochastic simulations indicate a “safe” debt level for Georgia in the range of 35–40 percent of GDP to ensure public debt remains below the debt limit with high probability when negative shocks materialize. - Alternative staff stochastic simulations (based on historical shocks) pointed to a “safe debt” level of about 45 percent of GDP under certain baseline assumptions; note caveats and optimistic assumptions. - Baseline assumptions cited in simulations: Georgia growing at 8 percent in nominal terms; foreign financing at 4 percent nominal; inflation at 3 percent; no significant changes in nominal exchange rate. - Risk: In a severe economic shock, public debt could exceed the 60 percent of GDP ceiling by 2019 with a 10 percent probability (per one simulation framing). - Consolidation implications: - A prudent safe debt level of 35–40 percent of GDP implies overall deficits of about 2.3–2.7 percent of GDP over the medium-term. - Convergence to a safe debt level of 35–40 percent would require consolidation of about 0.3–1.0 percent of GDP over the medium-term relative to a current overall deficit of 3.0 percent of GDP. - Under assumption of fixed revenue share (total revenue to GDP ratio of 28.4 for 2016), consistent level of expenditures estimated at around 31 percent of GDP. - Fiscal targets and buffers: - Medium-term fiscal targets should provide adequate buffers to absorb adverse shocks and fiscal risks without breaching fiscal rules. - Buffers sized using stochastic simulations and risk assessments, including exposures to SOEs, PPAs, and PPPs. - Reporting and MTBF enhancements: - Enhance ex-post and ex-ante discussion of compliance in the Basic Data and Directions (BDD) and annual Budget Execution Report, including reconciliation and explanations of changes in macroeconomic and fiscal forecasts between successive plans. - Strengthen MTBF: make the expenditure ceiling for the second year of the rolling four-year MTBF binding initially (targeted July 2019); undertake an MTBF review (July 2018) to identify enhancement needs. - Implementation timetable (selected items): - Include all flows and stocks of LEPLs in fiscal aggregates — December 2017 — MAFD. - Case-by-case SOE classification analysis — December 2018 (update every 3 years) — MAFD. - Include liabilities of LEPLs, subnational governments, and reclassified entities in GFS gross debt — End 2018 — MAFD. - Record SOE transfers/subsidies above-the-line as expenditures — March 2018 — Treasury + MAFD. - Make second-year expenditure ceiling binding initially — July 2019 — BD; MTBF review — July 2018 — BD. - Enhance BDD ex-post/ex-ante compliance discussion — July 2018 — MAFD. ### SOEs, PPAs/PPPs, elevated fiscal risks, and recommended safeguards - Observed exposures and magnitudes: - Government net fiscal exposure to existing PPAs in the energy sector is about 5 percent of GDP. - Total liabilities of high-risk SOEs accounted for 9 percent of GDP by end-2015. - FTE estimated gross exposure to SOEs by end-2016 at 17 percent of GDP; gross exposure to PPP and PPAs combined about 20 percent of GDP. - Table selected figures (percent of 2016 GDP): Total project value 77.2; Total net exposure 19.8; Legacy NPV exposure 4.9. Example projects: Nenskra PPA project value 6.2, net 6.2; Gardabandi TPP project value 5.4, net 2.4; Anaklia Deep Water Port project value 4.6, net 0.2. - Future risks and scenario-based “safe debt”: - Accounting for a shock in the range of 5–10 percent of GDP distributed evenly over a 6-year horizon points to a “safe” debt level of 35–40 percent of GDP. - Sensitivities: estimates sensitive to long-term growth, exchange rate depreciation, and financing conditions; higher depreciation or higher average interest rates require lower deficits. - PPP/PPA exposures and management: - New PPP/PPA activity raises fiscal risks beyond historical distributions. - Authorities developing new PPP/PPA management framework; Fund-supported program includes end-December 2017 structural benchmark to include a ceiling on overall PPP/PPA government exposure in a new PPP Law. - Recommendations for monitoring subnational governments: - Subnational expenditures equal 4.5 percent of GDP; own-source revenue about 30 percent of funding; local borrowing requires MoF approval. - MoF should closely monitor municipalities’ fiscal positions; inclusion of municipalities within the TSA is a major improvement. - Enhance fiscal reporting on subnational governments per FTE recommendations. ### PPP/PPA ceilings — guidance and design considerations - Role of ceilings: - Ceilings on PPP/PPA exposures can contain fiscal risks and prevent circumvention of main fiscal rules where PPPs/PPAs are not fully captured by national accounts or debt ceilings. - Ceilings should be guided by MTBF and debt sustainability analysis and reflect capacity to formulate and implement high-quality projects. - Choice of measure: - Ceilings can cover stock and/or annual flow of PPPs/PPAs. - Stock ceilings suit debt sustainability concerns; flow ceilings suit capacity-to-repay and liquidity concerns. - Simplicity and measurability: - PPP/PPA ceilings should be simple, unambiguous, credible, and verifiable. - Simpler measures initially: capital investment under contracts or sum of known government commitments plus simple measure of contingent liabilities (e.g., face value). - As valuation methods improve, ceilings can broaden to cover other expected costs of contingent liabilities. ### Institutional strengthening: forecasting, MTBF, fiscal risk disclosure, and oversight - Forecasting performance and recommendations: - Average absolute real GDP forecast error for current budget year since 2004 around 4 percent; improved since 2010 with current year forecast error around 2 percent. - Real GDP medium-term forecasts tend to be too optimistic; inflation forecast errors biased high (inflation lower than forecast). - Fiscal forecast errors decreased but remain (e.g., expenditures exceeded 30 percent of GDP limit in each of the last three years). - Recommendations: reconcile and explain forecast changes between BDD vintages; analyze sources of forecast errors regularly; strengthen collaboration between forecasting and revenue departments. - MTBF enhancements: - Introduced in 2004 with four-year projections; revisions to medium-term expenditure plans averaged absolute changes around 7 percent for second year and around 10 percent for third year historically. - Strengthening steps: make outer-year ceilings progressively binding; clarify baseline vs new policy initiatives; introduce multi-year planning and contingency margins; strengthen top-down cabinet endorsement of aggregate ceilings. - Fiscal risk management and disclosure: - Disclosure of SOE fiscal risks improved; 2017 FRS provides detailed analysis and links between government and SOEs. - Development of risk-based model for PPAs in electricity sector with World Bank; results to be included in future FRS vintages. - Recommendation: include more detailed compliance discussion in BDD and annual Budget Execution Report (Recommendation 4.1 — July 2018). - Recommendation 4.2: improve forecasting credibility (reconciliation of forecast vintages — July 2018; analyze forecast errors — 2018; strengthen MAFD-Revenue Department collaboration). - Recommendation 4.3: further develop MTBF (make second-year expenditure ceiling binding — 2019; MTBF review — 2018). ### Corrective mechanisms, escape clauses, and independent oversight (Box 2.2 highlights) - Corrective mechanisms: - Current law requires corrective plan in next-year budget if approved budget not within ceilings; law silent on actions when actual outturns are non-compliant—major credibility weakness. - Recommended: corrective mechanism should apply to actual outturns; require reporting to Parliament explaining deviations (permanent vs temporary) and a specified plan to return to ceilings; include in BDD and Budget Execution Report. - Escape clauses: - Current ELA allows temporary waivers if: (i) current year’s approved budget was within limits; or (ii) extraordinary expenditures due to military action or economic recession are needed. - Shortcomings: first criterion not linked to trigger cause; "economic recession" undefined; procedures for triggering and reporting unspecified. - Suggested revisions: adopt clear recession definition (example: two consecutive quarters of negative real year-on-year GDP growth); specify procedures for decision and require government report to Parliament within a timeframe. - Oversight arrangements: - PBO publishes opinion on draft budget and annual fiscal policy review but lacks mandate to publish full assessments and to verify compliance with rules. - SAO provides opinion on draft budget but does not formally verify compliance. - Recommended minimum ELA oversight provisions: assessment of macroeconomic and fiscal forecasts (including comparison with other forecasters) and formal assessment of compliance with each fiscal rule; PBO may be best placed for independent verification. - Periodic review: - Incorporate periodic review clause in ELA (suggested every 5 or 10 years) for relevance and appropriateness; potential future adjustments may include structural balance rule when appropriate. ### Methodology notes and EU oversight requirements (annex highlights) - Stochastic simulation methodology (Annex VI): - Approach: generate macroeconomic and fiscal shocks from estimated joint distribution; simulate debt paths using debt accumulation equation and fiscal reaction function. - Data: relatively short balanced annual dataset 1994–2016 used for Georgia. - Fiscal reaction function estimated from panel of 74 emerging market economies; country-specific estimates for Georgia considered imprecise. - Debt accumulation equation includes stock-flow adjustment shocks and share of foreign-currency-denominated debt 훼푡−1 explicitly in formula provided. - Outputs: probabilistic debt-path analysis; compute share of debt paths that cross a given debt limit by a target date. - Caveats: model validity depends on past relationships remaining relevant; structural breaks and model fit are important shortcomings. - EU requirements for independent oversight bodies (Annex V): - Council Directive on budgetary frameworks (November 2011, Chapter IV, Article 6b) requires “independent bodies or bodies endowed with functional autonomy” to monitor compliance. - Two-pack provisions: require independent macroeconomic forecasts to be “produced or endorsed” by independent bodies; advise on activation/operation of correction mechanism. - Institutional fragmentation guidance: multiple independent bodies allowed only if clear allocation of responsibilities and no overlap. *Source: cr18132 - IMF staff report (CR18132).* ### PREFACE _________________________________________________________________________________________ 6 ### PREFACE ### Mission and participants - A mission from the IMF’s Fiscal Affairs Department (FAD) visited Tbilisi, Georgia, during September 13 – 25, 2017 to provide technical assistance on enhancing Georgia’s fiscal rules framework. - Mission composition: Torben Hansen (FAD, head), Isabel Rial (FAD), Joao Jalles (FAD), Stephen Farrington (FAD expert), and Sami Ylaoutinen (FAD expert). - Ministry of Finance (MoF) interlocutors included Mr. Nikoloz Gagua (Deputy Minister); Mr. Giorgi Kakauridze (Deputy Minister); Mr. Tsotne Kavlashvili (Deputy Minister); Mr. Lasha Khutsishvili (Deputy Minister); Ms. Ekaterine Mikabadze (Head of Macroeconomic Analysis and Fiscal Policy Planning Department (MAFD)); Ms. Ekaterine Guntsadze (Head of Budget Department); Mr. Mamuka Baratashvili (Head of Tax and Customs Policy Department); Mr. Ioseb Skhirtladze (Head of Public Debt and External Financing Department); Mr. Irakli Katcharava (Head of Domestic Public Debt Instruments Development Division); Mr. David Gamkrelidze (Head of Cash Forecasting and Management Department, Treasury Service); Mr. Zurab Tolordava (Head of Accounting and Methodology Department, Treasury Service); and Mr. Giorgi Pataridze (Head of Administrative Department, Revenue Service). - Other meetings: Dr. Irakli Kovzanadze (Chairman of the Finance and Budget Committee of the Parliament of Georgia); Mr. Archil Mestvirishvili (Board Member, Deputy Governor, National Bank of Georgia); Ms. Tatia Khetaguri (Head of the Parliamentary Budgetary Office); Ms. Marika Natsvlishvili (Director, State Budget Analysis and Strategic Planning Department, State Audit Office of Georgia (SAO)). - Acknowledgements: Ms. Ekaterine Guntsadze and Ms. Natia Gulua for ongoing support; Ms. Khatia Chanishvili and Ms. Natia Jakhia for interpretation and translation services; Mr. Francois Painchaud (IMF Resident Representative in Georgia) for support. ### Purpose of the mission - Deliver technical assistance on enhancing Georgia’s fiscal rules framework in response to earlier findings, including those from the Fiscal Transparency Evaluation (FTE) conducted by FAD in late 2016. ### Key context and findings summarized in the Executive Summary (excerpts) - The Economic Liberty Act (ELA), adopted in 2011 and came into force in 2014, defines numerical upper limits: - State debt: 60 percent of GDP. - Budget balance (consolidated budget deficit): 3 percent of GDP. - Expenditure rule: expenditures plus the increase in non-financial assets of the consolidated budget to GDP should not exceed 30 percent. - Compliance history: - The debt and budget balance rules (BBRs) have been adhered to since their introduction. - Expenditures have exceeded the legislative limit, albeit by a small margin. - FTE findings and implications: - Gaps in reporting of general government revenue and expenditures relative to GFSM2014. - Gaps in assessment and reporting on compliance with fiscal rules. - FTE recommended a review of the fiscal rules framework; this report summarizes that review. ### Design of the fiscal rules — principal findings and recommended changes - Design considerations for Georgia: - Objectives: primary objective of fiscal sustainability and need for flexibility to respond to economic shocks. - Context: transitional nature of the Georgian economy and intention to gradually move towards the European Union fiscal governance framework. - Rules should be simple and ideally not pro-cyclical. - Assessment of current rules: - The current debt rule (DR) and budget balance rule (BBR) are appropriate in the Georgian context and compatible with realistic debt and deficit trajectories. - More complex deficit rules (e.g., structural balance rule) would be difficult to implement at this stage. - The specification of the expenditure rule (ER) as a percentage of GDP has issues, including potential pro-cyclical properties. - Recommendation for ER: - Replace the current ER in the ELA with nominal expenditure ceilings set on a rolling basis as part of the MTBF, consistent with the debt and BBRs and medium-term fiscal objectives. - Strengthen three elements of the ELA: - Corrective mechanisms: - Current corrective mechanism applies to budgetary plans but not to actual outturns. - Introduce ex-post reporting requirements on compliance and require the government to put forward a corrective plan in case of non-compliance. - Escape clause: - Abolish the criteria that allows the Parliament to approve a budget that is not compliant with the fiscal rules. - Adopt a clear definition of economic recession. - Specify who can decide to activate the escape clause and the associated reporting requirements. - Independent oversight: - Incorporate clear provisions on independent oversight and verification of the government’s compliance with the fiscal rules. ### Coverage and measurement — principal findings and recommended actions - Gaps and inconsistencies with GFSM2014 impede effectiveness: - General government revenue and expenditures currently exclude own-source revenue and associated expenditures by Legal Entities of Public Law (LEPLs) and public non-market producers classified as SOEs. - Some transactions between government units and SOEs are not recorded and reported as expenditures in line with GFSM2014. - These gaps mean the fiscal rules do not fully capture all general government activities. - Recommended alignment with GFSM2014: - Include all flows and stocks of LEPLs in the recording and reporting of fiscal aggregates. - Undertake a case-by-case analysis of the SOE sector, prioritizing SOEs classified as high-risk, to identify non-market producers to be reclassified into the general government sector. - Include in gross debt statistics the liabilities of LEPLs, subnational governments, and any public non-market entity that may be reclassified within the general government sector. - Record transactions with SOEs that are in practice transfers or subsidies as above-the-line expenditures. - Quantified implication: - Based on 2015 data, aligning coverage and measurement with GFSM2014 would have resulted in the overall deficit being 0.6 percentage points of GDP higher compared to the official data. ### Implementing the fiscal rules — principal findings, simulations, and recommended enhancements - Fiscal targets and buffers: - Medium-term fiscal targets should provide for adequate buffers to absorb potential adverse shocks and fiscal risks without breaching the fiscal rules. - Stochastic simulations point to a “safe” debt level for Georgia in the range of 35-40 percent, which would ensure public debt remains below the debt limit with a high probability if negative shocks materialize. - This safe debt range is comparable to current debt levels and consistent with overall deficits of about 2.3‒2.7 percent of GDP over the medium-term, broadly in line with the current Fund-supported program. - Reporting and MTBF enhancements: - Enhance reporting on fiscal rules through a more in-depth ex-post and ex-ante discussion of compliance in the government’s fiscal strategy paper, including reconciliation and explanations of changes in macroeconomic and fiscal forecasts between successive plans. - Strengthen the MTBF by making the expenditure ceiling for the second year of the rolling four-year MTBF binding initially and undertake a review of the MTBF to identify any needs for enhancement. ### Implementation timetable and responsibilities (selected items from Table 0.1) - Design of fiscal rules: - Replace ER with nominal expenditure ceilings as part of MTBF — 2018 — MoF (proposal). - Introduce ex-post reporting on compliance and corrective-plan requirement — 2018 — MoF (proposal). - Abolish parliamentary approval of non-compliant budgets unless escape clause triggered; adopt clear recession definition; specify voting rights for triggering escape clause — 2018 — MoF (proposal). - Incorporate independent oversight provisions — 2018 — MoF (proposal). - Incorporate periodic review clause for fiscal rules — 2018 — MoF (proposal). - Coverage and measurement: - Include all flows and stocks of LEPLs in fiscal aggregates — December 2017 — MAFD. - Case-by-case SOE classification analysis — December 2018 — MAFD; update every 3 years — Ongoing — MAFD. - Include liabilities of LEPLs, subnational governments, and public non-market entities in GFS gross debt — End 2018 — MAFD. - Record SOE transactions that are transfers/subsidies above-the-line as expenditures — March 2018 — Treasury + MAFD. - Implementing the fiscal rules: - Include in BDD and annual Budget Execution Report a more in-depth ex-post and ex-ante compliance discussion — July 2018 — MAFD. - Provide reconciliation and explanation of forecast changes between BDD vintages; analyze causes of forecast errors regularly; strengthen forecasting collaboration — End 2018 / End 2017 timings assigned — MAFD and Revenue Department as specified. - Make second-year expenditure ceiling of the rolling 4-year MTBF binding initially — July 2019 — BD; undertake MTBF review — July 2018 — BD. ### Introduction — contextual points - Georgia meets the standard of good or advanced practices against 18 of the 36 principles in the IMF’s Fiscal Transparency Code, and basic standards on a further 10 principles. - Key strengths: publication of detailed fiscal information in accordance with international standards; presentation of medium-term macroeconomic and medium-term forecasts and spending plans in the budget; production of macroeconomic scenarios and specific fiscal risk analyses. - Fiscal planning reforms implemented over the past decade include: MTBF introduced in 2004; Budget Code enacted in 2009; program budgeting introduced in 2012; establishment of an independent Parliamentary Budget Office (PBO) in 2014. - Constitutional and legal constraints on revenues: - Article 94 of the Constitution: no state tax can be introduced (except excise taxes) or increased without first being approved by a referendum, except in cases prescribed by organic law. - The ELA (an organic law) clarifies scope and exceptions; while not a numerical rule, it constitutes an important fiscal constraint reducing flexibility in fiscal policy formulation. *https://www.imf.org/-/media/files/publications/cr/2018/cr18132.pdf* ### Box 1.1. Constraints on Taxes in the Constitution and the ELA ### Box 1.1. Constraints on Taxes in the Constitution and the ELA ### Overview - The Constitution and the ELA require a referendum for the introduction of new general state taxes (except excise taxes), or an increase of the top rate of an existing general state tax. - General state taxes comprise the income tax, profit tax, value-added tax (VAT), customs duty, and excise taxes, but not the property tax, which is an exclusive local government revenue source and appears unaffected by the referendum requirement. ### Exceptions to the referendum requirement - New taxes can be introduced, or top rates increased, if they substitute for an existing tax without increasing the overall tax burden. - Adjustments to tax rates below the top rates can be made without a referendum. - Tax progressivity, the tax regime, or the methodology by which a tax is applied may not be objects of a referendum. - The government may request a temporary increase (of up to three years) of general state taxes without a referendum, after which the tax rates must be reinstated to their original levels (this provision has not been used so far). ### Fiscal implications and constraints - The constitutional provision does not constitute a fiscal rule because it does not provide a firm limit on fiscal aggregates (revenue); revenue could be increased without a referendum through: - introducing new excise taxes; - applying the exceptions; or - improved revenue collection. - Nonetheless, the provision constitutes an important fiscal constraint because the relationship between tax rates and tax revenue reduces flexibility in fiscal policy formulation, potentially forcing a disproportionate share of fiscal consolidation onto the expenditure side. ### Shortcomings from a tax policy perspective - Restricts discretionary countercyclical tax measures (for example, a transfer duty to stabilize a growing property price bubble). The ELA’s three-year grace period could provide some countercyclical scope, but only for the existing four taxes; sometimes a new tax may be needed. - Exemption of new excise taxes from the referendum requirement may encourage extensive use of excises to address revenue shortfalls, which is only sustainable if neighboring countries’ excise rate levels on alcoholic beverages, tobacco products, and fossil fuels are comparable. Exceeding neighboring levels could: - lead to illegal cross-border shopping, smuggling, and counterfeit goods; - erode custom-cleared imports and domestic excise collections; and - increase syndicate crime with lasting effects even after excise differentials are removed. - Reliance on excises—and inability to increase income-based taxes—removes a key distributional tax instrument from policy options. Georgia’s Gini Coefficient by total incomes at 0.39 could, inter alia, be addressed by making income taxation more progressive (i.e., a higher marginal personal income tax rate). The efficiency costs of redistribution could also be reduced with tax schedules that entail higher taxes for upper-income groups than for middle-income earners, but the provision in the ELA forecloses that policy option. - The restriction on introducing new taxes, or raising rates on a more permanent basis, may encourage labeling taxes as user charges, levies, or administrative fees to meet revenue targets. This could be done by placing more administrative units on a self-financing path through administrative fees as a substitute for budget financing. Unless legislation contains tight definitions to differentiate between taxes and user charges, a proliferation of such instruments could ensue. *Source: cr18132 - Box 1.1. Constraints on Taxes in the Constitution and the ELA* ### Box 2.2. Considerations for Expenditure Rule Coverage ### Box 2.2. Considerations for Expenditure Rule Coverage ### Corrective mechanisms - Fiscal rules frameworks often include corrective mechanisms that prescribe actions if fiscal outturns do not meet rules. Some frameworks are detailed/automatic (size, timeframe, measures); others require a procedural corrective plan without specifying size or timeline. - Georgia’s ELA corrective mechanism: - If the approved budget for the current year is not within specified ceilings for the deficit and expenditures, the government must submit to Parliament a budget proposal for the next year that includes a plan for returning to the ceilings within two years. - The law is silent on actions when actual outturns are not in compliance with the ceilings, a major weakness for credibility. - The corrective mechanism does not include the current DR. - Recommended content and reporting: - The corrective mechanism should apply to actual outturns, and related reporting requirements should be specified in the ELA. - While automatic corrective mechanisms are not recommended in the Georgian context, the government should be required to report to Parliament on compliance with fiscal rules, including: - Explanation of causes and nature of deviations (for example, permanent or temporary); and - A plan with specified measures for return to the ceilings. - Reporting and plan could be incorporated in the Basic Data and Directions (BDD) document (submitted to Parliament in July) and included in the annual Budget Execution Report. ### Escape clauses - Well-designed escape clauses permit flexibility for exceptional/unforeseeable shocks but must be rare and clearly specified in legislation to avoid circumvention. - Typical features: limited triggering factors, specification of trigger procedures, and/or a pre-determined transition path back to rules. - Georgia’s ELA allows temporary waivers in two cases: - The Parliament can approve a budget not in compliance with deficit or expenditure rule only if: (i) the current year’s approved budget was within the limits for deficit or expenditures; or (ii) extraordinary expenditures caused by military action or economic recession are needed. In both cases, the corrective mechanism would apply. - Shortcomings of Georgia’s escape clauses: - The possibility to approve a non-compliant budget because the current year’s approved budget was within limits is not linked to any trigger cause and risks impairing credibility. - "Economic recession" is not defined and could lead to circumvention (noted that this has not occurred so far in Georgia). - Returning to ceilings within a two-year period may be inappropriate in all cases and could lead to pro-cyclicality. - Procedures for who can trigger the escape clause (voting mechanism) and associated reporting requirements are not specified. - Suggested revisions: - Adopt a clear definition of economic recession; one example: two consecutive quarters of negative real year-on-year GDP growth. - Note: Such a criterion would be tight but allow deviation in exceptional circumstances; Georgia experienced such episodes in 2001, 2008, and 2009. - Specify procedures for decision to trigger the escape clause (e.g., cabinet decision based on a Minister of Finance proposal, or Parliament decision based on a government proposal). - Require the government to submit a report to Parliament within a certain timeframe detailing justification for triggering the escape clause. ### Oversight arrangements - Independent oversight enhances accountability and credibility; many frameworks create independent fiscal councils with an explicit legislative mandate to perform a “watchdog” role (assessments of fiscal performance and plans; some evaluate macroeconomic forecasts). - Current Georgian arrangements: - The PBO (independent since 2014) publishes an opinion on the draft state budget and an annual fiscal policy review but does not publish a full assessment of each element of the fiscal framework nor has a clear mandate to report on compliance with the rules. - The SAO provides an opinion on the draft budget published in October, but does not formally verify compliance with fiscal rules. - The SAO’s 2017 report highlighted breaches of the 30 percent expenditure ceiling in outturn data and noted the absence of any government commentary on this in its budget reports. - Recommended ELA provisions for oversight and verification (minimum): - An assessment of macroeconomic and fiscal forecasts, including against those of other forecasters. - A formal assessment of compliance with each of the fiscal rules. - The PBO may be best placed to serve this independent verification role. ### Revisions of the fiscal rules - Fiscal rules should be subject to periodic review to ensure continued relevance and appropriateness. - Periodic reviews (and revisions if warranted) ensure validity and effectiveness for debt sustainability and economic stabilization. - Examples of potential future adjustments: introduction of a structural balance rule at a later stage; adaptation to progress in EU association. - Suggested periodicity in the ELA: reviews every 5 or 10 years. ### Recommendations - Recommendation 2.1. Design of the fiscal rules: - The government should seek to replace, through an amendment to the ELA, the current ER with a requirement to set multi-year expenditure ceilings on a rolling basis as part of the MTBF, consistent with the debt and BBRs and the medium-term fiscal objectives (2018). - Recommendation 2.2. Corrective mechanisms: - Strengthen corrective mechanisms in the ELA by: - Introducing reporting requirements on compliance with the fiscal rules ex-post; and - Introducing a requirement that the government put forward a corrective plan to be included in the BDD in case of non-compliance (2018). - Recommendation 2.3. Escape clause: - Strengthen the escape clause in the ELA by: - Abolishing the criterion that allows Parliament to approve a budget not compliant with fiscal rules if the current year’s approved budget is within the limits; - Adopting a clear definition of economic recession, for example two consecutive quarters of negative real year-on-year GDP growth; and - Specifying who can decide to activate the escape clause, and the associated reporting requirements (2018). - Recommendation 2.4. Oversight arrangements: - Incorporate in the ELA clear provisions on independent oversight and verification of the government’s compliance with the fiscal rules (2018). - Recommendation 2.5. Revisions of the fiscal rules: - Incorporate in the ELA a clause that the fiscal rules be reviewed, and if warranted revised, on a periodic basis, for example every 5 or 10 years (2018). ### Coverage and measurement - Progress and remaining gaps: - Georgia adopted a cash based budget classification based on GFSM2001 in 2008 and publishes more frequent budget execution reports in accordance with GFSM2001. - Accounting reforms are being implemented and are expected to be completed in 2020 to provide a basis for GFSM2014 compilation. - Gaps remain in coverage and measurement: fiscal reports in principle cover central and subnational governments but exclude LEPLs and public non-market producers currently classified as SOEs. - Debt reporting focuses on central government only. - Some transactions between government units and SOEs are not accounted for or reported in line with international statistical standards. - The current Fund-supported program uses an “augmented budget deficit” which adds to net lending/borrowing all budget lending transactions. A. Institutional and transactional coverage — Limited coverage of Legal Entities of Public Law (LEPLs) - LEPLs profile: - Central government controls 496 LEPLs; an unspecified additional number are controlled by subnational governments. - LEPLs receive budget transfers and collect own revenue. - Coverage issues: - LEPLs controlled by subnational governments are not covered in GFS reports. - For LEPLs controlled by central government, coverage is partial: activities funded by transfers are included in GFS reporting to the extent funded by central government transfers; expenditures funded from LEPLs’ own revenues are not incorporated into main fiscal aggregates and thus not covered by fiscal rules. - Magnitude and staff estimates (based on data for 2015 compiled during the mission): - By end-2015, LEPLs own revenues represented about 12 percent of general government total revenue, and about 17 percent of total expenditures. - Staff estimates that inclusion of LEPLs in general government statistics would: - Increase revenue by 4.5 percent of GDP; - Increase expenditures by 4.3 percent of GDP; and - Decrease the general government deficit by 0.2 percent of GDP. - Authorities estimate LEPLs loans at end-2015 account for 2 million Lari. - Table 3.1. Georgia: Transactions of LEPLs Controlled by Central Government, 2015–16 (In percent of GDP) - 2015 / 2016 - Revenues: 9.5 - Of which: Own-revenues: 4.5 2.8 - Expenditures: 9.3 - Of which: Funded by own-revenues: 4.3 2.7 - Overall balance: 0.2 - Stocks: - Financial assets: n.a n.a - Liabilities: 0.0 0.0 - Sources: FTE staff estimates - Recommendations on LEPLs: - Consolidate LEPLs within general government in GFS reports. - Detailed data on transactions of LEPLs controlled by central government are available in a timely manner to allow consolidation. - For LEPLs controlled by subnational governments, include information in general government compilation on a best-effort basis, with revisions once higher-quality data become available. - The Treasury is working to improve quality of subnational LEPLs data in the context of the accounting reform strategy. B. Government non-market producer units classified as SOEs - Classification risk: - Classifying government entities as SOEs outside general government can significantly affect fiscal aggregates, especially for countries with fiscal rules. - Incorrect SOE classification risks underestimating the general government sector covered by fiscal rules, weakening fiscal discipline and debt sustainability. - Magnitude depends on size and nature of SOE activities; international experience shows reclassification can have significant impacts on deficits (example: 2013 EU cases; Portugal reclassified major transport SOEs into general government after financial difficulties). *Italic: Source — cr18132 - Box 2.2. Considerations for Expenditure Rule Coverage* ### 39. SOEs that do not satisfy the criteria to be a public corporation should be ### SOEs that do not satisfy the criteria to be a public corporation should be ### Classification of SOEs and delineation with general government - GFSM2014 delineation: classification depends on identifying market or non-market producers on a case-by-case basis. Only market-producers (provide most output at economically significant prices) should be classified as SOEs. - Market/non-market nature should be assessed through both qualitative and quantitative tests (see Box 3.1). - The quantitative “market test” often uses a 50 percent threshold comparing value of revenue sales and average production costs over at least 3 years. ### Box 3.1 — Delineation between General Government and SOEs (criteria) - General government: institutional units that are non-market producers; output intended for individual and collective consumption; financed by compulsory payments by other sectors. - SOEs: market producer units, classified outside general government. - Qualitative criterion (to be considered an SOE): - It should be an institutional unit (not an ancillary unit serving the government almost exclusively). - It should sell its output to both government and other customers; to be an SOE, the unit should sell most of its output to non-government units. If government buys more than 50 percent of output but through open competition (e.g., one tender procedures), the unit is considered an SOE; otherwise consolidate with general government. - Quantitative criterion (“market test”): - Value of revenue sales > 50 percent of average production costs over at least 3 years. - Sales: include revenue sales before taxes and excluding payments; include many non-sales revenues: property income (interest, dividends, rents); administrative fees; grants (transfers from government). - Production costs: include compensation to employees, use of goods and services, consumption of fixed assets, and other taxes on production; own capital formation excluded. - Recommendation: keep classification for at least three years and only reclassify if (i) criteria holds for more than three years; or (ii) clear expectations it will hold for several years. Quantitative criterion should be used with qualitative criteria. - Sources cited: GFSM2014, ESA2010. EUROSTAT uses a 50 percent threshold consistent with GFSM2014/ESA2010. ### Findings for Georgia (market test and sectorization) - FRS and FTE analysis suggest some SOEs in Georgia should be reclassified as general government units because they operate on a non-commercial basis and/or fulfill social functions (e.g., supplying electricity or gas free or at subsidized prices). - Staff estimates using FRS information indicate that 15 percent of the 65 largest SOEs under central government control would not pass the market test in 2015. - The FTE found limited information to apply full criteria to all SOEs controlled by central and subnational governments at mission time. - Subnational governments’ total liabilities accounted for 0.1 percent of GDP in 2015 (after consolidation). - Authorities’ estimate of LEPLs liabilities: approximately 2 million Lari at end-2015, close to zero in percent of GDP. ### Measurement and accounting of government transactions with SOEs - GFSM2014 treatment: - Transfers/subsidies to SOEs: recorded as expenditures (above-the-line), reduce overall deficit. - Loans and capital injections that generate financial assets for the government: recorded as financial transactions (below-the-line), not affecting overall deficit. - Government support forms (Box 3.2): - Providing a transfer/grant/subsidy: unrequited payment; capital transfers recorded as expenditures. - Providing a capital injection: addition to equity; to be classified as capital injection government should (i) expect to receive something of equal value in exchange (usually shares/debt instruments); (ii) expect a sufficient rate of return (dividends/interest); (iii) provide funds to a profitable SOE that has not shown a series of losses. - Providing loans: expected repayment per schedule; recorded below-the-line. - Georgia’s practices: - Two main forms referred to as budget lending (Table 3.3): - On-lending to SOEs: transfer of external funds borrowed by central government and on-lent to SOEs at similar financial conditions; managed by Public Debt and External Financing Department at MoF; memoranda of understanding specify repayment conditions. - Capital injections: transfers of budget resources to SOEs to (i) finance specific projects after negotiation with central government, or (ii) support loss-making SOEs in financial distress. - Table 3.3. Georgia: Composition of Budget Lending Operations, 2013–16 (Percent of GDP) - 2013: Total Budget Lending 1.0 100%; On-lending (external) 0.7 70%; Capital injections 0.3 30% - 2014: Total Budget Lending 0.9 100%; On-lending (external) 0.7 73%; Capital injections 0.3 27% - 2015: Total Budget Lending 1.4 100%; On-lending (external) 0.6 46%; Capital injections 0.7 54% - 2016: Total Budget Lending 1.4 100%; On-lending (external) 0.7 49%; Capital injections 0.7 51% - On-lending assessment: - On-lending appears to comply with loan definitions in GFSM2014; contractual with repayment conditions. - Table 3.4. Georgia: On-Lending Debt Service Payments, Planned vs. Actual, 2013–16* (Million Lari and percent) - Principal* (Plan, Actual, %): 2013: 26.8, 27.4, 102; 2014: 67.6, 55.4, 82; 2015: 78.7, 68.7, 87; 2016: 88.3, 85.7, 97; Av. % 92 - Interest* and commissions (Plan, Actual, %): 2013: 8.8, 9.6, 109; 2014: 23.2, 17.8, 77; 2015: 26.1, 25.4, 97; 2016: 28.7, 26.2, 91; Av. % 94 - * Including arrears. - For the average of the last three years, actual service of on-lending to SOEs represent 92 and 94 percent of planned repayments for principal and interest, suggesting reasonable probability of repayment. - The FTE found around a fifth of the loans have been restructured or reorganized, by extending grace periods around interest and principal repayments. - Capital injections classification issues: - Authorities currently account two transaction types as capital injections: (i) transfers earmarked to specific capital projects; (ii) transfers to SOEs under significant financial stress (loss-making for several years). - These transactions do not pass the “capital injection test” because the government is not expecting to receive a financial asset of equal value and does not expect a sufficient rate of return; shares of loss-making companies would be significantly lower than the transfer value. - Preliminary analysis suggests capital injections are capital transfers and should be recorded as expenditures above-the-line. ### Fiscal implications and observed magnitudes - In 2015: - About half of the 65 largest SOEs under government control were loss-making. - One-third were classified as experiencing some form of financial difficulty (high-risk). - Central government provided support to high-risk SOEs of about 0.8 percent of GDP, mostly as capital injections, despite around one third of them also having received capital injections in each of the preceding years. - Combined liabilities of the high-risk SOEs accounted for 9 percent of GDP. - Implication: Aligning coverage and measurement with GFSM2014 (including reclassifying some SOEs and recording capital injections as expenditures) would increase the overall deficit relative to official data for 2015. ### Coverage of public debt - Current debt ceiling institutional coverage is narrower than other fiscal rules: debt rule (DR) limited to central government, while budget balance and expenditure rules cover central and subnational governments. - Gross debt in Georgia is compiled on a cash basis; it does not include accounts payable and is compiled in nominal values. - Authorities should extend institutional coverage of gross debt to general government in line with GFSM2014 and the Guide for Public Sector Debt Statistics (2011), by adding: - (i) debt liabilities of subnational governments; - (ii) debt liabilities of LEPLs; - (iii) debt liabilities of any SOEs reclassified into the general government. ### Recommendations (explicit) - Recommendation 3.1. Legal Entities of Public Law (LEPLs): The MoF should include all flows and stocks of LEPLs in the recording and reporting of fiscal aggregates in line with the GFSM2014 standards (December 2017). - Recommendation 3.2. Classification of SOEs: The MoF should undertake a case-by-case analysis of the SOE sector, with priority given to those SOEs classified as high-risk, to identify SOEs that are non-market producers and should be reclassified into the general government sector according to GFSM2014 (December 2018). The assessment should be updated every three years to ensure that the sectorization of each unit remains valid (ongoing). - Recommendation 3.3. Coverage of public debt: The MoF should include in gross debt statistics the liabilities of LEPLs, subnational governments, and any public non-market entity that may be reclassified within the general government sector, in line with GFSM2014 and the Public Sector Debt Guide (2018). - Recommendation 3.4. Budget lending: The MoF should record budget lending to SOEs, including capital injections, in accordance with GFSM2014. Particularly, transfers or subsidies should be recorded as expenditures above-the-line (December 2017). ### Fiscal rules and “safe debt” analysis - Staff stochastic simulations based on historical distribution of shocks point to a “safe debt” level of about 45 percent of GDP for Georgia, close to current levels. - Under baseline assumptions (consistent with Fund-supported program): Georgia growing at 8 percent in nominal terms, foreign financing at 4 percent nominal, inflation at 3 percent, and no significant changes in the nominal exchange rate — simulations suggest public debt remains below the debt ceiling with high probability and is consistent with an overall deficit target close to 3.0 percent of GDP. - Risk: In a severe economic shock, public debt could exceed the 60 percent of GDP ceiling by 2019 with a 10 percent probability. - Caveat: Baseline assumptions may be optimistic — percentage of debt denominated in foreign currency and favorable average interest rates of foreign loans are assumed constant with no exchange rate volatility; Georgia remains highly vulnerable to exchange rate risks, which could affect debt dynamics as market borrowing increases and concessional borrowing declines. *IMF staff report (CR18132).* ### 55. Future risks could also be higher than those faced in the past. The probability ### Future risks could also be higher than those faced in the past. ### Fiscal risks from SOEs, PPAs, and PPPs - Government is engaging in new activities that may be associated with higher future fiscal risks, beyond what past-event-based probability distributions capture. - Key sources of elevated risk: - Exposure to power-purchase agreements (PPAs) of SOEs in the energy sector and other quasi-fiscal activities of SOEs. - Promotion and expansion of public-private partnerships (PPPs) across several sectors. - Potential reclassification of SOEs within the general government sector due to non-market nature, which could increase gross debt levels permanently. - FRS and FTE quantified exposures: - Government net fiscal exposure to existing PPAs in the energy sector is about 5 percent of GDP. - Total liabilities of high-risks SOEs account for 9 percent of GDP by end-2015. - FTE estimated Georgia’s gross exposure to SOEs by end-2016 at 17 percent of GDP; gross exposure to PPP and PPAs combined about 20 percent of GDP. - Table figures (selected): - Total project value in table: 77.2 (percent of 2016 GDP) - Total net exposure: 19.8 (percent of 2016 GDP) - Legacy NPV exposure: 4.9 (percent of 2016 GDP) - Individual project examples: Nenskra PPA project value 6.2, net 6.2; Gardabandi TPP project value 5.4, net 2.4; Anaklia Deep Water Port project value 4.6, net 0.2. ### Safe debt level, scenarios, and consolidation implications - A more prudent scenario suggests a “safe” debt level in the range of 35–40 percent of GDP given enhanced future risks. - Accounting for a shock in the range of 5–10 percent of GDP distributed evenly over a 6-year horizon points to a “safe” debt level of 35–40 percent of GDP. - Associated deficit and fiscal adjustments: - A safe debt level range of 35–40 percent of GDP would be consistent with overall deficits of about 2.3–2.7 percent of GDP. - These deficits are broadly in line with the Fund-supported program. - Simulations suggest convergence to a safe debt level of 35–40 percent of GDP would require a consolidation of about 0.3–1.0 percent of GDP over the medium-term relative to a current overall deficit of 3.0 percent of GDP. - Under assumption of a fixed level of revenue in percent of GDP (total revenue to GDP ratio of 28.4 for 2016), a consistent level of expenditures is estimated at around 31 percent of GDP. - Sensitivities: - Estimates are sensitive to assumptions on long-term growth, exchange rate depreciation, and financing conditions. - A higher depreciation rate or a higher average interest rate would require lower deficits to remain in the “safe” range. - A higher long-term growth rate would provide space for higher deficits while still guiding debt to a “safe” level. ### Reporting on fiscal rules and transparency recommendations - Current practice: - The BDD is presented to Parliament in early July each year and includes four-year forecasts for main macroeconomic variables and fiscal forecasts; updated BDD iterations are presented end-October and end-November alongside draft budget proposals. - Fiscal aggregates defined in the fiscal rules are included in a summary table in an annex to the budget, but the BDD does not discuss whether budget projections are in line with the fiscal rules. - Recommended enhancements to BDD and Budget Execution Report to increase transparency and credibility: - An evaluation of economic and fiscal outturns for the previous year against the government’s published forecasts and plans, including performance against the fiscal rules. - A more detailed presentation of key forecasting judgements, including a comparison with other forecasting institutions. - A discussion of the consistency of the projections with the fiscal rules and the margins for error in meeting the rules. ### Compliance risks from subnational governments - Scale and constraints: - Subnational government expenditures are equivalent to only 4.5 percent of GDP. - Subnational governments are highly reliant on central government transfers; own source revenue makes up only 30 percent of their funding. - Limits exist on local government borrowing from non-public entities; borrowing requires MoF approval. - Observations: - Local government borrowing is relatively low; central government on-lending from international financial institutions forms the largest portion. - Some individual municipalities run large deficits and liability ratios relative to their own source revenue, but these are negligible from an economy-wide perspective. - Recommendations: - MoF should closely monitor municipalities’ fiscal positions. - Recent inclusion of municipalities within the TSA is a major improvement for monitoring. - Fiscal reporting on subnational governments should be enhanced in line with the FTE recommendations. ### Enhancing budget institutions: forecasting, MTBF, and fiscal risk management - Macroeconomic and fiscal forecasting - Historical forecast accuracy: - Since 2004, the average absolute real GDP forecast error for the current budget year has been around 4 percent. - Accuracy has improved since 2010, with current year forecast error falling to around 2 percent. - Forecast biases and errors: - Real GDP forecasts for the medium term have tended to be too optimistic. - Inflation has been considerably lower than forecast. - Fiscal forecast errors for the current budget year have decreased but remain, evidenced by expenditures exceeding the 30 percent of GDP limit in each of the last three years. - Medium-term expenditure forecast errors are larger when policy lending is considered, reflecting upward revisions to policy lending plans during the budget year and between medium-term plans. - Recommended actions: - Budget documentation should explain changes between successive forecasts and medium-term plans (reconciliation of forecast vintages). - Systematic review and explanation of sources of forecast errors, focusing on persistent optimism or pessimism. - Greater cooperation between forecasting and revenue departments to improve understanding of revenue elasticities and forecast accuracy. - Medium-term budget framework (MTBF) - Status: - Georgia introduced an MTBF in 2004 based on four-year fiscal projections and expenditure plans; plans are updated and presented with draft budget laws in October and November. - Revisions to medium-term expenditure plans averaged absolute changes around 7 percent for the second year and around 10 percent for the third year over the past decade, with substantially smaller revisions since 2011. - Strengthening options: - Gradually develop MTBF into a more binding framework; consider making the ceiling for the first outer year binding and progressively binding additional outer years as the framework matures. - Clarify distinction between baseline estimates and new policy initiatives. - Introduce explicit multi-year planning and contingency margins. - Strengthen top-down nature of MTBF through early cabinet endorsement of aggregate expenditure ceilings and their application in the budget process. - Enhance reporting on the MTBF and strengthen strategic multi-year ministry budget planning. - Fiscal risk management and disclosure - Importance: - Rules-based fiscal frameworks require strong supporting institutions for macroeconomic and fiscal forecasting, medium-term planning, and fiscal risk management and disclosure. - Current progress: - Disclosure of fiscal risks from SOEs has improved substantially; the FRS accompanying the 2017 budget provides detailed analysis including links between government and SOEs, consolidated financial indicators, and factors affecting SOE financial positions. - FRS includes description of existing PPAs, preliminary estimates of government commitments related to PPAs, and main sources of risks. - Authorities are working with World Bank specialists to develop a risk-based model for PPAs in the electricity sector, with results expected to be included in future FRS vintages. - Policy action underway: - A new PPP/PPA management framework is being developed. - The Fund-supported program includes an end-December 2017 structural benchmark to include a ceiling on overall PPP/PPA government exposure in a new PPP Law. - Ceilings can be expressed as nominal caps, percentages of government capacity to repay, or percentages of GDP (country examples discussed). *Source: cr18132 - 55. Future risks could also be higher than those faced in the past.* ### 77. While ceilings on the size of PPPs/PPAs can be a useful tool for managing fiscal ### While ceilings on the size of PPPs/PPAs can be a useful tool for managing fiscal risks ### Role and guidance on PPP/PPA ceilings - Ceilings on the size of PPPs/PPAs can help contain fiscal risks and limit overall government commitments for PPPs to levels that are fiscally affordable. - There are no simple rules of thumb for the level at which PPP ceilings should be set. - Ceilings are not a substitute for medium-term planning and a strong public investment framework. - The assessment of the maximum size of a PPP/PPA program should be guided by the MTBF and the debt sustainability analysis and should also capture the capacity to formulate and implement high-quality projects. - The effectiveness of a PPP ceiling in supporting short-term budget affordability and long-term sustainability depends on the type of projects in the PPP portfolio: - Portfolios with a high share of government-funded PPPs/PPAs (for example, linked to some type of availability payments by government) will have larger short-term implications for budget affordability. - Portfolios mostly comprising user-funded PPPs/PPAs (for example, concessions) have smaller short-term budget implications. - Box 4.3 discusses key considerations in the design of PPP/PPA ceilings. ### Box 4.3 — Considerations for the Design of PPP/PPA Ceilings - In countries with a debt ceiling, a separate ceiling on PPPs/PPAs overall government exposure can support the effective implementation of the main fiscal rule. - Fiscal rules in most countries only apply to traditional government debt in the form of debt securities and loans. - National accounting and/or reporting systems may not be advanced enough to incorporate complex long-term contracts as PPPs/PPAs, resulting in practical exclusion from fiscal rules. - A separate PPPs/PPAs ceiling can help prevent the circumvention of the main fiscal rules through PPPs. - Ceilings can cover the stock and/or the annual flow of PPPs/PPAs. - Ceilings on the overall size of the PPP program (stocks) and the annual PPP-related payments (flows) can increase the predictability of the government’s exposure to PPPs and allow for ready implementation of affordability tests. - Choice of measure depends on the fiscal concern: - If the main fiscal concern is debt sustainability—either because debt is on an unsustainable trend or the current level is dangerously approaching a debt ceiling—a ceiling expressed in terms of stock of PPPs/PPAs tends to be more effective. - If the main fiscal concern is the government’s capacity to repay—either due to cash liquidity issues, or due to a large number of government-funded PPPs (for example, roads, prisons, or hospitals that require government payments for a period of 15–25 years)—then ceilings expressed in flows tend to be more effective in safeguarding long-term fiscal affordability. - A PPP/PPA ceiling should be simple and measurable. - The ceiling measure should be unambiguous, credible, and verifiable by independent experts. - Caution is advised in employing complicated measures based on, for example, the option value of the PPP portfolio. - If more complicated measures are used, clarity should be provided on parameters (e.g., the discount rate and probabilities of contingencies) and methods used. - Simpler methods may be preferable initially: - PPP/PPA ceiling could be based on the capital investment under these contracts, or the sum of known government commitments (e.g., availability payments) and a simple measure of contingent liabilities (e.g., face value). - As more sophisticated and reliable valuation methods are developed, ceilings can be broadened to cover other expected costs of contingent liabilities. ### Key recommendations (Section E) - Recommendation 4.1. Reporting on the fiscal rules: - The MoF should include a more in-depth discussion of compliance with the fiscal rules ex-post and ex-ante in the BDD and the annual Budget Execution Report (July 2018). - Recommendation 4.2. Macroeconomic and fiscal forecasts: The MoF should enhance the quality and credibility of macroeconomic and fiscal forecasts by: - Providing a reconciliation and explanation of forecast changes in successive vintages of the BDD (July 2018); - Analyzing the sources and causes of forecast errors on a regular basis to help improve judgements and forecast techniques (2018); and - Strengthening collaboration between the MAFD and the Revenue Department on forecasting of revenue. - Recommendation 4.3. Medium-Term Budget Framework: The MoF should further develop the MTBF by: - Gradually extending the binding nature of the MTBF, initially by making the expenditure ceiling for the second year of the rolling 4-year MTBF binding (2019); and - Undertaking a review of the MTBF to identify any needs for enhancement in some areas to become even more effective (2018). ### Annex highlights — Fiscal governance, correction mechanisms, escape clauses, and fiscal councils - EU fiscal governance: - The Stability and Growth Pact (SGP) requires general government deficits do not exceed 3 percent of GDP and public debt do not exceed 60 percent of GDP. - The Fiscal Compact (in force January 2013) requires signatories to give effect in national legislation to a structural balance budget rule, an automatic correction mechanism, and an escape clause, all by January 2014. - The structural BBR must limit annual structural deficits to a maximum of 0.5 percent of nominal GDP. - Debt reduction rule: countries with debt above the 60 percent of GDP limit are required to continuously reduce their debt levels by at least 1/20th of the distance between the current level and 60 percent of GDP until the latter is reached. - Expenditure benchmark: countries which have reached their MTO must keep annual growth of primary expenditure (excluding unemployment benefits) at or below long-term nominal GDP growth. - Reforms (Six Pack, Two Pack) strengthened reporting, surveillance, and enforcement; financial penalties can be imposed (up to 0.1 percent of GDP for non-implementation; fines for EDP non-compliance starting at 0.2 percent of GDP and can reach 0.5 percent of GDP). - Correction mechanisms — country examples: - Switzerland and Germany use “debt brakes” with notional accounts storing deviations from the structural BBR; thresholds and treatment differ: - Germany thresholds: 1.0 percent of GDP per ordinary law and 1.5 percent per constitution. - Switzerland threshold: 6 percent of expenditures. - Switzerland requires elimination of the excess amount within the next three annual budgets. - Slovak Republic triggers at debt-to-GDP ratios: 50 percent (minister explains and suggests measures), 53 percent (cabinet must pass package to trim debt and freeze wages), 55 percent (expenditures cut automatically by 3 percent and next year’s budgetary expenditures frozen except for co-financing of EU funds), 57 percent (cabinet must submit a balanced budget). - United States sequestration: automatic spending cuts if Congress enacts appropriations exceeding pre-set caps; sequesters can bias against capital spending. - Escape clauses — country examples and conditions: - Brazil (since 2000): Real GDP growth below 1 percent over four quarters, and natural disaster; invocation requires Congressional approval. - Colombia (since 2011): Extraordinary events threatening macroeconomic stability; suspension subject to favorable opinion of CONFIS. - Germany (since 2010): Natural disasters or unusual emergency situations outside government control with major financial impact; absolute majority of parliament needed; parliament must approve an amortization plan. - Jamaica (since 2010): Targets may be exceeded on grounds of national security, national emergency, or other exceptional grounds specified by the Minister subject to affirmative resolution. - Mauritius (since 2008): Temporary deviations in case of emergencies and large public investment projects. - Mexico (since 2006): If non-oil revenues are below their potential due to a negative output gap, a deficit equivalent to the shortfall is allowed. - Panama (since 2008): If real GDP grows by less than 1 percent, the non-financial public sector deficit ceiling can be relaxed to 3 percent of GDP in the first year, followed by a gradual transition to the original ceiling (1 percent of GDP) within 3 years. - Peru (since 2000): If real GDP declines or other emergencies declared by Congress, the deficit ceiling can be relaxed up to 2.5 percent of GDP; Executive must specify deficit and expenditure ceilings during the exception period; a minimum adjustment of 0.5 percent of GDP is required until the 1 percent deficit ceiling is reached. - Romania (since 2010): In case of government change, new government announces consistency of program with MTBF; if not consistent, MoF prepares revised MTBF for parliamentary approval and Fiscal Council review. - Slovakia (since 2012): Escape clauses for major recession, banking system bailout, natural disaster, and international guarantee schemes. - Spain (since 2002): Natural disasters or exceptional slowdown; exceptional budget deficits accompanied by a medium-term financial plan to correct within next 3 years (approved by majority vote of parliament). - Switzerland (since 2003): Government can approve by supermajority a budget deviating from the BBR in "exceptional circumstances" (natural disaster, severe recession, changes in accounting methods). - EU member states/euro area (since 2005): EDP may not be opened when the 3 percent deficit limit is exceeded only temporarily and exceptionally and the deficit is close to the limit; deadlines for excessive deficit correction can be extended in case of adverse economic developments. - WAEMU (since 2000): Temporary and pronounced shortfall of real GDP (at least 3 percentage points below the average of the previous three years) and budget revenue (at least 10 percentage points below the average of the previous three years average). - Fiscal councils — roles and observations: - Fiscal councils are independent public institutions promoting sustainable public finances through public assessments of fiscal plans and performance, and provision or evaluation of macroeconomic and budgetary forecasts. - Fiscal councils foster transparency and accountability and trigger reputational effects; mandate magnifies reputational impact and encourages government to “comply or explain.” - Around one quarter of emerging market economies and close to half of advanced countries with fiscal rules invite an independent body to verify compliance with fiscal rules. - Some countries institutionalize independent macroeconomic assumptions for the budget (e.g., Canada and the UK). - The remit of fiscal councils in the EU differs across countries, but there are common responsibilities (production/endorsement of macroeconomic projections, compliance monitoring, assessments, “comply or explain” mechanisms). *Source: IMF staff and excerpts from cr18132* ### Annex V. EU Requirements for Independent Oversight Bodies ### Annex V. EU Requirements for Independent Oversight Bodies ### EU legislative framework and obligations - Several European legislative acts require that fiscal rules be monitored by “independent bodies,” including fiscal councils. - Council Directive on budgetary frameworks (November 2011, Chapter IV, Article 6b) requires “independent bodies or bodies endowed with functional autonomy vis-à-vis the fiscal authorities of the Member States” to carry out “the effective and timely monitoring of compliance with the rules.” - The Treaty on Stability, Coordination, and Governance (TSCG) reaffirmed this obligation (Title 3, Article 3.2) without further specification. - The Directive entered into force in November 2011 as part of the “the six-pack” and applies to all EU27 member states. ### Two-pack provisions and additional details - The regulations forming the two-pack specify the council’s role in case of “significant deviation from the medium-term objective or the adjustment path towards it.” - The council should advise on the activation and operation of the correction mechanism (Chapter II, Article 4). - The council should provide an assessment about the circumstances allowing temporary deviations from targets (Chapter II, Article 4). - Two-pack requirement on forecasts: - Requires that "independent macroeconomic forecasts" be “produced or endorsed” by “independent bodies” in the context of budget preparation (Chapter I, Article 2.1.2). - Fiscal councils (FCs) can address biases in ministry forecasts by either providing unbiased forecasts themselves or exposing bias to the public. - Institutional fragmentation guidance: - Two-pack preamble (Article 7) allows more than one independent body only if there is a clear allocation of responsibility and no overlap of competencies over specific aspects of monitoring. - “Excessive institutional fragmentation of monitoring tasks should be avoided” (Article 7 of the preamble). ### Annex VI. Underlying Methodology Using Stochastic Simulations ### Purpose and high-level description - Methodology used to estimate the safe debt level via stochastic simulations; full description referenced in IMF (FAD How to Notes, 2017 – “How to Calibrate Fiscal Rules—A Primer”). - Approach: generate macroeconomic and fiscal shocks, simulate corresponding public debt paths using a debt accumulation equation and a fiscal reaction function. ### Macroeconomic shock generation - Joint dynamics of macroeconomic (non-fiscal) variables estimated from either: - a quarterly, unrestricted vector autoregressive model (VAR), or - a multivariate normal distribution at annual frequency. - Dataset: relatively short balanced dataset—limited annually to 1994–2016. - Estimated joint distribution of shocks on: - real domestic and foreign interest rates (푟푡 and 푟푡∗), - real GDP growth (푔푡), - exchange rate (푒푡). - Correlations between shocks are calibrated; random shock sequences are drawn from the estimated distribution. ### Fiscal reaction function - Estimated for a panel of 74 emerging market economies in a panel estimation with country and time effects. - Purpose: capture government response to public debt — a semi-elasticity of the primary balance to the debt ratio. - Estimated equation: - 푝푏푖,푡 = 푐푖 + ℎ푖 + 훽푝푏푖,푡−1 + 휁푑푖,푡−1 + 휀푖,푡 - where 푝푏푖,푡 denotes the ratio of the primary fiscal balance to GDP; 푑푖,푡−1 the gross public debt-to-GDP ratio at the end of the previous year; and 푐푖, ℎ푖 are the country and time fixed effects. - Fiscal policy shocks included: - 휀푖,푡 ~ 풩(0, 휎푖), with country-specific variance. - Note: Due to limited time series for Georgia, country-specific estimates are not precise and hence are unreliable for the authors’ purposes; a more general emerging markets sample was used instead. ### Debt accumulation and simulation - Debt trajectories combine macro-fiscal shocks, fiscal policy response, and the debt accumulation equation; stock-flow adjustment shocks included to account for contingent/implicit liabilities. - Debt accumulation equation used: - 푑푡 = −푝푏푡 + 푠푓푎푡 + ( (1+푟푡)(1−훼푡−1) + 훼푡−1(1+푟푡∗)푒푡/푒푡−1 )/(1+푔푡) 푑푡−1 - where 훼푡−1 is the share of foreign-currency-denominated debt. - Algorithm: generates a large number of random shock sequences over the 6-year forecasting period and computes corresponding debt paths for each sequence. - Outputs: probabilistic analysis of debt trajectories; compute the share of debt paths that cross a given debt limit at a certain date. ### Validity and key shortcomings - Validity conditioned on the quality of the statistical model used to produce forecasts. - Important shortcomings: - Relationships estimated using past data may not be relevant for the future if structural breaks are present. - Importance of satisfactory goodness-of-fit of the model. *Source: cr18132 - Annex V. EU Requirements for Independent Oversight Bodies* --- _Source: https://www.imf.org/-/media/files/publications/cr/2018/cr18132.pdf_