## wpiea2019218-print-pdf ## Source details **Canonical URL:** [wpiea2019218-print-pdf](https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019218-print-pdf.pdf) ## Other formats - [Markdown version](/-/media/files/publications/wp/2019/wpiea2019218-print-pdf.pdf.md) - [Structured JSON version](/-/media/files/publications/wp/2019/wpiea2019218-print-pdf.pdf.json) --- ### Dataset: coverage and sources - Analysis focuses on 173 countries over the period 1996 to 2016. - Dataset includes information on market access indicators, the quality of fiscal governance, and fiscal and macroeconomic variables. - Market access indicators: - Sovereign credit ratings from DataStream - Thomson Reuters for Fitch, Moody’s and Standard and Poor’s; ratings converted into a common numeric scale and averaged across agencies when a country is rated by more than one agency. If there were rating changes during the year, the average of changes within the year is taken. - Sovereign bond issuance data and option-adjusted spreads from DataStream - Thomson Reuters; 51 emerging or developing economies identified as having issued at least one sovereign bond between 1996 and 2016. Coverage verified against Presbitero et al. (2016) and IMF Sub-Saharan Africa Regional Economic Outlook reports. - Quality of fiscal governance: - Primary measure: Worldwide Governance Indicators (WGI) — Government Effectiveness indicator (Kaufmann et al., 2010). WGI available annually starting from 1996; standardized scores across six dimensions. - WGI Government Effectiveness reflects perceptions of the quality of public services, civil service quality and independence, quality of policy formulation and implementation, and credibility of government commitment to policies. - Fiscal and macroeconomic data from IMF World Economic Outlook, October 2018 vintage, and the World Development Indicators. ### Descriptive analysis — stylized facts linking fiscal governance and market access - Binned scatterplots show correlations between market access indicators and WGI Government Effectiveness score. - Sovereign ratings: - Strong, nonlinear relation between fiscal governance and whether a country had a sovereign rating. - All countries in the lowest decile were unrated in 2010; most countries with above median WGI scores had a sovereign rating. - Rated countries with above-median WGI scores tend to have investment grade ratings (BBB- or higher). - Sovereign bond issuance (sample restricted to emerging and developing economies): - Nonlinear relation: initial improvements in government performance increase probability of issuing a sovereign bond; beyond a point further improvement is not associated with higher probability of issuance (domestic issuance preference and clustering of small countries). - Sovereign bond spreads: - Countries with worse fiscal governance pay much larger spreads on external bonds; worst performers can pay spreads of up to 600 basis points on average. - Countries in the top third of the WGI Government Effectiveness index typically face spreads of less than 50 basis points. ### Key regression specifications and quantitative associations (WGI-based) - Specification: second-degree polynomial in the WGI Government Effectiveness score; logistic regressions for binary dependent variables and OLS otherwise; robust standard errors clustered by country. Controls include log GDP per capita (constant 2011 PPP), public debt/GDP, inflation, fiscal deficit, growth, current account balance, log population, primary commodities exporter indicator, and year fixed effects (except bond issuance regressions). - For countries in the lowest quartile of the WGI Government Effectiveness score: - A one-standard deviation improvement is associated with a 30 percent higher probability of having a sovereign credit rating. - Associated with an average rating improvement of 1.3 notches for countries that are rated. - Linked with a 20 percent higher likelihood of having issued at least one external sovereign bond since 1996. - Linked with an average decline in sovereign bond spreads of almost 190 basis points. - Selected regression coefficients and statistics from Table 1 (all controls included): - Dependent variable columns: Rating dummy; Average rating; Bond issuance; Bond spread. - Govt Effectiveness: - Govt Effectiveness: 1.66*** (Rating dummy), 1.12*** (Average rating), -0.33 (Bond issuance), -188.93*** (Bond spread) - Standard errors: (0.36), (0.10), (0.34), (61.11) respectively. - Govt Effectiveness^2: -0.53** (Rating dummy), 0.14*** (Average rating), -0.60*** (Bond issuance), 11.59 (Bond spread) - Standard errors: (0.22), (0.05), (0.21), (29.65) respectively. - Log(GDP p.c.) coefficients: 1.03*** (Rating dummy), 0.57*** (Average rating), 0.83*** (Bond issuance), 1.01 (Bond spread) - Standard errors: (0.29), (0.08), (0.26), (34.00) respectively. - Public Debt/GDP coefficients: -0.01 (Rating dummy), -0.01*** (Average rating), 0.02** (Bond issuance), 1.28*** (Bond spread) - Standard errors: (0.01), (0.00), (0.01), (0.41) respectively. - Year fixed effects: Yes for Rating dummy, Average rating, and Bond spread; No for Bond issuance. - Adjusted R^2 / pseudo-R^2: 0.50 (Rating dummy), 0.86 (Average rating), 0.15 (Bond issuance), 0.52 (Bond spread). - Sample sizes (N): 2,800 (Rating dummy), 1,901 (Average rating), 165 (Bond issuance), 533 (Bond spread). - Significance legend: *0.10, **0.05, ***0.01. - Robustness: interactions between fiscal governance and fiscal deficits and debt ratios considered; two-step estimators (Heckman) considered for average rating regressions conditional on being rated; results broadly similar. ### PEFA assessments — motivation, framework, and data - PEFA used to obtain granular, policy-relevant measures of public financial management (PFM). PEFA started in 2001; more than 600 completed diagnostic reports in over 150 countries as of end-2018. - Hypothesis: PEFA scores highly correlated with WGI Government Effectiveness and map into a market access–PEFA nexus (appendix Figure A.2 shows strong correlation). - PEFA framework: - 31 indicators scored between D (lowest) and A (highest), organized into seven groups/dimensions: - (i) Budget reliability - (ii) Transparency of public finances - (iii) Management of assets and liabilities - (iv) Fiscal strategy and budgeting - (v) Predictability and control in budget execution - (vi) Accounting and reporting - (vii) External scrutiny and audit - Focus on seven dimensions to identify concrete, risk-mitigating PFM areas that can improve market access. - 2016 PEFA framework referenced; previous frameworks (2005 and 2011) mapped into the 2016 format for consistency. - PEFA data used: 173 publicly-available PEFA reports for 89 emerging and developing economies between 2005 and 2016. ### Empirical specification for PEFA analysis - Pooled panel 2005–2016 estimated as: - Y_{i,t} = α + ∑_{j∈{A,B,C}} β_k^j PEFA_{k^j,i,t} + γ′ X_{i,t} + ε_{i,t} - Y_{i,t}: indicator of market access. - PEFA_{k^j,i,t}: score on one of seven PEFA dimensions. - X_{i,t}: controls including log GDP per capita (constant PPP), public sector debt/GDP, overall fiscal deficit, current account balance, inflation, real GDP growth, log population, and primary commodities exporter indicator. - Estimation methods: logistic regressions for binary dependent variables; pooled OLS for continuous dependent variables; robust standard errors clustered by country. ### Results — PEFA dimensions and market access outcomes - Sovereign credit ratings: - Higher PEFA scores across the seven dimensions associated with greater probability of having a sovereign credit rating. - All countries with an A score in asset and debt management, fiscal strategy and fiscal reporting categories had a sovereign credit rating. (No countries achieved an A in budget execution and external audit categories in the sample.) - Largest quantitative impact on probability of having a sovereign rating: transparency of public finances. - Improvements in accounting and reporting and asset and debt management associated with large improvements in sovereign ratings, especially at A scores. - Improvement from a D score to at least a C score in fiscal strategy associated with an improvement of just over one notch (example: equivalent of achieving a BB rating if baseline is BB-). - Conditional-average-rating regressions correcting for selection bias yield qualitatively similar results. - Table 2 diagnostics: Pseudo R2 values across PEFA dimensions: 0.48, 0.51, 0.46, 0.49, 0.47, 0.45, 0.45 with sample sizes N = 173, 173, 157, 172, 173, 170, 173 respectively. - Table 3 (average sovereign rating conditional on being rated): Pseudo R2 values 0.56, 0.56, 0.58, 0.62, 0.58, 0.58, 0.57 with N = 102 for each regression. - Sovereign bonds (probability of issuance and spreads): - Higher transparency of public finances has largest impact on likelihood of issuing a sovereign bond. - Improvement in budget credibility associated with higher likelihood of issuing an external sovereign bond. - Table 4 (probability of issuing at least one sovereign bond since 1996) diagnostics: Pseudo R2 values of 0.24, 0.24, 0.22, 0.23, 0.22, 0.23, 0.22 with N = 173, 173, 173, 172, 173, 173, 173. - Higher PEFA scores associated with lower sovereign bond spreads across all seven dimensions, with particularly large and statistically significant effects for: - transparency of public finances, - asset and liability management, and - budget accounting and reporting. - Specific magnitudes: - Moving from a C score to an A score in budget accounting and reporting associated with a -166.07*** basis point change in sovereign bond spread (standard error (27.49)). - Moving from C to A in transparency of public finances associated with -176.89*** basis points (standard error (55.18)). - Table 5 diagnostics (average sovereign bond spread net of year fixed effects): Adjusted R2 values across dimensions: 0.01, 0.15, 0.13, 0.09, 0.04, 0.03, 0.01 with sample sizes N = 64 for each regression. - Impact of broad-based improvements (simulations): - Moving from D to C across all seven PEFA dimensions associated with an average increase of two rating notches (example: BB- → BB+). - Achieving at least a B score on all seven dimensions associated with an additional one notch improvement on average. - Achieving an A score on all seven dimensions suggests an improvement of almost another notch beyond B. - Probability of issuing a sovereign bond: - Overall D across all seven dimensions: very low and statistically indistinguishable-from-zero probability of issuing a sovereign bond. - Overall C associated with a 40 percent probability of having issued at least one sovereign bond during 1996–2016. - Overall A associated with about a 65 percent probability of having issued at least one sovereign bond during 1996–2016. - Sovereign bond spreads: improvement from C to A across PEFA dimensions associated with a 200 basis point reduction in option-adjusted bond spreads. - Auxiliary analysis: share of market-based external debt increases with fiscal governance performance, suggesting spillovers to other market-based external financing. ### Policy implications and recommendations - Improving fiscal governance and public financial management enhances market access: higher likelihood of sovereign credit ratings, higher probability of issuing sovereign bonds, stronger credit ratings, and lower sovereign bond spreads. - Highest value-added PEFA areas for market access: - transparency of public finances, - fiscal reporting / accounting and reporting, - asset and debt management, - fiscal strategy. - Pragmatic sequencing: concentrate reform efforts on PEFA areas with largest impact on market access given vested interests and short-term reform costs. - Rationale for multilateral institution role: magnitude and persistence of benefits support capacity building, technical assistance, and program design support by multilateral institutions (including the IMF). ### Summary statistics (selected) - Sample observations: Obs 2,800 for most macroeconomic and fiscal series; Obs 1,901 for Sovereign Rating; Obs 173 for PEFA component scores. - Macroeconomic: - Population (millions): Mean 40.281, Std. Dev. 142.93, 10% 0.75, 25% 3.15, 50% 9.08, 75% 28.16, 90% 72.72, Obs 2,800. - GDP p.c. (2011 constant PPP): Mean 17,445, Std. Dev. 19,721, 10% 1,558, 25% 3,314, 50% 10,187, 75% 25,623, 90% 42,932, Obs 2,800. - GDP growth (%): Mean 4.12, Std. Dev. 4.91, 10% -0.13, 25% 2.00, 50% 4.01, 75% 6.24, 90% 8.45, Obs 2,800. - Inflation (%): Mean 6.38, Std. Dev. 15.34, 10% 0.44, 25% 1.80, 50% 3.92, 75% 7.82, 90% 12.70, Obs 2,800. - Current Account Balance (% GDP): Mean -2.11, Std. Dev. 10.71, 10% -12.31, 25% -7.07, 50% -2.53, 75% 2.32, 90% 8.65, Obs 2,800. - Fiscal: - Overall Fiscal Balance (% GDP): Mean -1.80, Std. Dev. 6.24, 10% -7.01, 25% -4.39, 50% -2.37, 75% 0.02, 90% 3.36, Obs 2,800. - Public Debt (% GDP): Mean 54.32, Std. Dev. 42.81, 10% 15.44, 25% 28.14, 50% 43.95, 75% 68.52, 90% 101.07, Obs 2,800. - Market access indicators: - Has Sovereign Rating (0-1): Mean 0.68, Std. Dev. 0.47, 10% 0, 25% 0, 50% 1, 75% 1, 90% 1, Obs 2,800. - Sovereign Rating: Mean 4.99, Std. Dev. 1.77, 10% 3.00, 25% 3.17, 50% 5.00, 75% 6.11, 90% 8.00, Obs 1,901. - Issued Sovereign Bond (0-1): Mean 0.42, Std. Dev. 0.49, 10% 0, 25% 0, 50% 0, 75% 1, 90% 1, Obs 2,800. - Sovereign Bond Spreads: Mean 170.19, Std. Dev. 195.91, 10% 13.90, 25% 38.70, 50% 110.90, 75% 268.70, 90% 381.00, Obs 533. - WGI: - Government Effectiveness: Mean 0.01, Std. Dev. 0.99, 10% -1.17, 25% -0.74, 50% -0.17, 75% 0.71, 90% 1.59, Obs 2,800. - PEFA component scores (numerical 1–4 where 1 = D, 4 = A; Obs 173): - Budget Credibility: Mean 2.51, Std. Dev. 0.79. - Transparency: Mean 2.40, Std. Dev. 0.65. - Asset and Debt Management: Mean 2.55, Std. Dev. 0.81. - Fiscal Strategy: Mean 2.61, Std. Dev. 0.67. - Budget Execution: Mean 2.92, Std. Dev. 0.61. - Accounting and Reporting: Mean 2.85, Std. Dev. 0.73. - External Audit: Mean 3.27, Std. Dev. 0.68. *Source: wpiea2019218-print-pdf* ### 2.1 Dataset ### 2.1 Dataset ### Dataset coverage - Analysis focuses on 173 countries over the period 1996 to 2016. - Dataset includes information on market access indicators, the quality of fiscal governance, and fiscal and macroeconomic variables. ### Market access indicators - Sovereign credit ratings obtained from DataStream - Thomson Reuters for Fitch, Moody’s and Standard and Poor’s. - Ratings converted into a common numeric scale and averaged across agencies when a country is rated by more than one agency. - If there were rating changes during the year, the average of changes within the year is taken. - Information on bond issuances, including option-adjusted spreads, taken from DataStream - Thomson Reuters. - Identified 51 emerging or developing economies that issued at least one sovereign bond between 1996 and 2016. - Sovereign bond issuance data coverage verified against Presbitero et al. (2016) and different editions of the IMF Sub-Saharan Africa Regional Economic Outlook report. ### Quality of fiscal governance - Primary measure: Worldwide Governance Indicators (WGI) — Government Effectiveness indicator (Kaufmann et al., 2010). - WGI available annually starting from 1996; governance described using standardized scores across six dimensions. - Government Effectiveness reflects: "perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies." - Other WGI indicators: voice and accountability; political stability; regulatory quality; rule of law; control of corruption. ### Fiscal and other macroeconomic data - Additional data sources: IMF World Economic Outlook, October 2018 vintage, and the World Development Indicators. ### Descriptive analysis — stylized facts linking fiscal governance and market access - Binned scatterplots (Figure 1) show correlations between market access indicators and WGI Government Effectiveness score. - Sovereign ratings: - Strong, nonlinear relation between fiscal governance and whether a country had a sovereign rating. - All countries in the lowest decile were unrated in 2010; most countries with above median WGI scores had a sovereign rating. - Rated countries with above-median WGI scores tend to have investment grade ratings (BBB- or higher). - Sovereign bond issuance: - Sample restricted to emerging and developing economies (most advanced economies issue domestically). - Nonlinear relation: at first, stronger government performance increases probability of having issued a sovereign bond; after a point, further improvement is not associated with higher probability of issuance (domestic issuance preference and clustering of small countries). - Sovereign bond spreads: - Countries with worse fiscal governance pay much larger spreads on external bonds. - On average, the worst performers can pay spreads of up to 600 basis points. - Countries in the top third of the WGI Government Effectiveness index typically face spreads of less than 50 basis points. - Controls added for robustness: - Log of GDP per capita in constant 2011 PPP dollars. - Public debt ratios (public debt/GDP). - Business cycle variables: inflation, fiscal deficit, growth, current account balance. - Country size: log of resident population. - Indicator for primary commodities exporter. - Year fixed effects (except in bond issuance regression which uses only cross-sectional variation). - Key quantitative associations (summary results reported from regressions illustrated in Table 1): - Specification uses a second-degree polynomial in the WGI Government Effectiveness score; estimations: logistic regressions for binary dependent variables and OLS otherwise; robust standard errors clustered by country. - For countries in the lowest quartile of the WGI Government Effectiveness score: - A one-standard deviation improvement is associated with a 30 percent higher probability of having a sovereign credit rating. - Associated with an average rating improvement of 1.3 notches for countries that are rated. - Linked with a 20 percent higher likelihood of having issued at least one external sovereign bond since 1996. - Linked with an average decline in sovereign bond spreads of almost 190 basis points. - Selected regression coefficients and statistics from Table 1 (all controls listed in the note are included in these regressions): - Dependent variables columns: Rating dummy; Average rating; Bond issuance; Bond spread. - Govt Effectiveness coefficient(s): - Govt Effectiveness: 1.66*** (Rating dummy), 1.12*** (Average rating), -0.33 (Bond issuance), -188.93*** (Bond spread) - Standard errors: (0.36) for Rating dummy, (0.10) for Average rating, (0.34) for Bond issuance, (61.11) for Bond spread. - Govt Effectiveness^2: -0.53** (Rating dummy), 0.14*** (Average rating), -0.60*** (Bond issuance), 11.59 (Bond spread) - Standard errors: (0.22) for Rating dummy, (0.05) for Average rating, (0.21) for Bond issuance, (29.65) for Bond spread. - Log(GDP p.c.) coefficients: 1.03*** (Rating dummy), 0.57*** (Average rating), 0.83*** (Bond issuance), 1.01 (Bond spread) - Standard errors: (0.29), (0.08), (0.26), (34.00) respectively. - Public Debt/GDP coefficients: -0.01 (Rating dummy), -0.01*** (Average rating), 0.02** (Bond issuance), 1.28*** (Bond spread) - Standard errors: (0.01), (0.00), (0.01), (0.41) respectively. - Year fixed effects: Yes for Rating dummy, Average rating, and Bond spread; No for Bond issuance. - Adjusted R^2 / pseudo-R^2: 0.50 (Rating dummy), 0.86 (Average rating), 0.15 (Bond issuance), 0.52 (Bond spread). - Sample sizes (N): 2,800 (Rating dummy), 1,901 (Average rating), 165 (Bond issuance), 533 (Bond spread). - Significance legend: *0.10, **0.05, ***0.01. - Robustness remarks: - Interactions between fiscal governance and fiscal deficits and debt ratios were considered; results are similar. - Two-step estimators (Heckman) were considered for average rating regressions conditional on being rated; results are similar. ### PEFA assessments — motivation and scope - PEFA framework used to obtain more granular, policy-relevant measures of public financial management (PFM). - PEFA started in 2001; harmonizes assessment of PFM across countries. - More than 600 completed diagnostic reports in over 150 countries as of end-2018. - Hypothesis: PEFA scores are highly correlated with WGI Government Effectiveness and map into a market access-PEFA nexus. - PEFA scores strongly correlated with WGI Government Effectiveness (reference to appendix Figure A.2). - PEFA framework details: - Includes 31 indicators scored between D (lowest) and A (highest). - Indicators organized into seven groups corresponding to key elements of a PFM system: - (i) Budget reliability — budgets feasible and implemented as intended. - (ii) Transparency of public finances — information comprehensive, consistent, and accessible. - (iii) Management of assets and liabilities — recording of assets/liabilities, fiscal risk identification, value for money of public investments. - (iv) Fiscal strategy and budgeting — budget consistency with fiscal policies and adequacy of macro/fiscal projections. - (v) Predictability and control in budget execution — effective budget implementation, audit and internal control processes. - (vi) Accounting and reporting — accurate records maintained and published timely. - (vii) External scrutiny and audit — independent review of public finances and follow-up on recommendations. - Focus on seven dimensions to identify concrete, risk-mitigating areas of PFM performance that can improve market access. - Note: 2016 PEFA framework referenced; previous frameworks (2005 and 2011) mapped into the 2016 format for consistency. ### Empirical specification for PEFA analysis - Data: 173 publicly-available PEFA reports for 89 emerging and developing economies between 2005 and 2016. - Estimated specification (pooled panel 2005–2016): - Y_{i,t} = α + ∑_{j∈{A,B,C}} β_k^j PEFA_{k^j,i,t} + γ′ X_{i,t} + ε_{i,t} - Y_{i,t}: indicator of market access. - PEFA_{k^j,i,t}: score on one of the seven PEFA dimensions. - X_{i,t}: control vector including log GDP per capita (constant PPP), public sector debt/GDP, overall fiscal deficit, current account balance, inflation, real GDP growth, log population, and primary commodities exporter indicator. - Estimation methods: - Logistic regressions for binary dependent variables (e.g., sovereign rating dummy, bond issuance dummy). - Pooled OLS for continuous dependent variables. - Robust standard errors clustered by country. *Source: wpiea2019218-print-pdf — sections 2.1 and 2.2 (Dataset and Descriptive analysis), and sections 3.1–3.2 (PEFA assessments and Empirical specification).* ### 3.3 Results ### 3.3 Results ### Sovereign credit ratings - Higher PEFA scores across the seven dimensions are associated with a greater probability of having a sovereign credit rating by one of the three main rating agencies. - All countries that achieved an A score in the asset and debt management, fiscal strategy and fiscal reporting categories had a sovereign credit rating. (Note: No countries in the sample achieved an A score in the budget execution and external audit categories.) - The category with the largest quantitative impact on the probability of having a sovereign rating was transparency of public finances. This indicator captures whether information on public finances is provided in an accurate, comprehensive, and timely fashion (including conformity of budget classification to international standards, comprehensiveness of budget information, timeliness and accuracy of intergovernmental transfers reporting, availability and use of public service delivery performance indicators, and timely public access to comprehensive information on public finances). - Improvements in accounting and reporting and asset and debt management performance are associated with economically and statistically large improvements in sovereign ratings, especially when countries achieve an upper score of A. - An improvement from a D score to at least a C score in fiscal strategy is associated with an improvement of just over one notch (example given: equivalent of achieving a BB rating if the baseline rating is BB-). - When estimating the impact on the average credit rating conditional on having a credit rating, results are qualitatively similar when correcting for potential selection bias (two-step estimator results available upon request). Key empirical notes from regressions reported: - Table 2 regressions control for log GDP per capita in constant 2011 PPP dollars, public debt ratios, log population, growth, inflation, current account balance, overall fiscal deficit and whether a country is a primary commodities exporter. - Table 2: Pseudo R2 values reported across PEFA dimensions include 0.48, 0.51, 0.46, 0.49, 0.47, 0.45, 0.45 with sample sizes N = 173, 173, 157, 172, 173, 170, 173 respectively. - Table 3 (average sovereign rating conditional on being rated) reports Pseudo R2 values of 0.56, 0.56, 0.58, 0.62, 0.58, 0.58, 0.57 with N = 102 for each regression. ### Sovereign bonds (probability of issuance and spreads) - Higher transparency of public finances had the largest quantitative impact on the likelihood of issuing a sovereign bond, consistent with transparency’s role in enabling sovereign credit ratings which facilitate bond issuance. - An improvement in budget credibility is associated with a higher likelihood of issuing an external sovereign bond. - Table 4 regressions (probability of issuing at least one sovereign bond since 1996) control for the same covariates as above. Table 4 reports Pseudo R2 values of 0.24, 0.24, 0.22, 0.23, 0.22, 0.23, 0.22 with N = 173, 173, 173, 172, 173, 173, 173. - For bond spreads, higher PEFA scores are associated with lower sovereign bond spreads across all seven dimensions, with particularly large and statistically significant effects for: - transparency of public finances, - asset and liability management, and - budget accounting and reporting. - Specific magnitudes reported: - Moving from a C score to an A score in the budget accounting and reporting dimension is associated with a 166 basis point reduction in sovereign bond spread (Table 5 reports C→A -166.07*** with (27.49) reported as the standard error). - The same improvement in the transparency of public finances category is associated with a similar size spread reduction (Table 5 reports C→A -176.89*** with (55.18) as the standard error). - Table 5 regression diagnostics: Adjusted R2 values across dimensions reported as 0.01, 0.15, 0.13, 0.09, 0.04, 0.03, 0.01 with sample sizes N = 64 for each regression. The dependent variable is the average sovereign bond spread by country at bond issuance, net of year fixed effects. ### Impact of broad-based improvements (simultaneous improvements across all seven PEFA dimensions) - Simulated across-the-board improvements (starting from baseline D) imply substantial gains in market access: - Moving from a score of D to a C across all seven PEFA dimensions is associated with an average increase of two rating notches (example: BB- → BB+). - Achieving at least a B score on all seven dimensions is associated with an additional one notch improvement on average. - Achieving an A score on all seven dimensions suggests an improvement of almost another notch beyond B. - Probability of issuing a sovereign bond (conditional on PEFA score): - Countries with an overall D score across all seven dimensions have a very low and statistically indistinguishable-from-zero probability of issuing a sovereign bond. - An overall score of C is associated with a 40 percent probability of having issued at least one sovereign bond during 1996–2016. - An overall score of A is associated with about a 65 percent probability of having issued at least one sovereign bond during 1996–2016. - Sovereign bond spreads: - An improvement from a C to an A score across PEFA dimensions is associated with a 200 basis point reduction in option-adjusted bond spreads (statistically and economically significant). - Evidence from auxiliary analysis (appendix): the share of market-based external debt increases with fiscal governance performance, suggesting improvements in fiscal governance that lower sovereign bond spreads may extend benefits to other market-based sources of external financing. ### Policy implications and recommendations - Improving fiscal governance and public financial management enhances market access: higher likelihood of sovereign credit ratings, higher probability of issuing sovereign bonds, stronger credit ratings, and lower sovereign bond spreads. - Practices with highest value-added for market access include improvements related to: - transparency of public finances, - fiscal reporting / accounting and reporting, - asset and debt management, and - fiscal strategy. - Because vested interests and short-term reform costs can impede reforms, concentrating reform efforts on the PEFA areas with the largest impact on market access is a pragmatic sequencing strategy. - The magnitude and persistence of benefits provide a clear rationale for a role by multilateral institutions (including the IMF) in fostering fiscal governance improvements through capacity building, technical assistance, and, where relevant, program design. ### Avenues for future research - Examine the impact of fiscal governance on other sources of financing such as external loans and foreign participation in domestic bond markets. - Investigate whether strong fiscal governance helps retain market access and achieve better macroeconomic outcomes during crisis periods. *Source: 3.3 Results (wpiea2019218-print-pdf)* ### PART 1: PEFA OVERVIEW 3 ### PART 1: PEFA OVERVIEW 3 ### Scope and institutions covered - The core PEFA methodology was initially focused on central government, including related oversight and accountability institutions, such as the legislature and supreme audit institutions. - PEFA has increasingly been used in the assessment of subnational government PFM performance. - The scope of the category of “central government”, as used in PEFA, is based on the classification structure developed by the International Monetary Fund (IMF) for Government Finance Statistics (GFS). - PEFA methodology refers to GFS terminology where possible but is adaptable to situations where other classifications and standards are used. - Other parts of government identified in GFS include different administrative tiers with separate accountability mechanisms and their own PFM systems, including subnational governments such as state, provincial, regional, and local governments, including districts and municipalities. ### PEFA indicator focus and coverage - The PEFA indicator set is focused on the financial operations of the entire level of government covered by the assessment. - Activities of central government implemented outside the budget are included in the coverage of a small number of indicators and are referred to as: - expenditure and revenue of extrabudgetary units, and - expenditure and revenue related to the extrabudgetary activities of budgetary units (for example, PEFA indicator (PI)-6). - Public corporations are referred to in PI-10. - Subnational governments with a direct relationship to central government are referred to in PI-7 and PI-10. - The measurement guidance explains how each indicator relates to GFS categories, where relevant. - PEFA examines operations outside of the government being assessed only to the extent that they have an impact on the fiscal performance of the central government. ### Figures referenced (descriptive) - FIGURE 1: PEFA pillars and the budget cycle — lists pillars including: - Accounting and reporting - Transparency of public finances - External scrutiny and audit - Predictability and control in budget execution - Budget reliability - Policy-based fiscal strategy and budgeting - Management of assets and liabilities - Figure A.2: PEFA and WGI score — plots PEFA, Standardized average score against WGI, Government effectiveness with axis range shown from -2 to 2 for PEFA and -2 to 2 for WGI in the figure labels. - Figure A.3: Government Effectiveness and Share of External Public Debt that is Market-Based — left panel examines relation between share of public external debt that was market-based in 2010 and the WGI Government Effectiveness score; right panel shows impact of PEFA score improvements on the same indicator. Note: sample restricted to emerging and developing economies. ### Conversion of sovereign credit ratings (Table B.1) — selected entries - Aaa -> AAAAAA -> Numerical Scale 8 - Aa1 -> AA+ -> AA+ -> Numerical Scale 7 - A1 -> A+ -> A+ -> Numerical Scale 6 - Baa1 -> BBB+ -> BBB+ -> Numerical Scale 5 - Ba1 -> BB+ -> BB+ -> Numerical Scale 4 - B1 -> B+ -> B+ -> Numerical Scale 3 - Caa1 -> CCC+ -> C -> Numerical Scale 2 - D -> DDDDD -> D -> Numerical Scale 1 ### Summary statistics (Table B.2) — key figures - Sample observations: Obs 2,800 for most macroeconomic and fiscal series; Obs 1,901 for Sovereign Rating; Obs 173 for PEFA component scores. - Macroeconomic - Population (millions): Mean 40.281, Std. Dev. 142.93, 10% 0.75, 25% 3.15, 50% 9.08, 75% 28.16, 90% 72.72, Obs 2,800 - GDP p.c. (2011 constant PPP): Mean 17,445, Std. Dev. 19,721, 10% 1,558, 25% 3,314, 50% 10,187, 75% 25,623, 90% 42,932, Obs 2,800 - GDP growth (%): Mean 4.12, Std. Dev. 4.91, 10% -0.13, 25% 2.00, 50% 4.01, 75% 6.24, 90% 8.45, Obs 2,800 - Inflation (%): Mean 6.38, Std. Dev. 15.34, 10% 0.44, 25% 1.80, 50% 3.92, 75% 7.82, 90% 12.70, Obs 2,800 - Current Account Balance (% GDP): Mean -2.11, Std. Dev. 10.71, 10% -12.31, 25% -7.07, 50% -2.53, 75% 2.32, 90% 8.65, Obs 2,800 - Fiscal - Overall Fiscal Balance (% GDP): Mean -1.80, Std. Dev. 6.24, 10% -7.01, 25% -4.39, 50% -2.37, 75% 0.02, 90% 3.36, Obs 2,800 - Public Debt (% GDP): Mean 54.32, Std. Dev. 42.81, 10% 15.44, 25% 28.14, 50% 43.95, 75% 68.52, 90% 101.07, Obs 2,800 - Market Access Indicators - Has Sovereign Rating (0-1): Mean 0.68, Std. Dev. 0.47, 10% 0, 25% 0, 50% 1, 75% 1, 90% 1, Obs 2,800 - Sovereign Rating: Mean 4.99, Std. Dev. 1.77, 10% 3.00, 25% 3.17, 50% 5.00, 75% 6.11, 90% 8.00, Obs 1,901 - Issued Sovereign Bond (0-1): Mean 0.42, Std. Dev. 0.49, 10% 0, 25% 0, 50% 0, 75% 1, 90% 1, Obs 2,800 - Sovereign Bond Spreads: Mean 170.19, Std. Dev. 195.91, 10% 13.90, 25% 38.70, 50% 110.90, 75% 268.70, 90% 381.00, Obs 533 - WGI - Government Effectiveness: Mean 0.01, Std. Dev. 0.99, 10% -1.17, 25% -0.74, 50% -0.17, 75% 0.71, 90% 1.59, Obs 2,800 - PEFA component scores (converted to numerical scale from 1 to 4, where 1 is the lowest "D" score, and 4 is the highest "A" score; Obs 173) - Budget Credibility: Mean 2.51, Std. Dev. 0.79 - Transparency: Mean 2.40, Std. Dev. 0.65 - Asset and Debt Management: Mean 2.55, Std. Dev. 0.81 - Fiscal Strategy: Mean 2.61, Std. Dev. 0.67 - Budget Execution: Mean 2.92, Std. Dev. 0.61 - Accounting and Reporting: Mean 2.85, Std. Dev. 0.73 - External Audit: Mean 3.27, Std. Dev. 0.68 *Source: wpiea2019218-print-pdf - PART 1: PEFA OVERVIEW 3* --- _Source: https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019218-print-pdf.pdf_