Fiscal Policy: Ready for The Next Shock?

Fiscal Policy: Ready for The Next Shock? Franziska Ohnsorge December 217

Duration of Global Expansions: Getting Older Although Not Yet Dying of Old Age 18 Global expansions (Number of years) 45 Expansions in the United States (Number of quarters) Post-196 average 12 Post-196 average 35 6 25 1976-1981 1983-199 1992-28 21 -Present 15 1982Q4-199Q3 1991Q1-21Q1 21Q4-27Q4 29Q2- Present Sources: International Monetary Fund, Kose and Terrones (215), National Bureau of Economic Research, U.S. Bureau of Economic Analysis, World Bank. Left Panel. Post-196 average is an average length of past three global expansions (excluding the current ongoing expansion). Right Panel. Post-196 average is an average length of seven historical episodes since 196 (excluding the current ongoing expansion). 2

Spillovers from Growth in Major Economies: Sizable Impact on the Rest of the World. Growth effect of 1-percentage-point decrease in U.S. growth (Percentage points) Impact of 1-percentage-point decrease in EM7 and G7 growth on growth in other EMDEs (Percentage points) -.5-1. -1-1.5-2. Other AEs EMDEs 1 year 2 years Sources: World Bank, Haver Analytics, Organisation for Economic Co-operation and Development. Left Panel. Cumulative impulse response of weighted average output growth in non-u.s. advanced economies (other AEs) and emerging market and developing economies (EMDEs) to a 1-percentage-point increase in growth in real GDP in the United States. Growth spillovers to other AEs and EMDEs based on a Bayesian vector autoregression of global GDP growth excluding the United States and other AEs or EMDEs, U.S. GDP growth, U.S. 1-year government bond yields plus J.P. Morgan s EMBI index and GDP growth in other AEs or EMDEs. The oil price is exogenous. Bars represent median values, and error bars 16-84 percent confidence bands. Other AEs consist of Euro Area (19 countries), Canada, Japan and the United Kingdom and 2 EMDEs are included. Data are for 1998Q1-216Q2. Right Panel. Cumulative impulse response after one year on output growth in 23 other AEs and 2 EMDEs to a 1-percent increase in VIX. Vector autoregressions are estimated for 1998Q1-216Q2 with two lags. The model for other AEs includes VIX, MSCI index for advanced economies (MXGS), U.S. 1-year government bond yields, aggregate real output and investment growth in 23 other AEs. The model for EMDEs includes VIX, MSCI emerging market equity price index, J.P. Morgan s EMBIG spreads, aggregate real output and investment growth in 2 EMDEs. G7 real GDP growth, U.S. 1-year government bond yields and MSCI world equity price index are added as exogenous regressors. -2-3 EM7 G7 1 year 2 years 3

Three Questions 1 What is fiscal space? Fiscal space is a government s ability to service its obligations. It has four distinct dimensions. 2 Why is fiscal space a key challenge in EMDEs (Emerging Market and Developing Economies)? It has narrowed over the past decade. 3 What does narrower fiscal space mean for fiscal policy? Fiscal stimulus is less effective. 5

Three Questions 1 What is fiscal space? Fiscal space is government s ability to service its obligations. It has four distinct dimensions. 2 Why is fiscal space a key challenge in EMDEs (Emerging Market and Developing Economies)? It has narrowed sharply over the past decade. 3 What does narrower fiscal space mean for fiscal policy? Fiscal stimulus is less effective. 6

Fiscal Space: Four Major Dimensions Government Debt Sustainability Fiscal Space Balance Sheet Composition External and Private Sector Debt Market Perception Source: Kose et al. (217) A Cross-Country Database of Fiscal Space. Note: The database contains a total of 28 indicators: 12 indicators from Government Debt Sustainability showing the longer-run capacity of the government to finance its obligations (including general government gross debt, overall fiscal balance, and sustainability gaps); six indicators from Balance Sheet Composition showing the structure of sovereign debt (including the shares of government debt held by nonresidents and in foreign currency, and maturity of debt); eight indicators from External and Private Sector Debt showing the potential fiscal impact caused by liabilities in the external and private sectors, for example contingent liabilities (including external debt, short-term external debt, and credit to the private sector); and two indicators from Market Perception showing the ability of government to access capital market and roll over debt (sovereign CDS spreads and long-term sovereign debt ratings). 7

Definition of Fiscal Space and Database: Fiscal Space is the Ability of a Government to Service Its Debt Country coverage: 41 advanced economies, 159 EMDEs Time coverage: 199-217 28 indicators: 12 indicators of government debt sustainability (including general government gross debt, overall fiscal balance, and sustainability gaps); 6 indicators of government debt composition (including the shares of government debt held by nonresidents and in foreign currency, and maturity of debt); 8 indicators of external and private sector debt (including external debt, short-term external debt, and credit to the private sector); 2 indicators of market perception (sovereign CDS spreads and long-term sovereign debt ratings). Source: Kose, Kurlat, Ohnsorge, and Sugawara (217) A Cross-Country Database of Fiscal Space. Data is available at 8 http://www.worldbank.org/en/research/brief/fiscal-space

Fiscal Space: Four Major Dimensions Correlations of Indicators of Fiscal Space: All Countries 1..8.6.4.2 (Average correlation coefficients) Range, with indicators of different dimensions With indicators of same dimension 1..8.6.4.2 Correlations of Indicators of Fiscal Space: EMDEs (Average correlation coefficients) Range, with indicators of different dimensions With indicators of same dimension.. Government debt sustainability Balance sheet composition External, private sector debt Market perception Government debt sustainability Balance sheet composition External, private sector debt Market perception Source: Kose et al. (217) A Cross-Country Database of Fiscal Space. Note: Correlations coefficients between different indicators of fiscal space, average within each dimension (red lines) and range of averages across dimensions (blue bars). From Table 3 of Kose et al. (217). 9

Fiscal Space in EMDEs: Has Deteriorated in Multiple Dimensions since 27 Fiscal Space Government Debt Sustainability Balance Sheet Composition External and Private Sector Debt Market Perception Less space Same space Less space Less space Source: Kose et al. (217) A Cross-Country Database of Fiscal Space. Note: The database contains a total of 28 indicators: 12 indicators from Government Debt Sustainability showing the longer-run capacity of the government to finance its obligations (including general government gross debt, overall fiscal balance, and sustainability gaps); six indicators from Balance Sheet Composition showing the structure of sovereign debt (including the shares of government debt held by nonresidents and in foreign currency, and maturity of debt); eight indicators from External and Private Sector Debt showing the potential fiscal impact caused by liabilities in the external and private sectors, for example contingent liabilities (including external debt, short-term external debt, and credit to the private sector); and two indicators from Market Perception showing the ability of government to access capital market and roll over debt (sovereign CDS spreads and long-term sovereign debt ratings). 1

Government Debt and Deficits in EMDEs: Deteriorated Over Past Decade General government gross debt Overall fiscal balance 8 7 6 EMDEs LIC 2 1-1 5-2 4 3-3 -4-5 EMDEs LIC 2 27 29 211 213 215 217-6 27 29 211 213 215 217 Source: Kose et al. (217). Note: Simple averages. LIC includes 28 economies with available data. 12

Government Debt and Deficits in some EMDEs: Especially in Some Countries General government gross debt Overall fiscal balance 12 1 EMDEs LIC 1 5 EMDEs LIC 8 6 4-5 2-1 27 29 211 213 215 217-15 27 29 211 213 215 217 Source: Kose et al. (217). Note: Sinple averages. LIC includes 28 economies with available data. Diamonds denote two individual EMDEs (blue) and two LICs (red) with the largest deteriorations between 27 and 217. 13

Fiscal Sustainability Gap: What Fiscal Consolidation Would Stabilize Debt? A summary metric of government debt dynamics. Compares a country s fiscal balance with the balance that stabilizes government debt at a level (under different assumptions of output growth, long-term interest rate, and target debt level). Assumptions of output growth and interest rate (median debt by country group as a target): Historical conditions: Median Stressed conditions: Growth: Median - one standard deviation; Interest rate: Median + one standard deviation A positive (negative) gap implies that government debt will go below (above) the target debt level if the gap is sustained. 14

Sustainability Gaps: Obvious Need for Improvements in Fiscal Balances 3 Sustainability gaps, by group 3 Sustainability gaps, by region Current conditions Stressed conditions -3-3 -6-9 Current conditions Stressed conditions All countries Advanced economies EMDEs Source: Kose et al. (217). Note: Simple average of latest data (for 217) by country group. A sustainability gap is defined as the difference between the actual fiscal balance and the debt-stabilizing balance that captures cumulative impact of sustained fiscal balance on debt stocks. A sustainability gap under current conditions is based on contemporaneous growth and interest rates. Under stressed conditions it is assumed that output growth is one standard deviation lower and interest rate is one standard deviation higher in computing a debt-stabilizing balance. Left Panel. Based on 17 countries (35 advanced economies and 72 EMDEs). Right Panel. Based on 1 countries, 14 countries, 2 countries, 8 countries, 5 countries, and 15 countries in East Asia and Pacific (EAP), Europe and Central Asia (ECA), Latin America and the Caribbean (LAC), Middle East and North Africa (MNA), South Asia (SAR), and Sub-Saharan Africa (SSA), respectively. -6-9 -12 EAP ECA LAC MNA SAR SSA 15

EMDE Balance Sheet Composition: Weakened in Some Aspects, Improved in Others over Past Decade 65 Government debt held by nonresidents, Concessional debt (Percent of government debt) 5 Government debt in foreign currency (Percent of government debt) 11 Sovereign debt average maturity (Years) 6 55 5 45 4 35 3 Nonresident share (EMDEs) Concessional debt (LICs) 27 29 211 213 49 48 47 46 45 29 211 213 215 217 Source: Kose et al. (217). Note: Simple average of EMDEs. Data for 217 are year-to-date. Left Panel. Data are for general government in all countries but Kenya and Uganda (public sector). Based on 35 economies. Center Panel. Data are for general government in all countries but Kenya and Uganda (public sector). Based on 16 economies. Right Panel. Average maturity (average time of principal repayment) of sovereign and quasi-sovereign instruments denominated in U.S. dollars. Based on 36 economies. 1 9 29 211 213 215 217 17

EMDE External Debt, Private Debt and Market Perception: Deteriorated over Past Decade 6 Total external debt stocks EMDEs LICs 5 Credit to the private sector EMDEs LICs 12 Long-term sovereign debt ratings (Index, 1 21 [best]) EMDEs LICs 5 4 1 4 3 8 3 2 2 27 29 211 213 215 217 1 27 29 211 213 215 217 Source: Kose et al. (217). Note: Simple average of respective country groups. Data for 217 are year-to-date. LICs (low-income countries) are as classified by the World Bank in June 217. Left Panel. Based on 127 EMDEs and 28 LICs. Center Panel. Domestic credit to the non-financial private sector provided by commercial banks and, if data are available, by other financial corporations. Based on 123 EMDEs and 21 LICs. Right Panel. Averages of foreign-currency long-term sovereign debt ratings by Moody s, Standard & Poor s, and Fitch Ratings. Each rating is first converted to a numerical scale ranging from 1 to 21 (higher, better rating). The scores below 12 are the non-investment grades. Based on 89 EMDEs and 5 LICs. 6 27 29 211 213 215 217 18

Private and Public Debt: Dimensions of Fiscal Space Amplify Each Other 16 Credit to the private sector, 26 and 215 215 1 Private debt, by level of sustainability gap, 216 (Share of EMDEs) 12 8 6 Moderate Elevated 8 4 High 4 Commodity exporters Commodity importers 26 4 8 12 16 Sources: International Monetary Fund, World Bank. Left Panel. Credit to the private sector in percent of GDP in 26 on the horizontal axis and in 215 on the vertical axis. Right Panel. Share of EMDEs with respective levels of private debt out of those with sizable negative or moderate negative sustainability gaps. A sustainability gap is defined as the difference between the actual fiscal balance and the debt-stabilizing balance. Sample includes 7 EMDEs where data on sustainability gaps and private debt are available in 216. Sustainability gaps are considered to be sizable negative when negative gaps are in excess of 1 percent of GDP and moderate negative when negative gaps are below 1 percent of GDP. High private debt is defined as private sector credit in the top quartile of the distribution among 7 EMDEs during 2-16 (53 percent of GDP). Elevated private debt is defined as private sector credit in the second highest quartile (32 53 percent of GDP). 2 Sizable negative Sustainability gap Moderate negative 19

Why Has Fiscal Space Narrowed in EMDEs? Because of More Countercyclical Fiscal Policy 1..8.6.4.2. -.2 -.4 -.6 -.8 Procyclicality of government consumption (Correlation coefficient) 198-99 2-7 28-14 EMDEs 198-99 2-7 EM 28-14 -1-2 -3-4 -5 Structural fiscal balance during recessions All events pre-28 Great Recession -3-2 -1 1 2 3 Years -1-2 -3-4 -5 Structural fiscal balance around the Great Recession Wide fiscal space Narrow fiscal space -3-2 -1 1 2 3 Years Sources: Huidrom, Kose and Ohnsorge (216), World Bank. Left panel: Correlations between the cyclical components of government consumption and GDP from an unbalanced panel of annualdata for 6 EMDEs, including 31 emerging markets (EM) with global financial market access. All correlations are statistically significantly different from zero and differences in correlations across time are also statistically significant. Positive (negative) correlations suggest procyclical (countercyclical) fiscal policy. Center panel: t= is the year of the trough of the contraction episode. All variables refer to the unweighted sample mean. The structural balance is defined as the difference between cyclically adjusted revenues and cyclically adjusted expenditures. Right panel: t= is the year of the trough of the contraction episode. All variables refer to the unweighted sample mean. These results are based on the data sample of the event study which includes the 21 EMDEs that experienced contractions during the 28-9. The median debt-to-gdp ratio in the full sample of 17 EMDEs is 44 percent. Countries with debt-to-gdp ratios above the median are considered to have narrow fiscal space, while those with debt-to-gdp ratios below the median are considered to have wide fiscal space. 2

Why Has Fiscal Space Narrowed in EMDE Oil Exporters? And The Oil Price Plunge 8 6 Government debt 8 4 Fiscal balance Past oil price plunges Interquartile range Latest plunge (t = 214) 4 2 Past oil price plunges Interquartile range Latest plunge (t = 214) t-3 t-2 t-1 t t+1 t+2 t+3 Sources: International Monetary Fund, World Bank. Note: Past oil price plunges include collapses in global oil prices in 1991, 1998, 21, and 28, as identified in the June 215 Global Economic Prospects. Simple averages, based on 36 EMDE oil exporters. Government debt refers to liabilities that require payments of interest and principal by governments and includes, for example, debt liabilities in the form of currency and deposits, debt securities, and loans. It is for general government; therefore, all transactions among government entities are consolidated. -4-8 Year t-3 t-2 t-1 t t+1 t+2 t+3 21

EMDE Government Debt and Deficits in Historical Context: Debt Still Lower, But Deficits Much Wider General government debt and deficits General government debt and deficits in 217 (Percent of countries) 14 12 1 8 6 4 2 217 2 EMDEs LICs EMDEs LICs Debt (left scale) Deficits (right scale) 4.5 4. 3.5 3. 2.5 2. 1.5 1..5. Source: Kose et al. (217). Left panel: Simple averages. LIC includes 21 economies with available data in 2 and 217. Red bars are for 2 (government debt) or for the 2-6 average (fiscal deficits). Right panel: Share of EMDEs and share of LICs with government debt in percent of GDP in 217 below levels in 2 ( Below 2 levels ) or 2 percentage points of GDP above levels in 2 ( Much above 2 levels ). Share of EMDEs and share of LICs with fiscal deficits in percent of GDP in 217belowlevels in 2 ( Below 2levels ) or 3 percentage pointsof GDP above average levels in 2-6 ( Much above 2levels ). Orange bars denotetheremainder. 1 9 8 7 6 5 4 3 2 1 Below 2 levels EMDEs LIC EMDEs LIC Debt Much above 2 levels Deficits 22

EMDE Government Debt and Deficits in Historical Context: Somewhat Weaker Now than on the Eve of Previous Financial Crises 15 Government debt Financial stress Interquartile range Current period (t = 216) 3 Fiscal balance 1-3 5 Year t-3 t-2 t-1 t t+1 t+2 t+3-6 -9 Financial stress Interquartile range Current period (t = 216) t-3 t-2 t-1 t t+1 t+2 t+3 Sources: International Monetary Fund, World Bank. Note: Unweighted averages. The dashed blue lines show interquartile ranges. Financial crises include all three types of crises: banking, currency, and debt crises. The red line is shown for reference and based on all EMDEs, although it is not a crisis episode. Crisis episodes are taken from Gourinchas and Obstfeld (212) and Laeven and Valencia (213). In any country, when consecutive crises are identified within the next five years, the one associated with the lowest real GDP growth is used. Left Panel. Government debt refers to liabilities that require payments of interest and principal by governments and includes, for example, debt liabilities in the form of currency and deposits, debt securities, and loans. It is for general government, and therefore, all transactions among government entities are consolidated. 23

Three Questions 1 What is fiscal space? Fiscal space is a government s ability to service its obligations. It has four distinct dimensions. 2 Why is fiscal space a key challenge in EMDEs (Emerging Market and Developing Economies)? Fiscal space has narrowed over the past decade in most dimensions. 3 What does narrower fiscal space mean for fiscal policy? Fiscal stimulus is less effective. 24

How Effective is Fiscal Policy? The Size of the Fiscal Multiplier 1$ increase in spending Multiplier Change in output? Y = C + I + G + NX Why ΔG ΔY? Because C, I, NX also respond to G 25

Large Literature on Fiscal Multipliers: Wide Range of Estimates Unconditional multipliers. A lot of research But magnitudes range widely (-1.1 to +3.8) Conditional multipliers. Multipliers depend on circumstances. Country-specific features Trade openness Open economies (-.5) < closed economies (+1.1) Exchange rate regimes Flexible exchange rate regimes (-.7) < fixed exchange rate regimes (+1.4) Economic conditions Business cycle Expansions (-1.1) < recessions (+3.8) Fiscal space Little fiscal space (-3) < ample fiscal space (+.7) Fiscal instrument Revenues versus expenditures Revenues (+.3 to +.6) < spending (+.3 to +1.8) Investment, consumption and transfers Non-investment spending (+.3 to +1) < investment spending (+.5 to +1.8) Sources: A review of the large literature to 29 is provided in Spilimbergo, Symansky and Schindler (29). Since then, fiscal multipliers have been estimated in Ilzetzki, Mendoza, and Vegh (213); Auerbach and Gorodnichenko (212 and 213); Nickell and Tudyka (214); Auerbach and Gorodnichenko (217); Riera-Crichton, Vegh and Vuletin (214); Corsetti et al. (212 and 213);, Gali, Lopez-Salido and Valles (27); and Huidrom, Kose, Lim and Ohnsorge (216). Multipliers by country-specific features and economic conditions are from Ilzetzki, Mendoza and Vegh (213); Auerbach and Gorodnichenko (212) and Huidrom, Kose, Lim and Ohnsorge (216). Multipliers by fiscal instrument are from Spilimbergo, Symansky and Schindler (29). 26

Fiscal Multipliers and Fiscal Space: Larger When Fiscal Space is Wider 1.4 1.2 1..8.6.4.2. -.2 -.4 Fiscal multiplier: On impact (Change in GDP per unit of government spending) More fiscal space 23.1 32. 38.5 42.8 47.4 54.6 63.8 76.7 Government debt (percent of GDP) Source: Huidrom et al. (216). Note: The graphs show the conditional fiscal multipliers for different levels of fiscal position at select horizons. These are based on estimates from the IPVAR model, where model coefficients are conditioned only on fiscal position. Government debt as a percentage of GDP is the measure of fiscal position and the values shown on the x-axis correspond to the 15 th to 85 th percentiles from the sample. Fiscal position is strong (weak) when government debt is low (high). Solid lines represent the median, and dotted bands are the 16-84 percent confidence bands..4.3.2.1. -.1 -.2 -.3 Fiscal multiplier: After two years (Cumulative change in GDP per unit of government spending) More fiscal space 23.1 32. 38.5 42.8 47.4 54.6 63.8 76.7 Government debt (percent of GDP) 27

Why Does Fiscal Space Matter for Fiscal Multipliers? Because of Consumer and Investor Concerns Ricardian channel: When a government with little fiscal space implements stimulus, households anticipate tax hikes sooner. They save pre-emptively. This dampens private consumption (Perotti 1999). Interest rate channel: When a government with little fiscal space implements stimulus, investors become concerned about sovereign credit risk and raise their risk premia. This raises economywide borrowing cost and dampens private investment (Corsetti et al. 212). 28

Fiscal Multipliers and Fiscal Space: Because of Consumer and Investor Concerns.3.2.1. -.1 -.2 -.3 -.4 -.5 Impulse response of private consumption (Cumulative percent deviation from baseline) Narrow fiscal space Ample fiscal space 1 2 3 4 5 6 7 8 9 11112131415161718192 Quarters Impulse response of CDS spreads (Cumulative basis point deviation from baseline) Source: Huidrom et al. (216). Note: The graphs show the conditional fiscal multipliers for different levels of fiscal position at select horizons. These are based on estimates from the IPVAR model, where model coefficients are conditioned only on fiscal position. Government debt as a percentage of GDP is the measure of fiscal position and the values shown on the x-axis correspond to the 15 th to 85 th percentiles from the sample. Fiscal position is strong (weak) when government debt is low (high). Solid lines represent the median, and dotted bands are the 16-84 percent confidence bands. 7 6 5 4 3 2 1-1 -2-3 Narrow fiscal space Ample fiscal space 1 2 3 4 5 6 7 8 9 11112131415161718192 Quarters 29

References Huidrom, Kose, Lim and Ohnsorge (216). Do Fiscal Multipliers Depend on Fiscal Positions? CEPR Discussion Paper DP11346. Huidrom, Kose and Ohnsorge (forthcoming). Challenges of Fiscal Policy in Emerging and Developing Economies. Emerging Markets Finance and Trade. Kose, Kurlat, Ohnsorge and Sugawara (217). A Cross-Country Database of Fiscal Space. World Bank Working Paper 8157. Huidrom, Kose, Lim and Ohnsorge (217). Why Do Fiscal Multipliers Depend on Fiscal Positions? Mimeo. January 215 Global Economic Prospect Report: Having Fiscal Space and Using It. June 217 Global Economic Prospect Report: A Fragile Recovery.

Thank you! Questions & Comments Franziska Ohnsorge fohnsorge@worldbank.org