IMF Fiscal Affairs and Research Departments Workshop on Fiscal Policy Washington, June 2, 2009 A couple of issues in automatic stabilisation: Efficiency effects and medium-term sustainability by: Carlos Martinez-Mongay(*) European Commission DG ECFIN (*) The views expressed in this presentation do not necessarily represent those of the European Commission European Commission 1 OUTLINE 1. Distortionary taxes, the size of governments and the trade-off between stabilization and efficiency 2. Asset booms, composition effects and the trade-off between stabilization and sustainability European Commission 2 1. Distortionary taxes, the size of governments and the trade-off between stabilisation and efficiency European Commission 3 A simple model for automatic stabilisation (European Commission (2001); se also Artis and Buti, 2000, Blanchard, 2000, Buti, Roeger, in’t Veld, 2001) A simple AD/AS model: (1) Yd = f1(d-pb) – f2(i-pe) + ed (2) Ys = Y* + w(p –pe) + es (3) d = d* - t(Y –Y*) f1 , f2 > 0 (1) IS-type schedule (2) Lucas-Pillips supply function d deficit b stock of public debt i interest p inflation (‘e’ expected) Y* potential output (Y – Y* is the output gap) ed, es, demand and supply shocks (3) d* cyclically-adjusted deficit t automatic stabilisers (sensitivity of the deficit to the output gap, Y-Y*) European Commission 4 Shocks don’t come with labels: demand, supply, temporary, permanent Y = Y* + (w ed + bes)/[w(1 + f1 t ) + b] p = p* + (ed – (1 + f1 t ) es)/[w(1 + f1 t ) + b] If automatic stabilisers are turned off (t = 0) Y = Y* + (w ed + bes)/(w + b); es = 0, Y = Y* + w ed /(w + b); ed = 0, Y = Y* + bes/(w + b); p= p* + (ed – es)/(w + b) es = 0, p= p* + ed/(w + b); ed = 0, p= p* – es/(w + b); If automatic stabilisers are turned on (t >0) es = 0 Y = Y* + w ed /[w(1 + f1 t ) + b]; ed = 0 Y = Y* + bes/[w(1 + f1 t ) + b]; es = 0 p = p* + ed /[w(1 + f1 t ) + b]; ed = 0 p = p* - (1 + f1 t ) es)/[w(1 + f1 t ) + b]; es permanent Y – (Y* + es )= -{(1 + f1 t ) es/[w(1 + f1 t ) + b]}; European Commission 5 Do governments stabilize output? Automatic stabilizers operate on the demand side. Progressive taxes and benefit (viz. unemployment) systems make surpluses grow more than output in expansions and less than output in recessions Automatic stabilizers also operate on the supply side: • Avoid under-investment in recessions • Keep levels of physical and human capital • Growth close to potential But distortionary taxes and benefit systems affect agent’s decisions to invest, work and save, which has an impact on economic efficiency. A policy dilemma? • • Need to induce more flexibility in markets (short-run adjustment) and enhance potential output (long-run adjustment) by streamlining benefit systems and lowering distorting taxes Efficiency gains may come at the cost of less stability (more output volatility) No policy dilemma if distortionary tax-benefit systems would hamper macroeconomic stability European Commission 6 Higher tax rates can lead to steeper aggregate supply curves (1) w = f ( L) T (2) (1 w (3) = w (4) u =u* T T ) = (1 ) 1 L w w (1 t ) 1 L (1 t ) (1 t ) (p p e ) (1 t ) European Commission 7 An extended AD/AS model with taxes in the supply curve Aggregate-demand curve (1) y d = f1d f2 i p e f3 p p * f4 y y * e d Aggregate-supply curve (2) y s = w t p p e e s Monetary rule: (3) where i = p y p = p 1 p * and y = y 1 y * Fiscal rule (SGP –automatic stabilisation) (4) d = ty European Commission 8 An extended AD/AS model with taxes in the supply curve Trade balance implies: consistency p p * = y y * f5 (6) y= where f5 f = f4 f3 f3* * 4 w t * w t d f 2 s f6 y e e rt rt rt rt = f2 w t 1 f6 f1t f2 . and f6 = f2 1 (f5 ) f4 f3f5 Inflation can be obtained by equating (6) to (w – t)p + es (i.e. (2) with pe = 0) European Commission 9 Under a demand shock (ed0, es=0). r w * *t * f6f6* f2 f y = e d t 1 2 t rt* w t w t 2 p ed = t w t 2 C 2 w * *t * f6*f6 2 f t 1 f f wf 1 6 2 1 rt* rt w * *t * f6f6* rt* w t where C = European Commission 10 Under a supply shock (es0, ed=0). r * r w t w * * t * f 6f 6* w * * t * f 6*f 6 y = f 2 e s t t f ( w t ) ( 1 f f t f ) 1 6 1 2 t rt* rt* 2 f 2e s (f2 B) 2 p 2 = f1 (w t ) t (w t ) 2 B 2 f2 where r *r w t w * *t * f6f6* B= t t rt* European Commission 11 Is there an efficiency-stabilization tradeoff? Table 1 Effect of a rise in taxes on output and inflation stabilisation output inflation Below tˆ Above tˆ demand shock + supply shock + + + Initial level of taxes Below tˆ Above tˆ Shocks Note: stabilising effect destabilising effect European Commission 12 Is there an efficiency-stabilization tradeoff? The cases of closed and small economies * * * * * * * ˆt = wf1 (1 f2 ) f6 . f2 (w t )(*1 f1 t f2 ) 2f1 2f1 rt tˆC = wf 1 (1 f 2 ) 2f1 wf1 (1 f 4 ) tˆS = 2f1 European Commission 13 Is there an efficiency-stabilization tradeoff? Empirical evidence Sample of 25 OECD countries over the period 1960-2000 (annual data). Joint estimation of output and inflation volatility (error terms are correlated; GMM) Output volatility: standard deviation of the output gap (% of the trend) over the period Inflation volatility: ratio between the standard deviation and the average of the GDP deflator Government size is measured as tax revenues (% of GDP) Other variables: Trade openness: Average of exports and imports in % of GDP Country size: Average of the GDP (PPS) Specialisation: Average share (%) of manufacturing GVA in total GDP Additional Instruments (determinants of government size): Per capita income (in 1995 PPS) Old-age dependency ratio: people over 65 in percentage of total population European Commission 14 Alternative ways of measuring the size of governments Table 3: Correlation between volatility, government size, openness and country size Output Inflation stability stability (1) (2) Inflation stability Total expend. Current expend. Total revenue Tax revenue Trade Openness Country Size (8) Total Current Total Tax Trade expend. expend. revenue revenue openness (3) (4) (5) (6) (7) 0.54* -0.45* -0.41* -0.50* -0.39* 0.99* -0.40* -0.36 0.95* 0.93* -0.44* -0.38 0.94* 0.91* 0.96* -0.03 -0.12 0.40* 0.34 0.46* 0.45* -0.35 -0.15 -0.21 -0.19 -0.29 -0.26 -0.71* * Significant at 5%. Asymptotic critical value between 2x(1/25)0.5 = 0.40 and 2x(1/22)0.5 = 0.43 (1) Standard deviation of the output gap in percentage of trend GDP over 19602000(a) (2) The ratio between the standard deviation and the average over 1960-2000(a) of the annual percentage change in the GDP deflator (3) Logarithm of the average total expenditures (% of GDP) over 1960-2000(a) (4) Logarithm of the average current expenditures (% of GDP) over 1960-2000(a) (5) Logarithm of the average total revenues (% of GDP) over 1960-2000(a) (6) Logarithm of the average tax revenues (% of GDP) over 1960-2000(a) (7) Logarithm of the average exports and imports (half % of GDP) over 19602000(a) (8) Logarithm of the average GDP (PPPs) over 1960-2000(a) (a) See footnotes in tables 1 and 2 for the exceptions. European Commission 15 Stabilization and the tax burden Table 5. Instrumental variable estimates for macroeconomic stability Tax Country revenues (1) size (2) Industry share (4) Adj. R² J-test (4) [Output-gap volatility(5)] OG-IV1(6) OG-IV3(6) -1.74 (0.26)*** -1.94 (0.37)*** -0.21 (0.07)*** -0.33 (0.05)*** 0.43 0.05 (0.03)** 0.51 2.98 [0.39] 5.45 [0.14] [Inflation volatility (7)] GD-IV1(6) GD-IV3(6) -0.22 (0.07)*** -0.23 (0.08)*** -0.03 (0.01)** -0.03 (0.01)** 0.20 0.002 (0.003) 0.19 3.48 [0.32] 4.24 [0.24] Heteroskedastic-consistent standard errors in parenthesis; *** significant at 1%; ** significant at 5%, * significant at 10%. All the regressions include an intercept not shown here (1) Logarithm of the average tax revenues (% of GDP) over 1960-2000(a) (2) Logarithm of the average GDP (PPPs) over 1960-2000(a) (3) Average share of manufacturing gross value added in total GDP over 1960-2000 (4) Test for over-identifying restrictions; p-value between ‘[]’ (5) Standard deviation of the output gap in percentage of trend GDP over 1960-2000(a) (6) The instruments are trade openness (% of GDP), per capita GDP (1995 PPPs) and the dependency ratio (% of total population). The three instruments are the logs of the averages over 1960-2000. (7) Ratio between the standard deviation and the average over 1960-2000(a) of the annual percentage change in the GDP deflator (a) See footnotes in tables 1 and 2 for the exceptions European Commission 16 The average effective tax rate on labor income LETR = (SSC + PITR*(LETB – NWRV))/LETB Where: SSC social security contributions plus taxes on payroll and workforce PITR*(LETB – NWRV) personal taxes on labour, with PITR = PIRV /(COEL + NOS – NWRV – CORV – PWRV) LETB is gross wages and salaries plus the imputed income of the self-employed European Commission 17 Labor taxes in the sample Table 6: Labor related tax burden over time and countries Period 1960-1980 Country Belgium Denmark Germany Greece Spain France Ireland Italy Luxembourg The Netherlands Austria Portugal Finland Sweden United Kingdom United States Japan Direct Social Labour taxes security taxes 1 2 3 11.62 11.31 34.28 24.06 1.81 32.35 11.33 13.03 35.21 3.30 7.67 14.94 3.71 8.76 17.45 6.20 15.18 29.86 9.41 4.88 14.89 6.14 11.38 22.90 15.36 11.53 30.16 14.45 16.05 36.90 10.23 11.20 29.09 4.86 5.30 14.40 13.35 6.78 31.17 18.42 9.06 41.23 15.12 6.33 24.94 13.29 4.60 18.35 8.30 4.72 13.15 Period 1980-2000 Difference Direct Social Labour taxes security taxes 1 2 3 16.83 16.26 42.90 28.70 2.55 41.32 11.87 17.90 40.71 6.27 11.94 23.40 9.26 13.01 28.54 8.90 20.11 39.30 13.70 6.85 24.65 13.93 14.10 34.88 17.06 12.54 33.53 13.58 18.29 40.92 12.42 16.23 37.31 8.66 9.87 24.08 17.25 12.63 41.42 21.18 14.33 47.67 15.95 7.83 25.35 13.09 7.00 22.00 11.12 8.94 20.14 Direct Social Labour taxes security taxes 1 2 3 5.21 4.95 8.62 4.64 0.74 8.97 0.54 4.87 5.49 2.97 4.26 8.46 5.55 4.25 11.09 2.69 4.93 9.45 4.29 1.97 9.76 7.79 2.72 11.98 1.70 1.01 3.37 -0.88 2.24 4.03 2.18 5.03 8.22 3.80 4.57 9.68 3.90 5.85 10.25 2.76 5.27 6.44 0.83 1.50 0.41 -0.20 2.40 3.65 2.82 4.22 6.99 1:Direct taxes in percentage points of GDP 2:Social security contributions in percentage points of GDP 3:Labour effective tax rates in % of the tax base European Commission 18 Is there an efficiency-stabilization tradeoff? Empirical evidence Table 7: Macroeconomic stability and the tax mix Output-gap volatility (1) Price volatility [GDP deflator] (2) Government Size (3) -2.44 -4.9 -2.58 -2.55 (0.56)*** (1.41)*** (0.62)*** (0.59)*** -0.31 (0.18)* Country size (4) -0.23 -0.34 -0.29 -0.29 (0.05)*** (0.08)*** (0.09)*** (0.09)*** -0.03 (0.03) -0.04 (0.02)* -0.04 (0.02)** -0.05 (0.02)** 0.07 (0.22) 0.21 (0.12)* 0.23 (0.11)* 0.23 (0.12)** Specialization (5) 1.43 (0.75)* Government Size* Time dummy (6) Government Size* Labor rate (7) Government Size* Direct taxes (8) 0.20 (0.09)** Government Size* Social Security (9) A-R² (10) J (11) 1.04 (0.46)** 0.85 (0.49)* 0.86 (0.49)* -0.17 -0.46 -0.43 (0.36) (0.14)*** (0.14)*** 0.05 (0.03)* 1.45 (0.81)* -0.16 (0.21) 0.03 (0.30) 0.05 (0.06) -0.03 (0.30) 0.34 3.76 [0.43] 0.64 7.79 [0.45] 0.56 8.53 [0.38] 0.56 8.80 [0.36] 0.00 (0.07) 0.27 3.76 [0.43] 0.39 7.79 [0.45] 0.39 8.53 [0.38] Models estimated by IV methods (instruments: trade openness, per capita GDP, dependency ratio. Heteroskedastic-consistent standard errors in parenthesis; *** significant at 1%; ** significant at 5%, * significant at 10%. All the regressions include an intercept not shown here. (1) Standard deviation of the output gap in percentage of trend GDP over 1960-2000(a). (2) The ratio between the standard deviation and the average over 1960-2000(a) of the annual percentage change in the GDP deflator. (3) Logarithm of the average tax revenues (% of GDP) over 1960-2000(a). (4) Logarithm of the average GDP (PPPs) over 1960-2000(a). (5) Average share of manufacturing gross value added in total GDP over 1960-2000. (6) The corresponding equation is estimated on a sample consisting of averages over the period 1960-1980 and the period 1981-2000. The resulting number of observation is 30. Government size is interacted with a dummy taking the value 1 over the second period and 0 otherwise. (7) The corresponding equation is estimated on a sample consisting of averages over the period 1960-2000. The resulting number of observation is 17. Government size is interacted with labor taxes in percentage of total labor costs. (8) The corresponding equation is estimated on a sample consisting of averages over the period 1960-2000. The resulting number of observation is 17. Government size is interacted with direct taxes in percentage of GDP. (9) The corresponding equation is estimated on a sample consisting of averages over the period 1960-2000. The resulting number of observation is 17. Government size is interacted with social security contribution in percentage of GDP. (10) Adjusted R². (11) Test for over-identifying restrictions; p-value between ‘[]’. 0.38 8.80 [0.36] European Commission 19 Is there an efficiency-stabilization tradeoff? Additional empirical evidence (Debrun et al, 2008) European Commission 20 Is there an efficiency-stabilization tradeoff? Additional empirical evidence (Debrun et al, 2008) European Commission 21 2. Asset booms, composition effects and the trade-off between stabilization and sustainability European Commission 22 Measuring automatic stabilizers in the presence of composition effects Tax revenue growth % (d) Tax base growth (c) (b) Nominal GDP growth (a) Trend GDP growth Period of analysis T0 T1 t a) Cyclical effect (growth at nominal GDP) (b) Composition effect (growth at base rate) (c) Discretionary measures (d) Others European Commission 23 Automatic stabilizers and asset booms. The case of Spain The longest expansion of the Spanish economy… Real GDP growth in Spain and in the Euro Area 10 8 6 % 4 2 0 -2 -4 -6 70 72 74 76 78 80 82 Spain 84 86 88 Euro area 90 92 94 96 98 00 02 04 06 08 10 Average diff. Spain-Euro area Source: AMECO European Commission 24 Automatic stabilizers and asset booms. The case of Spain … domestically led, especially by the housing sector (and consumption). SPAIN-INVESTMENT CONTRIBUTION TO GDP GROWTH 1.80 1.60 1.40 % 1.20 1.00 0.80 0.60 0.40 0.20 0.00 1971-1975 EQUIPMENT 19982-1991 DWELLINGS OTHER CONSTRUCTION 1995-2007 OTHER INVESTMENT Source: AMECO and own calculations European Commission 25 Automatic stabilizers and asset booms. The case of Spain The expansion 1995-2007 was underpinned by a substantive fall of the risk premium, and unprecedented demographic growth, … % % Interest rates, exchange rates and demography 2.0 12 1.8 10 1.6 8 6 1.4 4 1.2 2 1.0 0 0.8 -2 0.6 -4 Population Real short term interest rate (rhs) 2008 2007 2006 2005 2004 2003 2002 2001 2000 1999 1998 1997 1996 1995 1994 1993 1992 1991 1990 -10 1989 0.0 1988 -8 1987 -6 0.2 1986 0.4 Reer (rhs) European Commission 26 Automatic stabilizers and asset booms. The case of Spain …which induced a boom in asset markets 500 800 000 450 700 000 Index base 1995=100 400 600 000 350 500 000 300 250 400 000 200 300 000 150 200 000 100 100 000 50 0 0 1990 19911992 1993 19941995 1996 19971998 1999 20002001 2002 20032004 2005 20062007 2008 Housing starts in Spain (rhs) Spanish housing prices Euro area housing prices Spanish stock market (IBEX-35) Euro area stock market (Dow Jones Euro STOXX) 27 European Commission Automatic stabilizers and asset booms. The case of Spain Asset markets have also driven the downturn SPAIN-INVESTMENT CONTRIBUTION TO GDP GROWTH 0.50 0.00 -0.50 -1.00 % -1.50 -2.00 -2.50 -3.00 -3.50 -4.00 -4.50 2008 EQUIPMENT 2009 DWELLINGS 2010 OTHER CONSTRUCTION OTHER INVESTMENT Source: AMECO and own calculations European Commission 28 A tax-rich expansion, 6 5 % GDP 4 3 2 1 0 TOTAL IND. TAXES DIR. TAXES SOC.SEC. -1 1995-2007 1995-2008 3.5 2.0 3 1.5 2.5 2 % GDP 1.0 1.5 0.5 1 0.5 0.0 0 -0.5 INDIRECT TAXES HOUSING VAT OTHER VAT OTHER -0.5 DIR TAXES Diff %GDP 2007-1995 Diff %GDP 2008-1995 CORPORATE Diff %GDP 1995-2007 PERSONAL Diff %GDP 1995-2008 European Commission 29 Automatic stabilizers and asset booms. The case of Spain … apparently of a structural nature… General Government Tax Revenues (indirect and direct taxes and social contributions, ïn %GDP) 38.0 1.0 37.0 0.5 36.0 0.0 35.0 -0.5 34.0 -1.0 33.0 -1.5 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 32.0 Cyclical component (RHS) Cyclically-adjusted tax revenues Tax revenues European Commission 30 Automatic stabilizers and asset booms. The case of Spain …but probably just reflecting a significant divergence between national accounts and ‘actual’ tax bases, especially in indirect and corporate taxes 103 TAX BASES FOR INDIRECT TAXES TAX BASES FOR CORPORATE TAXES 350 800 101 300 700 200 93 150 91 100 89 87 50 600 Index 1991=100 95 Index 1980=100 250 97 500 400 300 200 100 19 80 19 82 19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04 20 06 20 08 Index 1980=100 99 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Private consumption (%GDP) Private consumption and dw ellings (%GDP) Dw ellings (%GDP)( (rhs) CORPORATE PROFITS NET OPERATING SURPLUS Source: AMECO and own calculations European Commission 31 Automatic stabilizers and asset booms. The case of Spain which led to large composition effects Decomposing changes of indirect taxes and selected categories of indirect taxes (% of GDP) 5.0 3.5 4.0 %of 1991 and 2006 GDP respectively 4.0 3.0 2.5 in % of 2006 GDP Decomposing changes of direct taxes and selected categories of direct taxes (% of GDP) 2.0 1.5 1.0 0.5 0.0 3.0 2.0 1.0 0.0 -1.0 -2.0 -0.5 -3.0 -1.0 1982-1991 Indirect taxes 82-91 Indirect taxes 95-07 VAT Housing 95-07 Other VAT 95-07 1995-2007 Composition Composition Others Personal 95-07 Corporate 95-07 Other 95-07 Others Discretionary Discretionary European Commission 32 Automatic stabilizers and asset booms. The case of Spain … explaining between 2/3 and ¾ of the accumulated increase in tax revenues % GDP Estimations of excess collection associated with the composition effect 6.0 5.0 4.0 3.0 2.0 1.0 0.0 4.4 3.0 5.0 3.6 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Estimation by econometric model Observed changes ∆REVt = d0 + α 1(REV – a - β GDP - t TREND)t-1+ δ1 ∆GDPt + δ2 ∆GDPt-1 + δ1 ∆(HOUSE/GDP)t-2 + δ2 ∆(MGVA/GDP)t + g1 I1t + … + gm Imt + εt (1) European Commission 33 Automatic stabilizers and asset booms. The case of Spain Enough fiscal consolidation in good times? 40 6 39 4 38 37 2 35 0 % GDP % GDP 36 34 -2 33 32 -4 31 30 -6 1995 1997 1999 CA tax revenues (lhs) 2001 2003 CA balance (rhs) 2005 2007 Interest Payment (rhs) European Commission 34 Automatic stabilizers and asset booms. The case of Spain A sustainable fiscal expansion? PUBLIC DEFICIT AND DEBT 65 5 3 60% 60 1 55 -1 50 -3% -3 45 -5 40 -7 35 -9 forecast -11 2006 2007 2008 Public surplus/deficit (% GDP, lhs) 2009 30 2010 Public debt (% GDP, rhs) European Commission 35 References Andrés, J. R. Domenech and A. Fatás, 2004, The Stabilising Role of Government Size, ISEAD, 2004/44/EPS Artis, M. J. and Buti, M., 2000, « Close to balance or in surplus » -A policy maker’s guide to the implementation of the Stability and Growth Pact, Journal of Common Market Studies, 38(4), 563-592. Blanchard, O., 2000, Commentary, Federal Reserve Bank of New York Economic Policy Review, 6(1), 69-74. Brunila, A., M. Buti and J. in’t Veld, 2001, Fiscal policy in Europe: how effective are automatic stabilisers?, European Economy, Economic Papers, N° 177. Buti, M., W. Roeger, J. in’t Veld, 2001, Stabilising output and inflation: Policy conflicts and coordination under a stability pact, Journal of Common Market Studies, 39, 801-828. Buti, M. C. Martinez-Mongay, K. Sekkat and P. van den Noord, 2003, Automatic Fiscal Stabilisers in EMU: A Conflict between Efficiency and Stabilisation, SESifo Economic Studies, VoL. 49, 123-140. (see also OECD Ecocomics Department Working Papers, N° 335). Debrun, X., J. Pisani-Ferry and A. Sapir, 2008, Government size and output volatility: Should we forsake automatic stabilisation? European Economy Economic Papers N° 316. Gali, J., 1994, Government size and macroeconomic stability, European Economy Review, 38, 117-132. Fatás, A. and I. Mihov, 2001, Government size and automatic stabilisers, Journal of International Economics, 55, 3-28. Martinez-Mongay, C. and K. Sekkat, 2005, Progressive Taxation, Macroeconomic Stabilization and Efficiency in Europe, European Economy Economic Papers, N° 233 Martinez-Mongay, C. J. L. Maza Lasierra and J. Yaniz Igal, 2007, Asset Booms and Tax receipts: The case of Spain, 1995-2006. Roeger, W. and J. in’t Veld, 2009, Fiscal Policy with Credit Constrained Households, European Economy Economic Papers, N° 357. European Commission 36