Austerity in the Aftermath of the Great Recession

Transcription

1 Austerity in the Aftermath of the Great Recession Christopher L. House University of Michigan and NBER. Christian Proebsting EPFL École Polytechnique Fédérale de Lausanne Linda Tesar University of Michigan and NBER. December 18, 2016

2 Real per Capita GDP US States 1

3 Research Design 1. Measuring austerity 2. Effects of austerity: Reduced-form estimates 3. Effects of austerity: Model-based estimates 2

4 Data Sample 28 European countries + US Unbalanced panel: Data sources: Eurostat, OECD, World Bank, BEA, National Statistical Agencies 3

5 Country Sample Country GDP Share Import Share XRT Regime Country GDP Share Import Share XRT Regime Belgium 2.6 % 31.3 % Euro Bulgaria 0.3 % 40.3 % Peg Germany 18.3 % 24.3 % Euro Denmark 1.7 % 26.9 % Peg Ireland 1.1 % 44.2 % Euro Estonia 0.1 % 42.2 % Peg Greece 1.9 % 25.5 % Euro Latvia 0.1 % 36.8 % Peg Spain 8.4 % 24.1 % Euro Lithuania 0.2 % 32.2 % Peg France 15.3 % 21.2 % Euro Czech Republic 1.0 % 37.1 % Floating Italy 12.5 % 22.0 % Euro Hungary 0.7 % 38.9 % Floating Cyprus 0.1 % 39.5 % Euro Poland 2.5 % 28.8 % Floating Luxembourg 0.2 % 56.9 % Euro Romania 0.9 % 28.2 % Floating Netherlands 4.1 % 21.0 % Euro Sweden 2.4 % 28.8 % Floating Austria 2.1 % 30.3 % Euro United Kingdom 15.4 % 23.9 % Floating Portugal 1.4 % 28.3 % Euro Norway 1.9 % 25.3 % Floating Slovenia 0.3 % 38.1 % Euro Switzerland 2.8 % 31.7 % Floating Slovak Republic 0.4 % 41.5 % Euro United States 91.5 % 13.3 % Floating Finland 1.3 % 27.5 % Euro RoW % 8.4 % Floating 4

6 Fiscal Austerity Fiscal measures Government purchases Total outlays Tax revenue and Tax rates Primary balance Identifying austerity Large changes in G or T (IMF) Narrative approach (Romer and Romer) SVAR approach (Mertens and Ravn) Forecast errors in G or T (Blanchard and Leigh) 5

7 Forecasting Government Purchase ln G i,t = ln G i,t 1 + g + γ ( ln ŶEU,t 1 ln Y i,t 1 ) + θ G ( ln Y i,t ln Ŷi,t) + ε G i,t g: EU-12 ave. GDP growth rate ( 93-05): 1.8% γ: GDP gap convergence: 3.8% θ G : Output elasticity: 0.38 (0.06) 6

8 Forecasting Government Purchases In-sample forecasts: t 2010 ln Ĝi,t = ln G i,t 1 + g + γ ( ln ŶEU,t 1 ln Y i,t 1 ) + θ G ( ) ln Y i,t ln Ŷi,t Out-of-sample forecasts: t > 2010 ln Ĝi,t = ln Ĝi,t 1 + g + γ ( ln ŶEU,t 1 ln Y i,t 1 ) + θ G ( ) ln Y i,t ln Ŷi,t 7

9 Forecasting Economic Performance GDP, Consumption and Investment ln Y i,t = ln Y i,t 1 + g Y + γ Y ( ln ŶEU,t 1 ln Y i,t 1 ) + ε Y i,t Inflation rate, Exchange rate, Net exports / GDP π i,t = π i, ε π i,t 8

10 Forecasts (1) 9

11 Forecasts (2) 10

12 Austerity & Economic Performance Average forecast error: X i, = t=2010 ( ln Xi,t ln X i,t ) Expressed in % of GDP for C, I and G X i, = α 0 + α G i, Γ controls i + ε i Controls: Forecast deviation of tax revenue in Debt-to-GDP prior to crisis (end of 2007) Change in TFP:

13 Austerity & Economic Performance X i, = α 0 + α G i, Γ controls i + ε i (1) (2) (3) (4) (5) Gov t. Purchases (0.36) (0.34) (0.31) (0.32) (0.31) Total Revenue (0.56) (0.49) Gov t Debt to GDP (0.02) (0.03) TFP (0.12) (0.14) R

14 Austerity & Economic Performance 13

15 All Countries By Currency Regime α R 2 α Fix α Fl R 2 GDP (0.31) (0.39) (0.57) Inflation (0.15) (0.20) (0.29) Consumption (0.27) (0.34) (0.49) Investment (0.23) (0.29) (0.42) NX to GDP (0.28) (0.34) (0.50) Exchange Rate (0.50) (0.46) (0.68) 14

16 Summary of Empirical Patterns 1. Low government purchases are correlated with (a)... low GDP, consumption, investment (b)... low inflation (c)... high net exports (d)... a depreciation of the exchange rate for floaters 2. Only weak relationship for most revenue measures 3. No evidence of strong spillover effects Spillover Analysis 15

17 Multi-Country DSGE Model Main Features Includes the 29 countries in the sample + RoW GHH preferences Rule-of-Thumb consumers Trade as in BKK (1994) Price rigidity (Calvo) Consumption, labor and capital taxes (const.) Financial accelerator (CF, BGG, Chicago Fed) Shocks: Government purchases, monetary policy, business credit spread 16

18 Households GHH preferences: U (c n, L n ) = 1 L 1+ 1 η n 1 1 c n κ n σ η 1 1 σ Rule-of-Thumb consumers A share χ of households spends a fixed share of their income on current consumption Aggregate consumption: c htm n,t C n Y n Y n,t C n,t = (1 χ) c n,t + χc htm n,t 17

19 18

20 18

21 18

22 18

23 18

24 18

25 18

26 18

27 Financial Market Imperferctions Nominal interest rate for business loans with (1 + i n,t ) F (λ n,t ) exp(ε F n,t) λ n,t = µ n,tk n,t NW n,t : Leverage ratio ε F n,t : Credit shock 19

28 Model Calibration Steady-state shares Relative country size Armington weights to match observed import shares Gov t purchases match observed GDP shares Effective tax rates (Mendoza et al., 1994) Share Rule of Thumb: 50% Elasticities Trade elasticity: 0.50 Price rigidity: 0.80 (quarterly) Taylor rule: φ GDP = 0.5, φ π = 1.5 Calibration 20

29 Model Simulation Government purchase shocks: As measured in empirical section Monetary policy shocks: Deviations from Taylor rule (intercept estimated for each monetary authority) Business credit spread shocks: Spreads between lending rates and central bank interest rates Simulated period: 2005:1-2014:4 Cross-sectional analysis: 2010:1-2014:4 21

30 GDP 22

31 Inflation 22

32 Net exports 22

33 Data Benchmark α α Fix α Fl α α Fix α Fl GDP (0.26) (0.31) (0.51) Inflation (0.13) (0.16) (0.25) Consumption (0.22) (0.27) (0.43) Investment (0.19) (0.23) (0.37) Net Exports over GDP (0.24) (0.27) (0.44) Exchange Rate (0.42) (0.37) (0.60) 23

34 Data G shocks only α α Fix α Fl α α Fix α Fl GDP (0.26) (0.31) (0.51) Inflation (0.13) (0.16) (0.25) Consumption (0.22) (0.27) (0.43) Investment (0.19) (0.23) (0.37) Net Exports over GDP (0.24) (0.27) (0.44) Exchange Rate (0.42) (0.37) (0.60) 23

35 Data Money shocks only α α Fix α Fl α α Fix α Fl GDP (0.26) (0.31) (0.51) Inflation (0.13) (0.16) (0.25) Consumption (0.22) (0.27) (0.43) Investment (0.19) (0.23) (0.37) Net Exports over GDP (0.24) (0.27) (0.44) Exchange Rate (0.42) (0.37) (0.60) 23

36 Model vs. Data Model matches cross-sectional dimension of data quite well! Rule-of-thumb restriction gives realistic consumption response Financial accelerator less successful in generating investment response Austerity shocks account for most of the cross-sectional variation in the data Money shocks important for response of inflation for floaters 24

37 Model Variation Data Bench- mark Low price rigidity Pass Mon Pol ZLB All GDP Inflation Fixed GDP Inflation Floaters GDP Inflation

38 Effect of a ZLB 26

39 Model Variation Data Benchmark No RoTh Sep Pref Hi Trade Elas No Trade All GDP Inflation Fixed GDP Inflation Floaters GDP Inflation

40 Model Mechanism Rule-of-Thumb consumers amplify NK multiplier through reductions in C Reductions in C lead to countervailing wealth effect on labor supply Kill off wealth effect through GHH preferences Openness lets G shocks leak out Low trade elasticity prevents exports to substitute for G 28

41 Shortfall as of 2014:4 Data Benchmark No Austerity ZLB ZLB + No Austerity Euro area GDP Inflation Floaters GDP Inflation GIIPS GDP Inflation EU10 GDP Inflation

42 Debt to GDP Debt / GDP (2009): EU10-80% GIIPS - 95% Did austerity reduce debt-to-gdp ratios? Static view: Simply sum up savings of reducing G Model: G affects GDP and hence, tax revenues Countries are linked through trade and monetary policy 30

43 Debt to GDP 31

44 Conclusion Observed divergent economic performance across Europe since Are policy reactions to blame? Measured austerity (shortfalls in G) associated with low economic activity, low inflation and positive net exports. Multi-country DSGE model with monetary responses, trade and foreign exchange regimes. Quantitative predictions of model consistent with cross-sectional and time-series patterns in the data. Output would have been 13% higher in GIIPS and 5% higher in EU10 by 2014 without austerity. 32

45 US States Back 33

46 Spillovers Following Auerbach and Gorodnichenko (2012): X i, = α 0 + α Gshock i + α Gshocki + ε i Domestic shock: Spillover shock: Gshock i = Y i Imp i Y i G i, Gshock i = N j i Imp i j Y j Y j Y i G j,

47 Spillovers Back 35

48 Description Parameter Value Source Preferences Discount factor (quarterly) β 0.99 Standard value 1 Coefficient of relative risk aversion σ 2 Standard value Frisch elasticity of labor supply η 1 Barsky et al. (1997) Share of rule-of-thumb consumers χ 0.5 Campbell and Mankiw (1989), Martin and Philippon (2014) Trade and Country Size Trade preference weights ωn j x OECD Trade in Value Added Dataset Trade demand elasticity ψ y 0.5 e.g. Heathcote and Perri (2002), Cravino (2014), Proebsting (2015) Country size N ny n x OECD Input-Output Tables Technology Capital share α 0.38 Trabandt and Uhlig (2011) Depreciation (quarterly) δ Average private investment share, X/Y = 0.197, Utilization cost a Del Negro et al. (2013) Investment adjustment cost f 2.48 Christiano et al. (2005) Elasticity of substitution between varieties ψ q 10 Standard value Price Rigidity Sticky price probability θ p 0.80 Alvarez et al. (2006) Financial Market Imperfections SS External finance premium F n(λ ss) x ECB, Global Financial Database and national sources Elasitcity external finance premium F ɛ 0.20 Brave et al. (2012) Persistence spread shock ρ F 0.99 Brave et al. (2012) Fiscal and Monetary Policy Gn Gov t purchases over final demand Yn x OECD and Eurostat Persistence government spending shock ρ G 0.93 Del Negro et al. (2013) Labor and Capital tax rates τ L, τ K x Own calculations based on National Tax Lists Consumption tax rates τ C x Own calculation based on Eurostat data Taylor rule persistence φ i 0.75 Taylor rule GDP coefficient φ GDP 0.5 Taylor rule inflation coefficient φ π 0.5 Back 36