Gernot Müller (University of Bonn, CEPR, and Ifo)

Exchange rate regimes and fiscal multipliers Benjamin Born (Ifo Institute) Falko Jüßen (TU Dortmund and IZA) Gernot Müller (University of Bonn, CEPR, and Ifo) Fiscal Policy in the Aftermath of the Financial Crisis Brussels, March 2-3, 212

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 1/33 1. Introduction Does the fiscal multiplier depend on the exchange rate regime? If so, how strongly? Exchange rate regime: float (with independent monetary policy) vs fixed rate (peg or monetary union w/o independent monetary policy) Multiplier: (short-run) increase of output triggered by exogenous increase of government spending by one percent of GDP

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 2/33 Issue remains vital in the aftermath of the crisis 24% Ireland Spain Greece UK 22% Government consumption (% of potential ouptut) 2% 18% 16% 14% 12% 23 24 25 26 27 28 29 21 211 212 213

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 3/33 Exchange rate regime impacts multiplier General insight: monetary policy matters for multiplier (Woodford 211, Christiano/Eichenbaum/Rebelo 211, Davig/Leeper 211) Mundell-Fleming model: public spending raises income and demand for domestic currency Peg: monetary policy accommodates demand, multiplier >> 1 Float: real appreciation crowds out net exports, multiplier =

Textbook treatment (Mankiw Macroeconomics ) Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 4/33

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 5/33 Evidence in support of MF-Model Estimates of multiplier in monetary union larger than consensus estimate of.5 1 US-states: 1.5 (Nakamura/Steinsson 211) Italian Provinces: 1.2 (Acconcia/Corsetti/Simonelli 211) Ilzetzki/Mendoza/Végh 211: estimate panel VAR for 44 countries, Blanchard-Perotti identification Float: Peg: 1.65 under peg (long-run)

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 6/33 Yet behavior of trade variables in response to higher government spending doesn t square with MF account Ilzetzki/Mendoza/Végh find weak evidence for MF channel Very short-lived exchange rate appreciation under float No significant movement of net exports Studies for US (and Canada, UK, Australia) find no evidence for real appreciation and/or (large) trade deficit Robust across identification schemes: Kim/Roubini 28, Monacelli/Perotti 21, Enders/Müller/Scholl 211, Corsetti/Meier/Müller 212

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 7/33 This paper 1. New time-series evidence using unique data set, which allows to control for anticipation Short-run multiplier under peg 1.2, under float.75 No evidence for MF transmission 2. New Keynesian small open economy model Can account for fiscal transmission under both exchange rate regimes Monetary policy is key, but transmission channel differs relative to MF

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 8/33 2. Vector autoregression (VAR) Bi-annual VAR model estimated on panel of OECD countries (1985:2 211:1): effects of unanticipated shocks to government spending Identification assumption: government spending predetermined Government spending is government consumption, not transfers With bi-annual observations more restrictive than Blanchard/Perotti 22, but not too restrictive (Born/Müller 212)

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 9/33 VAR model with spending forecast Vector of endogenous variables x i,t = [ g i,t fc t+1 i,t y i,t r i,t rx i,t nx i,t ] Reduced-form VAR model x i,t = µ i + K k=1 C k x i,t k + u i,t, Consider also country-specific time trends and country fixed effects Structural shocks: w t = Ωu t, with Ew t w t = I

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 1/33 Anticipation as a source of non-invertibility Shocks of structural model cannot be recovered from VAR Generic state-space representation of general equilibrium model z t+1 = Az t + Bw t+1 (1) x t+1 = Cz t + Dw t+1 (2) Rearranging: [ 1 (A BD 1 C )L ] z t+1 = BD 1 x t+1 If eigenvalues of (A BD 1 C ) smaller than 1, state can be written as function of past observables; obtain VAR( ) (Poor-man s invertibility condition of Fernández-Villaverde et al 27)

Classification of Ilzetzki/Reinhart/Rogoff 28 1. No separate legal tender 2. Pre announced peg or currency board arrangement P 3. Pre announced horizontal band that is narrower than or equal to +/-2% E 4. De facto peg G 5. Pre announced crawling peg 6. Pre announced crawling band that is narrower than or equal to +/-2% 7. De facto crawling peg 8. De facto crawling band that is narrower than or equal to +/-2% 9. Pre announced crawling band that is wider than or equal to +/-2% F 1. De facto crawling band that is narrower than or equal to +/-5% L 11. Moving band that is narrower than or equal to +/-2% O 12. Managed floating A 13. Freely floating T 14. Freely falling 15. Dual market in which parallel market data is missing Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 11/33

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 12/33 Peg sample Belgium Canada Czech Republic Denmark Finland France Germany Iceland Italy Luxembourg Netherlands Portugal Spain 1986 199 1995 2 25 21 Float sample Australia Canada Czech Republic Germany Iceland Japan Korea, Republic of Mexico New Zealand Norway Sweden Switzerland United Kingdom United States 1986 199 1995 2 25 21

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 13/33 Results for baseline specification Exogenous increase in government spending by one percent of GDP Compute impulse responses for 12 half years Report 9 percent confidence bounds Compare results for peg and float

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 14/33 Results for baseline model: peg 1.5 Government spending 2 Output Real exchange rate 1 1.5.5.5.2.1.1.2 4 8 12 Real interest rate.3 4 8 12 1.5.2.1.1.2 4 8 12 Net exports.3 4 8 12.5 4 8 12

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 15/33 Results for baseline model: float 1.5 Government spending 2 Output Real exchange rate 1 1.5.5.5.2.1.1.2 4 8 12 Real interest rate.3 4 8 12 1.5.2.1.1.2 4 8 12 Net exports.3 4 8 12.5 4 8 12

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 15/33 Results for baseline model: float vs peg 1.5 Government spending 2 Output Real exchange rate 1 1.5.5.5.2.1.1.2 4 8 12 Real interest rate.3 4 8 12 1.5.2.1.1.2 4 8 12 Net exports.3 4 8 12.5 4 8 12

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 16/33 Variable Horizon Peg Float Peg - Float Output 1 1.26.46.79 (.25) (.26) (.37) 3 1.19 (.36) 5 1. (.34) Real exchange rate 1.1 (.1) 3.3 (.11) 5.42 (.8) Real interest rate 1.2 (.1) 3.2 (.3) 5.1 (.1) Net exports 1.3 (.9) 3.2 (.7) 5.4 (.5).88 (.27).55 (.24).24 (.21).2 (.23).19 (.14).1 (.8).4 (.1).4 (.).2 (.6).2 (.7).4 (.6).32 (.47).46 (.42).14 (.23).32 (.26).61 (.17).3 (.12).6 (.3).5 (.1).2 (.1).4 (.1). (.8)

Sensitivity analysis Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 17/33

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 17/33 Sensitivity analysis Results are robust.

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 18/33 Pre-crisis sample: peg vs float 1.5 Government spending 2 Output Real exchange rate 1 1.5.5.5.2.1.1.2 4 8 12 Real interest rate.3 4 8 12 1.5.2.1.1.2 4 8 12 Net exports.3 4 8 12.5 4 8 12

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 19/33 Nominal rate specification: peg vs float 1.5 1.5.1.5.5 Government spending 4 8 12 Interest rate.1 4 8 12 Output 2 1.5 1.5 4 8 12 Inflation.3.2.1.1.2 4 8 12 1.5.5 Real exchange rate 1 4 8 12.2.1.1.2.3 Net exports 4 8 12

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 2/33 Alternative definition of peg: baseline vs tighter peg and EA countries 1.5 Government spending 2 Output Real exchange rate 1 1.5.5.5.2.1.1.2 4 8 12 Real interest rate.3 4 8 12 1.5.2.1.1.2 4 8 12 Net exports.3 4 8 12.5 4 8 12

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 21/33 3. Small open economy model New Keynesian economy (akin to Gaĺı/Monacelli 25) Imperfectly competitive firms produce country-specific varieties Pricing in producer currency, infrequent price adjustment Fraction of households (λ) without access to asset market (similar to Gaĺı/López-Salido/Vallés 27) Policies Exogenous government spending financed by debt and lump-sum taxes Monetary policy sets short-term interest rate to stabilize inflation or to fix exchange rate

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 22/33 Final goods and price indices C t = (1 ω) 1 σ +ω 1 σ ( [ 1 ( [ 1 Y H,t(j) ɛ 1 ɛ dj Y F,t(j) ɛ 1 ɛ dj ] ɛ ɛ 1 ] ɛ ) σ 1 σ ɛ 1 ) σ 1 σ σ σ 1 Price indices ( 1 P H,t = P t = [ (1 ω)ph,t 1 σ ) 1 P H,t (j) 1 ɛ 1 ɛ di + ωpf 1 σ ] 1 1 σ,t ( 1 P F,t = ) 1 P F,t (j) 1 ɛ 1 ɛ di Real exchange rate Q t = P te t P t

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 23/33 Firm s problem If allowed, adjust P H,t (j) so as to max E t k= ξ k Λ t,t+k Y t,t+k (j) (P H,t (j) Ω t,t+k ) Subject to demand function Y t,t+k (j) = Y t = (1 ω) ( ) PH,t (j) ɛ Y t+k P H,t+k ( PH,t P t ) σ C t + ω ( PH,t P t ) σ C t + G t And production function: Y t (j) = H t (j)

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 24/33 Household problem Asset holders (1 λ-fraction of population) solve subject to ( 1 γ C max E t β k A,t+k 1 γ H 1+ϕ ) A,t+k 1 + ϕ k= E t {ρ t,t+1 Ξ t+1 } Ξ t + P t C A,t = W t N A,t T t + Υ t Non-asset holders (λ-fraction of population) [ 1 γ C N,t max 1 γ H 1+ϕ ] N,t 1 + ϕ s.t. P t C N,t = W t H N,t T t

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 25/33 Fiscal policy Government consumption determined exogenously G t = (1 ρ)g + ρg t 1 + ε t Government budget constraint and tax rule: R 1 t D t+1 = D t + P H,t G t T t, T t = ψd t Monetary policy Floating exchange rate log (R t ) = log (R) + φ π (Π H,t Π H ) Fixed exchange rate log (R t ) = log (R t ) + φ E log (E t /E ), with φ E >

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 26/33 Model simulation Focus on log-linear approximation of equilibrium (around zero-inflation, zero-debt steady state) Analytical solutions for floating exchange rate and λ = Focus on numerical solution Calibration strategy Assign parameter values while matching impulse response functions (first 3.5 years after shock) under peg Assess model performance (informally) considering impulse responses under float

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 27/33 Model parameters Pre-set values Discount factor (steady state) β =.98 Risk aversion γ = 1 Import share ω =.35 Government-consumption share (steady state) G /Y =.2 Taylor-rule coefficient (float) φ π = 1.5 Tax-rule coefficient ψ =.21 Obtained by matching spending response Persistence spending peg ρ =.9 (.2) Persistence spending float ρ =.84 (.3) Matching IRF under peg Fraction of prices kept unchanged ξ =.75 (.5) Trade-price elasticity σ =.45 (.4) Constrained households λ =.32 (.39) Inverse Frisch elasticity ϕ =.62 (.56)

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 28/33 Effects of spending shock under peg: model vs VAR 1.5 Government spending 2 Output Real exchange rate 1 1.5.5.5.2.1.1.2 4 8 12 Real interest rate.3 4 8 12 1.5.2.1.1.2 4 8 12 Net exports.3 4 8 12.5 4 8 12

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 29/33 Effects of spending shock under float: model vs VAR 1.5 Government spending 2 Output Real exchange rate 1 1.5.5.5.2.1.1.2 4 8 12 Real interest rate.3 4 8 12 1.5.2.1.1.2 4 8 12 Net exports.3 4 8 12.5 4 8 12

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 3/33 Given empirical success, use model to address the following questions Why is the multiplier fairly small under peg? Intertemporal optimization implies declining expenditure of asset holders: PPP ensures that long-term rate rises with inflation Why is the multiplier relatively large under float?

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 31/33 Multiplier: the role of λ and φ π 1.15 1.1 1.5 Float ( = ) Peg ( = ) Float Peg 1 Multiplier.95.9.85.8.75 1 1.5 2 2.5

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 32/33 5. Conclusions Multiplier differs across exchange rate regimes regime, but Differences less stark than what MF model suggests Trade flows not in accord with MF model Variant of NK-model in line with VAR evidence Monetary policy accommodative under float Multiplier exceeds unity under peg only if λ >

Exchange rate regimes & multipliers 1. Introduction 2. Empirical analysis 3. Model 4. Conclusion 33/33 Fiscal policy in (the European) monetary union Potentially important stabilization tool Evidence for cross-country spillovers (Beetsma/Guiliodri/Klaasen 26) Austerity detrimental to economic activity, but Multiplier smaller than what MF suggests (or than in liquidity trap) Multiplier considerably smaller in the presence of sovereign risk (Corsetti/Kuester/Meier/Müller 212)