Shocks, frictions and monetary policy Frank Smets OECD Workshop Paris, 14 June 2007
Outline Two results from the Inflation Persistence Network (IPN) and their monetary policy implications Based on Altissimo, Ehrmann and Smets (2006) Why is interest rate setting different in the euro area and the US? Based on Sahuc and Smets (2006)
The New Keynesian Phillips curve t = γ pt 1 + ( 1 γ ) Et pt + 1 + κ( p mc p ) + t t u t Extrinsic inflation persistence: A key factor in determining the degree to which inflation inherits the persistence in its approximate determinants is the degree of price stickiness Intrinsic inflation persistence: ( mct pt Temporary shocks lead to a persistent response of inflation due to e.g. indexation or rule-of-thumb behaviour κ γ p t 1 )
IPN evidence : extrinsic persistence Prices change infrequently in the euro area and less often than in the United States Statistics Euro area US CPI Frequency 15.1 24.8 Average duration (months) 13.0 6.7 Median duration (months) 10.6 4.6 PPI Frequency 20.0 n.a. Surveys Frequency 15.9 20.8 Average duration (months ) 10.8 8.3 New Keynesian Phillips Curve Average duration (months) 13.5-19.2 7.2-8.4
IPN Evidence: Extrinsic persistence The difference in price stickiness with the US is consistent across sectors; However, the services sector is by far the most sticky sector: Unprocessed food food (oil industrial Processed Energy Non-energy Services Total Euro Area 28.3 13.7 78.0 9.2 5.6 15.8 United States 47.7 27.1 74.1 22.4 15.0 24.8
Price stickiness and competition Frequency of price reviews by degree of perceived competition a 80 70 60 50 40 30 20 10 0 Very low competition Low competition High competition Very high competition More than twelve Between four and twelve Less than four Frequency of price changes by degree of perceived competition a 50 45 40 35 30 25 20 15 10 5 0 Very low competition Low competition High competition Very high competition More than four Two or three One Incidental or none
IPN evidence: Intrinsic persistence Forward-looking behaviour is often dominant (e.g. Rumler, Jondeau and Lebihan, Gali, Gertler and Lopez-Salido), but there is a lot of uncertainty (e.g. Whelan) Estimates of the importance of the backward-looking component or the degree of intrinsic inflation persistence fall: in stable monetary policy regimes; with explicit measures of inflation expectations and allowing for learning Inflation expectations are not a source of inflation persistence in the euro area, because they are wellanchored to the inflation objective.
Conclusion: The NKPC t = γ pt 1 + ( 1 γ ) Et pt + 1 + κ( p mc p ) + t t u t Relatively high degree of extrinsic inflation persistence: high κ But relatively low intrinsic inflation persistence: low γ
Implications for monetary policy Embed NKPC in a fully structural model (Smets and Wouters, 2003). Analyse the response to a cost-push shock under different degrees of intrinsic and extrinsic inflation persistence, when the central bank minimizes inflation variability over the medium term. Small weight on output and interest rate variability; Optimal monetary policy under commitment
Monetary policy implications 1 0.2 0 Output gap 0.4 0.3 Interest rate Response to a cost-push shock under different degrees of intrinsic Inflation persistence (Smets-Wouters) -0.2 0.2-0.4 0.1-0.6 Low Medium High -0.8 0 5 10 15 20 0-0.1 0 5 10 15 20 1.5 Inflation 0.5 Real rate 1 0 0.5 0-0.5-0.5 0 5 10 15 20-1 0 5 10 15 20
Monetary policy implications 2 0.1 0 Output gap 0.4 0.3 Interest rate Response to a cost-push shock under different degrees of price stickiness (Smets-Wouters) 0.2-0.1 0.1-0.2 Low Medium High -0.3 0 5 10 15 20 0-0.1 0 5 10 15 20 1 Inflation 0.6 Real rate 0.4 0.5 0.2 0 0-0.2-0.5 0 5 10 15 20-0.4 0 5 10 15 20
Conclusions High price stickiness reduces the impact effect of a cost-push shock on the output gap, but increases the persistence of the output gap response following such shocks; This could partly explain why recessions are shallower, but more persistent: 1993-2004 Euro area United States Real GDP growth Standard deviation 1.46 1.96 First order autocorrelation 0.49 0.18
Conclusions: monetary policy implications Low intrinsic inflation persistence and high extrinsic persistence leads to a more gradual interest rate response to cost-push shocks Can this difference in the structure of the euro area and US economy explain the difference in interest rate setting?
Interest-rate setting in EA and US
Sahuc and Smets (2007) Use an estimated DSGE model (Smets-Wouters, 2003, 2007) for distinguishing between three hypotheses: Difference in estimated policy rule: Related to differences in objectives, monetary policy strategy and/or institutional set-up Differences in the structure of the economy: Interest rate sensitivity of demand, structural rigidities in goods and labour markets Differences in the source and persistence of economic shocks: E.g. different productivity developments
Smets and Wouters (2007) model
Smets-Wouters (2007) model
Smets and Wouters (2007) Seven identified structural shocks: productivity, labour, preference, investment, cost-push, exogenous spending, monetary policy Shocks are assumed to follow a first-order autoregressive process with the exception of the cost-push shocks. Model is estimated separately for the euro area and the United States over the period 1985:1-2004:4.
Has the policy reaction function changed with EMU? No! Excluding monetary policy shocks basically gives the same path
Is the policy reaction function different? No: EA US
Is it a different economic structure? There is some evidence that the euro area economy is more rigid; The main estimated structural differences are: The habit formation parameter is estimated to be higher 0.71 versus 0.63; The investment adjustment costs are higher: 8.3 versus 7.6; The degree of price stickiness is also higher, consistent with the IPN findings: 0.008 versus 0.012; However, those differences seem to have only a limited impact on the transmission process
Is it a different economic structure?
Is it a different economic structure? No.
It s different shocks
It s different shocks
It s different shocks
Conclusions Prices appear to be more sticky in the euro area than in the US increasing the output costs of maintaining price stability; Further progress in structural reform and in particular a deregulation of the services markets would increase the resilience of the euro area economy to cost-push shocks. Differences in interest rate setting between the euro area and the US appear, however, mostly to be due to differences in shocks, not in structure or policy.