MONETARY POLICY IN EMU

1
Estimating potential output growth in the euro
areaAlberto MussoDirectorate General
EconomicsEuropean Central BankPresentation for
AIECE Workshop on Estimating potential output in
the European Union29 October 2004
2
Structure of presentation

• Framework, measurement and policy issues
• Main model and basic estimates
• Developments in factors of production
• Conclusions

3
Definition and policy relevance

• Definition sustainable (i.e. non-inflationary)
  rate of growth of aggregate real output over the
  medium-term
• Monetary analysis assumption for potential
  output growth is an input to the derivation of
  the reference value for monetary growth
• Economic analysis measures of slack among
  indicators monitored to assess the risks to price
  stability in the short to medium run
• See ECB The Monetary Policy of the ECB, 2004

4
Role in ECB monetary policy strategy
potential output growth as input in the
derivation of the reference value
including slack indicators
5
ECB review of the reference value

• In 1998 medium-term trend real potential GDP
  growth estimated in the range 2-2.5 per annum
• Over 1999-2002 assumptions reviewed, but no new
  decisive evidence emerged pointing to a
  significant change in assumption
• In May 2003 Governing Council decided to no
  longer review assumptions on an annual basis but
  their validity still monitored regularly
• Any changes in assumptions will be communicated
  as soon as they become necessary
• See ECB Press Releases December 1998-2002 and May
  2003

6
Issues in the estimation of PO growth

• Measurement inevitably surrounded by considerable
  uncertainty, especially in real time and looking
  ahead, stemming from various sources
• model uncertainty
• parameter uncertainty
• data uncertainty
• From ECB perspective
• Relevant time horizon is medium run
• Euro area aggregate perspective
• See ECB Monthly Bulletin article entitled
  Potential output and output gaps concept, uses
  and estimates, October 2001

7
The approach of the ECB

• Given the uncertainty characterising unobserved
  PO an encompassing approach is warranted
• assess estimates from alternative methods and
  institutions
• assess developments in key factors of growth
• define trend PO growth as the range that
  captures medium term developments

8
A basic analytical framework

• Economic structure provided by use of a
  Production Function Approach combined with a
  generalised Phillips-curve
• Statistical inference made possible by applying
  Unobserved Components estimation techniques
• Advantages
• consistent with the definition of PO
• general and flexible
• allows for contribution of inputs to be
  estimated
• can get a well-specified statistical model
• can estimate confidence bands
• See T. Proietti, A. Musso and T. Westermann
  (2002) Estimating potential output and the
  output gap for the euro area a model-based
  production function approach, EUI working paper
  ECO 2002/09, May 2002

9
The production function approach

• Cobb-Douglas production function with total
  factor productivity, capital and labour as inputs
• Labour decomposed into working age population,
  the participation rate and the contribution from
  the unemployment rate
• Capital stock and working age population assumed
  to have a permanent but no transitory component
• Thus, measurement model consists in decomposing
  TFP, labour force participation and the
  unemployment rate into respective permanent and
  transitory component
• For evidence supporting the Cobb-Douglas
  production function for the euro area see A.
  Willman "Euro area production function and
  potential output a supply side system approach",
  ECB working paper 153, June 2002

10
The inflation equation

• The measurement model is augmented with a
  Phillips-type of relationship
• Inflation is modelled as a function of the output
  gap, lagged inflation, commodity prices and euro
  exchange rate
• Consistent with Gordons triangle model, where
  inflation has three main determinants excess
  demand, inertia, and supply shocks
• See R. Gordon (1998) Foundations of the
  Goldilocks Economy Supply Shocks and the
  time-varying NAIRU, Brookings Papers on Economic
  Activity, 2 1998

11
Statistical representation

• Trend and cycles modelled along the lines of the
  multivariate structural time series model of
  Harvey and Koopman (1997)
• Cyclical components are allowed to be correlated,
  while trends are assumed to be uncorrelated
• Trends modelled as local linear trends
  (stochastic trends)
• Cyclical components modelled as functions of
  AR(2) processes with complex roots
• See Harvey, A. and S. Koopman (1997)
  Multivariate structural time series models, in
  C. Heij et al. (eds), Systematic dynamics in
  economic and financial models, Chichester Wiley
  and Sons

12
Variants

• Multivariate variants considered
• A) common cycle model
• all cyclical components driven by the cycle in
  the capacity utilisation rate
• B) pseudo-integrated cycles model
• cyclical components driven by a combination of
  common and specific cycles
• C) hysteresys model
• trends function also of cycles

13
Estimation and inference

• System approach for the five variables
• All components estimated simultaneously within a
  multivariate unobserved components model that
  incorporates the relationships specified (PFA and
  PC)
• Model cast in state space form, and Kalman filter
  and associated smoother algorithms enable maximum
  likelihood estimation and signal extraction
• Computations performed using Ox 3.0 and SsfPack
  2.3

14
Work in progress

• Planned extensions
• more general production function (CES, other
  inputs)
• more flexible specification of inflation equation

15
Table 1 - Alternative estimates of euro area
average potential output growth
(Annual percentage changes, average over period
indicated)
Sources ECB computations and IMF World Economic
Outlook September 2004 OECD Economic Outlook
June 2004 European Commission Autumn Economic
Forecasts October 2004.
16
(No Transcript)
17
Comments on estimates

• Long run average of all estimates (1980-2004)
  tends to support the range of 2-2½ assumed by
  the ECB
• Potential output growth estimates tend to vary
  across method, although less so over longer
  periods
• Residual cyclicality from all methods
• Current and future estimates are more uncertain
  than historical estimates

18
Developments in factors of growth

• Growth accounting framework allows for
  alternative insightful decompositions
• 1) Output as function of
• TFP
• Labour supply
• Capital stock
• 2) Output as function of
• Labour productivity, function of TFP and capital
  intensity
• Labour utilisation, function of average hours
  worked, the unemployment rate and the
  participation rate
• Demographic components population and dependency
  ratio

19
(No Transcript)
20
(No Transcript)
21
Table 2 Developments in euro area labour
productivity
See ECB Monthly Bulletin article entitled Labour
productivity in the euro area aggregate trends
and sectoral patterns, July 2004.
22
Chart 4 Contributions to growth from labour
productivity per hour in the euro area
See ECB Monthly Bulletin article July 2004.
23
Summary of main developments by factor

• Gradual decline in productivity growth, accounted
  for by lower contribution from both capital
  deepening and TFP
• Increasing contribution from labour supply
  growth, resulting mainly from a decrease in the
  unemployment rate and slower decline in hours
  worked

24
Looking ahead short to medium run

• Projected decline in working age population
  growth is likely to have an adverse impact
• This may be counteracted by likely further
  increase in the participation rate (and possible
  further decline in the unemployment rate)
• Outlook for productivity growth more uncertain

25
Looking ahead medium to long run

• Demographic developments, as projected by both
  Eurostat and the United Nations, are likely to
  have a significant adverse impact in the long run
• Developments in other factors more uncertain
• Only a combination of compensatory measures can
  counteract the adverse impact of demographic
  developments on growth
• Further structural reforms are needed to sustain
  and possibly expand potential growth in the
  medium to long run

26
Overall conclusion

• Revision of assumption for medium term potential
  output growth is warranted only in the presence
  of clear and compelling evidence that structural
  changes are actually leading to a change in
  potential output growth
• The ECB will continue to closely monitor future
  developments