Evaluating measures of core inflation

1
Evaluating measures of core inflation

• Mark A. Wynne
• Federal Reserve Bank of Dallas
• Presentation to International Seminar on Core
  Inflation
• Rio de Janeiro, June 7-8 2001

2
Origin development of a concept

• 1970s oil commodity price shocks
• The Ex. Food Energy approach
• Eckstein (1981)
• 1990s revival of stochastic approach to index
  numbers
• Selvenathan Prasada Rao
• Bryan Cecchetti
• Quah Vahey

3
Two new approaches

• The common component in price changes
• Bryan Cecchetti
• The persistent component of price changes
• Quah Vahey

4
A basic framework I
5
A basic framework II
6
Use of core inflation in monetary policy

• As a target for monetary policy
• Reserve Bank of New Zealand (until 1999), Bank of
  England, Reserve Bank of Australia (until 1998),
  Thailand
• As an input to policy decisions
• Fed, ECB
• As part of communication strategy
• Bank of England, Sveriges Riksbank

7
What is the question to which core inflation is
the answer?

• Alternative measure of inflation
• Domain of headline measure too broad
• Answer to a counterfactual
• What if?
• A measure of trend, steady-state or inertial
  inflation

8
Criteria for evaluating measures of core inflation

• Computable in real time
• Forward looking
• Track record
• Understandable by the general public
• History invariant to new data
• Sound theoretical basis

9
Criterion 1 Computable in real time

• Essential for use of core inflation
• as target
• as input to policy
• as part of communication
• Satisfied by almost all popular measures
• Exceptions HP, BP, two-sided filters

10
Criterion 2 Forward looking

• Inherently forward-looking measures
• e.g. SVAR measures
• Measures with predictive power
• Stable long-run relationship with headline rate

11
Criterion 3Track record

• Ex. Food Energy
• produced by almost all countries
• Trimmed mean
• most popular of newer measures

12
Criterion 4 Understandable by general public

• Essential for measure used as target
• Important if measure used as part of
  communications strategy
• Reproducible

13
Criterion 5 History invariant to arrival of new
data

• True of
• traditional Ex. Food Energy measures
• trimmed mean
• Problematic for
• SVAR, DFI based measures
• other model-based measures

14
Criterion 6 Sound basis in economic theory

• Contrast with theory of cost of living index
• well developed well understood
• What is the appropriate theory for core
  inflation?
• Quantity theory perspective on inflation
• Long run neutrality of money

15
Concluding observations

• Core inflation a slippery concept
• No good theory, yet
• Tracking trend a useful criterion for
  discriminating between measures
• not without problems