Diapositiva 1

1
Joint UNECE/ILO Meeting on Consumer Price Indices
The interpretation of the divergences between
CPIs at territorial level Evidence from Italy
Biggeri L., Brunetti A. and. Laureti T
Italian National Statistical Institute (Istat),
Rome, Italy University of Tuscia, Viterbo,
Italy Geneva, 8-9 May 2008
2
Joint UNECE/ILO Meeting on Consumer Price Indices
Structure of the paper

• Introduction

2. The divergence between CPIs at territorial
level a method for decomposing and interpreting
it
3. The organisation of the decomposition analyses
on Italian data
4. Analysis of the results
5. Concluding remarks
Geneva, 8-9 May 2008
3
Joint UNECE/ILO Meeting on Consumer Price Indices

1. Introduction

The aim of the paper is twofold
Method for the decomposition CPIs divergences
Empirical analyses to show the usefulness of the
method
For the construction of CPIs most NSIs make use
of the Laspeyres formula
1
Considering the same basket of products and
services, the divergence between two CPIs in
different areas l and j
2
Geneva, 8-9 May 2008
4
Joint UNECE/ILO Meeting on Consumer Price Indices
2. The divergence between CPIs at territorial
level
From a different point of view, by considering
area j as reference area
5

• A comparison between two local CPIs depends on
• Elementary indices
• Weighting system

• Examine and decompose these divergences
• Understand which are the factors that cause them

Four equivalent decomposition formulae
Geneva, 8-9 May 2008
5
Joint UNECE/ILO Meeting on Consumer Price Indices
2. A method for decomposing and interpreting the
differences (a)
3

3bis
Factors influencing the Elementary Price Index
Effect
Factors influencing the Weight Effect
Geneva, 8-9 May 2008
6
Joint UNECE/ILO Meeting on Consumer Price Indices
2. A method for decomposing and interpreting the
differences (b)
6
The factors can give different results since they
are defined relating to different distributions
of elementary price indices and weights
difference between the two arithmetic means of
elementary price index distributions

• interesting interpretation from an economic point
  of view
• determining the price effect
• influencing the overall difference between the
  two CPIs considered.

Geneva, 8-9 May 2008
7
Joint UNECE/ILO Meeting on Consumer Price Indices
3. Data set description and organisation of
analyses on Italian data

• DATA DESCRIPTION
• Monthly CPIs for the whole nation for elementary
  aggregates
• 40,000 outlets
• 85 municipalities
• 400,000 elementary prices
• 540 representative products
• System of weights for 85 municipalities using
  household expenditure shares
• Period January 2002-December 2007
  

• CALCULATIONS
• analysis limited to December of the years 2002
  and 2007
• Similar basket of products and services
• Selection of 9 chief regional towns

• Considering
• the cities where it is reasonable to assume a
  different behaviour of the sellers and consumers
  and a different evolution of attitude regarding
  sale and purchase.
• The territorial location of the cities in the
  north, centre and south Italy

Geneva, 8-9 May 2008
8
Joint UNECE/ILO Meeting on Consumer Price Indices
4. Analysis of the results (a)
Table 2 Overall Differences between pairs of
cities
Comparing Naples with the other cities the
differences (all positive but with different
values) have a hidden meaning concerning the
degree of importance of the various factors
Geneva, 8-9 May 2008
9
Joint UNECE/ILO Meeting on Consumer Price Indices
4. Analysis of the results (b)
Joint UNECE/ILO Meeting on Consumer Price Indices
-0.639
0.0035
13.029
0.0441
Naples (j) Turin (l)
factors
Price Effect 0.516
CPI difference 0.885
Weight Effect 0.369
factors
12.335
0.002
0.034
Price change distribution
Correlation coefficient between elementary
indices and weights
Turin -0.01
Naples 0.056
Geneva, 8-9 May 2008
10
Joint UNECE/ILO Meeting on Consumer Price Indices
Joint UNECE/ILO Meeting on Consumer Price Indices
4. Analysis of the results (c)
0.596
0.0037
12.131
0.031
Naples (j) Florence (l)
factors
Price Effect 1.396
CPI difference 1.691
Weight Effect 0.295
factors
12.335
0.0018
0.0240
Price change distribution
Correlation coefficient between elementary
indices and weights
Florence 0.026
Naples 0.056
Geneva, 8-9 May 2008
11
Joint UNECE/ILO Meeting on Consumer Price Indices
4. Analysis of the results (d)
The evolutions of local CPIs is quite different
across Italy and influence the divergences
between the CPIs in different ways and degrees.

• The divergences in the evolution of the CPIs
  depend
• mostly on the price effects.
• in same cases on the different share of
  expenditures (two southern cities with all the
  other cities)
• on the characteristics of the corresponding
  distribution of the elementary price indices and
  on the value of the correlation between
  elementary indices and weights

Similar results from the comparisons for December
2007
Geneva, 8-9 May 2008
12
Joint UNECE/ILO Meeting on Consumer Price Indices
5. Concluding remarks
A simple method for calculating the decomposition
of the divergence between two CPIs to obtain a
measure of the importance of the factors that
affect it.
Understanding why the local CPIs often differ
among one another
Very interesting results
Further research

• improve the method of the decomposition of the
  divergences between the CPIs

• Analysis for CPIs by classes and groups of
  products to understand the importance of the
  different products in affecting the price and
  weights effects

• analyse the degree of the price and weights
  effects during periods with different inflation
  rates.

Geneva, 8-9 May 2008
13
Joint UNECE/ILO Meeting on Consumer Price Indices
Joint UNECE/ILO Meeting on Consumer Price Indices
Thank you for your kind attention!