Module 4: Index Numbers

Dr. A. N. Basugade
M.Sc. Ph.D
Head, Department of Statistics,
GopalKrishnaGokhaleCollege,Kolhapur
Email – [email protected]
Module 4: Index Numbers
Overview

Introduction

Uses of Index Numbers

Limitations of Index Numbers

Types of Index Numbers

Construction of Index Numbers

Tests of adequacy of Index Numbers

Dr. A. N. Basugade

Index numbers
Index numbers are mostly used in the fields of Economics
and business for making comparisons of prices, cost of
production, purchasing power of money etc. But the most
important use of Index Numbers is to compare the prices at
different time periods (or time intervals), e.g. Cost of living of a
person in 1998 as compared to that of 1997 can be measured
with the help of index Numbers.

Definition: An index number is a statistical device designed to

show changes in a variable or a group of related variables with
respect to time.

Dr. A. N. Basugade

Uses of Index Numbers:
1) Index numbers are useful for forecasting: These are tools
to businessman to predict the behavior of prices and demands.

2) It acts as Economic Barometers: I.N. measures ups and

down in the eco-nomic condition of a country. The I.N. of
price, out put, bank deposits, foreign Ex-change etc. shows a
nature and variation in general economic and business activity
of the country.

3) It helps in measuring purchasing power of money: I. N.

are used for com-paring the purchase power of money over a
period of time to analyze whether the standard of living has
increased or decreased.

4) It helps in formulating new policies: Many economic and

business policies are guided by I. N. to frame new policies.
Dr. A. N. Basugade

Limitations of Index Numbers
1) Sampling Error: while constructing I. N. it is not possible
to include all the commodities. Thus I. N. are based on sample
and hence error may occur while selecting the sample.

2) Selection of Base year create the problem in: I. N. the price

of years are compared with the prices of base year. But the
selected base year need not be a normal year i.e. It may be
affected by various factors like, strike, war depression etc. Thus
selection of normal base year is a problem for constructing I. N.

3) Selection of method of Construction I. N. : There are many

methods and formulae to construct I. N. and they give different
results and answers. Hence it creates the problem to select a
proper formula for a particular I. N.
Dr. A. N. Basugade
 4) Changes in habits and taste : In construction of I.N. we
assume that all people have same habits and tastes over a
period of time. But which is wrong as-sumption.

5) Comparison over longer period is not reliable : If the

time period in base year and current year is too long then the
conditions will not remain same and I. N becomes
meaningless.

Types of Index Numbers :

1) Price Index Numbers 2) Quantity Index Numbers
3) Value Index Numbers.

1) Price I. N.: The price I. N. shows the changes in the prices

of commodities produced or consumed in a given period with
reference to the base period.

Dr. A. N. Basugade
 2) Quantity I. N. : Quantity I. N. shows the changes in the
quantity of goods produced or consumed in a given period
with reference to base period.

3) Value I. N.: Value I. N. shows the changes in the values of

commodities in a given period with reference to base period
( Value - Price x Quantity )

Method of constructing Index Numbers.

A) Unweighted Index Number. B) Weighted Index Numbers.
each of two methods are further divided into two classes as
i) Aggregative Method ii) Average Relative Method

A) Unweighted Index Number. This is the simplest method of

constructing I.N.
Dr. A. N. Basugade
i) Simple aggregate Price I.N.: It is denoted by P01which is the
price I.N. of current year w. r. t. base year and is given by
P01 = (∑P1 / ∑ P0) x 100
Where, ∑P1 = sum of current year prices
∑P0 = sum of base year prices

ii) Simple average of Price Relative I.N.: It is given by

P01 = {[∑(P1 / P0)] x 100}/n ( average is A.M.)
P01 = antilog {log [∑(P1 / P0) x 100]/n}
Where, ∑(P1/ P0) = sum of ratio of current year prices to the
base year prices
n = number of commodities

Dr. A. N. Basugade
Example 1: Compute simple index number using aggregate & average relative
methods of 1995 by taking 1993 as base year for the following data.
Prices in Rs.
Commodities 1993 1995
A 10 10.50
B 02 02.75
C 04 04.50
D 03 03.25

Commodities Prices in Prices in (P1 / P0 )100 Log (P1 / P0 )

1993 1995
A 10 10.5 105.0 2.0212
B 2 2.75 137.5 2.1383
C 4 4.5 112.5 2.051
D 3 3.25 108.3 2.0346
n=4 19 21 463.3 8.2451

Dr. A. N. Basugade
i) Simple aggregate price I.N.
P01 = (∑P1/ ∑P0) x 100 = (21/19)*100 = 110.5

ii) Simple average of price relative I.N. (using A.M.)

P01 = {[∑(P1/P0)] x100}/n = 463.3/4 = 115.8

iii) Simple average of price relative I.N. (using G.M.)

P01 = antilog{log [∑(P1/ P0) x100]/n} = antilog(8.2451/4)
= antilog (2.0613) = 115.2

B) Weighted Index Numbers:

i) Weighted aggregate I.N.: There are four formulae to
obtain these I.N.
a) Laspeyre’s I.N b) Paasche’s I.N. C) Fisher’s Ideal I.N.
d) Marshall-Edgeworth I.N.

Dr. A. N. Basugade
a) Lasperye’s Price Index Number: For Lasperye’s formula, base
year quantities are taken as weights.
P01(La) = (∑p1 q0 / ∑p0 q0) x 100

b) Paasche’s Price Index Number: For Paasche’s formula, current

year quantities are taken as weights.
P01(Pa) = (∑p1 q1 / ∑p0 q1) x 100

c) Fisher’s Ideal Price Index Number: Fisher’s formula is the

geometric mean of Lasperye’s & Paasche’s index numbers.
P01(F) = [P01(La)* P01(Pa)](1/2)

d) Marshall-Edgeworth Index Number: For this formula, sum of

quantities of current year and the base year are taken as
weights P01 = [∑p1 (q0 + q1 )/ ∑p0 (q0 + q1) ] x 100
Dr. A. N. Basugade
Ex. 2.Compute all the weighted aggregate price index numbers from
the following data.

Com Prices Quauntities

modit
ies 1990 1992 1990 1992 p0q0 p0q1 p1q0 p1q1

p0 p1 q0 q1
A 4 5 3 4 12 16 15 20
B 6 6 5 6 30 36 30 36
C 9 10 6 7 54 63 60 70
D 10 12 8 8 80 80 96 96
Totals 176 185 201 222

a) Lasperye’s Price Index Number = P01(La) = (∑p1 q0 / ∑p0 q0) x 100

= (201/176)*100 = 114.20
b) Paasche’s Price Index Number = P01(Pa) = (∑p1 q1 / ∑p0 q1) x 100
= (222/185)*100 = 120

Dr. A. N. Basugade
c) Fisher’s Ideal Price Index Number = P01(F) = [P01(La)* P01(Pa)](1/2)
= [114.20* 120](1/2) =117.06
d) Marshall-Edgeworth Index Number =P01=[∑p1(q0 +q1 )/∑p0 (q0 + q1)] x 100
= [∑(p1q0 +p1q1 )/ ∑(p0q0 + p0q1)] x 100
= [ (201+222)/(176+185)]*100 = 117.17

Tests of adequacy of Index Numbers:
We have seen that there are a number of formulae for constructing index
numbers. Thus the problem would be to select a proper formula. The
following four tests are suggested to compare the adequacy of a formula. (1)
Unit Test, (2) Time Reversal Test, (3) Factor Reversal Test, (4) Circular Test.

1. Unit Test: It is natural to expect that the index number should be free from
units of measurement of quantities and the units of prices. All the above
formulae, except the aggregative index formula satisfy this test. In this respect
all of them are equally good.

2. Time Reversal Test: It states that if we calculate first P01 and then P10 by
interchanging base year and the current year the product of the principal
factors in the two index numbers should be equal to unity. Symbolically, P01 x
P10 = 1. Of the various index numbers studied above only the factor in Fisher's
index number satisfies this test.

Dr. A. N. Basugade
 3. Factor Reversal Test : It states that the product of the factors of
the index number of price and the index number of quantity should
be equal to value ratio. Let P01 be the price index number and Qol be
the quantity index number obtained by interchanging prices and
quantities of the respective years. With usual notation value ratio
means the ratio of the total expenditure for the current year to the
total expenditure for the base year, i.e., V01= ∑p1 q1 / ∑p0 q0
Hence, symbolically the factor reversal test demands that
P01 x Qol = ∑p1 q1 / ∑p0 q0 = V01
Only Fisher's index satisfies this test.
4. Circular Test: Another test of adequacy of index number is called
circular test. It is a sort of extension of time reversal test. If we
calculate an index number of year a with b as base year and of year
b with c as base what will be the index of year c with the base a.
Then the circular test states that if the index is a good one then the
product of the three indexes should be equal to unity.
Symbolically, P01 x Pl2 x P20 = 1
Dr. A. N. Basugade
 Most of the indexes do not satisfy the above test. Even
Fisher's index number does not satisfy the test. However, if the
quantities remain constant for all years then fisher's index
number also satisfies this test.
7. Weighted Average of Price Relatives
If the weights of all items are given explicitly or if the value
weights are calculated, then the index number can be
calculated either by using A.M. A.M. is used then index
number is given by
P01 = ∑PW / ∑W
where, P=( P1/P0) x100 and W= weight

It should be noted that when the value weights are calculated

(or given) as percentages of the total expenditure the sum of
all the weights will be obviously equal to 100.
Since, the group index number is essentially the price relative
the above formula is of the same form as the previous one.
If G. M. used then the index number given by
P01 = antilog{∑W log P/∑W }
Dr. A. N. Basugade
Summary:
At the end of this module student must be able to

Define various Index Numbers

State uses of Index Numbers

Describe limitations of Index Numbers

Construction of various Index Numbers

Apply tests of adequacy of Index Numbers

Dr. A. N. Basugade