Topic 4: Index Numbers

An Overview

This topic looks at how single figures called ‘index numbers’ can summarise large masses of
data. Time series deflation and composite price indices will also be examined.

The Concept of an Index Number

An index number is a single number which summarises a comparison between two sets of
figures. It measures the percentage change in the value of some economic commodity over a
period of time.

Base 100

Many ordinary calculations use 100 as a base. Statisticians take advantage of this fact by basing
index numbers to 100. The main advantage of base 100 is that it simplifies the calculation of
percentage changes.

Base Year

When comparing annual figures, it is necessary to select one of the figures as the base and
designate the figure relating to that year as 100.

A base year can, therefore, be defined as the year against which all other years are compared.

Index Number Symbols

The following symbols are used in connection with index numbers:

P0 = price of item(s) in base year (year 0)

P1 P2, etc. = price of item(s) in years 1, 2, etc.

Q0 = quantity of item(s) in year 0

Q1 Q2, etc. = quantity of item(s) in years 1, 2, etc.

W= weight

1
 One-Item Index Number

The calculation of index numbers is straightforward where only one item is involved. One year
is chosen as the base year, and the values for the other years are stated in proportion to the
value in the base year i.e.:

PRICE INDEX QUANTITY INDEX

Year 0 100 100
Year 1 (P1 / P0) x 100 (Q1 / Q0) x 100
Year 2 (P2 / P0) x 100 (Q2 / Q0) x 100

Example:

For the following data, calculate the Price Index and the Quantity Index (2020 is the base year):

YEAR PRICE QUANTITY PRICE QUANTITY INDEX

INDEX
2020 500 21,200 ? ?
2021 530 28,633 ? ?
2022 530 35,028 ? ?
2023 550 40,650 ? ?
2024 600 44,531 ? ?

Solution:

YEAR PRICE QUANTITY PRICE QUANTITY INDEX

INDEX
2020 500 21,200 100 100
2021 530 28,633 106 135.06
2022 530 35,028 106 165.22
2023 550 40,650 110 191.75
2024 600 44,531 120 210.05

Price Index
2020 500/500*100 = 100
2021 530/500*100 = 106
2022 530/500*100 = 106
2023 550/500*100 = 110
2024 600/500*100 = 120

Quantity Index
2020 21200/21200*100 = 100
2021 28633/21200*100 = 135.06
2022 35028/21200*100 = 165.22
2023 40650/21200*100 = 191.75
2024 44531/21200*100 = 210.05
2
Calculation of price and quantity relatives

2023 2024
Item Price Number sold Price Number sold

IPOD 120 40 150 60

TV 400 35 440 45

In 2024:

IPOD
Price Relative 150/120*100 = 125
Quantity Relative 60/40*100 = 150

TV
Price Relative 440/400*100 = 110
Quantity Relative 45/35*100 = 128.57

Example:

The following data relates to the production of cars from a particular assembly line over a
number of months

Month Production
Mar 142
Apr 126
May 128
Jun 104
Jul 108
Aug 146
Sep 158
Oct 137

Calculate sets of productivity relatives with March (base) =100

Solution:

March 142/142*100 = 100

April 126/142*100 = 88.73
May 128/142*100 = 90.14
June 104/142*100 = 73.24
July 108/142*100 = 76.05
Aug 146/142*100 = 102.82
Sept 158/142*100 = 111.27
Oct 137/142*100 = 96.48

3
Time Series Relatives

It is very useful to show the changes in the value of items over time. There are two distinct
ways in which relatives can be calculated:

1. Fixed base relatives:

Here each relative is calculated based on the same fixed time point. This approach is useful where
the basic nature of the commodity is unchanged over the whole period. Fixed based relatives tend
to be used to compare ‘like with like’, for example the price of cheese in a supermarket over a
six month period or weekly family expenditure on entertainment.

2. Chain base relatives:

This involves each relative been calculated with respect to the immediately preceding time point.
This approach can be used with any set of commodity values, but must be used when the basic
nature of the commodity changes over the whole time period. For example, a monthly index of
petrol costs for a salesman may vary with the type of car he drives. The model would therefore
change based on the type of car been driven.

Example 1: Calculate the Fixed and chain base set of relatives for a given time series.

Month Jan Feb Mar Apr May Jun

Production 4563 4245 4841 4644 5290 5166
Fixed base relative
Chain base relative

Note: Let March be the base period = 100

Solution:

Fixed Base Index

March = Base
Jan 4563/4841*100 = 94.26
Feb 4245/4841*100 = 87.69
Mar 4841/4841*100 = 100
Apr 4644/4841*100 = 95.93
May 5290/4841*100 = 109.27
June 5166/4841*100 = 106.71

Chain Base Index

Jan -
Feb 4245/4563*100 = 93.03
Mar 4841/4245*100 = 114.04
Apr 4644/4841*100 = 95.93
May 5290/4644*100 = 113.91
Jun 5166/5290*100 = 97.66

4
We can see that the fixed base index has risen considerably over the six months. It has
increased from 94.26 to 106.71. This represents a 13.21% increase in production.

Percentage increase = 106.71-94.26/94.26*100= 13.21%

The chain base index shows the change in values from month to month. We can see that there
were significant increases in production in both March and May.

Changing the base of fixed base relatives

With time series data it is sometimes necessary to change the base. This may be because the
original base time point is outdated and a more recent one is required.

In the example below a base of 1995 was originally used but due to this been outdated a more
recent one is needed.

Example:

Index 2018 2019 2020 2021 2022 2023 2024

1995=100 324 351 377 384 391 404 428

The procedure for changing the base of a time series of relatives is essentially the same as
calculating a set of relatives for a given time series of values. However, the procedure is provided
below and demonstrated using the above relatives

Step 1: Choose a new base time point (for example 2018) and identify the corresponding relative.
If we choose 2012 as the base year the corresponding relative will be 324

Step 2: Divide each relative in the set by the value of the relative identified above and multiply
the result by 100.

Table

2018 2019 2020 2021 2022 2023 2024

Old index 324 351 377 384 391 404 428
1995=100
New index 100 108 116 119 121 125 132
2018=100

5
Time Series Deflation

Time series deflation is a technique used to obtain a set of index relatives that measure the
changes in the real value of some commodity with respect to some given indicator.

Example:

Month Average Daily Wage (€) Retail Price Index

1 17.60 106.1
2 18.10 107.9
3 18.90 112.0
4 19.60 113.1
5 20.25 116.0
6 20.30 117.4
7 20.60 119.5
8 21.40 119.7

Real Value Index = Current value * Base indicator * 100

Base value Current indicator 1

Solution:

Month 1 17.60/17.60*106.1/106.1*100 = 100

Month 2 18.10/17.60*106.1/107.9*100 = 101.13
Month 3 18.90/17.60*106.1/112*100 = 101.73
Month 4 19.60/17.60*106.1/113.1*100 = 104.47
Month 5 20.25/17.60*106.1/116*100 = 105.24
Month 6 20.30/17.60*106.1/117.4*100 = 104.24
Month 7 20.60/17.60*106.1/119.5*100 = 103.92
Month 8 21.40/17.60*106.1/119.7*100 = 107.78

Overall, we can see that real wages have increased by 7.78% over the eight-month period. This
means that workers are now 7.78% better off overall. This shows an increase in their real wealth
when wages are adjusted for inflation. Workers and union groups should be broadly satisfied.

Final Table

Month Average Daily Wage (€) Retail Price Index Real Wage Index
1 17.60 106.1 100
2 18.10 107.9 101.13
3 18.90 112.0 101.73
4 19.60 113.1 104.47
5 20.25 116.0 105.24
6 20.30 117.4 104.24
7 20.60 119.5 103.92
8 21.40 119.7 107.78

6
 Weighted Aggregate Indices

Most index numbers relate to more than one item. For example, if we need an index for the cost
of living, more than one number is involved.

The formula for such a weighted aggregate price index is:

Σ(P1´ W)
x100
Σ(P0 ´ W)

Example:

Suppose that there are three items in the Irish cost of living index – bread, cheese and beer.

Item 2023 2024

Bread €0.45/loaf €0.50/loaf
Cheese €5/kg €10/kg
Beer €9/keg €7.50/keg

This data has the following difficulties:

• Two prices have gone up and one down
• The prices are for different units
• It is unclear as to which item is most important

Assuming that weights of 100, 10 and 1 are given to represent the importance of each item, the
following weighted cost of living index can be derived:

2023 2024
Item Weight Price Price x Price Price x Weight
Weight
Bread 100 0.45 45 0.50 50
Cheese 10 5.00 50 10.00 100
Beer 1 9.00 9 7.50 7.50
Total 104 157.50

157.50
Weighted Aggregate Index = ´ 100 = 151
104

7
 Choice of Price Indices

In practice, the selection of weights can be a difficult problem.

One solution to this is to use the actual quantities consumed as weights - such an index number
is called a quantity weighted index.

The question then arises as to which quantities to use - quantities in the base year or quantities
in the current year.

• A Laspeyres index uses base year quantities.

• A Paasche index uses current year quantities.

Laspeyres Price Index

This can be found as follows:

ΣP1Q0
Index = ´ 100
ΣP0Q0

Paasche Price Index

This can be found as follows:

ΣP1Q1
Index = ´ 100
ΣP0Q1

Quantity Indices

Although they are not the most important type of index, quantity indices can be used to show
how the amount of certain goods vary over time and location.

When looking at price indices, quantities emerged as the best weighting factor. The converse is
also true; prices are considered the best weights when calculating quantity indices:

ΣQ1P0
Laspeyres Quantity Index = ´ 100
ΣQ0P0

ΣQ1P1
Paasche Quantity Index = ´ 100
ΣQ0P1

8
Example:

Suppose that we are given the following quantity information about bread, cheese and beer in
the years 2023 and 2024.

Item 2023 2024

Bread 80,000 loaves 100,000 loaves
Cheese 10,000 kg 15,000 kg
Beer 1,000 kegs 3,000 kegs

Item 2023 2024

Bread €0.45/loaf €0.50/loaf
Cheese €5/kg €10/kg
Beer €9/keg €7.50/keg

Find the Laspeyres and Paasche price indices and quantity indices for 2018?

Solution:

P0 P1 Q0 Q1 PIQ0 P0Q0 P1Q1 P0Q1

2023 2024 2023 2024
Bread 0.45 0.50 80000 100000 40000 36000 50000 45000
Cheese 5.00 10.00 10,000 15000 100000 50000 150000 75000
Beer 9.00 7.50 1000 3000 7500 9000 22500 27000
Total 147500 95000 222500 147000

Laspeyres Price Index

147500/95000*100 = 155.26

Paasche Price Index

ΣP1Q1
Index = ´ 100
ΣP0Q1 222500/147000*100 = 151.36

Laspeyres Quantity Index

ΣQ1P0
Index = ´100 147000/95000*100 = 154.74
ΣQ0P0

Paasche Quantity Index

ΣQ1P1
Index = ´ 100 222500/147500*100 = 150.85
ΣQ0P1

9
Comparison of Laspeyres and Paasche price indices:

1. What the indices are calculating:

Laspeyres: compares base period expenditure with a hypothetical current period expenditure
at base period quantities.

Paasche: compares current period expenditure with a hypothetical base period expenditure at
current period quantities

2. When prices are rising:

Laspeyres: tends to over-estimate price increases

Paasche: tends to under-estimate price increases

3. Nature of the indices:

Laspeyres: is a ‘pure’ price index since it is comparing like with like.

Paasche: not considered ‘pure’ since the weights change.

4. Weights used in calculations:

Laspeyres: uses base period quantities as weights and can thus become out of date.

Paasche: uses current quantities and weights are thus always up-to-date.

5. Ease of calculations:
Laspeyres: needs only the base period quantities no matter how many periods the index is
being calculated for.

Paasche: needs new quantities for each time period. Quantities are normally more difficult
to determine than prices.

Generally, on balance, the Laspeyres index is favoured to the Paasche index.

Notes:
• Paasche price indices require actual quantities to be available for each year of the series.
• Different years can be compared with each other for a Laspeyres price index, but different
years can only be compared with the base year for a Paasche price index
• Paasche price indices use more up-to-date weights than Laspeyres price indices, and,
hence, the Laspeyres index becomes more outdated as time progresses.

10
This point can be seen clearly in the table below:

Laspeyres Index Paasche Index

Base Yr = 100 Base Yr = 100
ΣP1Q0 ΣP1Q1
Year1 : Year1 :
ΣP0Q0 ΣP0Q1
ΣP2Q0 ΣP2Q2
Year2 : Year2 :
ΣP0Q0 ΣP0Q2

A new set of quantities is needed for each Paasche index, but the one set of quantities will do
for many Laspeyres indices.

Fisher’s ‘Ideal’ Index

While Laspeyres and Paasche price indices can be difficult, so Fisher’s ‘ideal’ price index is a
compromise method that can sometimes be used to get an average of base and current period
weights.

It can be found as follows:

Fisher’s ideal price index = L.P.I ´ P.P.I

where L.P.I. = Laspeyres Price Index

P.P.I = Paasche Price Index

In effect, the Fisher index is the geometric mean of the Laspeyres and Paasche indices.

Similarly, Fisher’s ‘ideal’ quantity index can be found as follows:

Fisher’s ideal quantity index = L.Q.I´ P.Q.I

where L.Q.I. = Laspeyres Quantity Index

P.Q.I = Paasche Quantity Index

11
 Index Linking

One of the most common applications of index numbers is in the area of index linking. This
occurs when the value of an item is linked to the value of an index.

To compute the value of an index-linked item, compute the percentage change in the index and
apply this percentage change to the index linked item.

Index linking has many important applications:

• Producers typically index link their products to the rate of inflation in order to protect
their competitive position.

• Union officials claim that wage rates should rise by the rate of inflation in order to
protect workers real rate of pay.

Example:

At the start of a year, the Consumer Price Index stood at 340. A person invested £360 in an
index-linked savings bond at this time. If the index had risen to 360 at the end of the year, what
will now be the value of the savings bond?

Solution:

The index has risen by 20 from 340. This is a percentage increase of 5.88%. As the savings
bond is index-linked, it will increase by a similar percentage i.e.

5.88% of £360 = £21.17

Hence, the value of the bond at the end of the year will be £381.17 (£360 + £21.17).

Index Construction

The following factors should be considered when preparing indices:

1. The Purpose of the Index

For example, the Consumer Price Index is used to calculate the average price increases or
decreases of all Irish goods. It is consequently used as an index of Irish inflation.

2. Selecting the Items and Choosing Weights

Any items selected should be unambiguous and must have ascertainable values. The choice of
weights is a difficult issue since different sets of weights can give different index numbers. In
general, quantities are used as weights for different items.

3. Choice of a Base Year

Generally speaking, the year chosen should be a reasonably normal year and not be too distant.
12
The Consumer Price Index

The Consumer Price Index or CPI measures the overall change in the price of goods and services
that people typically buy over time. The CPI is published by the Central Statistics Office.
Approximately 51,000 prices are collected each month to help compile this data. These prices
are then compared to the previous month to identify changes and trends. The recent Covid-19
pandemic has had a significant impact on the CPI.

People tend to have different tastes and habits; therefore, the CPI incorporates these into its
analysis. The price of services is also included in the CPI analysis. These services include
hairdressing, taxi fares and insurance. The collection of goods and services is normally referred
to as the basket of goods and services. Approximately 612 items are included in this basket of
goods. The basket is split approximately between 55% services and 45% goods. Items such as
charity contributions, lottery and betting payments are not included in the analysis. The basket
does not refer to any one individual but refers to and represents the majority of households in
Ireland.

A household budget survey is used to ascertain the goods and services to be included in the
survey. The most recent Household Budget Survey was the 2022/2023 survey. Every county
within Ireland is included in the analysis. Weights are given to larger cities such as Dublin to
reflect their importance in the index. Foreign tourists on holiday in Ireland are also included in
the analysis.

Who uses the CPI?

• The Irish government

• The European Union
• Macroeconomists
• Researchers with an interest in economic trends
• It can be used to benchmark pay, social welfare, pensions, house insurance and other
contracts.
• Other special interest groups

The CPI is the official measure of inflation. Inflation is the rate at which your money loses its
ability to buy things. The CPI can be compared to wage increases to measure the real wealth of
an individual. It is essential to note that the CPI is a price index and not a cost-of-living index. A
cost-of-living index would incorporate items such as income, taxation and social welfare
payments into its analysis. The base reference for the current CPI is 2023, and the reference day
for pricing is the 2nd Tuesday in every month in order to ensure consistency.

Ireland witnessed a high amount of inflation in 2021, 2022 and 2023. This was driven by
increases in groceries, fuel, energy costs and transportation costs. At approximately 8.5% in
February 2023 Ireland had its highest level of inflation for almost forty years. This has led to a
cost-of-living crisis in Ireland with the Irish Government introducing a number of measures to
alleviate the difficulties which people are facing. The rapid increase in energy prices resulted in
the Irish Government providing energy credits. As of November 2024, Inflation has started to
fall considerably to 1.7%. The ECB have begun to decrease interest rates again in recent times
as inflation falls.

13
The harmonised index of consumer prices (HICP) is a measure of price changes within the
member states of the European Union. In September 2024, the rate of inflation was approximately
1.8%. The index is based on the harmonisation and standardisation of practices within the EU
and the purpose is to allow for the comparison of different consumer price trends in different
member states.

The weighted indices for the EU19 (Eurozone) and EU27 (27 member states) are calculated by
Eurostat. The HICP is viewed as a subset of the main CPI and certain items are excluded from
the HICP. Approximately 93% of HICP expenditure is included in the CPI. Certain items such
as mortgage interest, Local Property tax, buildings materials, union subscriptions, motor taxation
and house insurance tend to be excluded from the HICP.

The basket of goods in the Irish CPI is normally reviewed every five years. For example, in 2011
new items were included in the index such as assessment fees, computer repairs, bundled
telephone and internet. In 2016, among the new items added were champagne, e-cigarette refills,
avocados, melons and sweet potatoes. The Covid-19 crisis delayed the latest rebasing of the index
to 2023. Among the items introduced in 2023 were air fryers, wireless speakers and smart
watches. Digital Cameras, newspaper advertisements and MP4 players were removed from the
index in 2023.

The CPI has greater coverage and is therefore viewed as a superior measure for national purposes,
but for comparisons within Europe, the HICP based on harmonised practices is a more precise
measure. In the Republic of Ireland, the CPI is the official measure of inflation.

Limitations of Index Numbers

The use of index numbers is subject to the following limitations:

• Data used in the calculation of indices are rarely up-to-date, and often incomplete. For
example, weightings will become out-of-date over time as tastes and habits change, thus
making the index of little use.

• As no year can strictly be called a normal year, the index contains an element of
approximation.

• Index numbers can be misinterpreted by the uninformed layman. For example:

Suppose a chain base price index for a number of months yielded the values:
113, 109, 108, 105. This might be confused with a fixed base index and interpreted as a
decrease in prices, whereas in fact prices are still rising (although the rate of increase in
prices is falling).

• If a production index increases from 400 to 410, this might be interpreted as a 10%
increase in production (it is only a 2.5% increase).

• Indices calculated for a whole country can be seriously in error when they are compared
to regions within the country.

14
• New inventions, which arise during a period, create a difficulty for index numbers – if
they are not included in the base data, it is inconsistent to include them in the current
data.

15