STAT 458: SURVEY ORGANISATION AND MANAGEMENT

LECTURE 4: INPUT-OUTPUT (I-O) ANALYSIS

1.1. Introduction
Input-output analysis is a method of systematically quantifying the mutual
interrelationships among the various sectors of a complex economic system. Thus, an
Input-output analysis is an analytical technique used to analyse the inter-industry
relationships in order to understand the interdependencies and complexities of the
economy. It is also known as the Leontief Model because it developed by Prof Wassily
W. Leontif (1951), or inter-industry analysis since the fundamental purpose of the input-
ouput framework is to analyse the iter-dependencies of the industries in the economy. In
practical terms, the economic system to which it is applied may be as large as a nation
or even the entire world economy, or as small as the economy of a metropolitan area or
even a single enterprise.

In all instances the approach is essentially the same. The structure of each sector's
production process is represented by an appropriately defined vector of structural
coefficients that describes in quantitative terms the relationship between the inputs it
absorbs and the output it produces.

The Input-output analysis is based on the principle that the input of one industry is the
output of another industry and vice versa. The input represents the expenditure of the
firm and the output represents its receipt. Hence, the sum of the money values of input is
the total cost of a firm and the sum of the money value of the output is the total revenue
of the firm.

The interdependence among the sectors of the given economy is described by a set of
linear equations expressing the balances between the total input and the aggregate
output of each commodity and service produced and used in the course of one or several
periods of time. In its most basic form, the I-O model consist of a system of linear
equation, each one of which describes the distribution of an industry’s product throughout
the economy. The fundamental information used on an I-O analysis concerns the flows
of products from each industrial sector considered as a producer to each of the sectors
(itself and others) considered as consumers. This basic information from which an I-O
model is developed is contained in an inter-industries transactions table. The rows of such
a table describe the distribution of a producer’s output throughout the economy. The
column describes the composition of inputs required by a particular industry to produce
its output.

1
An I-O model is constructed from an observed data for a particular economy (e.g., an
economy of a nation, region or a state). The whole economy is divided into two main
sectors:
i. Inter-industry sectors
ii. Final demand sectors

Both these sectors are made up of sub-sectorial divisions. The total output of any inter-
industry sector is generally capable of being used as inputs by the other industry-sectors,
and itself, and by the final demand sectors.

Source: Leontief W. (1966). Input-output Economics. Oxford University Press

1.2. The Statistic Input-Output Model

1.2.1. Assumptions
i. There are no external economies or diseconomies of production
ii. No two products are produced jointly (i.e., each industry produces only one homogenous
product).
iii. Each producing sector satisfies the properties of linear homogenous production function
(i.e., the production of each sector is subject to constant returns to scale).
iv. The combination of inputs is employed in a rigidly fixed proportion. (i.e., there is no
substitution between different materials and no technological progress.
v. The data are flows of products from each of the sectors (as producers) to each of the
sectors (as buyers) and these inter-industry flows/transactions are measured for a
particular time period and in monetary terms.
1.2.2. Notations

One essential set of data for an I-O model are monetary values of the transactions between pairs
of sectors (from each sector 𝑖, to each sector 𝑗). These are usually denoted by 𝑥𝑖𝑗 . Sector 𝑗’s
demand for input from other sectors during the period (eg. Year) will have been related to the
amount of goods produced by sector 𝑗 over the same period (eg. Year).

2
Assume that the economy can be categorized into 𝑛 sectors. If we denote the total output of
sector 𝑖 by 𝑥𝑖 and the total final demand for sector 𝑖’s product by 𝑥𝑗 , then the simple equation
accounting for the way in which sector 𝑖 distribute its output through sales to other sectors and
the final demand is given by:

𝑥𝑖 = 𝑥𝑖1 + 𝑥𝑖2 + 𝑥𝑖3 + ⋯ + 𝑥𝑖𝑗 + ⋯ + 𝑥𝑖𝑛 + 𝑓𝑖

𝑛

𝑥𝑖 = ∑ 𝑥𝑖𝑗 + 𝑓𝑖 ⋯ ⋯ ⋯ (𝑬𝒒𝒏. 𝟏)
𝑗=1

The 𝑥𝑖𝑗 ′𝑠 represent inter-industry sales by sector 𝑖 to all sectors 𝑗, including itself when 𝑖 = 𝑗.
Similarly, there will be an equation like eqn. 1 that identifies (describes) sales of output of each
of the 𝑛 industries as follows:

𝑥1 = 𝑥11 + 𝑥12 + 𝑥13 + ⋯ + 𝑥1𝑗 + ⋯ + 𝑥1𝑛 + 𝑓1

𝑥2 = 𝑥21 + 𝑥22 + 𝑥23 + ⋯ + 𝑥2𝑗 + ⋯ + 𝑥2𝑛 + 𝑓2

𝑥3 = 𝑥31 + 𝑥32 + 𝑥33 + ⋯ + 𝑥3𝑗 + ⋯ + 𝑥3𝑛 + 𝑓3

(Eqn. 2)
⋮ ⋮ ⋮

𝑥𝑖 = 𝑥𝑖1 + 𝑥𝑖2 + 𝑥𝑖3 + ⋯ + 𝑥𝑖𝑗 + ⋯ + 𝑥𝑖𝑛 + 𝑓𝑖

⋮ ⋮ ⋮

𝑥𝑛 = 𝑥𝑛1 + 𝑥𝑛2 + 𝑥𝑛3 + ⋯ + 𝑥𝑛𝑗 + ⋯ + 𝑥𝑛𝑛 + 𝑓𝑛

Let
𝑥1 𝑓1
𝑥2 𝑥11 𝑥12 … 𝑥1𝑛 𝑓2
𝑥3 𝑥21 𝑥22 … 𝑥2𝑛 𝑓3
𝑥𝑖 = 𝑥𝒙 + 𝑥𝑖2
𝑖1 = ⋮ + ; 𝑿𝑥𝑖3=+𝑥⋯ 𝑥𝑖𝑗32+ ⋯
31+ 𝑥 … + 𝑥3𝑛
𝑖𝑛 +; 𝒇
𝑓𝑖 = ⋮ (Eqn. 3)
𝑥𝑖 ⋮ ⋮ ⋱ ⋮ 𝑓𝑖
⋮ [𝑥𝑛1 𝑥𝑛2 … 𝑥𝑛𝑛 ] ⋮
[𝑥𝑛 ] [𝑓 ]
𝑛

Then, equation 2, which is the distribution of each sector’s sales can be summarized in a matrix
form as:

𝒙 = 𝑿𝑰 + 𝒇 ----- (Eqn. 4)

3
Where 𝑰 is the identity matrix which is a column vector of one’s.
1
1
1
𝑰= ⋮ ;
1
⋮
[1]

Consider the information in the 𝑗 𝑡ℎ column of the 𝑥’s on the right handside. These elements are
sales to sector 𝑗 from other sectors (i.e., purchases of the product of the various producing sectors
in the country by sector 𝑗). Thus, the column represents the sources and magnitude of sector
𝑗′𝑠 input.

Now, by engaging production, a sector 𝑗 also pays for other items such as labour, capital as well
as other inputs. All these other inputs are put together and referred to as value added in sector
𝑗. The general input-output table is given as below:

Table 1: The general input-output table

Sectors Purchases by intermediate users Final Demand Total
(Sectors/industries) Output/Sales
𝟏 𝟐 𝟑 ⋯ 𝒏 𝑪 𝑮 𝑰 𝑬 by Sectors
𝟏 𝑥11 𝑥12 𝑥13 ⋯ 𝑥1𝑛 𝐶1 𝐺1 𝐼1 𝐸1 𝑋1
𝟐 𝑥21 𝑥22 𝑥23 ⋯ 𝑥2𝑛 𝐶2 𝐺2 𝐼2 𝐸2 𝑋2
𝟑 𝑥31 𝑥32 𝑥33 ⋯ 𝑥3𝑛 𝐶3 𝐺3 𝐼3 𝐸3 𝑋3
⋮ ⋮ ⋮ ⋮ ⋱ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮
𝒏 𝑥𝑛1 𝑥𝑛2 𝑥𝑛3 ⋯ 𝑥𝑛𝑛 𝐶𝑛 𝐺𝑛 𝐼𝑛 𝐸𝑛 𝑋𝑛
Value Added 𝑊1 𝑊2 𝑊3 ⋯ 𝑊𝑛
(𝑾)
Valued 𝑅1 𝑅2 𝑅3 ⋯ 𝑅𝑛
Added (𝑹)
Total 𝑋1 𝑋2 𝑋3 ⋯ 𝑋𝑛
Supply/Input
by Sectors

A fully general input-output table will have 21 industries or sectors according to the ISIC (A-U) and
the final demand sector can be expanded to show subdivisions of each category. For practical
reasons and depending on the purpose of the analysis, it is sometimes necessary to merge some
industries, thereby reducing the value of 𝑛, merge final demand into one column and similarly
provide only values for value added without other details. For example, let us assume a four (4)
sector economy in which there are three (3) industry sectors and one final demand sector. In that
case, the input-output table will be as follows:

4
Table 2: A simplified input-output table for a four (4) sector economy
Sectors Purchases by intermediate Final Total
users (Sectors/industries) Demand Output/Sales
by Sectors
𝟏 𝟐 𝟑 𝒇
𝟏 𝑥11 𝑥12 𝑥13 𝑓1 𝑅1
𝟐 𝑥21 𝑥22 𝑥23 𝑓2 𝑅2
𝟑 𝑥31 𝑥32 𝑥33 𝑓3 𝑅3
Labour(𝑳) 𝐿1 𝐿2 𝐿3
Total 𝐶1 𝐶2 𝐶3
Supply/Input
by Sectors

1.3. Technical Coefficient Matrix

Using the input-output table, a matrix of technical coefficient can be constructed as follows: From
the input-output table, we recall that:

i. 𝑿𝟏 is total output of sector 𝑖;

ii. 𝒙𝒊𝒋 is the output for sector 𝑖, used as input for sector 𝑗;
iii. 𝒇𝒊 is the total final demand for sector 𝑖’s product.

Now, the input coefficient of the product of sector 𝑖 going into sector 𝑗 (𝒂𝒊𝒋 )is defined as the
amount of sector 𝑖’s output consumed by sector 𝑗 per unit output of sector 𝑗 (i.e., for each unit
that sector 𝑗 produces, how much of input does sector 𝑖 contribute or for sector 𝑗 to produce a
unit of its output, how much of sector 𝑖’s output does sector 𝑗 use).

Mathematically, 𝒂𝒊𝒋 is given as:

𝒙𝒊𝒋
𝒂𝒊𝒋 = eqn. 5
𝑪𝒋

Thus,

𝒙𝒊𝒋 = 𝒂𝒊𝒋 × 𝑪𝒋 eqn. 6

The total set of input coefficients gives the technical coefficient matrix of the economy.

5
Accordingly, Table 2 can be presented as follows:

Table 3: A simplified input-output table for a four (4) sector economy

Sectors Purchases by intermediate Final Total
users (Sectors/industries) Demand Output/Sales
by Sectors
𝟏 𝟐 𝟑 𝒇
𝟏 𝑎11 𝐶1 𝑎12 𝐶2 𝑎13 𝐶3 𝑓1 𝑅1
𝟐 𝑎21 𝐶1 𝑎22 𝐶2 𝑎23 𝐶3 𝑓2 𝑅2
𝟑 𝑎31 𝐶1 𝑎32 𝐶2 𝑎33 𝐶3 𝑓3 𝑅3
Labour(𝑳) 𝐿1 𝐿2 𝐿3
Total 𝐶1 𝐶2 𝐶3
Supply/Input
by Sectors

Where;

𝑅1 = 𝑎11 𝐶1 + 𝑎12 𝐶2 + 𝑎13 𝐶3 + 𝑓1

𝑅2 = 𝑎21 𝐶1 + 𝑎22 𝐶2 + 𝑎23 𝐶3 + 𝑓2
𝑅3 = 𝑎31 𝐶1 + 𝑎32 𝐶2 + 𝑎33 𝐶3 + 𝑓3 Eqn. 7

𝑅𝑖 = ∑𝑛𝑗=1 𝑎𝑖𝑗 𝐶𝑗 + 𝑓1

Example 1:

Using the information in the input-output table below to answer the following questions:

(a) How many sectors does this economy have?

(b) Complete the table.
(c) Generate the technical coefficient matrix of the economy.
(d) Derive another input-output table consisting of the input coefficients.
(e) Interpret the values of the input coefficients.

Sectors Input by Industries Final Total Output

Demand
𝟏 𝟐 𝒇
𝟏 50 150 100 𝑅1
𝟐 100 250 150 𝑅2
Labour 150 100
Total Input 𝐶1 𝐶2

6
Solution to Example 1:

Based on the information in the input-output table we have:

(a) How many sectors does this economy have?

Ans: 3 sectors (2 inter-industry sectors and 1 final demand sector)
(b) Complete the table.

Sectors Input by Industries Final Total Output

Demand
𝟏 𝟐 𝒇
𝟏 50 150 100 300
𝟐 100 250 150 500
Labour 150 100
Total Input 300 500

(c) Generate the technical coefficient matrix of the economy.

Sectors Input by Industries

𝟏 𝟐
𝟏 0.1667 0.3
𝟐 0.3333 0.5
(d) Derive another input-output table consisting of the input coefficients.
Ans: Same as above
(e) Interpret the values of the input coefficients.

7
1.4. Impact Analysis

The relationship between the total output and the combined inputs of the product of each
sector as shown in the input-output table can be described by a set of n (i.e., number of
industries) linear equations as follows:

𝑥1 = 𝑥11 + 𝑥12 + 𝑥13 + ⋯ + 𝑥1𝑗 + ⋯ + 𝑥1𝑛 + 𝑓1

𝑥2 = 𝑥21 + 𝑥22 + 𝑥23 + ⋯ + 𝑥2𝑗 + ⋯ + 𝑥2𝑛 + 𝑓2

𝑥3 = 𝑥31 + 𝑥32 + 𝑥33 + ⋯ + 𝑥3𝑗 + ⋯ + 𝑥3𝑛 + 𝑓3

(Eqn. 8)
⋮ ⋮ ⋮

𝑥𝑖 = 𝑥𝑖1 + 𝑥𝑖2 + 𝑥𝑖3 + ⋯ + 𝑥𝑖𝑗 + ⋯ + 𝑥𝑖𝑛 + 𝑓𝑖

⋮ ⋮ ⋮

𝑥𝑛 = 𝑥𝑛1 + 𝑥𝑛2 + 𝑥𝑛3 + ⋯ + 𝑥𝑛𝑗 + ⋯ + 𝑥𝑛𝑛 + 𝑓𝑛

We also know that the input coefficients or technical coefficient is given by

𝒙𝒊𝒋
𝒂𝒊𝒋 = ⇒ 𝒙𝒊𝒋 = 𝒂𝒊𝒋 × 𝑿𝒋
𝑿𝒋

𝑥𝑖 = 𝑥𝑖1𝑎+ 𝑥×𝑖2𝑋+ for

Hence, substituting 𝑥𝑖3 𝑥+ ⋯ + 𝑥𝑖𝑗 + ⋯ + 𝑥𝑖𝑛 + 𝑓𝑖
𝑖𝑗 𝑗 𝑖𝑗 in (eqn. 8), we obtain

𝑋1 = 𝑎11 𝑋 + 𝑎12 𝑋2 + 𝑎13 𝑋3 + ⋯ + 𝑎1𝑛 𝑋𝑛 + 𝑓1

𝑋2 = 𝑎21 𝑋 + 𝑎22 𝑋2 + 𝑎23 𝑋3 + ⋯ + 𝑎2𝑛 𝑋𝑛 + 𝑓2

⋮ ⋮ ⋮ (Eqn. 9)

𝑋𝑛 = 𝑎𝑛1 𝑋 + 𝑎𝑛2 𝑋2 + 𝑎𝑛3 𝑋3 + ⋯ + 𝑎𝑛𝑛 𝑋𝑛 + 𝑓𝑛

Making the subject of each the equations in (eqn. 8), we get:

𝑓1 = 𝑋1 − (𝑎11 𝑋1 + 𝑎12 𝑋2 + 𝑎13 𝑋3 + ⋯ + 𝑎1𝑛 𝑋𝑛 )

𝑓2 𝑥=𝑖 =
𝑋2𝑥− + 21
𝑖1 (𝑎 𝑥𝑖2𝑋1++𝑥𝑖3
𝑎22+𝑋⋯ + 𝑥𝑖𝑗 𝑋+3 ⋯
2 + 𝑎23 ++⋯ 𝑥+𝑖𝑛𝑎+ 𝑓𝑖 𝑛 )
2𝑛 𝑋

⋮ ⋮ ⋮ (Eqn. 10)

𝑓𝑛 = 𝑋𝑛 − (𝑎𝑛1 𝑋1 + 𝑎𝑛2 𝑋2 + 𝑎𝑛3 𝑋3 + ⋯ + 𝑎𝑛𝑛 𝑋𝑛 )

𝑥𝑖 = 𝑥𝑖1 + 𝑥𝑖2 + 𝑥𝑖3 + ⋯ + 𝑥𝑖𝑗 + ⋯ + 𝑥𝑖𝑛 + 𝑓𝑖

Simplifying the Right Hand Side (RHS) of each equation in (eqn.10) we obtain:

Making the subject of each the equations in (eqn. 8), we get:

𝑓1 = 𝑋1 (1 − 𝑎11 ) − 𝑎12 𝑋2 − 𝑎13 𝑋3 − ⋯ − 𝑎1𝑛 𝑋𝑛 )

𝑓2 = −𝑎21 𝑋1 + 𝑋2 (1 − 𝑎22 ) − 𝑎23 𝑋3 − ⋯ − 𝑎2𝑛 𝑋𝑛 )

⋮ ⋮ ⋮ (Eqn. 11)

𝑓𝑛 = −𝑎𝑛1 𝑋1 − 𝑎𝑛2 𝑋2 − 𝑎𝑛3 𝑋3 − ⋯ + 𝑋𝑛 (1 − 𝑎𝑛𝑛 )

Now given the matrix notation

𝑥𝑖 = 𝑥𝑖1 + 𝑥𝑖2 + 𝑥𝑖3 + ⋯ + 𝑥𝑖𝑗 +𝑓⋯ + 𝑥𝑖𝑛 + 𝑓𝑖

𝑋1 1
𝑋2 𝑎11 𝑎12 … 𝑎1𝑛 𝑓2
𝑋3 𝑎21 𝑎22 … 𝑎2𝑛 𝑓3 1 ⋯ 0
𝑿 = ⋮ ; 𝑨 = 𝑎31 𝑎32 … 𝑎3𝑛 ; 𝒇 = ⋮ ; 𝐼 = [ ⋮ ⋱ ⋮]
𝑋𝑖 ⋮ ⋮ ⋱ ⋮ 𝑓𝑖 0 ⋯ 1
⋮ [𝑎𝑛1 𝑎𝑛2 … 𝑎𝑛𝑛 ] ⋮
[𝑋𝑛 ] [𝑓𝑛 ]

Where 𝐼 is the identity matrix.

(Eqn. 9) can be written as

𝑨𝑿 + 𝒇 = 𝑿 (eqn. 12)

Which can be rearranged as:

𝒇 = 𝑿 − 𝑨𝑿
𝒇 = (𝑰 − 𝑨)𝑿 (eqn.13)

Making X the subject gives:

𝑋 = (𝑰 − 𝑨)−𝟏 𝒇 (eqn. 14)

It should be noted that (eqn. 14) is the matrix representation of the Leontief’s Input-output
analysis where (𝑰 − 𝑨)−𝟏 is known as the Leontief Inverse or the total requirement matrix.

Also (𝑰 − 𝑨)−𝟏 can be explained as the amount of inputs needed from each sector to satisfy a
given demand of products from one or several sectors.

9
1.5. The multiplier Table

The column sums of the (𝑰 − 𝑨)−𝟏 are known as the Leontief multiplier or the output multiplier.
The values in the multiplier table represents how much output (sales) the row industry will
produce per a unit input (purchases) from the column industry.

Example 2:
Using the information in the input-output table below to answer the following questions:

(a) Complete the table.

(b) Generate the technical coefficient matrix of the economy.
(c) Find the Leontief inverse
(d) Find the multiplier table
(e) Interpret the values of the multiplier table.

Sectors Purchases by intermediate users Final Total

(Sectors/industries) Demand Output/Sales
by Sectors
Agriculture Manufacturing All 𝒇
others
Agriculture 3 44 14 28 𝑅1
Manufacturing 32 132 284 914 𝑅2
All others 10 211 374 1059 𝑅3
Total value 44 975 982
added
Total 𝐶1 𝐶2 𝐶3
Supply/Input
by Sectors

10
Solution to Example 2:
Using the information in the input-output table below to answer the following questions:

(a) Complete the table.

Sectors Purchases by intermediate users Final Total
(Sectors/industries) Demand Output/Sales
by Sectors
Agriculture Manufacturing All 𝒇
others
Agriculture 3 44 14 28 89
Manufacturing 32 132 284 914 1862
All others 10 211 374 1059 1654
Total value added 44 975 982
Total Supply/Input 89 1862 1654
by Sectors

(b) Generate the technical coefficient matrix of the economy.

𝑎11 𝑎12 𝑎13 0.0337 0.0236 0.0085
𝑎
[ 21 𝑎22 𝑎23 ] = [0.3596 0.3394 0.1717]
𝑎31 𝑎32 𝑎33 0.1124 0.1133 0.2261
(c) Find the Leontief inverse matrix ((𝑰 − 𝑨)−𝟏 )

The Leontief matrix is given as (𝐼 − 𝐴)

1 0 0 0.0337 0.0236 0.0085 0.966 −0.0236 −0.0085
(𝐼 − 𝐴) = [0 1 0] − [0.3596 0.3394 0.1717] = [−0.3596 0.661 −0.1717]
0 0 1 0.1124 0.1133 0.2261 −0.1124 −0.1133 0.774
Hence:

0.966 −0.0236 −0.0085 −1 1.05 0.04 0.02

(𝐼 − 𝐴)−1 = [−0.3596 0.661 −0.1717] = [0.64 1.60 0.36]
−0.1124 −0.1133 0.774 0.25 0.24 1.35
(d) Find the multiplier table

Sectors Purchases by intermediate users

(Sectors/industries)
Agriculture Manufacturing All others
Agriculture 1.05 0.04 0.02
Manufacturing 0.64 1.60 0.36
All others 0.25 0.24 1.35
Total 1.94 1.88 1.73
(e) Interpret the values of the multiplier table.
11
1.6. The multiplier Effect

The multipliers show the potential stimulus (i.e., change or impact) to the economy if the
output of that sector is increased by unit. Hence, we can analyse the impact that a change
in final demand in each sector will have on the whole economy.

For example, in example 2, assume the final demand in each sector doubles, what will
be the output in the agriculture, manufacturing and other sectors?
Recall from (eqn. 14) that output is given by:

𝑋 = (𝑰 − 𝑨)−𝟏 𝒇

If final demand for each sector doubles, then it implies 𝑓 = 2𝑓, since final demand is
denoted by 𝑓. Thus, we are required to find:

𝑋 = (𝑰 − 𝑨)−𝟏 𝟐𝒇

Hence, the total demand (output) vector will be as follows:

1.05 0.04 0.02 28
𝑋 = [0.64 1.60 0.36] × 𝟐 × [ 914 ]
0.25 0.24 1.35 1059

1.05 0.04 0.02 56

𝑋 = [0.64 1.60 0.36] × [1828]
0.25 0.24 1.35 2118
178.10
𝑋 = [3723.5]
3817.9
Hence, the output in the agriculture sector would be 178.10 million units, Manufacturing
will be 3723.5 million units, and other sectors will be 2118 million units.

12