CLARE GATHOGA KENYATTA UNIVERSITY

SCHOOL OF ECONOMICS
DEPARTMENT OF APPLIED ECONOMICS
EAE 201: ECONOMICS OF PUBLIC EXPENDITURE

COURSE INSTRUCTOR:

This course is the first part of public sector economics. It introduces students to welfare
economics and analysis of key issues in public expenditure policies that affect individual
behaviour and how the government sets policies. The second part of public sector economics;
taxation economics and policy will be taught at level 300. The course being an applied area
of economics requires a proper understanding of economic theory. This course provides a
basis for understanding other courses in public sector economics. Besides providing a
rationale for government programmes, the course includes discussions on how the public
sector activities impact on the individual economic agents and on the economy as well. By
the end of the course students are expected to be able to demonstrate good understanding of
how the government works and how to analyse public sector expenditure problems.

Continuous assessment tests and any other forms of course assessments will be given
according to schedules set by the instructor. All students are expected to demonstrate high
integrity during such exercises as those who attempt to cheat in their work will face severe
penalties. Schedule for end of semester examinations are determined by Kenyatta
University’s directorate of teaching programmes and examinations office.
Course grades are determined by cumulative performance over all assignments, continuous
assessment tests and a final exam. The distribution of marks for grading and the grading scale
are as follows:
CATs and Assignment 30%: End of semester examinations 70%.

NB: All students must attend all lectures as scheduled. There Will Be No Make-Up Tests

REFERENCES
1. Musgrave and Musgrave (1989): Public Finance in Theory and Practice. 5thed. New
York, McGraw-Hill
2. Artkinson, A.B and J.E, Stiglitz (1980): Lectures in Public Economics. New York,
McGraw-Hill
3. Stiglitz, J.E (2000): Economics of the Public Sector. 3rded. New York, Norton
4. Rosen H (2005): Public finance. 7th ed. McGraw-Hill
5. Any other relevant Public Sector Economic (Public Finance) books you may find in
the library.

1
COURSE CONTENT
1: INTRODUCTION TO PUBLIC SECTOR ECONOMICS
 Definition of Public Sector
 Importance of the Public Sector
 Measuring the size of public sector
2. OVERVIEW OF WELFARE ECONOMICS
 Meaning of welfare economics
 Pareto optimal criteria for social welfare
 Perfect competition and economic efficiency
 Theory of second best
3. MARKET FAILURE: THE RATIONALE FOR GOVERNMENT INTERVENTION
 Meaning of market failure
 Sources of Market failure
 Roles of Government
 Government Failure: Government Failure defined and Causes of Government
failure
4: THEORY OF PUBLIC GOODS
 Definitions of public; private and impure goods
 Other type of goods: merit goods, club goods
 Characteristics of public goods
 Efficient / optimal provision of public goods
 Free rider problem
5: EXTERNALITIES
 Definitions
 Types of externalities
 Consequences of externalities
 Corrective mechanisms (Private and Public responses)
6: PUBLIC CHOICE THEORY
 Public choice in direct democracy
 Public choice in representative democracy
7: THEORY OF PUBLIC EXPINDITURE
 Meaning of Public expenditure
 Uses of Public expenditure
 Principles of Public expenditure
 Theories of public expenditure growth
 Public expenditure in Africa: Income distribution; Poverty reduction; Social
insurance
8: EVALUATION OF PUBLIC EXPENDITURE PROGRAMMES
 Benefits and Costs from Public enterprises/projects
 Public enterprise pricing
 Evaluation Criteria
9: PUBLIC BUDGET
 Importance of budgeting
 Types of budgets
 Principles of budgeting

2
 TOPIC ONE:
INTRODUCTION TO PUBLIC SECTOR

1.1 Public Sector

Refers to those industries and services in a country that are owned and run by the state e.g.
education services, health services, infrastructure, security, etc.

1.2 Public Sector Economics

Refers to the study of the effects of public expenditure and taxes upon the economy and
incorporates public debt management. This branch of Economics seeks to answer the
following questions;
1. What functions should be performed by government.
2. Does financing of public expenditures through taxation reduce the growth potential
of the economy by increasing inefficiency?
3. How should an economy decide on how public expenditure should grow? Should
government increase expenditure?
4. Do public expenditures and taxation destroy incentives?
5. Does government action make people better off or worse off?
6. What happens to market economic behaviour when one tax is substituted for another
tax with equal yield? Will more private / public goods be produced?
7. Does the public sector allocate resources efficiently between public sector activities?

1.3 Reasons why the Public Sector is required

a. Government regulations and other measures are needed to secure competitive factor
and product markets which ensures efficient response to produce what consumers
want most in the cheapest way possible.
b. Government legal structure is needed to protect and enforce contractual arrangements
and the exchanges needed for competitive market operations.
c. In a market situation where decreasing cost is experience, competition is inefficient
and hence government regulations and other measures are needed to correct the
situation.
d. Public policy is required in order to secure high levels of employment, price level
stability, and desired rates of economic growth.
e. The production and consumption characteristics of certain goods and such that they
cannot be provided for through the market e.g. public goods. The problems of
externalities that lead to market failure is solved if such goods are provided by the
public sector.
f. Social values may require adjustments in the distribution of wealth and income that
result from markets for equity reasons. Such redistribution cannot be realized by
private agents, as no private agent has that power. The government therefore plays a
critical role in this.

3
1.4 Size of the Public Sector

The size of public sector refers to the degree of participation by the government in provision
of goods and services within the economy.

1.4.2 Measuring the Size of Public Sector:

There are four basic ways of measuring the size of the public sector
(i) Share of Government Expenditure in GDP
- This measures the share of total output of goods and services purchased by the
government. Transfer payments should not be included.

Total Government Expenditure

· Therefore SPS 
Gross Domestic Pr oduct
·
(ii) Share of total tax revenue in GDP
- Measures the country’s tax effort i.e. shares of gross income diverted from the
private income system into the public budget
Total Tax Re venue to the government
SPS 
Gross Domestic Pr oduct

- This measure provides the most convenient measure for global comparison of
the public sector activities.

(iii) The share of government contributions in national income

- National income measures the sum total of factor incomes (W, r, R, II), earned
during a given period. Hence to measure the size of the public sector, we get the
proportion of factor incomes that originated from the government economic
activities.
SPS  Total Factor earning in govt
National Income

(iv) The share of government contributions in personal income

- Personal income includes incomes received by households and contains three

government components;
(i) Transfer Payments (Rf)
(ii) Wage and salary earnings from public employment (w)
(iii) Interest receipts (r).

4
- The sum total of these represents the government contribution to the personal
income.

SPS  Total Govt Contribution to Personal Income

Personal Income
 w  Rf  r
PI

5
 TOPIC TWO

AN OVERVIEW OF WELFARE ECONOMICS

2.1 MEANING OF WELFARE ECONOMICS

Welfare Economics deals with the questions concerning the optimal allocation of inputs
among industries and the optimal distribution of commodities among consumers.One of the
criteria of social welfare is the Pareto-optimal criterion.

2.2 PARETO-OPTIMAL CONDITIONS

This is a situation in which it is impossible to make anyone better off without making
someone worse off. It requires the fulfilment of all the following marginal conditions for
optimal resource allocation.

2.2.1 Optimal allocation of commodities among consumers

The marginal rate of substitution (MRS) between any two commodities must be the same for
any two consumers

MRS1A, B = MRS2A, B i.e. MRS1 = MRS2

If the marginal rates of substitution were unequal, both consumers could benefit by trading.
For example, suppose that the first consumer regards an additional unit of product A as
K IS THE
having the same utility as two extra units of B, whereas the second consumer regards an MU OF
additional unit of product A as having the same utility as three extra units of product B. GOOD B
Then, if the first consumer trades one unit of product A for 2.5 units of product B from the
second consumer, both consumers are better off. ( FOR CONSUMER 1 AND 2 BY 4K AND 0.5K RESPECTIVELY

0 2’s Indifference curves

B2 Z
Edgeworth contract curve of exchange

B1 a

0
1’s Indifference
curve A1 A2

6
The edgeworth curve is a locus of points of tangency of the two consumer’s indifference
curves i.e. where the slopes of the indifference curves are equal.The pareto-optimal condition
implies that commodities (A & B) should be distributed in such a way that consumers are on
their contract curve since the contract curve is composed of points where the marginal rates
of substitution are equal for the consumers.

Points Z is therefore inefficient since redistribution of the commodities such as to reach any
point between aandb increases the utility of both consumers.A movement to point a increases
the utility of individual 2 without reducing the utility of individual 1. A movement to point b
increases the utility of individual 1 without reducing the utility of individual 2. Thus all the
points froma tob represent improvements in social welfare compared to point Z.

2.2.2 Optimal allocation of inputs among producers.

The marginal rate of technical substitution between any two inputs must be the same for any
pair of producers.

MRTS 1K, L = MRTS2K, L

If this condition does not hold, it is possible to increase total production merely by
reallocating inputs among producers. For example, suppose that for the first producer, the
marginal product of input L is twice that of input K, whereas for the second producer the
marginal product of input L is three times that of input K. If the first producer gives one unit
of input L to the second producer in exchange for 2.5 units 0ofB’s isoquant
input K, both firms can expand
their output. If MPL is M1 for the first producer and M2 for the second producer, then the
output of the first producer is reduced by M1 units because of its loss of the unit of input L,
but it is increased by (2.5 x M1/2) units because of its gain of the 2.5 units of input K. The
consequence being that on the balance its output increases by M1/4 units because of the
trade.Similarly, the output of the second producer is increased by M2 units because of its gain
of the one unit of input L, but it is decreased by (2.5 x M2/3) units because of its loss of
the2.5 units ofinput K. On the balance, its output increases by M 2/6 units because of the trade.

Z
KB

b Contract curve of
production
a

A’s isoquant 0 LA LB

7
The Pareto-optimal condition implies that inputs should be allocated so that producers are on
their contract curve of production, since the contract curve is made up of points at which
MRTS are equal for producers.Point Z is inefficient since reallocation of the given K and L
between the producers of A and B such as to reach any point from a to b inclusive, results in
the increase of at least one commodity without a reduction in the other.

2.2.3 Optimal allocation of inputs among producers and optimal allocation of

Commodities among consumer

The marginal rate of substitution between any two commodities must be the same as the
marginal rate of transformation between these two commodities for any producer.

MRS1A,B=MRS2A, B=MRTA, B

Output B

P
T
B
Indifference curve

Output A
O A P1

PP1is the product transformation curve showing the maximum amount of good A that can be
produced, given various output levels for good B. The MRT is the slope of the product
transformation curve. It shows the number of units of good B that the society must give up in
order to obtainan additional unit of good A.

To maximize the consumer’s satisfaction, production must take place at point T where the
output of A is OA and the output of good B is OB. T is the point on the product
transformation curve that is on the consumer’s highest indifference curve. Since the product
transformation curve is tangent to the indifference curve at point T, it follows that the MRT
equals the MRS at this point.Therefore, the MRT = MRS if the consumer satisfaction is
maximized.

8
2.3. COMPETITION AND ECONOMIC EFFICIENCY

The basic argument is that perfectly competitive economy satisfies the three sets of
conditions for welfare maximization. The other argument is that it prevents undue
concentration of power and the exploitation of consumers. The proof that perfect competition
leads to the fulfilment of the three marginal conditions of welfare maximisation is provided
as follows:

2.3.1 Marginal rate of Substitution (MRS) between any pair of commodities must be the
same for all consumers
- Under perfect competition, each consumer chooses purchases so that the MRS
between any pair of commodities is equal to the ratio of the prices of the pair of
commodities;
P
MRS A,B  A
PB

- Under perfect competition, prices are the same for all buyers and hence price
ratios. It follows therefore, that the MRS between any pair of commodities must
be the same for all consumers.

Example

If every consumer can buy bread at Kshs. 25 per loaf and butter at Kshs. 100, then
each one will arrange purchases so that the MRS of butter for bread is four. Thus
the MRS will be the same for all consumers (4 for everyone)

2.3.2 MRTS between any pair of inputs must be the same for all producers.
- This condition is met in perfect competition since producers will choose the
quantity of each input so that the MRTS between any pair of inputs is equal to
the ratio of the prices of the pair of inputs.
- Since input prices, and thus price ratios, are the same for all producers under
perfect competition, it follows that the MRTS must be the same for all
producers.

Example

If every producer can buy labour services at Kshs. 300 an hour and Machine
tool services at kshs. 600 an hour each will arrange the quantity of its inputs so
that the MRTS of machine tool services for labour is two. Therefore the MRTS
will be the same for all producers (2 for each of them)

2.3.3 MRPT must be equal to MRS for each pair of goods.

- Marginal Cost (MC) is the additional cost of producing an extra unit of a good. In
order to get how many units of good A must be given up to get an extra unit of B,
we divide the MCB by the MCA .This measures the MRT.
- Therefore, MRT equals the ratio of the MC of good B to the MC of good A.

9
 MC A
MRT 
MC B

- In perfect competition, price of a good equalsits MC, and MRS is equal to price
ratios. Therefore:

MC A PA
MRT    MRS A,B
MC B PB

Therefore all the three conditions are satisfied under perfect competition, but not under
monopoly because:

- Monopolist produces less than the perfectly competitive industry. Thus too few
resources are devoted to the industries that are monopolized and too many
resources are devoted to competitive industries.
- Monopoly therefore wastes resources because its actions result in an over-
allocation of resources to competitive industries and an under-allocation of
resources to monopolistic industries.
- The society is then less well off.

2.4 INTERTEMPORAL CHOICE

This relates to consumption over time i.e. how one allocates income earned in different time
periods to consumption. When one is young she borrows, income earnings are highest during
a person’s middle age; she can therefore lend for example in the form of investing in a
retirement fund, and earnings fall after retirement. Income is therefore earned in an uneven
pattern and individuals attempt to smooth out their consumption through borrowing and
lending so that their consumption varies less than their income.

The opportunity cost of consuming income (y) this year means forgoing consumption of
y(1  r ) next year. The price of present consumption is therefore (1 + r) units of future
1
consumption. Alternatively, the price of future consumption is units of present
(1  r )
consumption (r is the interest rate). The interest rate is the premium for earlier availability of
goods.The consumer maximizes utility subject to the wealth. Wealth is the present value of
current and future income. An increase in interest rate represents an increase in the price of
present consumption. According to Fisher () People are impatient;that is, they prefer present
consumption to the same amount of future consumption. This means that a given level of
income y will generate less utility if it is consumed in the future rather than in the present
hence there is a time rate of preference.

Dynamic Consistency Propertycan be imposed on inter-temporal utility function. The

property requires that the marginal value of consumption in period i in terms of forgone
consumption in period j be independent of the date. It should only be dependent on the
consumption levels in the two time periods.

10
The Fisher Separation theorem

The consumer will achieve the highest indifference level by first choosing that income stream
that maximize wealth, and then rearranging consumption so as to maximize utility. This
happens only when the lending rate is the same as the borrowing rate.

X2 g (x1, x2) = k

B
u (x1, x2)
A
x2
w  x1 
1 r
0 X1

Assume that an individual can produce a combination of consumption pointg(x1, x2) = k.

With efficient capital market S1, the bundle produced need not be the bundle consumed. The
consumer maximizes utility by first maximizing wealth and then by borrowing or lending (at
some interest rate) moves along the budget (wealth) line to some point of tangency B.At point
A, the consumer maximizes utility subject to a wealth constraint.When borrowing rate
exceeds the lending rate, the Fisher Separation theorem may not hold and maximization of
wealth may not lead to utility maximization.

The optimum Inter-temporal Condition

The marginal temporal rate of transformation between every pair of factors and products as
well as the marginal temporal rate of substitution between every pair of factors and between
every pair of products must be equal to the rate of interest on riskless securities.It relates to
borrowing and lending between producers in the absence of risk or uncertainty and implies
that the rate of interest at which an individual producer is willing to lend a given amount of
capital must equal its marginal productivity to the borrowing producer.

11
CONSUMER AND PRODUCER SURPLUSES

Consumer’s surplus – The difference between what a consumer is willing to pay for a good
and what the consumer actually pays when buying it. It is defined by the area bounded by the
demand curve for a good and the market price of the good.

Welfare losses can be measured by consumer’s surplus.The analysis of welfare losses is an

attempt to have a money measure of the loss in utility incurred from selling commodities at
prices other than marginal cost.

Suppose in the following figure P0is the marginal cost (MC) of producing Q and Q1units of Q
are sold at P1.

A
B
P1

D C
P0 MC

MB

0 Q1 Q0

The traditional analysis asserts that the benefit from consuming Q0 is the area OACQ0. At
price P1, total benefits are supposed to be OABQ1. The difference Q1BCQ0 is partitioned into
DBC and DCQ0Q1.

The later area is an amount of income spent on other commodities, presumed to be sold at
marginal cost. The remaining area DBC is called deadweight loss, a money measure of the
loss due to the price distortions P1- P0. These distortions could be due to excise taxation or
monopolistic sales.

Producer’s surpluson the other hand measuresthe total benefit or value that producers
receive (enjoy) if they produce units at a cost less than the market price i.e. benefits that low
cost producers enjoy by selling at the market price.

12
 P
Consumer surplus
S

P0 MC
Producer surplus

NB: Levying a tax increases price paid by the consumers thereby reducing the consumer’s
surplus. At the same time, it reduces the net price to the producer thereby reducing the
producer’s surplus. The end result is a deadweight loss i.e. welfare loss.

P
St

S
A
P1

P0 B C

PN
D
D
Q
Q1 Q0

The loss in consumer’s surplus is BAC while loss in the producer’s surplus is BCD. The
total BAC + BCD is the loss of welfare due to increased price. It represents the Deadweight
loss.

13
THE THEORY OF THE SECOND BEST

The attainment of a Pareto optimumrequires the fulfilment of all the three marginal
conditions for optimal resource allocation.

(a) The marginal rate of substitution between any two commodities must be the same for any
two consumers for optimal allocation of commodities among consumers
(b) The marginal rate of technical substitution between any two inputs must be the same for
any pair of producers for optimal allocation inputs among producers
(c) The marginal rate of substitution between any two commodities must be the same as the
marginal rate of transformation between these two commodities for any producer for
optimal allocation of inputs and optimal allocation of commodities

The theory of second best is concerned with a situation in which one or more of the
conditions cannot be met. It asks whether, under such circumstances, it is still desirable to
fulfil the other conditions that can be met.For example, in a case where there are a number of
monopolies in an economy and that these monopolies are not fulfilling the conditions,
suppose that the country can break up some but not all, of the monopolies and force them to
fulfil the conditions.The theory of the Second best asks: Can we be sure that this will
increase welfare?The principal conclusion is NO, since some of the conditions for efficiency
remain unfulfilled, there is no assurance that the reduction in the number of unfulfilled
conditions will result in increased welfare.

Implications of theory of second best to Welfare Economics

The theory implies that if some parts of the economy are misbehaving in the sense that they
are not fulfilling the conditions, there is no reason to believe that welfare would be greater if
other parts of the economy were to be convinced (or forced) to fulfil these conditions.The
theory of the Second best states “It is not true that a situation in which more, but not all of the
optimum conditions are fulfilled is necessarily, or is even likely to be superior to a situation
in which fewer are fulfilled.This shows that piece-meal attempts to force fulfilment of the
optimal conditions can easily be a mistake.

Reference

Lipsey O R and K. Lancaster (1956) “The General Theory of Second Best “Review of
Economic Studies P.11 – 32.

14
 TOPIC THREE

MARKET FAILURE: RATIONALE FOR GOVERNMENT INTERVENTION

3.1 Introduction
A market is an exchange institution that serves the society by organising economic activity.
The market use prices to communicate to the wants and limitations of the society so as to
bring about coordinated economic decisions in the most efficient manner. The power of a
perfectly functioning market rests in its decentralised process of decision making and
exchange. There is no central planner needed to allocate resources. Rather, prices ration
resources to those who value them the most and in doing so, individuals are swept along by
the invisible hand of the market to achieve what is best for the society collectively. Optimal
private decisions based on mutually advantageous exchange lead to optimal social outcomes.

Markets can fail if prices do not communicate the society’s desires and constraints accurately.
Prices often understate the full range of services provided by an asset or simply do not exist
to send a signal to the market place about value of the asset. Market failures occur when
private decisions based on the prices or lack of them does not generate an efficient allocation
of resources. Inefficiency implies that resources could be reallocated to make at least one
person better off without making any one else worse off. A wedge is therefore driven
between what individuals want privately and what the society wants collectively.

With market failure, there is a tendency for it to produce too much of some goods and an
insufficient amount of others. In extreme cases, certain goods will not be produced at all.
Government intervenes in an economy on the grounds of market failure.

3.2 Sources of Market Failure

The principal types of market failure are discussed below.

3.2.1 Public goods

A public good is one whose consumption benefits more than one person or firm. Some of
these goods are non-rival, in the sense that providing the good to one person necessarily
allows the good to be provided to every other person at no additional cost.

Other goods are only partially non-rival, in the sense that the quality of the benefit provided
to each person diminishes as the number of people to whom it is provided rises. These goods
are said to be congestible, and are much more common than non-rival goods. Examples of
congestible goods include parks and recreational facilities, police and fire protection, and
roads and bridges. Every public good involves a violation of assumption (i).

Some public goods also have the property that, if they are provided to one person, they are
automatically made available to everyone. Such goods are said to be non-excludable.

15
A pure public good is both non-rival and non-excludable, and hence violates assumptions (i)
and (ii). Competitive firms are unable to provide sufficient quantities of these goods. Non-
excludability means that the firms are unable to set a fee for the use of the public goods that
they provide, and hence can only cover their costs if the users make voluntary payments.
This situation gives rise to the free rider problem.

Each user is confronted with the following choice: he/she can contribute to the provision of
the public good and enjoy its benefits, or he/she can keep his/her money in his/her pocket and
enjoy its benefits anyway. Not surprisingly, people faced with this choice prove to be
reluctant to part with their money. Total contributions are relatively small, so only a small
quantity of the public good is ultimately provided. Every person would be better off if
everyone could be forced to give a little more. Governments, when they finance the provision
of public goods through taxes, are therefore engaging in a socially beneficial form of
coercion.

While the under-provision of public goods takes its most dramatic form when the public good
is “pure”, the provision of less-than-pure public goods is also problematic. If a good is non-
rival but excludable, a private provider of that good can only remain in business by charging
the users a positive price. This practice results in the exclusions of some potential users. The
provider’s interests are at odds with those of society, because society’s welfare is maximized
by excluding no one.

If the good is congestible and excludable, by contrast, society’s welfare is maximized by

excluding some users from each facility. Decisions about exclusions and the facility size must
then be made simultaneously.

3.2.2 Externalities
An externality can occur when a person’s utility is affected by another person’s consumption
or by a firm’s production activities. An externality can also occur when a firm’s profits are
affected by another firm’s production activities or by an individual’s consumption. However,
not all such interactions constitute externalities. An externality only occurs when appropriate
monetary compensation is not made. Appropriate compensation induces the generator of the
externality to take into account the effects of his/her actions on others, so that he/she curtails
harmful activities and extends beneficial ones.

Example 1
You are harmed if your neighbour throws noisy parties that prevent you from sleeping. Your
neighbour is not required to compensate you for the harm done to you, so he doesn’t take
your interests into account: the parties are long and loud. An externality is present here.

16
Example 2
The small stores at a shopping mall benefit from the presence of a large department store.
The department store draws customers to the mall, creating additional business for the small
stores. The leases signed by the stores reflect this benefit: the department store often pays no
rent, and the small stores pay higher rent than they would pay in the absence of the
department store. This arrangement shifts the burden of rent from the department store to the
small stores, so that the department store implicitly receives compensation. No externality
occurs.

3.2.3 Common property resources

A common property resource is a good which is not owned by anyone. Individuals acquire
ownership of a common property resource simply by taking it. Self-interested individuals are
likely to take as much as they can as quickly as they can.

In the case of renewable common property resources, the rush to be first can lead to the
exhaustion of the resource. Many fisheries have been commercially depleted, and others are
threatened with depletion. The commercial values of the whale, the rhinos, the elephant, and
the sea turtle are great enough to threaten these species with extinction.

3.2.4 Co-ordination failures

Co-ordination failures are a particular form of externality, and therefore involve a violation of
assumption 1. In an efficient market economy, market prices convey everything that each
economic agent needs to know about every other economic agent. Consider, for example, the
markets for consumer goods. Each consumer believes that he/she can buy at the prevailing
price as much of each good as he/she likes. Each firm believes that it can sell at the prevailing
price as many units of goods as it wants. The transactions that consumers and firms want to
make depend only upon prices, and they are able to carry them out.

Keynesian economics argues that this picture of the workings of the market economy is
deficient. The quantity of goods that consumers want to buy is determined by their income,
which is in turn determined by the quantity of labour that they can sell. Similarly, the quantity
of labour that firms buy is determined by the quantity of goods that they can sell. A recession,
the Keynesians argue, is a situation in which consumers do not buy goods because they
cannot sell labour, and firms do not buy labour because they cannot sell goods. If this view is
correct, each agent’s behaviour is influenced by quantities as well as prices.

In addition, an agent’s decision to trade in a market might be influenced by his/her estimate

of the probability that other agents will trade in that market. Multiple equilibria are then
possible. There could be an equilibrium in which few people trade because few people are
expected to trade, and another in which many people trade because many people are expected
to trade. Since trading is mutually beneficial, welfare is higher when more people trade.

17
3.2.5 Imperfect competition
A competitive firm expands its production until the price of the last unit of output is just
equal to the market price of the resources needed to produce that unit. If the other firms in the
economy are also competitive, the market price of these resources is just equal to the market
value of the other goods that could have been produced with these same resources. In these
circumstances, consumers learn about their options by examining prices. If a good’s price is
high, they are warned that consumption of this good requires them to forgo other goods that
they themselves believe to be valuable. If a good’s price is low, they are told that the
consumption of this good requires them to forgo something, but not something of any great
value. Consumers use these signals to decide which goods they should consume: specifically,
they consume expensive goods sparingly and cheap goods freely. This mechanism (Adam
Smith’s “invisible hand”) causes the economy’s limited resources to be allocated to the
production of the goods that consumers most want.

This mechanism tends to break down if some firms are large enough to appreciably affect the
prices at which goods are bought and sold. The most extreme case is monopoly, in which
there is only one seller of a particular good. The price set by the monopolist is greater than
the good’s marginal cost of production. Consumers respond by buying fewer units of the
good than they would if its price reflected its marginal cost of production.

Perfect competition is the only form of market organization under which a good’s price is
certain to be equal to its marginal cost. Hence, any violation of assumption 4 is likely to
cause the free market outcome to diverge from the competitive outcome.

The presence of imperfect competition suggests that something has prevented sustained
competition among firms.

One possibility is that production is characterized by increasing returns to scale, meaning that
output more than doubles when the use of all factor of production is doubled. The largest firm
is then able to produce and sell goods more cheaply than its competitors, and will eventually
drive them out of business. Once it is alone in the market, it will restrict its output and raising
its selling price.

A second possibility is that entry into the industry involves such high set-up costs that
potential competitors are unable to raise the necessary financial capital.

It might also be that a necessary patent is possessed by only one firm, ensuring its position as
a monopolist.

3.2.6 Asymmetric information

The price system is an important mechanism because it is decentralized, that is, because
every economic decision is made by the people or firms directly affected that decision. Each
firm knows which goods can be produced at its manufacturing plants, and the various ways in

18
which these goods can be produced, and it makes decisions on exactly these issues. Each
consumer knows his/her own tastes better than anyone else, and decides which goods and
services he/she will purchase. All parties base their own actions on their own information and
their knowledge of market prices. Consequently, they do not need to communicate detailed
information about tastes and production processes to each other.

The value of the price system lies precisely in its ability to exploit information that is not
widely known, but it can only do so if two essential kinds of information are known to
everyone.

First, the market participants must be equally well informed about the nature of the good
being traded. Otherwise, an efficient producer might match the low prices of his/her
more efficient competitors by substituting low quality components for higher quality
components. If consumers were unable to discover the difference between these
products, the less efficient producers would not be driven from the market.

Second, the market participants must be equally well informed about the circumstances under
which the good is traded.

There are many situations in which the participants are not equally well informed, and the
market outcome in these situations might not be efficient. Trades which are mutually
advantageous might not take place, and trades that take place might not be mutually
advantageous. For example, situations arise when you deal with doctors, lawyers, stock
brokers, and garage mechanics. They have better information than you do, but you might be
uncertain as to whether they are advancing your interests or their own.

Market failure can occur when one person in a transaction does not have full information
about either the actions or the type of the second person. Type implies unknown quality of a
good or the hidden characteristics of an agent such as inherent intelligence.

Example
A seller knows more about the quality of a product than a buyer.
A worker has a better idea of how much she could produce than her employer does.

With asymmetric information, if too many a low-quality items are offered for sale, it makes it
difficult for owners of the high-quality items to sell their products.
(Read markets for Lemon Model). Without complete information, markets will be incomplete
and will fail to allocate resources efficiently.

Types of asymmetric information problems

(i) The moral hazard/incentive problem:
This arise when the actions of the person are unobservable to the second person (Also
known as a hidden action problem). Equilibrium in a market involving hidden action
leads to some form of rationing. This is where firms would like to provide more than

19
 they do, but they are unwilling to do so since it will change the incentives of their
customers. For example, in the Insurance Industry, firms discriminate among users
depending on the choices they have made that influence the probability of damage
and need people to take care. But the more the insurance taken, the less the care taken
rationally.

(ii) Adverse selection problem: exists where one side of the market can’t observe the
“type” or quality of the good on other side of the market. Also known as hidden
information problem.

Equilibrium in a market involving hidden information will typically involve too little
trade taking place because of the externality between the “good” and the “bad” types.

The government can compel people for example to purchase insurance if it is possible
to make everyone better off. This is a good case of government intervention.

However, there are costs to government intervention and economic decisions made by
the governmental decree may not be cost-effective as those made by private firms.

Furthermore, there may be purely private solutions to the adverse selection problems.
For example, providing health insurance as a fringe benefit can help to eliminate the
adverse selection problem. The insurance company can base its rates on the average
over the rest of employees and is assured that all employees must participate in the
programme, thus eliminating the adverse selection.

3.2.7 Information Failure

Efficiency requires that information be freely disseminated or that the only charge be for
actual cost of transmitting the information. The private market provides inadequate supply of
information. For example, information on weather can be inadequate if provided by a private
agent.

Regulations on information disclosure are unnecessary, irrelevant (because consumers pay

little attention to the information the law requires firms to disclose) and costly to both the
government to administer effectively. However, they are still useful because information is a
public good that must be provided by the government (e.g. weather, climatic change, etc).
Giving information to one or more individuals does not reduce the amount others have.

3.2.8 The tax system

An efficient tax system is one that does not distort economic decision-making and has low
administrative and compliance costs.

20
However, taxation leads to market failure because it raises the price of the taxed commodity
thereby reducing the quantities that is demanded on the market. At equilibrium therefore less
is demanded.

The tax prevents market interaction among buyers and sellers from automatically equating
MSC and MSB as required attain efficiency.

The total excess burden of a tax is an additional cost to society over and above the amount
paid as tax. The excess burden measures the loss in net benefit from private use of resources
that results when a price-distorting tax prevents markets for taxed goods and services from
attaining efficient output levels.

3.2.9 Income distribution

Pareto efficient says nothing about the distribution of income. Therefore, competitive
markets may give rise to a very unequal distribution of income which may leave some
individuals with insufficient resources on which to live.

Even if the competitive economy is efficient, the distribution of income to which it gives rise
may be viewed as undesirable. Hence the government activity in this case is to alter the
distribution of income. To get more equity, some amount of efficiency must be sacrificed.

3.2.10 Unemployment, inflation and disequilibrium

The presence of unemployment and inflation is evidence that the markets have failed to
produce full employment. The failure of the system to adjust to a full-employment level of
output may be due to:-
(i) An inelastic investment function
(ii) A highly elastic liquidity preference function
(iii) Failure of the investment and savings functions to intersect at a positive rate of
interest
(iv) Downward rigidity of costs and prices which keep the price level from falling thereby
pre-stalling a possible upward shift in the consumption function which might
otherwise result as the real value of money balances rises.

The government has a role to carry out economic stabilization by use of the fiscal and
monetary policies.

3.2.11 Merit goods and demerits goods

The Pareto optimum assumes that individuals are the best judges of their own welfare.
However, allocative inefficiency will occur if individuals undervalue the personal benefits
derived from consumption of a commodity. Individuals may not act in their own best interest.

21
It is argued that evaluating each individual’s welfare by their own perceptions provides an
inappropriate or inadequate criterion for making welfare judgements.

Even if fully informed, consumers may make bad decisions. For example, individuals
continue to smoke even though it is bad for them. Individuals fail to wear seat belts, even
though wearing seat belts increases the chances of survival from an accident. In these cases,
the government need to intervene. Goods that the government compels individuals to
consume, such as seat belts, primary education are called merit goods.

The view that the government should intervene because it knows what is in the best interest
of individuals better than the individuals themselves do is known as paternalism.

The disadvantage with the paternalistic role of government is that special interestgroups will
attempt to use government to further their own view about how individuals should act or
what they should consume.

Demerit goods are those that government wishes to deter their consumption by imposing
penalties. For example, smoking causes cancer and since individuals who get cancer may be
treated in public hospitals or financed by public funds, smoking imposes a cost on non
smokers. Hence the way forward is to impose a tax so that smokers pay the full costs.

The paternalistic role of the government is important in a number of areas such as

government policies toward drugs, liquor (prohibition) and compulsory primary schooling
among others.

3.3 THE ROLE OF GOVERNMENT AND GOVERNMENT FAILURE

3.3.1 Role of Government

The primary role of the government is to provide the legal framework within which all
economic transactions occur.

The government produces public goods and services. In addition it affects production
decisions of the private firms through regulations, taxes and subsidies. The reasons are
varied ranging from dissatisfaction with particular actions of the firm, concern for monopoly
power, special interest groups and the failure of the private sector to provide certain goods.
Examples are electricity generation, etc. The government production is measured by the
value-added, the difference between the value of output and the value of materials purchased.
While the value added for private goods is measured by price, the value-added for public
goods and services are valued at the cost of their inputs.

The government may subsidize production of certain private goods. For example, maize
production, coffee production, etc. Sometimes subsidization is done through tax credit. For
example, if the government allows the individual to reduce her tax payments by the amount
of the education expenditure. The government may also subsidize goods and services through

22
hidden actions. For example, it may restrict the importation of some foreign goods or impose
a tariff on imports which raises prices of those goods in the domestic market. The domestic
producers of competing goods are helped. It also provides subsidies through low-interest
loans and loan guarantees.

Government regulates businesses in an attempt to protect workers, consumers, and the

environment, to prevent anti-competitive practices, and discriminations. Government also
affects financial sector through a variety of regulations and activities and sometimes
establishing bodies to carry out credit and insurance activities.

The government purchases goods and services in order to provide national defence, to
maintain infrastructure, to provide security, fire protection, parks, health services, research,
building houses and urban development projects, maintaining the natural environment etc.

Government redistributes income in the following ways:

(i) Public assistance programmes that provide benefits to those poor enough to
qualify or in terms of disaster. This involves providing them with cash or
providing payment for special services or commodities (in kind benefits)
(ii) Social insurance is another way. This provides benefits to the retired,
disabled, unemployment and the sick.
(iii) Indirect effects of the tax system through progressive taxation, subsidy
programmes and quotas, spending on goods and services. For example
subsidies to the urban bus transport may help the poor.
Therefore the government functions (Public policy objectives) are:
1. Allocation function - The provision of goods and services or the process by which
total resource use is divided between private and public goods and services and by
which the mix of public goods is chosen.
2. The re-distribution function – refers to the adjustment in income and wealth to assure
conformity with what the society considers as fair or just.
3. The stabilization function – refers to the use of budgetary policy as a means of
maintaining high employment, reasonable degree of price level stability and an
appropriate rate of economic growth with an allowance for the effects on trade and
balance of payment.
4. Regulatory role of the government – it enacts and enforces laws of contract. This
ensures that market trades and private exchanges take place smoothly. It also
administers the system of law and justice that regulates individual’s behaviour.

3.3.2. Government failure

The major limitation of government intervention is that it does not necessarily lead to pareto
efficient allocation of resources. When such intervention is not pareto efficient, we have
government (non-market) failure. Examples of government failure abound in many
developing countries. Government failure may be due to the following:

23
 i) The pursuit of self-interest amongst politicians and civil servants rather than operating
on behalf of citizens, which leads to misallocation of resources.
ii) Electoral pressures leading to inappropriate government spending and tax decisions
iii) A tendency to look for short-term solutions to economic problems rather than making
considerable analysis of long-term considerations
iv) Regulatory capture which occurs in situations where the industries under the control
of a regulatory body begin to move policy option so that the outcome is in their favour
v) Disincentive effects created by measures designed to reduce income inequalities or
loss of business competitiveness caused by the introduction of national minimal wage
or the working families’ tax credit.
vi) The environmental impact of government price support for farmers
vii) Imperfect or limited information hence the consequences of many actions are
complicated and difficult to predict.
viii) Disjunction between non-market activities cost and revenue resulting in redundant
and rising cost
ix) Derived externalities
x) Distributional inequities
xi) Limited control over private market responses: The government has very limited
control over the consequences of its actions.
xii) Limited control over bureaucracy – Parliament, Government and local government
design some government agency. The agency may also be responsible for ensuring
that the regulations are enforced.
xiii) Limitations imposed by political processes – choosing among possible actions
through the political process would raise additional difficulties. This is because
Government actions affect many persons, but they are decided upon by only limited
group i.e. elected representatives. The decision makers have to ascertain the
preferences of their constituencies and they have to find some way of reconciling or
making choices among conflicting preferences. Government may therefore act in an
inconsistent manner.
xiv) Distortionary taxes - If the government had perfect information about the
characteristics of each individual in our society, there is a strong argument that it
would not impose distortionary taxes. Individuals who can more easily pay taxes
should pay more in taxes than those who cannot easily pay. If the government could
ascertain who had greater abilities, and who therefore was in a better position to pay
taxes, it would simply impose high lump-sum taxes on those individuals. The
government can base its tax only on observable variables such as income and
expenditure. The choice is either to have a uniform lump-sum tax or to have a tax that
depends on easily measured variables, such as expenditures or wages and such taxes
are distortionary. For example, an income tax does not always succeed in taxing those
who ought to be taxed. This is because it treats equally the individual who has low
ability but works hard and the individual who is of high ability, and takes it easy,
provided the two have the same income.
Therefore, taxation can easily:-
- Distort economic decision-making
- Lead to excessive administrative and compliance costs
- Inconsistent with macroeconomic policy

24
 TOPIC FOUR
THEORY OF PUBLIC GOODS

4.1 Definitions

A pure public good is a good which is non-rival and non-excludable in consumption.

 Non-rival consumption implies that once the good is provided, the additional resource
cost of another person consuming the good is zero.
 Non-excludable means that it is not possible to prevent anyone from consuming the
good once provided because it is impossible (preventing one from enjoying clean air)
or it very expensive to do so.

A pure private good is a good which is rival in consumption and excludable. Only those who
pay for the good enjoys utility from it

An impure good (club good) is a good which is non-rival in consumption but excludable.

Non-excludability characteristics of public goods

In the case of pure private good a set of property rights define the ownership of the good.
The individual who possess the property right has the sole claim to enjoy the benefits of the
good and can therefore exclude others from doing so. In the case of a pure public good
technical feature of excludability begin to breakdown.

- First, it is generally difficult to exclude individuals from enjoyment of public good. If

for example a geographical area is provided with defence services which diverts an
attack from abroad it becomes extremely difficult to exclude anyone who lives in the
country from being defended. Similar example is found in street lighting.

For pure public good the degree of exclusion depends upon the technical
characteristics of the good and the resources available to the producer to enforce the
exclusion. In general, however, there is no perfect exclusion. So an optimal amount
of exclusion is a decision to be made by a producer.

- Second reason why the exclusion principle breaks down is that, while it may be
technically feasible to exclude, the application of exclusion device may be very
expensive. That is, the cost of exclusion can outweigh any advantages to be obtained
from its application. A pure public good is the one for which exclusion is either
technically not feasible and if feasible, the cost of enforcing the exclusive device is
too prohibitive to apply.

With non-excludability, there is no incentive for a profit maximizing producer to

supply the public good because once he produces it he cannot exclude individuals

25
 from consuming it and hence he is unable to charge a price. Individuals wishing to
consume the benefits of a pure public good could, however, form a private co-
operative. They could agree to contribute to the cost of supplying the public good.
Such an arrangement might be feasible for a small group of individuals, but as the
group grows in size the possibility of individuals becoming free riders increases and
the private voluntarily arrangement fails.

Non-rivalness in consumption
Definition of a pure public good implies that it is non-rival in consumption. This means that
that it does not cost anything for an additional individual to enjoy the benefit of public goods.

- Non-rivalness arises from the indivisibility of public goods. That is, adding one or
more persons (up to a capacity constraint) does not add to the marginal cost.
Formally there is zero marginal cost for an additional individual to enjoy a good.
Non-rivalness thus implies, one individual access to the commodity does not reduce
another individual’s benefit because these benefits are available to all without
interference.

- A perfect solution in this case would require a zero price because marginal cost equals
zero. This means that revenues will not cover losses and so a private profit
maximizing producer will not supply such a commodity. The market, in other words,
will not allocate such goods efficiently thus the market failure.

Goods can be classified into four cases according to their consumption and excludability
characteristic.

Exclusion

Consumption Feasible Not feasible

Rival 1. Rival and excludable 2. Rival and non excludable

Non Rival 3. Non rival but excludable 4. Non rival and non excludable

Characteristic

Case 1: Private good: Rival in consumption and excludable e.g. a loaf of bread or clothing.
One only consumes the goods after paying for it. Whoever does not pay is excluded. Benefits
are internalised.

Case 2: A good that is rival in consumption but non-excludable.

Example is travelling on a crowded street, traffic jam during rush hours. In this case there is
market failure due to non-excludability or high cost of exclusion.

Case 3: A good that is non-rival in consumption but excludable.

26
Examples are clubs, watching a movie, swimming, education, crossing a bridge that is not
crowded.

Case 4: A pure public good

Examples are, air purification, national defence, street lighting, e.t.c. The good is non-rival in
consumption and non-excludable.

Important aspects of public goods Definition

i. Even though everyone consumes the same quantity of the good, this consumption
need not be valued equally by all. Each individual is the best judge of his or her
welfare or utility.
ii. Consider those house cleaning in an apartment with many college
roommates, which has a public good characteristic to it – everyone
benefits from a clean bathroom, and it is hard to exclude everyone
from these benefits. Yet some students care about cleanliness much
more than others.

(ii) Classification as a public good is not an absolute; it depends on market conditions and
the state of technology.
iii. Think about the services provided by a lighthouse. Once the beacon is
lit, one ship can take advantage of it without impinging on another
ship’s ability to do the same. Moreover, no particular vessel can be
excluded from taking advantage of the signal. Under these conditions,
the lighthouse is a pure public good.
iv. Now suppose that a jamming device were invented that made it
possible to prevent ships from obtaining the lighthouse signal unless
they purchased a special receiver. In this case, the non excludability
criterion is no longer satisfied, and the lighthouse is no longer a pure
public good.
v. A scenic view can be considered a pure public good when there are a
small number of people involved. But as the number of sight seers
increases, the area may become congested. The same “quantity” of the
scenic view is being “consumed” by each person, but its quality can
decrease with the number of people. Hence, the non rivalness criterion
is no longer satisfied.
vi. In many cases, then, it makes sense to think of “publicness” as a matter
of degree. A pure public good satisfies the definition exactly.
Consumption of an impure public good is to some extent rival or
excludable.

27
(iii) A commodity can satisfy one part of the definition of a public good and not the other.
That is, non-excludability and non-rivalness do not have to go together.
vii. For example, consider the streets of Nairobi during rush hour. In most
cases, non-excludability holds, because it is not feasible to set up
enough toll booths to monitor traffic. But consumption is certainly
rival.
viii. On the other hand, many people can enjoy a huge seashore area
without diminishing the pleasure of others. Despite the fact that
individuals do not rival each other in consumption, exclusion is quite
possible if there are only a few access roads.
ix. Characterization of a commodity depends on the state of technology
and also on legal arrangements. In order to deal with the road
congestion problem, a technology now exists which involves the use of
radio waves to identify passing cars and automatically charge tolls to
drivers’ charge accounts. One can use such technology to charge cars
as they enter congested city streets and the streets would become
excludable.

(iv) A number of things that are not conventionally thought of as commodities have public
good characteristics.
x. An important example is honesty. If each citizen is honest in
commercial transactions, all of the society benefits due to the reduction
of the costs of doing business. Such cost reductions are characterized
both by non-excludability and non-rivalness.
xi. Similarly, Thurow (1971) argues that income distribution is a public
good. If income is distributed “fairly”, each person gains satisfaction
from living in a good society, and no one can be excluded from having
that satisfaction. Consumption of the income distribution is non-rival
and non-excludable, and therefore it is a public good.
xii. Certain types of information provide other examples of a public good.

(v) Private goods are not necessarily provided exclusively by the private sector. Neither
are public goods exclusively provided by the public sector
xiii. There are many publicly provided private goods – rival and excludable
commodities- that are provided by governments e.g. medical services
and housing are two examples of private goods sometimes provided
publicly. Similarly, public goods can be provided privately. E.g.
individuals donating money to maintain the beautiful flowers in public
spaces.
xiv. Therefore, the label private or public does not by itself tell us anything
about which sector provides the item.

28
4.2 Efficient / optimal provision of public goods

One goal of the public sector or government is to ensure the efficient allocation of resources
to both public and private goods. Some goods and services have to be supplied by the public
sector, since it would be difficult, if not impossible, to have them supplied by the private
sector in a market economy.

There are several approaches to the pure theory of public goods, but all are based on a fairly
common set of assumption and facts. The most fundamental fact that can be identified is the
existence of certain goods that, when “consumed” or enjoyed by one person, do not reduce
the enjoyment (technically) of someone else who wishes to e.g. national defence). This is in
contract to most other goods (private goods), where, for example, if there is one apple
between two people, A and B, consumption of the apple by A means that is not available to B
for consumption.

Formal definitions of pure public goods are manifold.

1. Paul Samuelson has referred to them as goods that have the property that “each
individual’s consumption of such a good leads to no subtraction from any other
individual’s consumption of that good.”
2. Richard Musgrave has stated that such goods “Must be consumed in equal amount
by all.”
3. John Head has put it slightly differently by saying that “once produced, any given
unit of the good can be made equally available to all.”

The non-rival nature of social good consumption makes it virtually impossible to optimally
price such goods even if provided by government.

In this section we shall derived the conditions for efficient provision of a public good, and
then compare them with those of private good.

Assume that there are two individuals, A and B each having a conventional downward
sloping demand curve for some public goods. That is, if the public goods could be sold in
units at a price, each individual would demand more as the price is reduced. These two
demand curves are shown as DA and DB in the right side of figure 2.1. Drawing such
demand curves is based on the unrealistic assumption that consumers volunteer their
preferences, and such curves have therefore been referred to as “Pseudo-demand curves”.

29
Price of
DB DA + B SOCIAL GOODS
Social
goods
DA
E MC = S
P1

V
J

DB
DA
O Z Z1 Quantity of
Social goods

Price of PRIVATE GOODS

DB
Private
goods

E
P2 MC = S

DA + B

DB
DA
O F G H Quantity of Private
goods

30
In the case of public good, once a unit is produced, is consumed equally by all. Thus no one
person can vary the quantity to be taken. Therefore to derive the total demand for public
good, the demand curves are added vertically, not horizontally, as would be the case for a
private good.

Suppose a quality OZ of the public good is made available to A. It is also made available in
the same quantity to B. What we want to know is how much would A and B, together, be
willing to pay for OZ of the public good. To get this add ZJ (what individual A is willing to
pay for OZ) and ZH (what individual B is willing to pay for OZ) to obtain ZE. This is done
for each and every amount of the public good, giving us the total demand curve DA + DB
(addition is vertical). This tells us how much A and B, together, would be willing to pay for
various amounts of the pure public good.

It should be noted that for each amount of public good supplied, individual B is willing to pay
more than individual A for that amount.

Since it takes resources to produce the public good, we introduce for convenience a constant
marginal cost (supply schedule) in our diagram, and from its intersection with the total
demand curve, obtain the equilibrium quantity and price, in this case OP1 and OZ.

Thus we see that the sum of the marginal evaluation by each person for the public good
equals the prices, which equals marginal cost, thereby meeting the conditions of optimal
pricing, that is, that price equal marginal cost.

 MB i  MC  P

In the case of private goods as can be depicted in the left side of figure 2.1, in a perfectly
competitive market situation, prices are fixed (market determined) and the consumers would
only vary amount they consume. So the total demand curve for private good (DA + DB) is
obtained by horizontal addition of DA and DB. That is, adding the quantities which A and B
purchase at any given price. Given the supply schedule, the equilibrium is determined at E,
the intersection of market demand and supply. At the equilibrium, Price equals OP2 and
output OH, with OF purchased by A and OG by, B, where OF + OG = OH.

In a perfectly competitive market, equilibrium requires that the price of a good be equal to
each consumer’s marginal evaluation, which in turn is equal to marginal cost. The consumer
theoretically adjusts his purchase to achieve this, since he faces a fixed price.

In the public goods, case, the price does not equal each consumer’s marginal evaluation and
the consumer cannot vary his purchase to achieve this.

NB: For a non-excludable public good, however, there may be incentives for people to hide
their two preferences. Individual A may falsely claim that the public good means nothing to
him. If he can get individual B to foot the entire bill, he can still enjoy the benefits from the
public goods and yet have more money to spend on private goods. In the figure presented
above, at output OZ1, individual B meets the entire Price (Z, V). This incentive to let other
people pay while you enjoy the benefits is known as the free rider problem. Hence, there is a
31
good chance that the market will fall short of providing the efficient amount of the public
goods. No automatic tendency exists for markets to reach the efficient allocation.

Even if consumption is excludable, market provision of a public good is likely to be

inefficient. Recall the fact that Pareto efficiency requires that Price (P) = Marginal cost (MC).
Because a public good is non rival in consumption by definition, the marginal cost of
providing it to another person is zero. Hence efficiency requires a price of zero. But if the
entrepreneur charges everyone a price of zero, then we cannot stay in business.

Is there a way out?

Suppose that the following two conditions hold

(i) The entrepreneur knows each person’s demand curve for the public good and
(ii) It is difficult or impossible to transfer the good from one person to another.

Under these two conditions, the entrepreneur could charge each person an individual price
based on willingness to pay a procedure known as perfect price discrimination. From our
example, because individual B values the public good most, we would pay a higher price and
thus the entrepreneur would still be able to stay in business.

Perfect price discrimination may seem to be the solution until we recall that the first
condition requires knowledge of everybody’s preference. But of course, if individuals’
demand curves were known, there would be no problem in determining the optimum
provision of public good as was earlier demonstrated.

4.3 Free rider problem

In economics, collective bargaining, and political science, free riders are actors who
consume more than their fair share of a resource, or shoulder less than a fair share of the costs
of its production. The free rider problem is the question of how to prevent free riding from
taking place, or at least limit its negative effects.

Because the notion of 'fairness' is controversial, free riding is usually only considered to be
an economic "problem" when it leads to non-production or under-production of a public
good, and thus to Pareto inefficiency, or when it leads to the excessive use of a common
property resource.

A common example of a free rider problem is defense spending: no person can be excluded
from being defended by a state's military forces, and thus free riders may refuse or avoid
paying for being defended, even though they are still as well guarded as those who contribute

32
to the state's efforts. Therefore, it is usual for the government to avoid relying on volunteer
donations, using taxes and conscription instead.

In the labour union context, a free rider is an employee who pays no union dues or agency
fees, but nonetheless receives the same benefits of union representation as dues-payers. For
example, under law, unions owe a duty of fair representation to all workers they represent,
regardless of whether they pay dues. Some jurists have questioned the fairness, if not the
legality, of this practice.

Free Riding is also a term used by brokerages when a client purchases shares beyond his or
her means. In other words free riders are those that purchase shares that do not pay for them.
The brokerage also places these trades in “Good Faith” a trade to be placed when funds are
not available in the account.

Example

Imagine a street in which 25 people are living. There is a chance to install a CCTV system to
improve security, the cost of which is $2,500. Each person may be prepared to pay $100 or
more for the benefit of extra security. However, since if the system is installed everyone will
still benefit; it is quite possible that some people on the street will refuse to pay.

Despite the fact they may be prepared to contribute $100, they will claim that they are not
prepared to pay, and instead hope that others in the street will be prepared to pay for the
system anyway, and they receive the benefit for no personal expense.

The result is that it is possible no system will be installed, an example of Market Failure. This
is despite the fact that allocative efficiency would be improved

33
 TOPIC FIVE

EXTERNALITIES

5.1 Meaning of Externality

Externalities occur when one person's actions affect another person's well-being and the
relevant costs or benefits are not reflected in market prices. An externality exists if the
consumption or production activities of one individual or firm affect another person’s utility,
or firm’s production function, so that the conditions of a pareto - optimal resource allocation
are violated. It does not work through a market price but rather through its impact on the
production of utility or profit. The set of markets is incomplete in that there is no exchange
institution where the person pays for the external benefits or pays for imposing the external
costs. The benefits or costs of the third party are not considered by either buyers or sellers of
an item whose production or use results in an externality. In the presence of externality the
market prices do not accurately reflect either all the marginal social benefit or all the
marginal social costs of traded items.

5.2 Types of Externality

5.2.1 Negative externality(External costs)

These are costs to third parties other than the buyers and sellers not reflected in the market
price e.g. damage done by pollution to persons and their property. When polluting, factory
owners may not consider the costs that pollution imposes on others. A negative externality
therefore arises when one person's actions harm another.

5.2.2 Positive externality (external Benefit)

This occurs where one individual’s actions confer benefit upon others. Hence are benefits to
third parties, other than the buyers or the sellers of a good or service, not reflected in market
price.Buyers and sellers of the goods that when sold result in positive externalities, do not
consider the fact that each unit produced provides benefits to others. For example, when an
individual cleans up the yard where he or she lives, the neighbours benefit from the clean
environment which is a positive externality. Similarly, the purchase of smoke alarms and
fire proof materials is likely to benefit those other than the buyers and sellers, by reducing the
risk of the spread of fire.

5.2.3 Pecuniary externalities

These are effects of increases (or decreases) in the price of a good on existing consumers as a
result of changes in the demand and supply of a good.

34
Important Observations from the definitions
1. Public goods can be viewed as a special kind of externality. Specifically, when an
individual creates a positive externality with full effects felt by every person in the
economy, the externality is a pure public good. At times, the boundary between public
goods and externalities is a bit fuzzy. Suppose that I install in my backyard a device for
electrocuting mosquitoes. If I kill the whole community’s mosquitoes, then I have, in
effect, created a pure public good. If only a few neighbours are affected, then it is an
externality. Although positive externalities and public goods are quite similar from a
formal point of view, in practice it is useful to distinguish between them.
2. Government action is itself a source of externalities, since routinely government
charges some people (taxpayers) and benefit others without much direct connection
between the two. The best we can hope for is that one externality may balance another, as
it does in the case of free provision of public goods. Perhaps this balancing act should be
no surprise, since we live in an imperfect world. But this balancing sets a pretty high
standard for government, and some economists (and others) doubt that real governments
are capable of meeting the standard.
3. External costs and benefits are the costs and benefits that decision-makers do not take
account of, so market decisions on the allocation of resources do not reflect the external
costs and benefits. But, of course, external costs and benefits are only part of the total
costs and benefits of any decision. The costs and benefits that decision-makers do take
account of, because they pay the costs and enjoy the benefits are called privatecosts and
benefits.
4. Social costs are the sum of private and external costs: Cost = Private Cost + External cost
5. Social benefits are the sum of private and external benefits: Social Benefit = Private
Benefit + External benefit

Let's sum up this terminology in a table:

External and Social Costs and Benefits: Terminology

External Private Social

Beneficiary doesn't
Benefits beneficiary pays total of both
pay

loser isn't
Costs loser is compensated total of both
compensated

Social costs and benefits are so-called because they are the total costs and benefits for
everybody in society. Economists do not think of society as an independent actor that can
gain benefits or suffer costs. Rather, the social costs are the sum total of all costs to

35
individuals in society, regardless of whether the costs are paid by the person who decides
whether they will be incurred. Similarly, social benefits are the sum total of all benefits to
individuals in society, regardless of whether the beneficiaries decide how much benefit will
be produced.

5.3 Consequences of externalities: Externalities and efficiency

5.3.1 Positive externality
When a positive externality is present, prices do not fully equal the marginal social benefit of
a good or service. The marginal social benefit (MSB) is the sum of the marginal private
benefit (MPB) received by consumers and the marginal external benefit (MEB) to others:
MSB = MPB + MEB. The marginal conditions for efficiency are met when MSB = MSC.

Example
Suppose inoculation against a disease results in a positive externality. Others benefit by
reducing the probability that they will contract a contagious disease.

Z
Price 45
S=MSC
Benefits
30 V

25 U

10 H
MPB+MEB=MSB

D=MPB

10 12
Inoculations per year

If inoculation services are sold in a competitive market, the market equilibrium will be point
U at which MPB = MSC.

Suppose that at point U, 10 m inoculations are sold at Kshs. 25 each. If the MEB is Kshs. 20
for each inoculation, it follows that at point U, the actual MSB exceeds the Kshs. 25 price
each consumer uses in deciding whether to be inoculated. The actual MSB = MPB + MEB =
25 + 20 = 45

Since at 10 million inoculations MSB > MSC, more of the inoculations should be consumed
until MSB = MSC. The efficient annual output corresponds to point V, at which 12m
inoculations would be consumed per year.

36
This result shows that with positive externality competitive markets will produce a
commodity at levels that are less than the socially optimal.

At V, the MSC is Kshs. 30. To get to this point the price of inoculations would have to
decrease to Kshs. 10, which corresponds to point H on the market demand curve and the
increase in net benefits would be area UZV.

The distance between MSB and the MPB curve decreases because MEB declines with output.
Thus with positive externalities, if marginal value declines with output, competitive markets
fail to perform efficiently only at low levels of output.

5.3.2 Negative externalities

When a negative externality exists, the price of a good or service does not reflect the full
MSC of resources allocated to its production. Suppose for example, that in the production of
an item each unit of output results in cost to parties other than the buyers or the sellers of the
product. Neither the buyers nor the sellers of the good consider these costs to third parties.
The marginal external cost (MEC) is the extra cost to third parties resulting from production
of another unit of the good or service. The MEC is part of the MSC of making a good
available. However, it is not reflected in the price of the good because producers base their
decisions on the marginal private costs (MPC) of producing an item.

To obtain the marginal social cost, the marginal external cost (MEC) of output must be added
to the marginal private costs (MPC): MSC = MPC+ MEC and the condition for efficiency is:
MSC (MPC + MEC) = MSB

Example

Assume that a paper industry operates under perfect competition. The current price is Kshs.
100 per ton, and the industry produces 5 million tons per year at that price as shown in the
figure below.

4.5 5

37
 MPC+MEC=MSC

S=MPC
G
Price
110
10
Benefit B
105
Cost
100 A

D+MSB

0
4.5 m 5m Tons of Paper

The demand curve D is based on marginal benefits that buyers receive from each ton of
paper, which is also the MSB. The supply curve is based on the marginal cost (MC) actually
incurred to produce additional units. But MC does not include all the costs incurred in
producing extra units of paper. Suppose that a MEC of Kshs. 10 is associated with each ton
of paper produced (i.e. from emissions/pollutants). The negative externality stems from the
fact that dumping industrial wastes in the stream decreases its usefulness to other users. The
MEC is not considered in the producer’s choice of output. To obtain the MSC of paper,
MEC must be added to MPC for each possible output. Because MEC = 10, the MSC curve is
above the MPC curve. The distance between the MPC curve and MSC curve is Kshs. 10.

The competitive market equilibrium corresponds to point A at which MPC = MSB. But the
efficient equilibrium will be at point B rather than at point A. At point B: MSC = MSB. The
market equilibrium output of 5 million tons is inefficient because its MSC equals 110 per ton
at point G, while its MSB is only Kshs. 100 per ton at point A (MSC > MSB). This indicates
that too much is being sold in the competitive market relative to the efficient amount.

A gain is net social benefit equal to the area ABG is possible by reducing annual output from
5million tons to 4.5million tons. To realize this, the price of paper would have to increase to
Kshs. 105 per ton to induce consumers to cut back consumption by 0.5 tons per year.

This result shows that with negative externality competitive markets will produce a
commodity at levels that are more than the socially optimal.

38
5.4CORRECTIVE MECHANISMS

5.4.1 Public Responses to Externalities

(a) Internalizing a Positive Externality

Internalization of an externality occurs when the MPB or cost of goods and services are
adjusted so that the users consider the actual marginal social cost or benefits of their
decisions.

A corrective subsidy is a means of internalizing positive externalities. This is a payment

made by the government to either buyers or sellers of a good so that the price paid by
consumers is reduced. The payment must equal the MEB of the good or service. For
example, in the case of inoculation, the payment is Kshs. 20.

S=MSC
Z
45
R
30 V
25 Subsidy payment
X
10
Y D’=MPB+20=MSB
D=MPB

10 12
0

A corrective subsidy to consumers increases the demand for inoculations and achieves the
efficient output. After subsidy payments are received by consumers, the net price of an
inoculation falls to Kshs. 10, inducing them to purchase the efficient amount of 12m per year.

The area RVXY (shaded) represents the total subsidy payments at the efficient output.

Examples of corrective subsidies include the provision of certain government services at

levels below the MC of such services. For example the government provides subsidies to
students in public universities.

(b) Internalizing Negative Externalities

A corrective tax is used to adjust the MPC of a good or service in such a way as to internalize
the externality. The tax must be equal to the MEC per unit of output. It makes sellers of the

39
product pay a fee equal to the MEC per unit of output sold. For example in the case of the
paper factory a corrective tax would force producers to reduce output from 5 million tons of
paper to 4.5 tons

S’=MPC+T=MSC
P, B, C
S=MPC
G
110
F B Net gains in well-being
105

100 A

95
H J
D=MSB

4.5 5

The corrective tax equals the MEC. In this case, the tax is set at Kshs. 10 per ton of paper
which is the MEC. This tax on each unit produced will be treated as an increase in the MPC
of production T. Supply shifts upwards from S to S1 where S1 reflects the MSC. The point
corresponding to market equilibrium shifts from A to B. The market price increases to 105
per ton and the equilibrium quantity consumed declines from 5m tons to 4.5m tons per year
which is exactly equal to the efficient annual output.

A tax of ksh. 10 per ton will collect Kshs. 45m (10 x 4.5m) of revenue per year at the
equilibrium output of 4.5m tons. This is represented by the area FBJH.

After the tax is imposed, the annual value of pollution costs to alternative users of the stream
declines. Initially these costs were 5m x 10 = 50m per year.

However, because annual output declined to 4.5m tons after the tax, the annual cost of
pollution declines to 45m.

The gain in efficiency resulting from the corrective tax is area GAB. This area measures the
increase in net social benefits when annual output of paper is reduced to the point at which its
MSB equals its MSC.

40
Important Observation

A corrective tax does not reduce the pollution in the stream to zero. It merely raises the cost
of using the stream to reflect the marginal damage done to alternative users of the stream.
The tax should be designed to force the producer to compare the MB of dumping waste in the
stream with the MEC of emitting untreated waste by adding MEC to producers MPC.

Uses of tax revenue collection

1. To compensate other users of the stream for damages that remain after the externality is
internalized by the corrective tax.
2. The producers could be compensated for the loss of their free right to dump
3. Revenue collected to go towards a reduction in other taxes or an increment in government
services

Political support for enactment of a corrective tax

(i) Polluters/products, employees and consumers will vote against it since they are not
involved in alternative use and will not be compensated for losses.
(ii) Recreational and commercial users of the stream, as well as tax payers will vote in
favour to the extent to which they have few interests in production or consumption of
the item (paper)

Therefore, internalization of the externality will result in income redistributive effects, which
influences the willingness of individuals involved to support the idea.

(c) Creating a market

As we emphasized previously, the inefficiencies associated with externalities can be linked to
the absence of a market for the relevant resource. This suggests that another way for the
government to enhance efficiency is to sell to the producers permits to pollute. By doing so,
the government creates a market for clean air or water that otherwise would not have
emerged.

Under this scheme, the government announces it will sell permits to spew Z* units of
pollutants into the environment (This is the quantity of pollutants associated with output Q*).
Firms bid for the right to own these permissions to pollute, and the permissions go to the
firms with the highest bids. The fee charged is that which clears the market, so the amount of
pollution equals the level set by the government. The price paid for permission to pollute
measures the value to producers of being able to pollute.

41
The permit approach is illustrated in the figure below. The horizontal axis measures the
number of rights to produce sulphur oxides, and the vertical measures the price of these
rights.

Market for Pollution Right

P1

DZ

Z* Rights to produce (Parts per 100 million) per year

The government announces it will auction off Z* pollution rights. In effect, the supply of
pollution rights is perfectly vertical at Z*. The demand for pollution rights, DZ, is downward
sloping. The equilibrium price per unit is P1. Those firms that are not willing to pay P1 for
each unit of pollution they produce must either reduce their output or adopt a cleaner
technology.

Incidentally, the scheme also works, if, instead of auctioning off the pollution rights, the
government assigns them to various firms that are then free to sell them to other firms. The
market supply is still perfectly vertical at Z*, and the price is still P1. Nothing changes
because a given firm is willing to sell its pollution rights provided the firm values these rights
at less than P1

However, even though the efficiency effects are the same as those of the auction, the
distributional consequences are radically different. With the auction, the money goes to the
government; with the other scheme, the money goes to the firms that were lucky enough to be
assigned the pollution rights.

42
In any case, in this simple model, the permit and the Pigouvian tax both achieve the efficient
level of pollution. Implementing either one requires knowledge of who is polluting and in
what quantities.
How is one to choose between them?
Cropper and Oates (1992) argue that the permit has some practical advantages over the tax
scheme.
- The permit scheme reduces uncertainty about the ultimate level of pollution.
- If the government is certain about the shapes of the private marginal cost and marginal
benefit schedules, then it can safely predict how a Pigouvian tax will affect behaviors.
But if information about these schedules is poor, it is hard to know how much a particular
tax will reduce pollution.
- If lack of information forces policymakers to choose the pollution standard arbitrarily, it
is more likely to be obtained with a system of pollution permits.
- Assuming that firms are profit maximising, they will find the cost-minimizing technology
to attain the standard.
- When the economy is experiencing inflation, the market price of pollution rights should
be expected to keep pace automatically, while changing the tax rate could require a
lengthy administrative procedure.
- On the other hand, one possible problem with the auctioning scheme is that incumbent
firms might be able to buy pollution licenses in excess of the firms’ cost-minimizing
requirements to deter other firms from entering the market. Whether such strategic
behavior would occur is hard to predict.

(d) Regulation
Under regulation, each polluter is told to reduce pollution by a certain amount or else face
legal sanctions. In our model, Bart would simply be ordered to reduce output to Q*.
Regulation is likely to be inefficient when there are multiple firms that differ from each other.
To see this, consider two firms, X and Z, each of which emits carbon dioxide (CO2), a
chemical that is thought to contribute to global warming.

43
 S
(MPCX + d) = (MPCZ + d)

d MPCX = MPCZ

MBZ
MBX
0 Z* X* X1 = Z1 Q per year

In the figure, output of the firms is measured on the horizontal axis and dollars on the
vertical. MBX is the marginal benefit schedule for X and MBZ the schedule for Z. For
expositional ease only, X and Z are assumed to have identical MPC schedules and profit-
maximizing outputs X1 = Z1.

Suppose it is known that the marginal damage at the efficient level of total output is d dollars.
Then efficiency requires that each firm produces at the point of intersection of its marginal
benefit curve with the sum of its marginal private cost curve and d. The efficient outputs are
denoted X* and Z*. The crucial observation is that efficiency does not require the firms to
reduce their CO2 emissions equally. The efficient reduction in production of Z exceeds that
of X. Here this is due to different MB schedules, but in general, each firm’s appropriate
reduction in output depends on the shapes of its marginal benefit and marginal private cost
curves. Hence, a regulation that mandates all firms to cut back by equal amounts (either in
absolute or proportional terms) leads to some firms producing too much and others too little.

5.4.2. Private/Market Responses to Externalities

In the presence of externalities, an inefficient allocation of resources emerges if nothing is
done about it. This section discusses the circumstances under which private individuals,
acting on their own, can avoid the existence of externalities. This they achieve through
devising ways of excluding those who do not pay

While most people are unaware of it, markets often solve public goods and externalities
problems in a variety of ways.

1. Charging those who benefit from the services: Businesses frequently solve free-rider
problems by developing means of excluding non-payers from enjoying the benefits of a
good or service. DSTV for instance, scramble their transmissions so that nonsubscribers

44
 cannot receive broadcasts. Both throughout history and today, private roads have
financed themselves by charging tolls to road users. Other supposed public goods, such
as protection and fire services, are frequently sold through the private sector on a fee
basis.
2. Public goods can also be provided by being tied to purchases of private goods. Shopping
malls, for instance, provide shoppers with a variety of services that are traditionally
considered public goods such as: lighting, protection services, benches, and rest-rooms.
Charging directly for each of these services would be impractical. Therefore, the
shopping mall finances the services through receipts from the sale of private goods in the
mall. The public and private goods are "tied" together. Private condominiums and
retirement communities also are examples of market institutions that tie public goods to
private services. Monthly membership dues are used to provide a variety of public
services.

- Lighthouses are one of the most famous examples that economists give of public
goods that cannot be privately provided. Economists have argued that if private
lighthouse owners attempted to charge ship-owners for lighthouse services, a free-
rider problem would result. Yet lighthouses off the coast of nineteenth-century
England were privately owned. Lighthouse owners realized that they could not charge
ship owners for their services. So they didn't try to. Instead, they sold their service to
the owners and merchants of the nearby port. Port merchants who did not pay the
lighthouse owners to turn on the lights had trouble attracting ships to their port. As it
turns out, one of the economics instructors' most commonly used examples of a public
good that cannot be privately provided is not a good example at all.

3. Contractual arrangements can sometimes be used to overcome other public goods and
externalities problems. For example, if the research and development activities of one
firm benefit other firms in the same industry, these firms may pool their resources and
agree to a joint project (antitrust regulations permitting). Each firm will pay part of the
cost, and the contributing firms will share the benefits. In this context economists say that
the externalities are "internalized."

Contractual arrangements sometimes fail to solve public goods and externalities

problems.

- The costs of bargaining and striking an agreement may be very high.

- Some parties to the agreement may seek to hold out for a better deal, and the
agreement may collapse.
- In other cases it is simply too costly to contract and deal with all the potential
beneficiaries of an agreement. A factory, for instance, might find it impossible to
negotiate directly with each affected citizen to decrease pollution.

45
4. Bargaining (Coase Theorem)

The root cause of the inefficiencies associated with externalities is the absence of property
rights. When property rights are assigned, individuals may respond to the externality by
bargaining with each other.
Example
Suppose property rights to the river are assigned to Bart. Assume that it is costless for Lisa
and Bart to bargain with each other. Is it possible for the two parties to strike a bargain that
will result in output being reduced from Q1?
Bart would be willing to not produce a given unit of output as long as he received a payment
that exceeded his net incremental gain from producing that unit (MB – MPC). On the other
hand, Lisa would be willing to pay Bart not to produce a given unit as long as the payment is
less than marginal damage done to her, MD. As long as the amount that Lisa is willing to
pay Bart exceeds the cost to Bart of not producing, the opportunity for a bargain exists.
Algebraically, the requirement is that MD > (MB – MPC).

MSC = MPC + MD

MPC

MD

MB

0 Q* Q1 Q per year

The above figure indicates that at output Q1, MB – MPC is zero, while MD is positive.
Hence, MD exceeds MB – MPC, and there is scope for a bargain. Similar reasoning indicates
that the payment Lisa would be willing to make exceeds MB – MPC at every output level to
the right of Q*.
In contrast, to the left of Q*, the amount of money Bart would demand to reduce his output
would exceed what Lisa would be willing to pay. Hence, Lisa pays Bart to reduce output just
to Q*, the efficient level. We cannot tell without more information exactly how much Lisa
will end up paying Bart. This depends on the relative bargaining strengths of the two parties.

46
Regardless of how the gains from the bargain are divided, however, production ends up at
Q*.

Now suppose that Lisa is the one assigned the property rights to the stream. The bargaining
process now consists of Bart paying for Lisa’s permission to pollute. Lisa is willing to accept
some pollution as long as the payment is greater than the marginal damage (MD) to her
fishing enterprise. Bart finds it worthwhile to pay for the privilege of producing as long as
the amount is less than the value of MB – MPC for that unit of output. Reasoning similar to
the foregoing suggests that they have every incentive to reach an agreement whereby Lisa
sells Bart the right to produce Q*.
Two important assumptions played a key role in the preceding analysis
- The costs to the parties of bargaining are low
- The owners of resources can identify the source of damage to their property and legally
prevent damages.
Under these assumptions, the efficient solution will be achieved independently of who is
assigned the property rights, as long as someone is assigned those rights. This result, known
as the Coase theorem (after Nobel laureate Ronald Coase), implies that once property rights
are established, no government intervention is required to deal with externalities (Coase,
1960). In other words,Coase (1960) in his theorem pointed out that if the assumption of zero
transaction costs condition is maintained, the set of markets can be expanded beyond normal
private goods to include non-market goods as long as institutional constraints to assigning
well-defined property rights are removed. The Coase theorem asserts that disputing parties
will work out a private agreement that is pareto efficient, regardless of the party to whom
unilateral property rights to the non-market good are assigned initially. As long as these legal
entitlements can be freely exchanged, government intervention is relegated to designating and
enforcing well–defined property rights.

5. Private Property Rights

Externalities can be solved by defining individual property rights in the appropriate

economic resource. Cleaning up a polluted lake, for instance, involves a free-rider problem if
no one owns the lake. The benefits of a clean lake are enjoyed by many people, and no one
can be charged for these benefits. Once there is an owner, however, that person can charge
higher prices to fishermen, boaters, recreational users, and others who benefit from the lake.
Privately owned bodies of water are common in the British Isles, where, not surprisingly,
lake owners maintain quality.

47
Well-defined property rights can solve public goods problems in other environmental areas,
such as land use and species preservation. The buffalo neared extinction and the cow did not
because cows could be privately owned and husbanded for profit. Today, private property
rights in elephants, whales, and other species could solve the tragedy of their near extinction.
In Africa, for instance, elephant populations are growing in Zimbabwe, Malawi, Namibia,
and Botswana, all of which allow commercial harvesting of elephants. Since 1979
Zimbabwe's elephant population rose from 30,000 to almost 70,000 today, and Botswana's
went from 20,000 to 68,000. On the other hand, in countries that ban elephant hunting—
Kenya, Tanzania, and Uganda, for example, there is little incentive to breed elephants but
great incentive to poach them. In those countries elephants are disappearing. The result is that
Kenya has only 16,000 elephants today versus 140,000 when its government banned hunting.
Since 1970, Tanzania's elephant herd has shrunk from 250,000 to 61,000; Uganda's from
20,000 to only 1,600.

Property rights are a less effective solution for environmental problems involving the air,
however, because rights to the air cannot be defined and enforced easily. It is hard to
imagine, for instance, how market mechanisms alone could prevent depletion of the earth's
ozone layer. In such cases economists recognize the likely necessity of a regulatory or
governmental solution. However, externalities such as air pollution involve million of people
(both polluters and pollutees). It is difficult to imagine them getting together for negotiations
at a sufficiently low cost. Further, even if property rights to air were established, it is not
clear how owners would be able to identify which of thousands of potential polluters was
responsible for dirtying their airspace and for what proportion of the damage each was liable.

The Coase Theorem is most relevant for cases in which only a few parties are involved and
the sources of thee externality are well defined. Even when these conditions hold, the
assignment of property rights is relevant from the point of view of income distribution.
Property rights are valuable; if Lisa owns the stream it will increase her income relative to
Bart’s, and vice versa.

6. Mergers
One way to deal with an externality is to “internalize” it by combining the involved parties.
For simplicity, imagine there is only one polluter and one pollutee, as in the Bat-Lisa
scenario. As stressed already, if Bart took into account the damages he imposed on Lisa’s
fishery, then a net gain would be possible. In other words, if Bart and Lisa coordinated their
activities, then the profit of the joint enterprise would be higher than the sum of their
individual profits when they don’t coordinate. In effect, by failing to act together, Bart and
Lisa are just throwing away money! The market, then, provides a strong incentive for the
two firms to merge – Lisa can buy the factory, Bart can buy the fishery, or some third party
can buy them both. Once the two firms merge, the externality is internalized – it is taken into
account by the party that generates the externality. For instance, if Bart purchased the
fishery, he would willingly produce less output than before, because at the margin doing so
would increase the profits of his fishery subsidiary more than it decreased the profits from his
48
factory. Consequently, the external effects would not exist, and the market would not be
inefficient. Put another way, an outside observer would not characterize the situation as an
“externality” because all decisions would be made within a single firm.

7. Social Conventions
Unlike firms, individuals cannot merge to internalize externalities. However, certain social
conventions can be viewed as attempts to force people to take into account the externalities
they generate. For example, schoolchildren are taught that littering is irresponsible and not
“nice”. If this teaching is effective, a child learns that even though he bears a small cost by
holding on to a candy wrapper or a banana peel until he finds a garbage can, he should incur
this cost because it is less than the cost imposed on other people by having to view his
unsightly garbage. Think about the golden rule, “Do unto others as you would have others do
unto you”. A (much) less elephant way of expressing this sentiment is “Before you undertake
some activity, take into account its external marginal benefits and costs.” Some moral
precepts, then, induce people to emphasize with others, and hence internalize the externalities
their behavior may create. In effect, these precepts correct for the absence of missing
markets.
NB:
1. The presence of externalities often requires some kind of intervention to achieve
efficiency. Implementing any environmental policy entails a host of difficult technical
issues. No policy is likely to do a perfect job. However, most economists prefer market-
oriented solutions. They are more likely to achieve efficient outcomes than direct
regulation.
2. The imperfections of market solutions to public goods problems must be weighed against
the imperfections of government solutions. Governments rely on bureaucracy and have
weak incentives to serve consumers. Therefore, they produce inefficiently. Furthermore,
politicians may supply public "goods" in a manner to serve their own interests, rather than
the interests of the public. For example, government often creates a problem of "forced
riders" by compelling persons to support projects they do not desire.
3. Private solutions to public goods problems, when possible, are usually more efficient than
governmental solutions. In the case of public goods, there are two reasons why markets
do not give an efficient allocation of resources: indivisibility and externality. Since it is
not practically possible for people to be charged for the benefit from the lighthouse (for
example), these benefits are "external." In addition, the costs of the public good is
indivisible, a condition that could create problems in itself. But -- when government
provides the public good, for free -- the externality just balances the indivisibility and free
provision is efficient.

49
 TOPIC SIX
PUBLIC CHOICE THEORY
6.1 Introduction
Application of economic principles to the analysis of political decision making is a field of
study referred to as Political economy. Political economy models assume that individuals
view government as a mechanism for maximizing their self-interest. Two points are
important regarding this assumption:
- The pursuit of self-interest does not necessarily lead to inefficient outcomes. Under
certain conditions, the market place harnesses self-interest to serve societal end. The
question is, “what if something performs that function in the political market?”
- While the maximization assumption may not be totally accurate, just as in more
conventional settings, it provides a good starting point for analysis.

Public choice theory examines direct democracies and how well they translate the preferences
of their members into collective action as well as the complications that arise when decisions
are made not by individuals themselves but by their elected representatives.

6.2 Important concepts

Public Choice is one met through political interaction of many persons according to
established rules. The supply of a public good through political institutions, require
agreement on the quantity of the public good and the means to finance it.

Political Equilibrium is an agreement on the level of production of one or more public

goods, given the specified rule for making the collective choice and the distribution of tax
shares among individuals.

A tax share (also known as tax price) is a pre announced levy assigned to citizens. It is equal
to a portion of the unit cost of goods proposed to be provided by the government to a voter.
Therefore, it represents the price per unit of a government supplied good. The sum of the tax
shares must be equal to the average cost (AC) of providing the public good, in order to avoid
budget surpluses or deficits. This implies that at the Political equilibrium:
N

t
i 1
i  AC .

Determinants of Political Equilibrium

1. Public / voting rule itself which refers to the proportion of Yes vote in relation to the
number of votes required for approval of the issue.
2. The Average and marginal cost of the public good.
3. Information available to the voters on the costs and benefits associated with the issue.

50
4. The distribution of tax shares among voters and the way in which extra taxes vary with
extra output of the goods provided.
5. The distribution of benefits among voters (whether it fair or not?)

6.3 PUBLIC CHOICE IN DIRECT DEMOCRACY

Democratic societies use various voting procedures to decide on public expenditures. They
include
1. Unanimity rules
2. Majority voting rules

A: Unanimity Rules
The free rider problem can lead to a disturbing situation where because people are selfish,
public goods are under-provided, even though everyone could be made better off if they were
provided in efficient amounts. This suggests that, in principle, if a vote were taken on
whether to provide an efficient quantity of the good, consent would be unanimous as long as
there was a suitable tax system to finance it.

Lindahl (1919-1958) in the early 20 th century proposed a procedure designed to elicit

unanimous agreement. To understand Lindahl’s procedure, assume there are two individuals,
Amandu and Eveli, and one public good, rockets for fireworks (r).
- Suppose Amandu is told that his share of the cost of rocket provision will be 30 per
cent. Then if the market price per rocket is Pr, Amandu’s price per rocket is 0.3 x Pr
(0.3Pr). Given this price, the prices of other goods, his tastes, and his income, there is
some quantity of rockets that Amandu wants to consume. More generally, let SA denote
Amandu’s share of the cost of rocket provision.
- For any particular value of SA, Amandu demands some quantity of rockets. As his tax
share increases and rockets become more expensive for him, he demands a smaller
quantity. This can be illustrated as in the figure 1. In the figure, the horizontal axis
measures the quantity of rockets. Amandu’s tax share is measured by the vertical
distance from point O. The curve D Ar shows how the quantity of rockets demanded by
Amandu decreases as his tax share increases.
- In the same way, define SE as Eveli’s share of the cost of rockets so that SA + SE = 1.
When SE goes up, the quantity demanded by Eveli decreases. In the figure, Eveli’s tax
share increase as we move down along the vertical axis from 0’. Thus, the distance 00’
is 1. Her demand schedule is denoted D Er . It slopes upward because upward
movements along the vertical axis represent a lower price to her.
- An obvious similarity exists between the role of tax shares in the Lindahl model and
market prices in the usual theory of demand. But there is an important difference.
Instead of each individual facing the same price, each faces a personalized price per unit

51
 of public good, which depends on his or her tax share. The tax shares are referred to as
Lindahl prices i.e. the tax share that an individual must pay per unit of a public good.

- Equilibrium is a set of lindahl prices such that at those prices each person votes for the
same quantity of the public good.

- In the figure Amandu’s equilibrium tax share is OS* and Eveli’s is O’S* at which both
parties agree that r* rockets should be provided. At the equilibrium, OS* + O’S = 1

0’
D Er

S*

D Ar

0 r* r per year
Figure 1: Lindahl’s model for unanimity voting

Feasibility of Unanimity Rules:

The Lindahl model shows the tax shares and level of public good provision to which
everyone agrees. The big question is how to reach the equilibrium. Imagine that an
auctioneer announces some initial set of tax shares. On the basis of their respective demand
schedules, Amandu and Eveli vote for the number of rockets they want. If agreement is not
unanimous, the auctioneer announces another set of tax shares. The process continues until
Amandu and Eveli unanimously agree on the quantity of rockets (r* in the figure). The
determination of the quantity of public goods, is then quite similar to the market process and
just like in the market outcome, one can prove that the allocation is Pareto efficient.

As a practical method for providing public goods, Lindahl’s procedure has two main
problems:
52
(i) It assumes people vote sincerely. However, if Amandu can guess the maximum amount
that Eveli would spend for rockets rather than do without them, he can try to force her to
that allocation. Eveli has the same incentives. Strategic behaviour may therefore prevent
Amandu and Eveli from reaching the Lindahl equilibrium.
(ii) It may take a lot of time to find the mutually agreeable tax shares. In this example, there
are only two parties. In most important cases, many people are involved and getting
everyone’s consent involves enormous decision-making costs.

Indeed, although unanimity rules guarantee that no one will be “exploited”, they often lead
to situations in which no decisions are made. Historically, when organizations adopted a
unanimity rule, it was often expressly because the participants wanted to make sure that no
actions were taken.

B: Majority Voting Rules

Because unanimity is difficult to attain, voting systems not requiring it may be desirable.
With a majority voting rule, one more than half of the voters must favour a measure for it to
gain approval.
Consider a community with 3 voters; James, John and Jane; who have to choose among three
levels of security provision; A, B and C.
Level A is 3 Security Guards
Level B is 5 Guards
Level C is 8 Security Guards

Suppose the voters’ preferences are depicted as in the table below with each column showing
how each voter ranks his/her choice.

Table 1: Preferences of voters over different levels of security

Choice James John Jane

First A C B
Second B B C
Third C A A

- If an election were held on whether to adopt A or B, James will vote A, while John and
Jane will vote B. Hence B wins by a vote of 2 to 1.
- If election is held between B and C, B would win by a vote of 2 to 1. (James and Jane
vote B).
- Therefore, level B wins any election against its opposition and is therefore the option
selected by majority rule.
In this example the selection of B is independent of the order in which the votes are taken.

53
However, Majority decision rules do not always yield such clear cut results. The following
example demonstrates that majority voting do not always yield clear cut results.

Suppose the preferences of the voters were instead as follows

Table 2: Alternative Preferences of voters over different levels of security

Choice James John Jane

First A C B
Second B A C
Third C B A

If paired elections to determine the most preferred level of security,

 An election between A and B; A wins by 2 to 1 votes
 An election between B and C; B wins by 2 to 1 votes
 An election between A and C; C wins by 2 to 1 votes

This is a discontented result, since if A is preferred to B and B is preferred to C, then A

should be preferred to C, by the transitivity assumption. But in the third election, Just the
opposite occurs. This shows that although each voter’s preferences are consistent, the
community’s preferences are not. This phenomenon is referred to as voting paradox. It shows
that with majority voting, community preferences can be inconsistent even though each
individuals preferences are consistent.

In such a situation, the ultimate public outcome depends crucially on the order in which the
votes are taken.
- If the first election is between A and B, the winner A runs against C, then C wins.
- But if the first election is between B and C, the winner B runs against A, then A wins.
Under such conditions, the ability to control the order of voting - the agenda - confers great
power. The process of organizing the order of votes to assure a favourable outcome is
referred to as agenda manipulation.

Another related problem is that paired voting can go on forever without reaching a decision.
After the election between A and B, A wins. If C challenges A, then C wins. If B then
challenges C, B wins. The process can continue indefinitely, a phenomenon called cycling:a
situation where paired voting on more than two possibilities goes on indefinitely without a
conclusion ever being reached.

54
The Cause of Voting Paradox

Clearly, majority voting need not lead to these problems. After all, the elections associated
with table 1 went smoothly. Why the difference? It turns on the structure of individual
preferences for various levels of security provision. Consider again the people in table 2.
Because James prefers A to B to C, it follows that A gives him more utility than B, and B
more than C. The schedule denoted James in figure 2 depicts this relationship. The schedule
labelled John and Jane do the same for the other voters.

Utility
John

James

Jane

0 A B C Level of security provision

Figure 2: Graph of voter preferences

A peak in an individual’s preferences is a point on the graph of an individual’s preferences at
which all the neighbouring points have lower utility. A voter has single-peaked preferences
if, as she moves away from the most preferred outcome in any and all directions, her utility
consistently falls, but has double-peaked preferences if, as she moves away from the most
preferred outcome, utility goes down, but then goes back up again.

James has a single-peak at point A, Jane has a single-peak at point B; and John has two
peaks, one at A and one at C. Thus John’s preference is the source of the voting paradox. If
John had a set of single peaked preferences, then majority voting rule would lead to a
consistent decision. In table 1, each voter has single peaked preferences. More generally, if
all voters’ preferences are single peaked, no voting paradox occurs.

Causes of Multi-peaked Preferences

Because multi-peaked preferences can throw a wrench into majority voting, it is important to
know whether they are likely to be important as a practical matter.
(i) Presence of a private substitute for a publicly provided good.
Consider again John in table 2, whose preferences have two peaks. He prefers either large or
very small security provision to a quantity in the middle. Although such preferences are not
necessarily irrational, they do seem a bit peculiar. Perhaps John believes that moderate

55
numbers of security guards provide little if any real protection, so that unless security
provision is very large, there might as well be close to nothing.

For example consider voters choosing amount of expenditure levels for a public park – a
good for which there are private substitutes. Assume that in the presence of small or medium
public park expenditures, voter Jerry will join a private country club, but given large
expenditures he will use the public park. Provided that Jerry’s tax burden increases with park
expenditure, he prefers a small to a medium park, since neither of these options benefits him
and thus he prefers the option with a smaller tax burden. But his most preferred outcome
might be the large expenditure public park depending on the associated tax burden compared
to the country club membership fee. The argument shows that Jerry may prefer either the
small or large public park to the medium sized one – thus preferences are double peaked.
This suggests that when there are private substitutes for a publicly provided good, a multi-
peaked pattern like John’s in figure 2 can easily emerge.

(ii) When issues cannot be ranked a long a single dimension

When issues cannot be ranked along a single dimension e.g. voters selecting use of a resource
which is to be used for different purposes and the purposes are not in a single dimension,
multi-peaked preferences are also a serious possibility. Suppose that a community is trying
to decide how to use a vacant building. Choice A is an abortion clinic, choice B is an adult
bookstore, and choice C is an Army recruitment office. Unlike the choice among different
levels of park expenditure or levels of security provision, here the alternatives do not
represent more or less of a single characteristic. Multi-peaked preferences can easily emerge.

The Median Voter Theorem

Let us now return to the simple case in which all alternatives being considered represent
smaller or greater amounts of a characteristic. People rank each alternative on the basis of this
characteristic. An example is how much of some public good to acquire.
We define the median voter as the voter whose preferences lie in the middle of the set of all
voters’ preferences: half the voters want more of the good than the median voter, and half
want less. The median voter theorem states that: “as long as all preferences are single
peaked, the outcome of majority voting reflects the preferences of the median voter”. (With
an even number of voters, there may be a tie between two median voters, which must be
broken arbitrarily).

To demonstrate the theorem, assume there are five voters: Donald, Daisy, Huey, Dewey and
Louie. They are deciding how large the expenditure on a party to give together should be,
and each of them has single-peaked preferences over size of the party expenditure. Suppose
the most preferred level for each voter is as given the table below
Table 3: Preferred level of Party expenditure

56
 Voter Donald Daisy Huey Dewey Louie

expenditure($) 5 100 150 160 700

- Because preferences are single peaked, the closer an expenditure level is to a given
voter’s peak, the more he or she prefers it.
- A movement from zero party expenditure to $5 would be preferred to no money by all
voters.
- A movement from $5 to $100 would be approved by Daisy, Huey, Dewey, and Louie
- From $100 to $150 would be approved by Huey, Dewey, and Louie
- Any increase beyond $150, however, would be blocked by at least three voters: Donald,
Daisy and Huey.
- Hence, the majority votes for $150. But this is just the amount preferred by Huey, the
median voter.
- The election results mirror the median voter’s preferences.

Therefore, when all preferences are single peaked, majority voting yields a stable result, and
the choice selected reflects the preferences of the median voter. However, when some voters’
preferences are multi-peaked, a voting paradox can emerge. Because multi-peaked
preferences may be important in many realistic situations, majority voting cannot be
depended on to yield consistent public choices. Moreover, even when majority voting leads
to consistent decisions, it may not be efficient in the sense that overall benefits exceed costs.

C: Logrolling
A possible problem with simple majority voting is that it does not allow people to register
how strongly they feel about the issues. Whether a particular voter just barely prefers A to B
or has an enormous preference for A has no influence on the outcome. Logrolling systems
allow people to trade votes and hence register how strongly they feel about various issues.
Suppose that voters Smith and Jones prefer not to have more missiles, but this preference is
not strongly felt. Brown, on the other hand, definitely wants more missiles. With a logrolling
system, Brown may be able to convince Jones to vote for more missiles if Brown promises to
vote for a new road to go by Jones’s factory.
Vote trading is controversial.
1. The proponents argue that:
- Trading votes leads to efficient provision of public goods.
- They also emphasize its potential for revealing the intensity of preferences and
establishing a stable equilibrium.

57
 - The compromises implicit in vote trading are necessary for a democratic system to
function. According to a sociologist James Q. Wilson (2000), vote trades are essential to
finding some way to balance competing interests, each of which is defended by a
legislator who owes little to any other legislator hence are an essential way of achieving
what force and language cannot produce.

An example of situations in which logrolling leads to desirable outcome

Suppose a community is considering three projects, a hospital, a library, and a swimming
pool. The community has three voters, Melanie, Rhett, and scarlet. Table 4 shows their
benefits for each project. (A minus sign indicates a net loss: that is, the costs exceed the
benefits).
Table 4: Logrolling can improve welfare

___________________Voter_____________________________________________
Project Melanie Rhett Scarlet Total Net Benefits
Hospital 200 -50 -55 95
Library -40 150 -30 80
Pool -120 -60 400 220

The first thing to notice about the table is that the total net benefit for each project is positive.
Thus, by definition, the community as a whole would be better off if each project were
adopted. But what happens if the projects are voted on one at a time? Melanie votes for the
hospital because her net benefit is positive, but Rhett and Scarlet vote against it because their
benefits are negative. The hospital therefore loses. Similarly, the library and the swimming
pool go down in defeat.

Vote trading can remedy this situation. Suppose Melanie agrees to vote for the library if
Rhett consents to vote for the hospital, Melanie comes out ahead by 160 (=200 – 40) with the
trade; Rhett comes out ahead by 100 (= 150 – 50). They therefore strike the deal, and the
hospital and library pass. In the same way, Melanie and Scarlet can make a deal in which
Melanie gives her support for the pool in return for Scarlet’s vote for the hospital. This
would also improve social welfare. Thus logrolling allows all three measures to pass, a
desirable outcome.

2. Opponents of logrolling stress that it is likely to result in special-interest gains not

sufficient to outweigh general losses. Large amounts of wastes can be incurred.
An example of situations in which logrolling leads to undesirable outcome

58
Assume we have the same three voters and three projects under considerations as in table 4,
but now the various net benefits are as depicted in table 5.
Table 5 Logrolling can also lower welfare

Voter
Project Melanie Rhett Scarlet Total Net Benefits
Hospital 200 -110 -105 -15
Library -40 150 -130 -10
Pool -270 -140 400 -10

Every project has a negative net benefit. Each should therefore be rejected, as would be the
case if the projects were voted on one at a time. However, with logrolling, some or all of
these inefficient projects could pass.
Suppose Melanie offers to support the library in return for Rhett’s vote for the hospital. The
deal is consummated because both of them come out ahead – Melanie by 160 (=200 – 40)
and Rhett by 40 (=150 – 110). With the support of Melanie and Rhett together, both projects
pass. In the same way, Rhett and Scarlet can trade votes for the pool and the library, so both
of those projects are adopted. This demonstrates how with logrolling, a majority of voters can
form a coalition to vote for projects that serve their interests, but whose costs are borne
mainly by the minority. Hence, although the benefits of the projects to the majority exceed
the costs, this is not true for the society as a whole. Therefore, while logrolling can
sometimes improve on the results from simple majority voting, this is not necessarily the
case.

6.4 Arrow’s Impossibility Theorem

We have shown that neither simple majority-voting nor logrolling has entirely desirable
properties. Many other voting schemes are also flawed. An important question is whether
there is any ethically acceptable method for translating individual preferences into collective
preferences free of any difficulties. But this depends on what “ethically acceptable” means to
an individual. According to Nobel laureate Kenneth Arrow (1951), in a democratic society, a
collective decision-making rule should satisfy the following criteria:
1. It can produce a decision whatever the configuration of voters’ preferences. Thus, for
example, the procedure must not fall apart if some people have multi-peaked
preferences.
2. It must be able to rank all possible outcomes
3. It must be responsive to individuals’ preferences. Specifically, if every individual
prefers A to B, then society’s ranking must prefer A to B.

59
 4. It must be consistent in the sense that if A is preferred to B and B is preferred to C,
then A is preferred to C.
5. Society’s ranking of A and B depends only on individuals’ rankings of A and B.
Thus, the collective ranking of defence expenditures and foreign aid does not depend
on how individual rank either of them relative to research on a cure for AIDS. This
assumption is sometimes called the independence of irrelevant alternatives.
6. Dictatorship is ruled out. Social preferences must not reflect the preferences of only a
single individual

Taken together, these criteria seem quite reasonable. Basically, they say that society’s choice
mechanism should be logical and should respect individuals’ preferences. Unfortunately, the
stunning conclusion of Arrow’s analysis is that in general it is impossible to find a rule that
satisfies all these criteria. A democratic society cannot be expected to make consistent
decisions. This result, sometimes called Arrow’s Impossibility theorem, thus casts doubt on
the very ability of democracies to function.

Naturally, the theorem has generated debate, much of which has focused on whether other
sets of criteria might allow formation of a social decision-making rule. It turns out that if any
of the six criteria is dropped, a decision-making rule that satisfies the other five can be
constructed. But whether or not it is permissible to drop any of the criteria depends on one’s
views of their ethical validity.

Arrow’s theorem does not state that it is necessarily impossible to find a consistent decision-
making rule. Rather, the theorem only says one cannot guarantee that society will be able to
do so. For certain patterns of individual preferences, no problems arise, e.g. when members
of society have identical preferences. Some have suggested that the real significance of
Arrow’s theorem is that it shows the need for a virtual uniformity of tastes if a democracy is
to function. They then argue that many institutions have the express purpose of moulding
people’s tastes to make sure that uniformity emerges, e.g. through mandatory public
education.

A very different view is that Arrow’s theorem does not really have much to say about the
viability of democratic processes. Another Nobel Prize winner, James Buchanan (1960),
believes that the inconsistencies of majority voting have beneficial aspects:
Majority rule is acceptable in a free society precisely because it allows a sort of
jockeying back and forth among alternatives, upon more of which relative unanimity
can be obtained … it serves to ensure that competing alternatives may be
experimentally and provisionally adopted, tested, and replaced by new compromise
alternatives approved by a majority group of ever-changing composition. This is the
democratic choice process.

60
Another important question raised by Arrow’s theorem concerns the use of social welfare
functions. Recall that a social welfare function is a rule that evaluates the desirability of any
given set of individuals’ utilities. In a democratic society, the social welfare function must be
chosen collectively. But Arrow’s theorem says that it may be impossible to make such
decisions, and hence we cannot assume that a social welfare functions really exists.

However, if it does not exist, how can economists use the social welfare function to rank
alternative states? Some economists have therefore rejected the function’s use. They argue
that it is merely a way of introducing value judgments and not a representation of “society’s”
preferences. As such, a social welfare function does not isolate the correct allocation of
resources. However, most economists believe that the function is an important tool. It may
not provide “the” answer, but it can be used to draw out the implications of alternative sets of
value judgments. With this interpretation, the social welfare function provides valuable
insights.

61
 TOPIC SEVEN
THEORY OF PUBLIC EXPENDITURE

7.1 Introduction
Public expenditure can be defined in different ways as:
a) The expenditure of central and local government;
b) The combined government expenditure plus disbursements out of the National
Insurance (social security) Fund; or
c) The total government expenditure as in (ii) plus expenditure of the public
corporations.
The size of the public expenditure will depend on the definition adopted and will differ
accordingly. This can give rise to confusion when comparisons are made over a period of
time or internationally. Thus, if public-expenditure is defined in terms of what the central
government and local authorities spend; it will appear smaller than when expenditure by
public corporations is also included.

The basis on which public expenditure once defined is analysed, but this does not affect the
total figure which represents the absorption of resources by the public sector. The analysis
can be undertaken-on the following basis:
i). Spending authority: Central government, local authorities, public corporations,
ii). Economic category: Current expenditure account (expenditure on goods, services,
transfer payments), capital account (investment),
iii). Programme: defence, agriculture, housing.

The total figure for public expenditure, whichever basis is used, should be the same and
represents the absorption of resources by the public sector.

7.2 Public Expenditure Categories

General government expenditure includes the spending of all levels of government (national
and local). Cash expenditures for different years have to be made directly comparable by
taking account of inflation over the period. There resulting standardized figures are real
expenditures.

There are two main categories:

(a) Exhaustive expenditures: These refer to government purchases that consume or exhaust
the purchasing power of the money it spends, hence ‘exhaustive expenditures’. They
include purchases of inputs, such as labour, used in its own production of goods and
services and purchases of outputs from the private sector (for example cleaning services).
They also include investment in fixed assets such as machinery, land and buildings. Put
simply, exhaustive expenditures use productive resources.
(b) Transfer payments/expenditures: These refer to expenditures where the government does
not purchase factors of production or use resources. Instead the money is transferred
from taxpayers to recipients. They include subsidies to private sector firms (for example
those receiving grants supporting investment in economically depressed regions),

62
 payments of interest (to those from whom the government has borrowed), loans granted
by the government (for example to tenants buying their council houses) and overseas aid
(for example in support of economic development programmes). The largest component
of transfer payments is welfare benefits such as income support, unemployment benefit
and state pensions.

7.3 Canons of Public Expenditure

These are principles proposed to govern the public expenditure decisions. They include;
1) Canon of economy – utmost care must be taken to avoid wasteful usage of public funds.
2) Canon of sanction – no public funds should be used without proper authorization and
funds should be used only for the purpose for which they were sanctioned.
3) Canon of benefit – public expenditure should be incurred only if it is beneficial to the
society as a whole. The benefits can be through income distribution or production.
4) Canon of surplus – the government should avoid deficit budgeting. It should be prudent
and aim at meeting its current expenditure needs out of its current revenue. It should not
overspend and run into debts.

7.4 Observations on The Growth Of Public Expenditure

Growth in the size of the public sector relative to GDP and a proportionate shift towards
transfer payments are characteristic features of developed countries. Any comprehensive
economic theory of public expenditure growth has to explain the long-term proportionate
growth of public spending within the economy and the relative shift from exhaustive to
transfer payments. No single economic theory is able to do this, nor even explain changes in
exhaustive expenditures alone. Instead there are two broad groups of models:
(i) Macro models of public expenditure growth attempt to account for the long-term
growth of government expenditure.
(ii) Micro models of public expenditure growth attempt to explain changes in particular
components of government expenditure, whether caused by increasing demand for
individual services or by changes in their cost structures. These models incorporate
demand-side and/or supply-side factors. On the demand-side, public expenditure
increases because citizens want more public spending as their incomes rise; if not all
citizens then at least those with political power and influence. On the supply-side, it is
those who work within the public sector (rather than the general public) who want
more public spending, including government bureaucrats.

7.5 Macro Models of Public Expenditure

(a) Development Models: Musgrave-Rostow’s Theory
This theory sees public expenditure as a prerequisite of economic development, its level
being directly related to the stage of development which a country has reached. The early
stage of development is viewed as the period of industrialization during which the population
moves from the countryside to the urban areas. To meet the needs that result from this, there
is a requirement for significant infrastructural expenditure in the development of cities. The
typical rapid growth experienced in this stage of development results in a significant increase

63
in expenditure and the dominant role of infrastructure determines the nature of expenditure.
In this stage, public investment as a proportion of the total investment of the economy is high.

In the middle stages of development, the infrastructural expenditure of the public sector
becomes increasingly complementary with expenditure from the private sector.
Developments by the private sector, such as factory construction, are supported by
investments from the public sector, such as the building of connecting roads. As urbanization
proceeds and cities increase in size, so does population density. This generates a range of
externalities such as pollution and crime. An increasing proportion of public expenditure is
then diverted away from spending on infrastructure to the control of these externalities.

Finally, in the developed phase of the economy, there is less need for infrastructural
expenditure or for the correction of market failure. Instead, expenditure is driven by the
desire to react to issues of equity and human capital. This results in transfer payments, such
as social security, health, and education, becoming the main items of expenditure. Once such
forms of expenditure become established, they are difficult to ever reduce. They also increase
with heightened expectations and through the effect of an aging population.

(b) Wagner’s Organic State Theory

This theory is primarily concerned with the explanation of the growth of the share of GNP
taken up by the public sector. This theory, popularly known as Wagner’s law, states that as
per capita income in an economy grows, the relative size of the public sector will grow also.
This is because
(i) the state would need to expand administration and law and order services,
(ii) there would be an increase concern for distributional issues and
(iii) there would be a greater need to control private monopolies and other forms of
market failures.
Thus the state grows like an organism reflecting changes in society and economy and making
decisions on behalf (and to the benefit) of its citizens.

The bases of the theory are:

(i) Growth of the economy results in complexities, hence requiring introduction of laws, and
development of legal infrastructure.
(ii) The process of urbanisation and increased externalities associated with it.
(iii) The goods supplied by the public sector have a high income elasticity of demand, for
example education, recreation and health. Hence economic growth raises income, leading
to an increase in demand for these products. Hence public expenditure rises as a
proportion of income.
Wagner’s law proposes that it is income that explains expenditure. Unlike Keynes who
asserted that it was expenditure that explained the level of income.

(c) The Peacock and Wiseman’s (1967) Political Constraint Model is based upon a
political theory of public expenditure determination, namely that governments like to

64
 spend more money, that citizens do not like to pay more taxes, and that governments
need to pay some attention to the wishes of their citizens.

The model assumes that there is some tolerable level of taxation which acts as a
constraint on government behaviour. As the economy (and thus income) grows, tax
revenue at constant rate would rise thereby enabling public expenditure to grow in line
with GNP.

During period of social upheaval (war, famine or some large scale social disaster),
however, this gradual upward trend in public expenditure would be distorted (displaced
upward). In order to finance the increase in public expenditure, the government may be
forced to raise taxation level: a policy which would be regarded as acceptable to the
electorate during period of crises. This is called the displacement effect. Besides, there is
also the inspection effect. This arises from people’s keener awareness of social problems
during the period of upheaval.

The government, therefore, expands its scope of services to improve these conditions,
and because the people’s perception of tolerable levels of taxation does not return to its
former level, the government is able to finance these higher levels of expenditure
originating in the expanded scope of government and debt charges.

The net result of these two effects is occasional short- term jumps in public expenditure
within a rising long-term trend.

(d) Keynesian Theory

This theory is an economictheory based on the ideas of Keynesian economics that promoted a
mixed economy, in which both the state and the private sector were considered to play an
important role. Keynesian economics sought to provide solutions to what some economists
believed to be the failure of laissez-faireeconomic liberalism, which advocated that markets
and the private sector operated best without state intervention.

In Keynes's theory, macroeconomic trends could overwhelm the micro-level behaviour of

individuals. Keynesians asserted the importance of aggregate demand for goods as the
driving factor of the economy, especially in periods of downturn. From this they argued that
government policies could be used to promote demand at a macro level, to fight high
unemployment and deflation. This was in contrast to supply-side economics.

Keynes believed that the government was responsible for helping to pull a country out of a
depression. If the government increased its spending, then the citizens were encouraged to
spend more because more money was in circulation. People would start to invest more, and
the economy would climb back up to normal. A central conclusion of Keynesian economics
was that there was no strong automatic tendency for output and employment to move toward
full employment levels (Keynes, 2002).

65
(e) Monetarist Theory
This is a School of thought that stresses the primary importance of money supply in
determining nominal GDP and the price level. The "Founding Father" of Monetarism was
Milton Friedman. Monetarism was a theoretical challenge to Keynesian economics.
Monetarists argued convincingly that the high rates of inflation were due to rapid increases in
the money supply. They argued that the economy could be complicated, but stabilization
policy did not have to be complicated. The key to good policy was to control the supply of
money.

Because Monetarists disliked big government and tended to trust free markets, they did not
like government spending and believed that fiscal policy was not helpful. Where it could be
beneficial, monetary policy could do the job better. Excessive government spending only
interfered in the workings of free markets and could lead to bloated bureaucracies,
unnecessary social programmes, and large deficits. Automatic stabilizers were sufficient to
stabilize the economy (Gordon, 1974).

(f) Crowding Out Theory

This theory states that government intervention result in to private investment activity being
reduced; that is ‘crowded out’. There are two forms of crowding out:
(i) The direct crowding out, where public sector production uses resources that could
otherwise be used by the private sector. If the public sector replaced the private sector, it
could be expected to constrain economic growth. This displacement effect occur directly
as the public sector uses tax revenues to buy resources that would otherwise be used by
the private sector. The challenges are that government statistics exaggerate the growth of
public sector employment by counting part-time jobs as full-time and also the assumption
that public and private sector outputs ware mutually exclusive, whereas many would
appear to be complementary (Bailey, 2002).
(ii) Indirect crowding out which occur if public expenditure, taxation and borrowing cause
disincentives to productive effort, namely to work and to invest. Disincentives to invest in
other productive resources could occur if borrowing lead to higher interest rates or
inflation. The sale of government debt in order to finance public sector borrowing could
lead to a rise in interest rates, which indirectly crowds out private investment. This would
happen if either the increased public sector borrowing lead to higher interest rates or the
private sector investment was highly interest-elastic, that resulted in the fall in private
investment (Bailey, 2002).
At best, macro models can only explain the long-term rising trend of public expenditure.
They are essentially aggregative (macro) in approach, providing insufficient explanation of
the changing composition of public expenditure.

66
 START HERE TODAY

7.6 REASONS FOR THE GROWTH OF PUBLIC EXPENDITURE IN DEVELOPING

COUNTRIES
Various factors – political, social and economic – have contributed to the growth of public
expenditure and the growth of the public sector. The following are some of the major factors:
a) The abandonment of the laissez-fairedoctrine. As the climate of public opinion
changed new theories began to emerge and old ones were abandoned; among the
latter was the doctrine of laissez-faire. The self-correcting mechanism of an economic
system that the classical economists believed in appeared to have failed.
Unemployment, which to them was a theoretical impossibility, not only proved
possible, but became a major international problem. During the Great Depression of
the 1930s over 20 per cent of the insured population of the UK was unemployed. The
theory of governmental non-intervention could no longer command support. There
was a pressure of public opinion on governments to provide, relief for the
unemployed and to create jobs. In order to do so, public expenditure was increased.
b) The advent of Keynesian economics. One book, The General Theory of Employment,
Interest and Money (1936), by John Maynard (later Lord) Keynes, had a profound and
pervasive influence on economists and on governments for many generations. His
arguments that the government not only could but should use public expenditure as a
tool of economic policy to manage a national economy so as to counteract
unemployment, found ready acceptance in a world that had not yet recovered from the
Great Depression. The Keynesian prescription was to inject money into the economic
system. If the people were not spending, then it was up to a government to do so.

This required an expansive fiscal policy, in which a government would deliberately

aim at a Budget deficit by spending more money than it raised in taxation. To cover
the difference (deficit) the government would borrow. The 'Multiplier' effect of public
expenditure would counteract unemployment. Such fiscal policy was attractive to the
governments and popular with the public. By increasing public expenditure, a
government was seen to be doing something about unemployment whilst the public
were getting something (additional state benefits) for nothing, as it appeared, since
there was no increase in taxation. Government therefore had an incentive to increase
public expenditure and they did. What is more the policy appeared to work,
unemployment began to fall. But to what extent the increase in economic activity can
be attributed to governments' conversion to Keynesian economics, and to what extent
it was the result of rearmament on which major countries embarked at the time when
the General Theory was published, is debatable. Increased expenditure on defence
was a response to the threat of war. As such it was a political measure but it did inject
money into the economy and therefore had economic consequences.
c) Wars and social crises, such as severe and prolonged unemployment had resulted in
the growth of public expenditure.
d) Increase in the range of economic activities by the state. Emergence of political
philosophies, social attitudes and economic theories that advocated extension of the
activities of the states prepared the way for governments to expand public
expenditure.

67
e) Psychological conditioning of the general public, during a period of war and social
crisis, to a greater government intervention and higher levels of expenditure and
taxation made it easier for governments in subsequent periods to retain and to expand
their activities.
f) Post-war reconstructionof countries' economies involved governments in planning,
allocation of resources and in financing some of the projects.
g) Economic development, according to some economists, has considerable impact on
the level of public expenditure. Before a developing country can industrialize, it has
to invest in transport, water and power supplies, sanitation, education and other basic
social projects to reach a 'take-off point. In this early stage of development a high
proportion of total investment will have to be made by the government, since the
projects do not offer any, or foreseeable, return to investors. Once the country has
reached a more advanced stage of economic and social development, private
investment expands alongside public investment but, because of the imperfections of
the market, government intervention grows and with it public expenditure.
h) Growth of national incomeis related to the level of government economic activity.
Some economists, Wagner among them, had argued that an increase in national
income results in an increase in public expenditure on economic welfare. The richer a
country the more resources, in theory, are available to the government.
i) Increased public expectation. It can, however, be argued that, although it cannot be
statistically proved, an indirect relationship exists between the growth uf national
income and public expectation of an improved standard of living, and hence public
expenditure. Governments are likely to-be under pressure to increase provision of
public goods and services so as to increase the standard of living in general and of the
poorest members of society in particular.
j) Extension of the franchise. The increased expectation of a higher standard of living
has manifested itself through the 'Ballot Box' in a vote for higher public expenditure.
The extension of the franchise had given the right to vote at elections to people who
had previously been excluded from a choice of a political party to govern them and
from influencing the policies that they wished the government to pursue.

Under a progressive system of taxation it is those in the lower income groups who
stand most to gain from the extension of provision of goods and services by the public
sector. They have least to lose from increases in taxation required to finance higher
expenditure since they may not be liable to any direct taxes. The gradual extension of
the franchise has put pressure on governments to respond to the electorate's demands.

The extension of the franchise in Europe did not follow a uniform pattern and come
simultaneously. For example, women in Liechtenstein only voted for the first time in
1986. It can be argued that votes for women have resulted in votes for higher public
expenditure. It has been suggested that as women are more directly involved in their
children's education and safeguarding the health of the family, they are likely to
demand improvements in state provision of social services, through the ballot box.

68
k) The establishment of the welfare state. This has created a base for the long term
growth of public expenditure.
l) Socialism.Socialist parties, committed by their ideology to the extension of the public
sector, won general elections and formed governments after the Second World War in
a number of countries, including the UK. Implementation of the policies set out in the
election manifestos furthered the development of mixed economies and contributed to
the growth of public expenditure.
m) Nationalization. The state takeover of private enterprises has increased public
expenditure in two ways, firstly by a government paying compensation to former
owners and secondly by subsidizing loss-making nationalized industries.
n) New technology and science. Some new technological developments in such fields as
atomic energy, aerospace and computers are so costly that in some countries they can
only be financed by the state or with substantial aid from government funds. Scientific
advances have enabled doctors to prolong life and reduce suffering, but in some cases
at an enormous cost to governments' health programmes by creating ever-increasing
demands.
o) Creation of super national organizations. The United Nations, NATO, European
Community and other multinational organizations that are responsible for the
provision of public goods and services on an international basis, have to be financed
out of funds subscribed by member states, thereby adding to their public expenditure.
p) Foreign aid. Acceptance by the richer industrialized countries of their responsibility
to help the poorer developing countries has channeled some of the increased public
expenditure of the donors into foreign aid programmes.
q) Increased complexity of national economies. As economies develop they become
more complex and the interests of various groups within a society come into conflict.
This has led to the proliferation of public bodies whose costs, arising out of their
coordinating, regulatory, administrative or judiciary functions are borne by
governments.
r) Inflation. A general increase in prices has been an international phenomenon during
the 1970s-1980s. Inflation increased the cost of all the activities of the public sector
and was thus a major factor in growth in money terms of public expenditure in many
countries.
s) Demographic changes. Since public expenditure is intended to benefit the people of a
country, it could therefore be expected that an increase in total population would
result in higher public expenditure. But other demographic trends such as changes in
the structure of the population (age and sex) and its geographical distribution also
have to be taken into account. The overall effect of the various trends on public
expenditure may be such that they cancel each other out, thus the extent to which the
growth of population has led to growth of public expenditure depends on the specific
conditions in different countries.

69
7.7 RESTRAINTS TO THE GROWTH OF PUBLIC EXPENDITURE

Some of the factors in the growth of public expenditure that we have discussed are of a
temporary nature, others contribute to structural changes that result in an increasing financial
commitment by governments on a permanent basis, but the ability to spend is not unlimited.
The following are the four main restraints:
a) Resources. In the long run, public expenditure cannot exceed the resources of a country.
b) Taxable capacity. This imposes a ceiling on the government's revenue from taxation and
thereby on an increase in public expenditure that is financed out of it.
c) Limit to borrowing. For a time public expenditure can outstrip revenue either as a matter
of necessity or of fiscal policy and the deficit can be financed out of loans. But there is a
limit to how much money lenders at home and abroad will be prepared to make available
to any government.
d) Public opinion.The final major restraint is the growth of public opinion. The level of
public expenditure in a democratic society will depend on the size of the public sector
that people want and are willing to pay for through taxation.

7.8 CONSEQUENCES OF THE GROWTH OF PUBLIC EXPENDITURE

Political, social and economic consequences are interrelated. They cannot therefore be easily
isolated and compartmentalized. Some are, however, more identifiable than others and are
listed below:
a) A political consequence of the growth of public expenditure is the increased size of the
public sector and hence of the power of the state.
b) A social consequence of the extension of the welfare system is to allay the fear of
deprivation that is consequent to unemployment, sickness and old age. The need for
people to provide for themselves is reduced.
c) Development of a welfare mentality is likely to increase people's dependence on
government support and to lead to the creation of what politicians and social
commentators call the 'underclass' in a society. Its members caught in the poverty trap
may lack the means, ability, resourcefulness and incentive to break out.
d) An economic consequence is an increase in taxation or borrowing or both, to finance
rising expenditure.
e) A disincentive effect on work and enterprise may result from an increase in taxation
required to finance provision of public goods and services but economists disagree on
this.
f) National debt will increase as a result of borrowing and this will affect the rates of interest
and supply of capital to industry.
g) The rate of economic growth may be adversely affected by the; transfer of resources from
use in manufacturing in the private sector to the public sector for provision of social
services.
h) The productive capacity and export potential of an economy may be reduced. Public
goods and services, such as social security benefits, are not exportable and do not earn
foreign currency.

70
i) The balance of payments, will suffer if exports are reduced and when interest payments
on the money that the government had borrowed abroad, or repayment of capital, become
due.
j) The prosperity of a country may, however, be increased if public expenditure is on
projects that further economic development. If this happens then the balance of payments
may improve.
k) The standard of living of the people in general and of some groups in particular can be
increased by the provision of public goods and services.
l) Inflation resulting from the injection of public spending into the income flow of a country
adversely affects not only the standard of living but the whole economy
m) Stabilization of the economy may result from the use of public expenditure to counteract
inflation and deflation.
n) The level of employment may rise, but if the effect of increased public spending is
inflationary, employment will be likely to fall.
o) A more egalitarian society can be achieved by narrowing the difference in the level of
consumption among its members by means of state benefits financed out of progressive
taxation.
p) Increased efficiency in provision of public goods and services as governments put greater
emphasis on value for money in an attempt to curb growing public expenditure.

NB: This list of favorable and adverse effects that may follow an increase in public
expenditure is by no means conclusive. Whether its consequences will be beneficial or not
will depend on the existing level of expenditure, the purpose for which the additional money
is used, the way that the expenditure is financed and the specific circumstances of a particular
country.

71
 CHAPTER EIGHT
EVALUATION OF PUBLIC EXPENDITURE PROGRAMMES:
COST-BENEFIT ANALYSIS (CBA)
START HERE
8.1 Introduction
Cost benefit analysis (CBA) is a principle basis of appraising public projects. It is the most
appropriate method for appraising projects from national point of view. CBA involves
enumeration, comparison and evaluation of benefits and cost. It implies weighing of returns
against the costs involved in a project. CBA therefore purports to describe and quantify the
social advantages and disadvantages of a policy or public project in terms of a common
monetary unit. The objective function is expressed as:

Net Social Benefits (NSB) = Benefits – costs: where the benefits and costs are measured in
terms of shadow or accounting prices of inputs and outputs.

To undertake CBA, one must identify the potential benefits and costs involved in a proposed
project. There are different types of costs and benefits in a project

8.2 Types of Benefits

Benefits refer to the additions to the flow of National output accruing from a project. The
different categories are as follows:

Real vs. Nominal Benefits

In CBA, we are concerned with real not nominal benefits. Consider a “river valley project”
which increases irrigational facilities to cultivators. If the state levies heavy betterment levy
on the cultivators, then the benefit of increased irrigational facilities to the cultivators is
nominal because the benefits goes to the treasury. But, if the same project besides increasing
irrigational facilities, raises the productivity of land per acre, and leads to a number of other
external economies whereby the level of real income of the farmers increase, then there
would be real benefits to the cultivators. Therefore with real benefits there is real increase in
output in the economy.

Direct vs. Indirect Benefits

Direct benefits are those benefits which are immediately and directly obtainable from a
project. These are also referred to as Primary benefits. They refer to increase in utility or
production as a result of consumption of outputs from the project. For example, in the River
Valley project there could be Flood control, improved irrigation and navigation facilities,
development of fisheries and generation of hydro electric power. These are direct benefits.
They are the products that are the reason for the investment.

Indirect benefits on the other hand are side effects (or secondary benefits). They are values
added to the direct benefits as a result of the activities stemming from or induced by the
project. For example, the river valley project apart from delivering its primary products

72
leading to primary benefits may also lead to employment of people, development of a market
centre, improvement of road transport, acquisition of skills and Managerial talents by local
people, etc.

Tangible and Intangible benefits

Some of the benefits from a project are tangible while some are intangible. Tangible benefits
are those that can be computed and measured in terms of money while the intangible are
those that cannot be measured in terms of money. The latter enter into individual valuations
for which there is neither a market nor a price. They may be positive (e.g. scenic beauty and
recreational value of the river project) or negative (e.g. resettlement of people).

SKIP TO Pp 77
8.3 Types of Costs

Project Costs
These refer to the value of resources used in constructing, maintaining and operating the
project. Therefore relates to costs of labour, capital, intermediate goods, etc, including
allowance for induced adverse effects.

Associated Costs
These refer to value of goods and services needed beyond those included in the cost of a
project, to make intermediate products or services of the project available for use or sale. For
example, in the River Valley Project, the farmers’ cost of producing the irrigated crop and
other costs such as cost of digging up the project.

Real vs. Nominal Cost

There is also a distinction between real and nominal costs. Suppose the developer in the river
valley project borrows from the people of the area resources to dig the canal, then there is no
real sacrifice from the people and the cost will be considered nominal. However, if the
people are asked to dig the canal themselves, then that constitutes real cost for them.

Direct (Primary) vs. Indirect (Secondary) Cost

Primary (direct) costs are those incurred for construction, maintenance and execution of a
project.

Secondary (Indirect) Cost are value of goods and services incurred to provide indirect
benefits of a project e.g. hospitals, schools, etc for the people working at the project site.
They also include cost of processing the immediate products of the project

Other categories of benefits and Costs

(a) Pecuniary benefits and costs
These arise because of changes in relative prices which occur in secondary markets as the
economy adjusts itself to the provision of the goods and services as a result of execution of
the project. The change in prices imply that gains (benefits) or losses (costs) accrue to some

73
individuals, but these are offset by gains or losses experienced by others. Therefore they do
not reflect net gains or costs to the society as a whole. They must therefore not enter into
projects’ valuation.

(b) Inside vs. outside

Inside - benefits and costs, which accrue within the jurisdiction in which the project is
undertaken.
Outside - benefits and costs, which accrue outside the jurisdiction in which the project is
undertaken, e.g. flood control project, which benefits people downstream. They
constitute spillovers from one jurisdiction to another.

Example 1: Irrigation Project

Real Benefits Costs

Direct Tangible: Increased farm output Cost of pipes
Intangible: Beautification of area Loss of wilderness

Indirect Tangible: Reduced soil erosion Diversion of water

Intangible: Preservation of rural society Destruction of wildlife

Pecuniary: Relative improvement in position of farm equipment industry.

Example II: Education Project

Real Benefit Costs
Direct Tangible: Increased future earnings Loss of students' earnings,
Teachers, salaries, cost of
buildings and books

Intangible: Enriched life foregone leisure time.

Indirect Tangible: Reduced cost of crime prevention.

Intangible: More intelligent electorate

Pecuniary: Relative increases in teachers incomes.

74
8.4 Measurement of costs and benefits
After identifying all the benefits and costs involved in a public project, there is need to give a
monetary value for each. The following are guidelines for valuation:

(i) Valuation of intangible non-market items: The value of intangible benefits and costs
are derived through a political process by voting e.g. defense, clean air etc.
(ii) Cost saving: estimates benefits in terms of costs saved e.g. reduced hospital costs,
reduced school dropouts.
(iii) Tangible costs and benefits: Tangible benefits are measured by the price, which the
output or service fetches in the market. Tangible costs are measured by the price
which must be paid for the product or by the cost which must be incurred in foregoing
the alternative private use of resources.
(iv) Shadow price: This refers to the true or intrinsic prices which reflect the scarcity of
resources in the economy i.e. it is the time value for a factor or a product also referred
to as the underlying social marginal cost of a good

Circumstances when shadow pricing should be used

(a) Imperfect market - Prices and costs don’t reflect true social valuations and hence
adjustments are needed
Example: If market costs of a given product is Ksh. 1000, but in a competitive
market it would cost Ksh. 800. The opportunity cost is Ksh. 800 hence in its price
of Ksh. 1000, the Ksh. 200 becomes a pecuniary gain to the monopolist but not a
real cost to the society.
(b) Taxes - market prices may include taxes. But taxes do not reflect a social cost (it
is a transfer from purchases to government) and should therefore be disallowed
in computing the cost of the project
(c) Unemployment: - the cost to be accounted for in public resource use is the lost
opportunity for putting the resources to alternative uses. This reason breaks
down if the resources are otherwise unemployed and because the opportunity
cost is zero.
(d) Developing economies - developing economies experience unemployment,
labour cost is too low, hence does not reflect their time social cost. Some even
overvalue their currency hence need for shadow pricing.

8.5 Discounting cash flows

Discounting the cash flows means translating future benefits streams into present values. It is
important because future benefits are less valuable than the present ones. To realize public
investment resources are withdrawn from private use. The opportunity cost of resources
withdrawn from the private sector should be measured in terms of the present value of private
consumption foregone. Future consumption losses (due to forgone investment) are similarly
discounted to their present values. Suppose B0 , B1 , B2 ,..., B N are given as the benefits in year
0, 1, 2,.., n, then
B0 B1 B2 Bn
Pr esent Value ( PV )  0
 1
 2
 ... 
(1  r ) (1  r ) (1  r ) (1  r ) n
N
Bn
Therefore PV   n
t  0 (1  r )

75
 1
where; r = discounting rate and = discount factor.
(1  r) n

Discounting means that we discount benefits to their base year equivalent.

Example:
Suppose there are two projects whose cash flows are as follows:

Projcet Year
0 1 2 3
A +1000 +500 +600 0
B +2361.60 +1000 1000 1000

With a discount rate of 10%

tn Bn  1000  500  600  0

PV A  
t 0 (1  r ) n (1  .1) 0 (1  .1) 1 (1  .1) 2 (1  .1) 3

= 1000(1.00) +500(0.909) + 600(.826) + 0(.751)

= 1000 + 454.5 + 495.60 + 0
= $1950

PVB = 2361.6(1.00) + 1000 (.909) + 1000 (.826) +1000 (.751)

= 2361.60 + 909 + 826 + 751
= $4847.60

Choice of Discount Rate

1. Private rate of discount rate

The discount rate used should be equal to the time preference of consumers in the private
sector provided that this may be derived from observed market rates. The rationale for using
private rate of discount is that it reflects consumer choice between present and future
consumption. In a perfect market, this interest rate will be equal to the marginal efficiency of
investment.

Difficulties in the use of private discount rate

1. Capital market imperfections such as differential access to credit and investment
institutions make different individuals be confronted with different costs and returns to
borrowing and lending. Hence no single rate may be obtained.
2. Uncertainty of future level of interest rate hence short and long term rates in the capital
market differ.

76
3. Some investment projects are risky than others hence gross rates of return differ by the
amount of risk premiums.
4. Income tax- lenders pay income tax on their capital. Proper measure of their time
preference is therefore the net or after-tax rate of return and not the gross market rate.
5. Macro-policy – the presence of unemployment and inflation lead to a need for
intervention policies to rectify the situation. Therefore, the market does not reveal correct
level of interest by which consumer time preference is reflected.

2. Social rate of discount

Due to the above limitations, some economist advocate for the use of a social rate of
discount, and this is the time preference rate for future generation. It measures the rate at
which society is willing to trade off present consumption for future consumption. The social
rate of discount may be lower than the private rate.

Reasons for this substitution (Private Rate to Social Rate)

1. Individuals underestimate the importance of savings and overestimate that of present
consumption. Hence consumer’s time discount is too high and the government should
correct this by applying a lower rate.
2. People are too greedy and do not care sufficiently about the welfare of those to come.
They therefore do not save more in order to leave future generation with larger capital
stock. Hence the government may use a low rate to offset this.
3. People do care about future generation and will derive pleasure from contributing to
their welfare. But they do not act as a group hence the government should use a lower
rate of discount in order to increase investment.
4. With technical progress raising future productivity the capital stock needed to sustain
the consumption standard may fall, calling for a higher discount rate.
5. The equilibrium rate of return to capital should be equal to the growth rate of the
economy which in turn equals the growth rate of the population.

8.6 Project / Programme Evaluation Criteria START HERE

1. Net Present Value Criteria

Net Present Value (NPV) is equal to the sum of the present value of all the cash flows
associated with a project.

t
CFo CF1 CFt CFt
NPV =
(1  r ) 0

(1  r ) 1
 ........... 
(1  r ) t
=  (1  r )
t 0
t

Where;
NPV = Net Present Value
CFt = Cash flow occurring in period / year t (t = 0, i --- n)
t = Life of the project
r = Opportunity Cost of capital

77
Example I
Suppose the following information represent cash flows in different time periods from a
project, in which the cost of capital (r) is 10%, calculate the NPV

Year Cash flow

0 (1,000,000)
1 200,000
2 200,000
3 300,000
4 300,000
5 550,000

NPV = -1,000,000 + 200,000+ 200,000 + 300,000 + 300,000 + 550,000

(1 + 0.1)0 (1 + 0.1)1 (1 + 0.1)2 (1 + 0.1)3 (1 + 0.1)4 (1 + 0.1)5

= 118,750
118,912.65

78
Example 2
Consider the previous example on deciding between Machine X and Y

Cash flow
Year Machine X Machine Y Discount factor P.V for P.Vfor
1 Machine X Machine Y
(15%); t
(1  r) (20,000) (20,000)

0 (20,000) (20,000) 1.00 (20,000) (20,000)

1 10,000 12,000 .870 8,700 10,440
2 8,000 10,000 .756 6,048 7,560
3 7,000 8,000 .658 4,606 5,264
4 5,000 5,000 .572 2,860 2,860
5 5,000 5,000 .491 2,455 2,4 55
NPV 4,669 8,579

Machine Y is chosen because it has the highest net present value.

Generally NPV is computed using the formula;
B  C 0 B1  C1 B  Ct
NPV  0 0
 1
 t
(1  r ) (1  r ) (1  r ) t

t n
Bt  C t

t 0 (1  r ) t

*Decision Criteria (Rule)

 NPV > 0 (+ ve) – the project is viable – accept. A positive NPV implies that the project
earns an excess return
 NPV < 0 (- ve) – the project is not viable – reject
 NPV = 0 Indifferent; Thus consider other factors such as social and political factors.
NPV of zero signifies that the benefits of the project are just enough to:-
- Recoup the capital invested and
- Earn the required return on the capital invested

Context within which NPV criteria may be used

(i) Accept – Reject
Faced with a single project, the NPV rule dictates that it should be accepted if the NPV
exceeds zero and rejected if it is less than zero. If NPV = 0, the decision maker would be
indifferent between undertaking the project or not.

(ii) Ranking
Where the decision maker has a series of investments all with positive NPVs and yet the
budget is fixed, she needs to rank projects in order of desirability and work down the list until
the budget is exhausted.

79
(iii) Mutual Exclusion
Where the choice is between projects, the general rule is to select the project offering the
highest NPV.

Observations
(a) Strict exclusivity – implies that either project may be undertaken or neither.
However, where there is variable degree of exclusivity such that a total commitment
to the other project, and some combination is possible, it is essential to consider any
such combinations and compute their NPVs. The combined projects should then be
treated as if they were extra projects. Exclusively still applies i.e. one may be able to
undertake one project or the other or some combinations being guided by the rule of
the highest NPVs.
(b) In the presence of capital rationing, it is necessary to normalize projects so that they
are comparable.

Advantages of NPV Criteria

(i) It recognizes time value of money.
(ii) Considers all cash flows over the entire life of the project.
(iii) Ease to compare/rank projects.
(iv) The NPV of various projects can be added. This ensures that a poor project (with a
negative NPV) will not be accepted just because it is combined with a good project
(with positive NPV) since NPV (A + B) = NPV (A) + NPV (B)

Disadvantages of the NPV Criteria

(i) NPV depends on the rate of discount chosen and tends to favour large projects, which
are capital intensive.
(ii) Given a discount rate, it favours projects whose returns (benefits) accrue in the early
life.
(iii) It may not give satisfactory results when the project compared involve different
amounts of investment.
(iv) Ranking of projects on NPV is influenced by the discount rate.
(v) NPV is an absolute measure and does not appeal to businessmen who think in terms
of rate of return.

2. Benefit Cost Ratio (BCR)

There are two ways of looking at the Benefit Cost Ratio:

(i) Benefit cost ratio (BCR) = PVB

I
OR

(ii) Net Benefit Cost Ratio (NBCR) = NPV = PVB-I = BCR - 1

I I

Where; BCR = Benefit Cost Ratio

PVB = Present Value of Benefits

80
 I = Initial investment
NBCR = Net Benefit Cost Ratio
NPV = Net Present Value

Example
If cash flows for a project are as follows

Year 0 1 2 3 4
Benefits 100000 25000 40000 40000 50000

With a discount rate of 12%

25000 40000 40000 50000

  
(1.12)1 (1.12) 2 (1.12) 3 (1.12) 4
(i) BCR   1.145
100000

(ii) NBCR = BCR – 1 = 1.145 - 1 = 0.145

Decision Rules
BCR > 1 Accept NBCR > 0 Accept
BCR = 1 Indifferent NBCR = 0 Indifferent
BCR < 1 Reject NBCR < 0 Reject

Merit
Since the criterion measures NPV per shilling of outlay, it can discriminate better between
large and small investments hence preferred to NPV criterion.

Short comings
- Under unconstrained conditions, it will accept and reject the same projects as the NPV
- When the capital budget is limited in the current period, it may rank projects correctly in
the order of decreasingly efficient use of capital. However, its use is not recommended
because it provides no means for aggregating, several smaller projects into a package that
can be compared with a large project.
- When cash outflows occur beyond the current period, the BCR criterion is unsuitable as a
selection criterion.

8.7 Strengths and Limitations of CBA

Strengths
- Provides a framework for structuring information and considering trade-offs.
- Helps make strategic choices about program priorities.
- Helps weed out the least desirable alternatives
- Can identify areas in which uncertainty is greatest and further research is desirable.

81
- Can increase explicitness in decisions and thereby elevate the level of public debate and
the usefulness of public participation.
- Enhances consistency among decisions.
- Can help assess the cumulative effects of regulations on groups, industries future
generations, geographical areas, etc.
- Can improve the credibility of government by showing how decisions are made and that
they are rational.

8.8 Limitations

a) Difficulties in cost assessment: Cost estimates are made on the basis of the choice of
techniques, locations and prices of factor services used. But market prices, particularly
those of factors of production form an imperfect guide to resource allocation in under-
developed economies, because there exist fundamental disequilibria which are reflected
in the existence of massive underemployment at present level of wages, the deficiency of
funds at prevailing interest rates and the shortage of foreign exchange at current rates of
exchange. Hence use of shadow or accounting prices has been suggested. These shadow
prices reflect intrinsic values of factors of production.
b) Difficulties in Benefit Assessment, due to the element of uncertainty in a new project as
to the correct estimation of future prices, demand and supply of its products, and
difficulty in assessing external economies.
c) Arbitrary Discount Rate – the assumed social rate of discount for any project is likely to
be arbitrary. This might not make it possible to effectively calculate net present value of
benefits.
d) Neglects joint Benefit and costs arising from a project. It is difficult to evaluate and
calculate such benefits separately.
e) Ignores opportunity costs

f) Adjustment for Risk and Uncertainty problem also arises. This is done in three ways:-
 At the time of calculating the length of project life
 the discount rate and
 by making due allowance in benefit and costs
g) The problem of externalities. Side effects of a project are difficult to calculate in this
analysis. There may be technological and pecuniary spillovers (externalities) of a river
valley project such as the effects of flood control measures or a storage dam on the
productivity of land at other places in the vicinity. It is difficult to calculate such external
effects.
h) Traditional CBA focuses only on efficiency but other factors e.g. administrability,
distribution of impacts, and promotion of technological innovation, may be of equal or
great importance in decisions.
i) CBA takes technological as given and cannot anticipate technological breakthroughs that
reduce costs.
j) CBA may lead to short sighted and undesirable decisions of agencies, preservation of the
democratic system “irrational” but strongly held views of the public etc are ignored.
k) Does not consider distribution of costs and benefits hence inequitable decisions are made.

82
 TOPIC NINE

THE PUBLIC BUDGET

9.1 The budget and reasons for budgeting:

Is a statement describing the estimated receipts and proposed expenditures and disbursements
under various heads.

The government plans to undertake various economic and other activities and pursue certain
policies. All these have their financial implications in form of collecting the necessary
revenues and incurring expenditures.

Therefore, a budget enables the authorities to decide about each individual item of revenue
and expenditure in the overall context to the total plan.

The government cannot go ahead with decisions regarding taxation, borrowing, expenditure
and other fiscal measures at random. All these decisions and policies are interrelated and
should form a part of the overall set of objectives which the government plans to pursue.

It should show the financial accounts of the previous year, the budget and revised estimates
of the current year and the budget estimates of the coming year.

The estimates of the coming years are split up into two parts:
- Those based upon the assumption that the current year’s taxes and their rates, and
expenditure policy, would continue and
- Those based upon the proposed changes therein.

The budget then becomes both a description of the fiscal policies of the government and the
financial plans corresponding to them.

In the budget, some revenues may be earmarked for certain specific expenditures. For
example, certain projects might be financed through public borrowings and collections
through betterment levies or special assessments fees.

On the expenditure side, certain amounts may be contractual and legal payments which
cannot be avoided such as interest payments on loans, repayment of loans, payments arising
out of satisfaction of certain court decrees, amounts falling due for payment on account of
salaries, pensions, provided funds among others.

9.2 A budget may be presented in parts (split) due to the following reasons:
(i) If the budget is divided into layers: where each layer (department, sub-
government) of the government prepares, passes and implements its own budget.
(ii) Political cause: when the existing executive government may or may not continue
for the full year on account of the fact that election is due. In this case, a lame
duck budget is presented (A budget which covers only a part of the year).
(iii) Economic cause: This results in supplementary budget. This is necessitated by
unforeseen emergencies such as war, natural calamities, terrorism etc that calls for
extra expenditure.

83
 (iv) The revenue receipts may fall short of the expected one. This may call a
supplementary budget.

9.3 Three issues have been raised on the annual budgeting in Kenya
(i) Busy seasons of the economy: Since agriculture still contribute largely to the country’s
GDP, the fiscal year should start with the busy sectors of the economy.
(ii) Annual practice of preparation, presentation and passage of the budget is a waste of
time. Expenditure against amounts sanctioned starts with a time lag while preparation
for the next annual budget starts soon after. Frequent changes do not allow a judicious
assessment of the effects of various tax measures. The broad features of the tax
structure should be left unaltered for a few years at a time and only minor changes
should be allowed from year to year.
(iii) Secrecy surrounding the budget proposal preparation. Budget proposals are
unnecessarily kept secret till their actual presentation to the parliament. This practice
causes a lot of uncertainty and speculation, which is damaging to proper planning of
economic activities by each individual/agent in the economy. For example, there is no
reason why tax proposals should not be debated widely to assess their impact on
savings, capital formation among others, before their passage. This will cause less
speculation and enable the public to participate in the budget deliberation in a more
constructive manner. They will be better placed to modify the proposals to suit the
needs of the economy.

9.4 Principles of a good budget

(i) It should be accompanied by an account of the performance of the fiscal policies
and programmes of the government during the previous year. This provides a
basis for deciding what was to be done, what has been accomplished and what
more should be aimed at and what direction.
(ii) Should be accompanied by an analytical description of the current economic
situation of the country and the position of treasury. This enables the legislature
and the public to see the relevance or otherwise of the budget proposals and help
the legislature in taking more objective and rational stands.
(iii) Proposals should be clearly comprehensive so that correct judgment can be
formed as to the way in which the budget is expected to function in the coming
year. Therefore, detailed budget proposals must accompany the proposals under
major heads of receipts and expenditure. Various statements should be used to
highlight particular aspects of the budget.
(iv) An economic and functional classification of the budget will depict its role in the
working of the economy in a clearer way than only the financial proposal under
various departments.
(v) Financial estimates should be as close to the actual estimates as possible under
normal circumstances.
(vi) It should reflect the overall-policy and purpose of the government and should be
so designed as to help the society more nearer to the chosen goal.

9.5 The Types of budgets

a. A legislative budget and an executive budget:
A legislative budget: one which is prepared and adopted by the legislature directly
or through its committees.
An executive budget: one which is prepared by the executive branch of the
government. It is normally passed and adopted by the legislature but the initiative

84
 is in the hands of the executives. The executive budget is preferred to the
legislative budget due to the following reasons:
(i) The executive is better equipped to estimate properly probable
receipts and the required expenditures.
(ii) It is the executive, who will be responsible for the execution of the
budget. It is not good to ask the executive execute a budget, which
did not originate from them.
(iii) The executive can be held directly responsible for any
shortcomings and lapses.

b. Multiple Budget and unified budgets:

A multiple budget is one divided into parts and presented in such away as to make it
possible to evaluate the specialized functions of the government. The true results of the
fiscal operation of government budget get scattered and have to be traced out through
various documents.

Unified Budget: one whose parts are presented as a whole (unified). Its total effect can
be seen and which is more relevant. A long with this budget, important sub-portions may
be classified and presented separately.

c. Conventional/Administrative budget and cash budget:

Conventional/Administrative budget: revenues and expenditures are shown on accrual
basis and those flows of funds which do not belong to the government are excluded. It
harbours two deficiencies as far as its role in the flow of funds in the economy and
therefore its effects on the economy are concerned.
(i) It runs on accrual basis: As receipts and payments falling due in a period
generally differ from the ones actually made, there is a distortion of the actual
picture of the flow of funds caused by government activities.
(ii) It always presents an inadequate picture of government activities since the actual
handling of cash receipts and payments have their profound role in the flow of
funds and the effect thereof.

Cash Budget: all flows of funds to and from government of actual payment basis are
shown, inclusive of funds, which are not owned by the government. It is larger than the
administrative budget and is a better representation of reality.

d. Revenue Budget and capital budget

Revenue Budget: covers those items, which are of recurring nature. It consists of revenue
receipts both tax and non-tax and the expenditure met out of revenue receipts. Non-tax
revenues include; currency, coinage and mint, interest receipts, dividends, profits,
revenue from general services, social services. It forms current spending which refers to
expenditures for services that are consumed within the year.

Capital Budget: covers those items, which are in the nature of acquiring and disposing of
capital asset. Capital account receipts include; market loans, borrowings from the
reserves, sale of treasury bills, loans from foreign governments and others. Capital
disbursements include expenditure on acquisition of various physical assets like land,
buildings, machinery and debentures and loans to state government and other bodies.

85
 Capital spending refers to expenditures for durable items that yield services over a long
time. The accounting procedure is to keep separate budgets for current and capital
expenditures. Maintaining a separate capital budget can provide a more accurate picture
of an organization’s financial status. This is because acquisition of the asset does not
contribute to an organization’s deficit. Only the annual depreciation of durable assets is
included in the current budget, and not their entire purchase price. The proceeds from the
sale of assets are treated as equivalent to tax revenues, and so count toward reducing the
deficit. Another argument in favour of capital budgeting is that it reminds people that
borrowing is not necessarily a bad thing. Just as a prudent household may go into debt to
purchase capital assets like a house or a car, a prudent government can borrow to finance
the purchase of long-lived assets.

Opponents of capital budgeting point out that, for governments, it is particularly difficult
to distinguish between current and capital expenditure. Are educational and job-training
programmes a current expense or an investment in human capital that will yield future
returns? Is a missile an investment (because it will last a long time), or a current
expenditure (because it is not reusable)? Such ambiguities could lead to political
mischief, with every proponent of a new spending programme claiming it wants an
investment and therefore belong in the capital budget. In fact, advocates of transfer
programmes such as food stamps often promote them as “investments”, because
enhancing the diets of the poor today makes them more productive tomorrow. Critics
assert that classifying transfer payments as investments renders meaningless the
distinction between capital and current spending.

Arguments for the division of budget into capital and revenue account:
(i) Every economic unit must distinguish between current expenses and those
incurred for acquiring of capital assets. Current expenses are equivalent to
consumption, while acquiring capital asset is not. It is only when these capital
assets depreciate that real expenditure takes place in the sense of consumption.
The government should not count the amounts spent on the acquisition of capital
assets as part of current expenses. Only the depreciation part should be counted.
(ii) Through such division, the government can follow a good working rule i.e.
deciding that all the current expenses would be met through taxation while all the
capital expenses would be met through borrowing.

e. Economic and functional classification of the budget

Economic classification of the budget: classification of the expenditure and the mode of
its financing in terms of economic categories (expenditure on wages, salaries, receipts
from market borrowing etc). It enables the gathering of a variety of economic
information relating to generation of savings, investment, consumption, creation of
financial assets and liabilities etc from various budget items.
Functional classification: refers to the types of tasks, which the government performs, or
the services, which it provides such as economic services, social services, defence
services and the like. It covers only the expenditure and not the receipts
Strengths of this classification:
1. It cuts across the departments and agencies and presents a picture of what the
government is doing in various spheres such as capital accumulates, health, education
etc. Hence helps in arriving at various decisions.

86
 2. Provides information that is required for formulation of fiscal policies since there is
correspondence between the purpose and the account head

Possible limitations of economic and functional classification of budgets

- It may not reveal the repercussions (or indirect effects) of the budgetary activities of
the government. And no precise technique is available to perform these tasks.

- It does not provide relevant information since there is no detailed breakup and
regulatory devices.

- Does not reflect the impact of fiscal operations on income and wealth inequalities or
the impact on many important growth variables like entrepreneurship, institutional
framework and regional disparities.

- Need an integrated classification of the budgets of all public authorities put together
and covering the entire economy.

- Accounts may be highly aggregative. A more detailed breakdown would enhance

their useful.

- It may use arbitrary and unrealistic benchmark ratios, which have no factual data.

9.6 Performance and Programme Budgeting System (PPBS)

This is when a budget covers both performance and programme.
Programme budgeting: involves laying down the sequence of steps for executing the project
along with the expenditure of resources involved at each stage.
Performance budgeting: devised tests for comparing actual with the expected results and
thereby assessing the performance efficiency of the project.

PPBS ensure success of the implementation of the budget proposals which depends upon.
 Inherent soundness of the proposals
 The efficiency of the government machinery
 The adequacy and
 Integrity of the government machinery.

The formulation of the budget proposals should be directly related to the extent to which they
can be implemented. Alternative expenditure proposals should therefore be weighted in terms
of their respective costs and benefits. The PPBS involves the rules of managerial efficiency,
and flexibility in formulation and execution of expenditure policy of the government.

Rationale of PPBS
(i) Recognize that economic problems arise out of scarcity of resources as compared
with the need by the society.
(ii) The choice and efficient execution of the chosen project and steps of achieving
efficiency is emphasized.

87
 (iii) Recognizes the fact that in the conventional budgeting and policies, there are no
automatic regulators which would indicate to the executive and legislature as to
when the usefulness of a particular project has ceased.
(iv) Without PPBS, haphazardness may be experienced leading to ad hoc solutions.
(v) PPBS puts emphasis on forward programming which leads to effective assessment
and coordination between budgeting provisions and policies over a span of time.

Limitations of PPBS
1. Multiplicity: a given objective may be the target of more than one programme and
a single programme may aim at more than one targets.
2. Quantification: not possible to quantitatively measure results of many
programmes.
3. Flexibility: less flexibility available in economic activity from the political point
of view and even data limitation.
4. Institutional Framework: no uniform format suitable for every economy.
5. Analytical skills: skills required to prepare PPBS may be lacking hence applied
only to a limited extent.
6. Accounting system: budget accounts may not be of the type to permit the needed
formulation of PPBS. It needs functional classification of the budget and
accounting system.

9.7 Zero-base budgeting (ZBB)

It involves examination of the very rationale of an expenditure item under consideration. The
aim is to guard against wastage in public expenditure. It involves detailed investigation of
each item of expenditure to see whether it is really needed or it should be revised or done
away with.
If a section is not able to justify its existence, it should be closed down. If its existence is
justified, the optimum level of its operations and the corresponding budgetary provisions
must be defended. In ZBB, no section is essential. It must prove itsworthiness.

9.8 IDEAL BUDGET CYCLE

1. Review and analysis: The results of budget planning and implementation of the
previous fiscal year’s budget proposal are taken as the background for the new
budget planning.
2. Plan strategically:
 Estimate macroeconomic indicators and revenues
 Set out broad policy objectives and expenditure priorities
 Analyse the medium-term fiscal and economic effects of proposed policies
 Set expenditure targets.
3. Mobilize and allocate resources
 Prepare the budget by linking budget allocations to the cost estimates of the
delivery of government services and the accomplishment of the policy
objectives.
 Encourage bidding by agency work programmes
 Adjust civil service to the agency work programmes

88
4. Implement efficiently
 Collect revenues, prioritise before spending, release funds through a transparent
treasury system, deploy personnel, and undertake activities.
5. Monitor and account
 Supervise government activities and the overall implementation of the budget.
 Monitor personnel and accounting information
 Account for and monitor commitment and expenditure.
6. Evaluate audit and publish
 Assess the effectiveness and the value for money of government programmes
and seek accountability.
 Audit government accounts and the financial and managerial performance of
the public sector
 Publicize the results of evaluation and audit.

89
CLARE GATHOGA
EES 200: MATHEMATICS FOR ECONOMISTS II

FACILITATOR: DOROTHY NGINA

LECTURE NOTES

Page 1 of 167
 LECTURE ONE: EQUILIBRIUM ANALYSIS

1.1 Introduction.

Welcome to our first lesson of Mathematics for Economist II. In this lesson we start by discussing

equilibrium analysis. The general objective of this lesson is to understand how to compute market

equilibrium and the national income equilibrium. In this lecture the learner will learn how to get

equilibrium values in simultaneous equations. Knowledge of solving simultaneous equations will

be very useful in this course especially when dealing with the general equilibrium.

1.2 Lesson Learning Outcomes

By the end of this lesson, you will be able to:

1.2.1 Compute the equilibrium price and quantity in the partial market model.

1.2.2 Compute the equilibrium price and quantity in the general market model.

1.2.3 Compute the equilibrium in the national income model

1.2.1 Equilibrium in the partial market model

Equilibrium: A constellation of selected interrelated variables so adjusted to one another that

no inherent tendency to change prevails in the model which they constitute. It is important to

explain some few terms that are in this definition

i) Selected variables: - There could be other variables that have not been included. If the

model is enlarged to include more variables, the equilibrium may no longer hold.

ii) Interrelated: - All the variables included must be simultaneously be in a state of rest.

Page 2 of 167
 iii) Inherent – when defining the equilibrium, the state of rest is based only on the

balancing of internal forces of the model. The external forces are therefore assumed to

be constant.

For example: Assume a demand function is given as: 𝑃 = 50 − 2𝑄,

50 𝑎𝑛𝑑 2 are parameters of the equation and are assumed to be exogenous, hence they do not

change. If any of the external factors change, a new equilibrium will prevail.

Equilibrium therefore can be simply defined as a situation characterized by lack of tendency to

change.

An equilibrium does not imply an ideal state of affairs. For example, the market price for a bag of

wheat could be Kshs. 4,000 which may be low to some of the farmers or is too high to the millers

who are buying it. But at this price, the total supply of wheat to the market is equal to the total

demand of wheat in that market.

We will analyze the partial market equilibrium using both a linear model and a non-linear model.

a) Partial market equilibrium - Linear model

This is a model of price determination in an isolated market/ single market. Only one

commodity is considered. Equilibrium in the market for a single commodity occurs when the

quantity demanded (Qd) is equal to the quantity supplied (Qs) in the market. For equilibrium to

occur, the price that the consumers are willing to pay is equal to the price that the sellers are willing

to accept.

Conditions for equilibrium

No excess demand in the market i.e

Page 3 of 167
𝑄𝑠 − 𝑄𝑑 = 0 𝑡ℎ𝑖𝑠 𝑤𝑖𝑙𝑙 𝑖𝑚𝑝𝑙𝑦 𝑡ℎ𝑎𝑡

𝑄𝑠 = 𝑄𝑑

The quantity demanded function and quantity supplied function are functions of price. If price is

below the specified level nothing will be supplied. The demand function can be given as:

𝑄𝑑 = 𝑓(𝑃) 𝑤ℎ𝑖𝑐ℎ 𝑐𝑎𝑛 𝑏𝑒 𝑔𝑖𝑣𝑒𝑛 𝑎𝑠:

𝑄𝑑 = 𝑎 − 𝑏𝑃

This indicates that the demand function is negatively sloped. While the supply function will be

given as:

𝑄𝑠 = 𝑓(𝑃) 𝑎𝑛𝑑 𝑐𝑎𝑛 𝑏𝑒 𝑔𝑖𝑣𝑒𝑛 𝑎𝑠:

𝑄𝑠 = −𝑐 + 𝑑𝑃

In the demand and supply functions, a, b, c and d are positive constants. The supply function is

positively sloped. At equilibrium: quantity demanded is equal to quantity supplied, that is:

𝑄𝑑 = 𝑄𝑠

The equilibrium price and quantity can be determined graphically by drawing both the demand

and supply function on the same graph. Both the quantity demanded and quantity supplied will be

plotted on the Y-axis while the price will be plotted on the X-axis.

Page 4 of 167
The equilibrium price and quantity are P* and Q* respectively. It will be the price and quantity

when the supply curve intersects with the demand curve.

Determination of the equilibrium price and quantity algebraically

If we have the demand and supply functions of the commodity we can determine the equilibrium

values of P and Q algebraically.

The demand function is given as:

𝑄𝑑 = 𝑓(𝑃)

Page 5 of 167
𝑄𝑑 = 𝑎 − 𝑏𝑃

While the supply function will be given as:

𝑄𝑠 = 𝑓(𝑃)

𝑄𝑠 = −𝑐 + 𝑑𝑃

At equilibrium:

𝑄𝑑 = 𝑄𝑠

If we equate the demand and supply functions, we get:

𝑎 − 𝑏𝑃 = −𝑐 + 𝑑𝑃

𝑎 + 𝑐 = 𝑑𝑃 + 𝑏𝑃

𝑃(𝑑 + 𝑏) = 𝑎 + 𝑐

𝑎+𝑐
𝑃∗ = 𝑑+𝑏

𝑃∗ should be greater than zero since all the constants a, b, c and d are positive.

To get the equilibrium quantity, we can replace the value of 𝑃∗ either in the demand or supply

function.

Let’s replace it in the demand function.

𝑄𝑑 = 𝑎 − 𝑏𝑃 ∗

𝑎+𝑐
= 𝑎−𝑏[ ]
𝑑+𝑏

Page 6 of 167
 𝑎𝑑 + 𝑎𝑏 − 𝑎𝑏 − 𝑐𝑏
=
𝑑+𝑏

𝑎𝑑 − 𝑐𝑏
𝑄∗ = 𝑓𝑜𝑟 𝑄 ∗ 𝑡𝑜 𝑏𝑒 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒, 𝑎𝑑 𝑠ℎ𝑜𝑢𝑙𝑑 𝑏𝑒 𝑔𝑟𝑒𝑎𝑡𝑒𝑟 𝑡ℎ𝑎𝑛 𝑐𝑏 𝑖. 𝑒
𝑑+𝑏

𝑎𝑑 > 𝑐𝑏

Intersection between demand and supply function should occur above the X-axis for

𝑃∗ 𝑎𝑛𝑑 𝑄 ∗ 𝑡𝑜 𝑏𝑒 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒.

Example

The demand and the supply function for a good are given as:

𝑄𝑑 = 200 − 2𝑃

While the supply function will be given as:

𝑄𝑠 = −20 + 2𝑃

Required:

Determine the equilibrium price and quantity

At equilibrium, 𝑄𝑑 = 𝑄𝑠

If we equate the demand and supply functions, we get:

200 − 2𝑃 = −20 + 2𝑃

220 = 2𝑃 + 2𝑃

4𝑃 = 220

Page 7 of 167
 220
𝑃∗ = = 55
4

𝑄 ∗ = 200 − 2[55]

= 200 − 110

𝑄 ∗ = 90

Exercise

The demand function is given as:

𝑃 = 100 − 2𝑄𝑑

While the supply function is given as:

𝑃 = 10 + 7𝑄𝑠

Determine the equilibrium price and quantity

b) Partial market equilibrium – Non-linear model

The demand and supply functions for a commodity are given as:

𝑄𝑑 = 6 − 𝑃2

𝑄𝑠 = 10𝑃 − 5

For equilibrium

𝑄𝑑 = 𝑄𝑠

6 − 𝑃2 = 10𝑃 − 5

Page 8 of 167
 𝑃2 + 10𝑃 − 11 = 0

Since the resulting equation is a quadratic equation, then we can use any one of the methods that

are available for solving such equations: Factorization, the quadratic formula, completing the

square method, and so on.

For instance, if we use the factorization method, we obtain:

𝑃(𝑃 + 11) − 1(𝑃 + 11) = 0

(𝑝 − 1)(𝑝 + 11) = 0

𝑃 = 1 𝑜𝑟 𝑃 = −11

We can only deal with positive prices hence equilibrium price is 1 while the equilibrium

quantity will be:𝑄𝑑 = 6 − (1)2 = 5

Self-test question

Determine the equilibrium prices and quantities in the following two problems:

i) 𝑄𝑑 = 3 – 𝑃2 , 𝑄𝑠 = 6𝑃 – 4

ii) 𝑄𝑑 = 8 − 𝑃2 , 𝑄𝑠 = 𝑃2 − 2

General Market Equilibrium

Rarely will the demand for a product be a function of own price alone. Prices of other related

commodities such as complements and substitutes will affect its. General market equilibrium deals

Page 9 of 167
with a case whereby demand or supply of a commodity is a function of its own price and also

prices of related commodities.

Numerical Example

The demand and supply functions of 2 commodities are given as:

𝑄𝑑1 = 18 − 3𝑃1 + 𝑃2 𝑄𝑑2 = 12 − 𝑃1 − 2𝑃2

𝑄𝑠1 = −2 + 4𝑃1 𝑄𝑠2 = −2 + 3𝑃2

At equilibrium:

Qd1 = 𝑄𝑠1 𝑎𝑛𝑑

Qd2 = 𝑄𝑠2

18 − 3𝑃1 + 𝑃2 = −2 + 4𝑃1 ⇒ 7𝑃1 − 𝑃2 = 20

12 − 𝑃1 − 2𝑃2 = −2 + 3𝑃2 ⇒ 𝑃1 + 5𝑃2 = 14

This gives us two simultaneous equations that can be solved simultaneously to obtain:

116 78
𝑃1 = 𝑎𝑛𝑑 𝑃2 = 𝑇ℎ𝑒 𝑒𝑞𝑢𝑖𝑙𝑖𝑏𝑟𝑖𝑢𝑚 𝑞𝑢𝑎𝑛𝑡𝑖𝑡𝑖𝑒𝑠 𝑓𝑜𝑟 𝑔𝑜𝑜𝑑 1 𝑎𝑛𝑑 2 𝑤𝑖𝑙𝑙 𝑏𝑒
36 36

98
𝑄1 = 𝑎𝑛𝑑 𝑄2 = 4.5
9

In the general market, equilibrium requires that all the markets be in equilibrium simultaneously.

For each market i, equilibrium requires that:

𝑄𝑑𝑖 − 𝑄𝑠𝑖 = 0

Page 10 of 167
 Or

𝑄𝑑𝑖 = 𝑄𝑠𝑖

We will get a set of prices, Pi and Qi for each market that will ensure the market is in equilibrium.

𝟐 𝒄𝒐𝒎𝒎𝒐𝒅𝒊𝒕𝒚 𝒄𝒂𝒔𝒆

𝑄𝑑1 = 𝑄𝑠1

𝑄𝑑1 = 𝑎0 + 𝑎1 𝑃1 + 𝑎2 𝑃2

𝑄𝑠1 = 𝑏0 + 𝑏1 𝑃1 + 𝑏2 𝑃2

𝑄𝑑2 = 𝑄𝑠2

𝑄𝑑2 = 𝑐0 + 𝑐1 𝑃1 + 𝑐2 𝑃2

𝑄𝑠2 = 𝑑0 + 𝑑1 𝑃1 + 𝑑2 𝑃2

Solving the 1st set of equations:

𝑎0 + 𝑎1 𝑃1 + 𝑎2 𝑃2 = 𝑏0 + 𝑏1 𝑃1 + 𝑏2 𝑃2

(𝑎0 − 𝑏0 ) + (𝑎1 − 𝑏1 )𝑃1 + (𝑎2 − 𝑏2 )𝑃2 = 0 − − − − − − − −(𝑖)

(𝑎𝑖 − 𝑏𝑖 ) = 𝛼𝑖

From the second set we get:

(𝑐0 − 𝑑0 ) + (𝑐1 − 𝑑1 )𝑃1 + (𝑐2 − 𝑑2 )𝑃2 = 0 − − − − − − − (𝑖𝑖)

Page 11 of 167
 (𝑐𝑖 − 𝑑𝑖 ) = 𝛾𝑖

If we replace each (𝑎𝑖 − 𝑏𝑖 ) in equation (i) with αi and each (𝑐𝑖 − 𝑑𝑖 ) in equation (ii) with

𝛾𝑖 we get 2 simultaneous equations:

𝛼1 𝑃1 + 𝛼2 𝑃2 = −𝛼0

𝛾1 𝑃1 + 𝛾2 𝑃2 = −𝛾0

The two equations are then solved simultaneously to get equilibrium price and quantities.

𝛼2 𝛾0 − 𝛼0 𝛾2
𝑃1∗ = − − − − − − − −𝑒𝑞𝑢𝑖𝑙𝑖𝑏𝑟𝑖𝑢𝑚 𝑝𝑟𝑖𝑐𝑒 𝑜𝑓 𝑔𝑜𝑜𝑑 1
𝛼1 𝛾2 − 𝛼2 𝛾1

𝛼0 𝛾1 − 𝛼1 𝛾0
𝑃2∗ = − − − − − − − −𝑒𝑞𝑢𝑖𝑙𝑖𝑏𝑟𝑖𝑢𝑚 𝑝𝑟𝑖𝑐𝑒 𝑜𝑓 𝑔𝑜𝑜𝑑 2
𝛼1 𝛾2 − 𝛼2 𝛾1

For the price to be meaningful, 𝛼1 𝛾2 − 𝛼2 𝛾1 ≠ 0. In addition, the numerator should have the same

sign as the denominator.

The equilibrium quantities will be obtained by replacing these values of equilibrium prices in the

demand or supply equations of the relevant commodities.

Equilibrium conditions for an n-commodity market

If the quantity demanded of good i is given as:

𝑄𝑑𝑖 = (𝑃1 , 𝑃2 , 𝑃3 , . . 𝑃𝑛 )𝑤ℎ𝑒𝑟𝑒 𝑖 = 1,2 … . . 𝑛

𝑄𝑠𝑖 = (𝑃1 , 𝑃2 , 𝑃3 , . . 𝑃𝑛 ) 𝑤ℎ𝑒𝑟𝑒 𝑖 = 1,2 … . . 𝑛

Equilibrium requires that for the n-markets to be in equilibrium simultaneously,

Page 12 of 167
 𝑄𝑑𝑖 = 𝑄𝑠𝑖

Equilibrium in National – Income Analysis (for a closed economy)

𝑌 = 𝐶 + 𝐼0 + 𝐺0 𝐶 = 𝑎 + 𝑏𝑌 𝑎 > 00< 𝑏 < 1

Y, C are endogenous variables while I0 and G0 are exogenous variables. In order to get the

equilibrium national income, we substitute the expression for C in the model

𝑌 = 𝑎 + 𝑏𝑌 + 𝐼0 + 𝐺0

𝑌 − 𝑏𝑌 = 𝑎 + 𝐼0 + 𝐺0

𝑌(1 − 𝑏) = 𝑎 + 𝐼0 + 𝐺0

a+ I0 + G0
Y* =
(1- b)

This is the expression for the equilibrium national income.

We can get the equilibrium level of consumption by replacing the equilibrium value of national

income in the consumption function.

a+ I0 + G0
𝐶 = 𝑎 + 𝑏[ ]
(1- b)

𝑎(1 − 𝑏) + 𝑏(a+ I0 + G0 )
=
1−𝑏

𝑎 − 𝑎𝑏 + 𝑏a+ bI0 +bG0 )

=
1−𝑏

𝑎 + 𝑏( I0 + G0 )
𝐶∗ =
1−𝑏

Page 13 of 167
This is the expression for the equilibrium consumption.

Example

The national income model for a closed economy is given as:

𝑌 = 𝐶 + 𝐼0 + 𝐺0 𝑤ℎ𝑒𝑟𝑒 𝐶 = 40 + 0.8𝑌 𝐼0 = 50 𝐺0 = 100

Determine the equilibrium values of the endogenous variables within the model.

The Endogenous variables are: Y and C

To solve for the equilibrium values of Y we substitute for C in to national income equation

𝑌 = 40 + 0.8𝑌 + 50 + 100

𝑌 = 190 + 0.8𝑌

𝑌 − 0.8𝑌 = 190

0.2𝑌 = 190

𝑌 ∗ = 950

To obtain the value of equilibrium consumption, we substitute for the value of Y* into the

consumption function

𝐶 ∗ = 40 + 0.8(950)

𝐶 ∗ = 40 + 0.8(950)

𝐶 ∗ = 800

Page 14 of 167
1.3 Self-Assessment Questions

Page 15 of 167
1.4 E-references

1. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical Economics,

4th edition, McGraw Hill, Singapore
2. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
3. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau

Page 16 of 167
 LECTURE TWO: DIFFERENTIATION OF NON-ALGEBRAIC FUNCTIONS

2.1 Introduction

This lecture will guide the learner on how to differentiate exponential and logarithmic functions.

In addition, economic applications will be explored.

2.2 Lesson Learning Outcomes

By the end of this lecture the learner should be able to:

2.2.1 Differentiate exponential functions.

2.2.2 Differentiate logarithmic functions.

2.2.3 Understand the economic applications of non-algebraic functions.

2.2.1/2.2.2 Differentiation of Non-Algebraic Functions (Exponential Functions and

Logarithmic Functions)

An algebraic function is any function that can be expressed in terms of polynomials or roots of

polynomials. For example:

4𝑋 2 +5𝑋−2
𝑌= is an algebraic function.
3𝑋+2

A non-algebraic function is any function where the independent variable appears in the power,

as a logarithm or a trigonometric function.

The following are examples of non-algebraic functions:

Y = sin 𝑋

Page 17 of 167
 𝑌 = 𝑙𝑜𝑔5 𝑋

2 −7𝑋
𝑌 = 𝑒 2𝑋

In this course we will concentrate on the exponential and logarithmic functions because they

have more applications in economics.

Rule of differentiating Non-Algebraic Functions

a) The Natural Exponential Function rule

𝐼𝑓 𝑦 = 𝑒 𝑥

𝑑𝑦
= 𝑦 = 𝑒𝑥
𝑑𝑥

If 𝑦 = 𝑒 𝑓(𝑥)

𝑑𝑦
= 𝑓′(𝑥). 𝑒 𝑓(𝑥) - Chain Rule
𝑑𝑥

Examples

i) 𝑦 = 𝑒 4𝑥

𝑑𝑦
= 4𝑒 4𝑥
𝑑𝑥

ii) 𝑦 = 𝑒 −𝑥

𝑑𝑦
= −𝑒 −𝑥
𝑑𝑥
3 −7𝑥 2
iii) 𝑦 = 5𝑒 2𝑥

𝑑𝑦 3 2
= 5 (𝑒 2𝑥 −7𝑥 ) ∗ (6𝑥 2 − 14𝑥)
𝑑𝑥

𝑑𝑦 3 2
= 5 (6𝑥 2 − 14𝑥)(𝑒 2𝑥 −7𝑥 )
𝑑𝑥

b) Exponential function Rule with Arbitrary Bases

Page 18 of 167
𝑦 = 𝑎𝑥

𝑑𝑦
The easiest way of finding the derivative of such a function with respect to x (𝑑𝑥 ) is presented in

the following steps:

1. Take the logarithm of the function to the base “e”

Ln 𝑦 = 𝑥 𝐿𝑛 𝑎

2. Find the derivative of both sides with respect to x

1 𝑑𝑦
= (1) Ln a
𝑦 𝑑𝑥

𝑑𝑦
3. Making 𝑑𝑥 the subject of the formula, we have:

𝑑𝑦
= y Ln a = 𝑎 𝑥 𝐿𝑛 𝑎
𝑑𝑥

Examples

Find the derivatives of the following with respect to x

i) 𝑦 = 3𝑥

1. 𝐿𝑛 𝑦 = 𝑥 𝐿𝑛 3

1 𝑑𝑦
2. = (1) Ln 3
𝑦 𝑑𝑥

𝑑𝑦
3. = 𝑦 𝐿𝑛 3 = 3𝑥 𝐿𝑛 3
𝑑𝑥

3 −𝑥 2
ii) 𝑦 = 53𝑥

4. 𝐿𝑛 𝑦 = (3𝑥 3 − 𝑥 2 )𝐿𝑛 5

1 𝑑𝑦
5. = (9𝑥 2 − 2𝑥) Ln 5
𝑦 𝑑𝑥

𝑑𝑦 3 −𝑥 2
6. = 𝑦 (9𝑥 2 − 2𝑥) 𝐿𝑛 5 = 53𝑥 (9𝑥 2 − 2𝑥)𝐿𝑛 5
𝑑𝑥

Page 19 of 167
 c) The Natural Logarithmic function rule

𝑦 = 𝐿𝑛𝑥

𝑑𝑦 1
=
𝑑𝑥 𝑥

In general, the rule becomes:

𝑌 = 𝐿𝑛 𝑓(𝑥)

𝑑𝑦 𝑓′ 𝑥
= 𝑓(𝑥)
𝑑𝑥

Examples

i) 𝑦 = 𝐿𝑛(𝑎𝑥)

𝑑𝑦 𝑎 1
= 𝑎𝑥 = 𝑥
𝑑𝑥

ii) 𝑦 = 𝐾 𝐿𝑛𝑥

𝑑𝑦 𝐾
=
𝑑𝑥 𝑥

iii) 𝑦 = 𝐿𝑛 𝑥 2

𝑑𝑦 𝑓′(𝑥) 2𝑥 2−1 2
= = =
𝑑𝑥 𝑓(𝑥) 𝑥2 𝑥

iv) 𝑦 = ln(3𝑥 3 − 2𝑥 2 )

𝑑𝑦 9𝑥 2 − 4𝑥
=
𝑑𝑥 3𝑥 3 − 2𝑥 2

𝑑𝑦 9𝑥 − 4
=
𝑑𝑥 3𝑥 2 − 2𝑥

v) 𝑦 = 𝑡 3 𝐿𝑛𝑡 2

Get the derivative of 𝑌 with respect to 𝑡

Use the product rule

Page 20 of 167
 𝑑𝑦 2 (𝐿𝑛𝑡 2 )
𝑡 3 . 2𝑡
= 3𝑡 + 2
𝑑𝑡 𝑡

= 3𝑡 2 𝐿𝑛𝑡 2 + 2𝑡 3+1−2

= 3𝑡 2 𝐿𝑛𝑡 2 + 2𝑡 2

= 𝑡 2 (3𝐿𝑛𝑡 2 + 2)

d) Other logarithmic functions rule

If a function is given as:

𝑙𝑛𝑋
a) 𝑌 = 𝑙𝑜𝑔𝑎 𝑋 , 𝑖𝑛𝑡𝑜 𝑛𝑎𝑡𝑢𝑟𝑎𝑙 𝑙𝑜𝑔𝑎𝑟𝑖𝑡ℎ𝑚𝑠, 𝑡ℎ𝑒 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛 𝑏𝑒𝑐𝑜𝑚𝑒𝑠: 𝑌= 𝑙𝑛𝑎

𝑑𝑦 1 1
= 𝑙𝑛𝑎 𝑋
𝑑𝑥

Example

i) 𝑌 = 𝑙𝑜𝑔10 𝑋
𝑑𝑌 1 1
= 𝑙𝑛10 . 𝑋
𝑑𝑋

ln 𝑓(𝑋) 𝑑𝑦 𝑓 ′ (𝑋)
b) If 𝑌 = 𝑙𝑜𝑔𝑎 𝑓(𝑋) 𝑌= = 𝑙𝑛𝑎 𝑓(𝑥
𝑙𝑛𝑎 𝑑𝑋

Example

i) 𝑌 = 𝑙𝑜𝑔5 (2𝑋 3 + 3𝑋)

ln(2𝑋 3 +3𝑋)
𝑌= 𝑓 ′ (𝑋) = 6𝑋 2 + 3
𝑙𝑛5

𝑑𝑦 6𝑋 2 +3 6𝑋 2 +3
= =
𝑑𝑋 𝑙𝑛5(2𝑋 3 +3𝑋) 1.61(2𝑋 3 +3𝑋)

e) Differentiation of complex products and quotients

Sometimes we may encounter products and quotients of functions whose derivatives may turn out

to be difficult to compute by means of the product and quotient rules. One efficient approach of

dealing with such functions is to make use of logarithms.

Page 21 of 167
We demonstrate how to do this using the following examples:

Examples

Find the derivatives of the following:

(1+𝑥 3 )
1. 𝑦 = (𝑥+5)(3𝑥−2)

Obtain the natural log of each term

𝐿𝑛 𝑦 = 𝐿𝑛 (1 + 𝑥 3 ) − 𝐿𝑛 (𝑥 + 5) − 𝐿𝑛 (3𝑥 − 2)

Obtaining the derivatives of both sides we get

1 𝑑𝑦 3𝑥 2 1 3
= (1+𝑥 3 ) − (𝑥+5) − 3𝑥−2
𝑦 𝑑𝑥

Multiply both sides by y to obtain

𝑑𝑦 3𝑥 2 1 3
= {(1+𝑥 3 ) − (𝑥+5) − 3𝑥−2}*y
𝑑𝑥

Substitute for y

𝑑𝑦 3𝑥 2 1 3 (1+𝑥 3 )
= {(1+𝑥 3 ) − (𝑥+5) − 3𝑥−2}*((𝑥+5)(3𝑥−2) )
𝑑𝑥

2. 𝑦 = 𝑎3𝑥+5 𝑒 5𝑥+9

𝐿𝑛 𝑦 = (3𝑥 + 5)𝐿𝑛 𝑎 + (5𝑥 + 9)𝐿𝑛 𝑒

𝐿𝑛 𝑒 = 1

Rewrite as follows:

Page 22 of 167
 𝐿𝑛 𝑦 = (3𝑥 + 5)𝐿𝑛 𝑎 + (5𝑥 + 9)

1 𝑑𝑦
= 3 𝐿𝑛 𝑎 + 5
𝑦 𝑑𝑥

𝑑𝑦
= {3 𝐿𝑛 𝑎 + 5} ∗ 𝑦
𝑑𝑥

𝑑𝑦
= {3 𝐿𝑛 𝑎 + 5} ∗ (𝑎3𝑥+5 𝑒 5𝑥+9 )
𝑑𝑥

2.2.3 Economic applications of non-algebraic functions.

a) Growth Rate of Economic Variables

i) Growth Rate of Single Functions

If Y is a function of time given as:

𝑌 = 𝑓(𝑡)

Method 1: The rate of growth of Y with time is given as

𝑑𝑌/𝑑𝑡 𝑀𝑎𝑟𝑔𝑖𝑛𝑎𝑙 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛 𝑓 ′ (𝑡)

𝑔= = =
𝑌 𝑡𝑜𝑡𝑎𝑙 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛 𝑓(𝑡)

As long as the function f(t) is continuous and differentiable.

Method 2: The growth rate of Y can also be obtained through making use of logarithms

This is computed by the following steps:

1. Take logarithms of the function of both sides

𝐿𝑛 𝑌 = 𝐿𝑛 𝑓(𝑡)

2. Differentiate the function with respect to t to obtain

𝑑𝑦 𝑓 ′ (𝑡)
(𝐿𝑛 𝑓(𝑡)) =
𝑑𝑡 𝑓(𝑡)

Page 23 of 167
Example One

The population size in a country denoted as N is defined by the following function:

𝑁 = 𝑁0 𝑒 𝑎𝑡

Determine the rate of growth in population size.

Using Method 1

𝑑𝑁⁄𝑑𝑡
𝑔=
𝑁

𝑑𝑁
= 𝑎𝑁0 𝑒 𝑎𝑡
𝑑𝑡

𝑎𝑁0 𝑒 𝑎𝑡
𝑔= =𝑎
𝑁0 𝑒 𝑎𝑡

Example Two

The growing value (V) of an asset is given by the exponential function. Find the rate of growth

of the value of the asset.

𝑉 = 𝐵𝑒 𝑟𝑡

Where: B = Constant

r = rate of growth

t = time

Solution

Using Method 2

Taking the natural logs of the function:

𝐿𝑛 𝑉 = 𝐿𝑛 𝐵 + 𝑟𝑡 𝐿𝑛 𝑒

Computing the derivative of the function with respect to t we obtain:

Page 24 of 167
 Growth rate = r

ii) Growth rate of product of functions

Consider two functions 𝑝(𝑡) and 𝑞(𝑡), where both functions are known to be growing with

time. The product of the two functions is given by:

𝑦 = 𝑝. 𝑞

Where: 𝑝 = 𝑝(𝑡) and 𝑞 = 𝑞(𝑡) and 𝑦 = 𝑦(𝑡)

All these functions are functions of time. The growth rate of y (𝐺𝑦 ) is obtained by the following

two step procedure.

Step 1:

Take logarithms of the equation to the base e (Ln)

𝐿𝑛𝑦 = 𝐿𝑛 𝑝 + 𝐿𝑛 𝑞

To obtain the growth rate of y, we obtain the derivative of Ln y with respect to time ‘t’

𝑑 𝑑 𝑑
𝐺𝑦 = {𝐿𝑛 𝑦} = {𝐿𝑛 𝑝} + {𝐿𝑛 𝑞} = 𝐺𝑝 + 𝐺𝑞
𝑑𝑡 𝑑𝑡 𝑑𝑡

The results tell us that the growth rate of the product of two functions is equal to the sum of

the growth rates of the individual functions.

Example

The price (P) of tea harvested from a plantation is growing at 3% per annum, while the quantity

(Q) supplied is growing at 1.8% per annum. Calculate the annual growth rate of the total

revenue (R) from the sale of the tea.

Solution

Page 25 of 167
 𝑅 = 𝑃. 𝑄

Where: 𝑃 = 𝑃(𝑡), 𝑄 = 𝑄(𝑡), 𝑅 = 𝑅(𝑡)

𝐿𝑛 𝑅 = 𝐿𝑛 𝑃 + 𝐿𝑛 𝑄

𝑑 𝑑 𝑑
𝐺𝑅 = {𝐿𝑛 𝑅} = {𝐿𝑛 𝑃} + {𝐿𝑛 𝑄} = 𝐺𝑃 + 𝐺𝑄
𝑑𝑡 𝑑𝑡 𝑑𝑡

𝐺𝑅 = 3% + 1.8% = 4.8%

iii) Growth rate of quotient of functions

We shall consider two functions of time: u and v whose quotient is given by:

𝑢
𝑦=
𝑣

Where 𝑢 = 𝑢(𝑡); 𝑣 = 𝑣(𝑡); 𝑦 = 𝑦(𝑡)

All the three functions are functions of time. The growth rate of y is derived by applying the

following procedure:

Step 1

Take Logarithms of y to obtain:

𝐿𝑛 𝑦 = 𝐿𝑛 𝑢 − 𝐿𝑛 𝑣

Step 2

Find the growth rate of y by finding the derivative of 𝐿𝑛 𝑦 with respect to time.

𝑑 𝑑 𝑑
𝐺𝑦 = {𝐿𝑛 𝑦} = {𝐿𝑛 𝑢} − {𝐿𝑛 𝑣} = 𝐺𝑢 − 𝐺𝑣
𝑑𝑡 𝑑𝑡 𝑑𝑡

Page 26 of 167
 The results tell us that the growth rate of the quotient of two functions is equal to the difference

of the growth rates of the individual functions i.e the growth rate of the numerator minus the

growth rate of the denominator.

Example

The national income (Y) of a developing country is growing at an annual rate of 1.3% while

the population (P) is growing at the rate 2.5% per annum. Find the growth rate of the per capita

income (PCY)

Solution

𝑌
𝑃𝐶𝑌 =
𝑃

Where: 𝑌 = 𝑌(𝑡); 𝑃 = 𝑃(𝑡); 𝑃𝐶𝑌 = 𝑃𝐶𝑌(𝑡)

The growth rate of PCY is obtained by the following procedure:

Step 1

Find the logarithm of both sides of the function

𝐿𝑛 𝑃𝐶𝑌 = 𝐿𝑛 𝑌 − 𝐿𝑛 𝑃

Step 2

Find the growth rate of PCY by finding the derivative of 𝐿𝑛 𝑃𝐶𝑌 with respect to time.

𝑑 𝑑 𝑑
𝐺𝑃𝐶𝑌 = {𝐿𝑛 𝑃𝐶𝑌} = {𝐿𝑛 𝑌} − {𝐿𝑛 𝑃} = 𝐺𝑌 − 𝐺𝑃
𝑑𝑡 𝑑𝑡 𝑑𝑡

= 1.3% − 2.5% = 1.2%

b) Elasticity of Economic Functions:

Consider the univariate function

𝑦 = 𝑓(𝑥)

Page 27 of 167
Using the concept of logarithms, the elasticity of y with respect to x is obtained using the following formula:

𝑑(𝐿𝑛 𝑦)
𝜀𝑦,𝑥 =
𝑑(𝐿𝑛 𝑥)

Example

Find the elasticity of demand for the following hyperbolic demand function

𝑄 = 𝑎𝑃−𝑏

Solution

Taking logarithms on both sides we obtain:

𝐿𝑛 𝑄 = 𝐿𝑛 𝑎 − 𝑏 𝐿𝑛 𝑃

To obtain the elasticity of Q with respect to P by finding the derivative of 𝐿𝑛 𝑄 with respect to

𝐿𝑛 𝑃

𝑑(𝐿𝑛 𝑄)
𝜀𝑄,𝑃 = = −𝑏
𝑑(𝐿𝑛 𝑃)

2.3 Self-Assessment Questions

a) The initial value of wine is W0 = 100, the rate of growth is 3%, and the growing value of
wine is given by the following equation.
Wt = W0er√t
Required: Calculate the value of wine at the end of 9th year.
b) Suppose that the population of Nairobi County is growing according to the function

N(t) = 2,400,000e0.03t, where N(t) is population size at time t; t is the number of years.

What time is required for the population to double?

2.4 E-references

Page 28 of 167
1. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical
Economics, 4th edition, McGraw Hill, Singapore
2. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
3. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau

Page 29 of 167
 LECTURE THREE: PARTIAL DIFFERENTIATION

3.1 Introduction

This lecture will guide the learner on how to differentiate multivariate functions and its

applications in economics especially in consumer and producer theories.

3.2 Lesson Learning Outcomes

By the end of this lecture the learner should be able to:

3.2.1 Get the total differential of a multivariate function.

3.2.2 Understand the rules of partial derivatives to multivariate functions.

3.2.3 Understand the concept of second order partial derivatives

3.2.4 Understand the economic applications of partial derivatives in economics

3.2.1 The concept of total differential of a multivariate function

A function of two or more independent variables is called a multivariate function. The total change

of the dependent variable will be as a result of the changes resulting from a change of each of the

independent variable.

For example, when we consider the change of utility of a household which consumes two types of

food, ugali and rice, we can represent it as:

U = f (X1, X2)

Where U = total utility derived

X1 = units of ugali consumed

Page 30 of 167
 X2 = units of rice consumed

When analyzing the total change in the utility of the household, due to change in number of units

of each food item consumed, it will be composed of two parts.

i) Change in utility due to change in units of ugali consumed

ii) Change in utility due to change in units of rice consumed

Symbolically: the total change in utility is given as:

𝜕𝑈 𝜕𝑈
𝑑𝑈 = 𝑑𝑋1 + 𝑑𝑋
𝜕𝑋1 𝜕𝑋2 2

The sum of the two parts is called the total differential of the utility function. i.e

𝑑𝑈 = Total differential of utility. It is the total change in utility as a result of change in number of

units of ugali and number of units in rice consumed.

Each of the parts is called a partial differential of the utility function with respect to gods 1 and 2

respectively.

𝜕𝑈
𝑑𝑋1 = Partial differential of utility function with respect to X1
𝜕𝑋1

. It measures the change in utility due to amount of good 1(Ugali) consumed

𝜕𝑈
𝑑𝑋2 = Partial differential of utility function with respect to X2. It measures the change in utility
𝜕𝑋2

due to the amount of good 2 (rice) consumed.

In general, if a function is given as:

𝑌 = 𝑓(𝑋1 , 𝑋2 , 𝑋3 , … … … … . . 𝑋𝑛 )

Page 31 of 167
The total differential is given as:

𝜕𝑌 𝜕𝑌 𝜕𝑌 𝜕𝑌
𝑑𝑌 = 𝑑𝑋1 + 𝑑𝑋2 + 𝑑𝑋3 … … … … … … . . + 𝑑𝑋
𝜕𝑋1 𝜕𝑋2 𝜕𝑋3 𝜕𝑋𝑛 𝑛

𝜕𝑌
Where 𝜕𝑋 𝑑𝑋𝑖 represents the change in dependent variable Y due to change in variable i
𝑖

If we are interested in knowing how Y changes when Xi changes, the other variables are assumed

to be constant.

Example1

𝑌 = 4𝑋1 + 10𝑋2 𝑋12 + 3𝑋23

Find the total differential of the function

𝜕𝑌 𝜕𝑌
𝑑𝑌 = 𝑑𝑋1 + 𝑑𝑋
𝜕𝑋1 𝜕𝑋2 2

𝜕𝑌
= Partial derivative of Y with respect to X1 when X2 is held constant.
𝜕𝑋1

𝜕𝑌
= Partial derivative of Y with respect to X2 when X1 is held constant.
𝜕𝑋2

𝑑𝑌 = (4 + 20𝑋2 𝑋1 ) 𝑑𝑋1 + (10𝑋12 + 9𝑋22 )𝑑𝑋2

Example 2

Get the total differential of the function below:

𝑍 = 5𝑋𝑌 + 2𝑋 2 𝑌 – 8𝑌 2 𝑋

Solution

Page 32 of 167
 𝑑𝑍 = (5𝑌 + 4𝑋𝑌 – 8𝑌 2 )𝑑𝑋 + (5𝑋 + 2𝑋 2 – 16𝑌𝑋)𝑑𝑌

Partial differentiation

This is applied to multivariable functions when we are analyzing the change in the dependent

variables as a result of change in one of the independent variables when the other independent

variables are held constant.

For example, a firm produces 2 goods 1 and 2. If we want to know the change in the total cost of

the firm when the production of good1 is increased when production of good 2 is held constant,

partial derivatives of the cost function with respect to change in good 1 will be useful.

3.2.2 Rules of partial differentiation

Whenever we differentiate multivariate function with respect to one of the independent variable,

the other variables are treated as constants.

Example

Y = 15X1 + 8X22

𝜕𝑌
= 15 (8x22 behaves like a constant hence its derivative with respect to X1 is 0)
𝜕𝑋1

𝜕𝑌
= 16𝑋2 (15x1 is a constant. Its derivative with respect to X2 is 0)
𝜕𝑋2

If:

Y = 2X1 X2 + 3X12 + 12X22

Page 33 of 167
 𝜕𝑌
= 2𝑋2 + 6𝑋1 (12X22 behaves as constant hence its derivative with respect to
𝜕𝑋1

X1 is 0, but in 2X1X2, X2 is in a multiplicative function, hence its retained as a constant)

𝜕𝑌
= 2𝑋1 + 24𝑋2
𝜕𝑋2

1. Sum and difference rule.

If Y = f(X1 ,X2) ± g (X1,X2)

Then:

𝜕𝑌
= 𝑓 ′ (𝑋1 ) + 𝑔′ (𝑋1 )
𝜕𝑋1

𝑆𝑖𝑚𝑖𝑙𝑎𝑟𝑙𝑦

𝜕𝑌
= 𝑓 ′ (𝑋2 ) + 𝑔′ (𝑋2 )
𝜕𝑋2

Example

𝜕𝑦 𝜕𝑦
Get 𝜕𝑥 𝑎𝑛𝑑 in the following function
1 𝜕𝑥2

i. Y = 8X1 + 5X12 X2

𝜕𝑌
= 8 + 10𝑋1 𝑋2
𝜕𝑋1

𝜕𝑌
= 5𝑋1 2
𝜕𝑋2

ii. Y = 15X22 X13 – 2X12 X2

Page 34 of 167
 𝜕𝑌
= 45X22 X12 – 4X1 X2
𝜕𝑋1

𝜕𝑌
= 30X2 X13 – 2X12
𝜕𝑋2

Practice question

Y = 5X12 + 3X1 X2 – 4X13 X22

𝜕𝑦 𝜕𝑦
Get 𝜕𝑥 𝑎𝑛𝑑
1 𝜕𝑥2

2. Product rule

If we have a function Y given as:

Y= f(X1 ,X2) . g (X1,X2)

The partial derivatives will be given as:

𝜕𝑌 𝑑𝑓 𝑑𝑔
= 𝑑𝑋 . 𝑔(𝑋1 , 𝑋2 ) + 𝑑𝑋 𝑓(𝑋1 , 𝑋2 )
𝜕𝑋1 1 1

𝑆𝑖𝑚𝑖𝑙𝑎𝑟𝑙𝑦

𝜕𝑌 𝑑𝑓 𝑑𝑔
= 𝑑𝑋 . 𝑔(𝑋1 , 𝑋2 ) + 𝑑𝑋 𝑓(𝑋1 , 𝑋2 )
𝜕𝑋2 2 2

Example:

𝑌 = (8𝑋 12 + 2𝑋2 ). (4𝑋23 + 3𝑋1 𝑋2 )

𝜕𝑌
= 16𝑋1 (4𝑋23 + 3𝑋1 𝑋2 ) + 3𝑋2 (8𝑋 12 + 2𝑋2 )
𝜕𝑋1

= 64X1X23 + 48X12 X2 + 24X2 X12 + 6X22

Page 35 of 167
 = 64X1 X23 + 72X12 X2 + 6X22

𝜕𝑌
= 2(4𝑋23 + 3𝑋1 𝑋2 ) + (12𝑋2 2 + 3𝑋1 )(8𝑋 12 + 2𝑋2 )
𝜕𝑋2

= 8X23 + 6X1 X2 + 96X22 X12 + 24X23 + 24X13 + 6X1 X2

= 32X23 + 12X1 X2 + 96X22 X12 +24X13

3. Quotient rule

𝑓(𝑋 ,𝑋 )
𝑌 = 𝑔(𝑋1 ,𝑋2 )
1 2

𝑑𝑓 𝑑𝑔
𝜕𝑌 . 𝑔(𝑋1 , 𝑋2 ) − 𝑓(𝑋1 , 𝑋2
𝑑𝑋1 𝑑𝑋1
=
𝜕𝑋1 (𝑔(𝑋1 , 𝑋2 )2 )

𝑑𝑓 𝑑𝑔
𝜕𝑌 . 𝑔(𝑋1 , 𝑋2 ) − 𝑓(𝑋1 , 𝑋2
𝑑𝑋2 𝑑𝑋2
=
𝜕𝑋2 (𝑔(𝑋1 , 𝑋2 )2 )

Example:

8𝑋 21 .𝑋2
𝑌 = 4𝑋 2
1 + 3𝑋2

𝑓(𝑋1 ) = 64𝑋1 2 𝑋2 𝑔(𝑋1 ) = 4

𝑓(𝑋2 ) = 8𝑋 12 𝑔(𝑋2 ) = 6𝑋2

𝜕𝑌 16𝑋1 𝑋2 . (4𝑋1 + 3𝑋2 2 ) − 4(8𝑋 12 . 𝑋2 )

= 2
𝜕𝑋1 (4𝑋 + 3𝑋 2 )
1 2

𝜕𝑌 64𝑋1 2 𝑋2 + 48𝑋1 𝑋2 3 − 32𝑋 12 . 𝑋2 )

= 2
𝜕𝑋1 (4𝑋 + 3𝑋 2 )
1 2

Page 36 of 167
 𝜕𝑌 32𝑋1 2 𝑋2 + 48𝑋1 𝑋2 3
= 2
𝜕𝑋1 (4𝑋 + 3𝑋 2 )
1 2

𝜕𝑌 8𝑋 12 . (4𝑋1 + 3𝑋2 2 ) − 6𝑋2 (8𝑋 12 . 𝑋2 )

= 2
𝜕𝑋2 (4𝑋 + 3𝑋 2 )
1 2

2
𝜕𝑌 32𝑋 13 . − 24𝑋 12 𝑋2
= 2
𝜕𝑋2 (4𝑋1 + 3𝑋2 2 )

4. Chain rule - derivative of a function of 2two independent variables raised to power

𝑌 = [𝑓(𝑋1 , 𝑋2 )]𝑛

𝜕𝑌
= 𝑛. [𝑓(𝑋1 , 𝑋2 )]𝑛−1 . 𝑓′𝑋1
𝜕𝑋1

𝜕𝑌
= 𝑛. [𝑓(𝑋1 , 𝑋2 )]𝑛−1 . 𝑓′𝑋2
𝜕𝑋2

Example

Y = (8X22 + 3X1 X2)4

𝜕𝑌
= 4. [8𝑋22 + 3𝑋1 𝑋2 ]3 .( 3𝑋2 )
𝜕𝑋1

𝜕𝑌
= 4. [8𝑋22 + 3𝑋1 𝑋2 ]3 .( 16𝑋2 + 3𝑋1 )
𝜕𝑋2

3.2.3 First order and second order partial derivative

i) First order partial derivative

If we have a multivariate function given as:

Page 37 of 167
𝑌 = 𝑓(𝑋1 , 𝑋2 )

Then:

𝜕𝑌 𝜕𝑌
𝑎𝑛𝑑 are called the first order partial derivatives with respect to X1 and X2 respectively.
𝜕𝑋1 𝑑𝑋2

When the first order partial derivatives are differentiated again with respect to X1 and X2 we get

the second order partial derivatives and the cross partial derivatives.

ii) Second order partial derivatives:

Given : 𝑌 = 𝑓(𝑋1 , 𝑋2 )

Then:

𝜕𝑌 𝜕𝑌 𝜕𝑌
𝑐𝑎𝑛 𝑏𝑒 𝑤𝑟𝑖𝑡𝑡𝑒𝑛 𝑎𝑠: = 𝑓𝑋1 while 𝜕𝑋 = 𝑓𝑋2
𝜕𝑋1 𝜕𝑋1 2

The second order partial derivative of Y with respect to X1 is given as:

𝜕 2𝑌
= 𝑓𝑋1 𝑋1
𝜕𝑋1 2

While the second order partial derivative of Y with respect to X2 is given as:

𝜕 2𝑌
= 𝑓𝑋2 𝑋2
𝜕𝑋2 2

𝜕𝑌 𝜕𝑌
In addition we get the partial derivatives of with respect to 𝑋2 𝑎𝑛𝑑 with respect to 𝑋1.
𝜕𝑋1 𝜕𝑋2

These are given as:

𝜕2 𝑌
= 𝑓𝑋1 𝑋2 This tells us how 𝑓𝑋1 changes when X2 changes.
𝜕𝑋1 𝜕𝑋2

Page 38 of 167
And

𝜕2 𝑌
= 𝑓𝑋2 𝑋1 tells us how 𝑓𝑋2 changes when X1 changes.
𝜕𝑋2 𝜕𝑋1

These two second order partial derivatives are known as the cross partial derivatives.

According to the young’s theorem, the cross partial derivatives should be equal. i.e.

𝑓𝑋1 𝑋2 = 𝑓𝑋2 𝑋1

Example:

Given a function

Y = 3X12 + 2X23 + 6X1 X2

Get the first and the second order partial derivatives with respect to X1 and X2.

𝜕𝑌 𝜕𝑌
= 6𝑋1 + 6𝑋2 while = 6𝑋22 + 6𝑋1
𝜕𝑋1 𝜕𝑋2

𝜕 2𝑌
= 𝑓𝑋1 𝑋1 = 6
𝜕𝑋1 2

𝜕 2𝑌
= 𝑓𝑋2 𝑋2 = 12𝑋2
𝜕𝑋2 2

𝜕 2𝑌
= 𝑓𝑋1 𝑋2 = 6
𝜕𝑋1 𝜕𝑋2

𝜕 2𝑌
𝑎𝑛𝑑 = 𝑓𝑋2 𝑋1 = 6
𝜕𝑋2 𝜕𝑋1

The young’s theorem is proved:

Page 39 of 167
𝑓𝑋1 𝑋2 = 𝑓𝑋2 𝑋1 = 6

3.3 Self-Assessment Questions

a) Get the partial derivatives of the following functions:

i) Y =f (X1 X2) = [18X12 + 3X2 X1 – X22]4

8𝑋12
ii) 𝑦 = (2𝑋1 +3𝑋2 𝑋1 )3 )

iii) Y = (4X1+ 8X22)2 (9X22 + 2X1 X2)

b) Get the first order, second order and cross partial derivatives of the following functions.

i) Y = 8X1 + 15X12 X2 + 3X23

ii) Y = 2X2 + 3XW + 7W3

3.4 E-references

(i) Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical Economics, 4th
edition, McGraw Hill, Singapore
(ii) Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
(iii)Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau

Page 40 of 167
 LECTURE FOUR: ECONOMIC APPLICATIONS OF PARTIAL DERIVATIVES

4.1 Introduction

In this lecture we shall learn how partial derivatives and the second order partial derivatives are

applied in Economics.

4.2 Lesson Learning Outcomes

By the end of this lecture the learner should be able to:

4.2.1 Get the marginal functions and their nature in multivariate economic functions

4.2.2 Compute the price and income elasticity of demand.

4.2.1 Marginal functions and their nature in multivariate economic functions

In Economics, most functions are multivariate. We therefore need to look at how these functions

change due to change in one of the independent variable when others are held constant.

a) Utility function and the marginal utility

Assume a consumer derives satisfaction from consuming 2 goods 1 and 2. The utility function

can be given as

U =f (Q1, Q2)

Where U = Total utility

Q1 = Amount of good 1 consumed

Q2 = Amount of good 2 consumed.

Page 41 of 167
We assume that this utility function is continuous, differentiable and 2nd order partial derivatives

can be obtained.

If each of the 2 goods has positive marginal utility, then

𝜕𝑈 𝜕𝑈
>0 >0
𝜕𝑄1 𝜕𝑄2

𝜕𝑈
= marginal utility of good 1
𝜕𝑄1

𝜕𝑈
= marginal utility of good 2
𝜕𝑄2

Total differential of the utility function

𝜕𝑈 𝜕𝑈
𝑑𝑈 = 𝑑𝑋1 + 𝑑𝑋
𝜕𝑋1 𝜕𝑋2 2

𝑑𝑈 = 𝑀𝑈𝑋1 𝑑𝑋1 + 𝑀𝑈𝑋2 𝑑𝑋2

If you recall what you learnt in micro economics I, an indifference comes is convex to be origin

and its downward sloping. Along an indifference curve, the utility level is constant and does not

change. Hence, 𝑑𝑈 = 0

0 = 𝑀𝑈𝑋1 𝑑𝑋1 + 𝑀𝑈𝑋2 𝑑𝑋2

The slope of an indifference curve is given as:

𝑑𝑋2 −𝑀𝑈𝑋2
=
𝑑𝑋1 𝑀𝑈𝑋2

Page 42 of 167
 Q2

Indifference map

Q1

If the slope of the marginal utility with respect to a good is positive, the utility function displays

decreasing marginal utility with respect to that good and if negative, it displays diminishing

marginal utility. Most normal goods display decreasing marginal utility hence;

𝜕𝑀𝑈(𝑄1 )
is expected to be less than 0.
𝜕𝑄1

𝜕𝑀𝑈(𝑄1 ) 𝜕 2 𝑈
=
𝜕𝑄1 𝜕𝑄1 2

Example one

A utility function is given as:

U=Q½ Q4/3

i)
Determine the marginal utility of good 1 and 2
ii)
Determine the nature of marginal utility for each of the goods.

Page 43 of 167
Solution

𝜕𝑈 4/3
i) = 0.5𝑄1−0.5 𝑄2 = marginal utility of good 1
𝜕𝑄1

𝜕𝑈 4 1/3
= 3 𝑄10.5 𝑄2 marginal utility of good 2
𝜕𝑄2

𝜕(𝑀𝑈𝑄 ) 4/3
4/3 −0.25𝑄2
ii) 1
= −0.25𝑄1−1.5 𝑄2 = Given that 𝑄1 𝑎𝑛𝑑 𝑄2 are non-negative, it
𝜕𝑄1 𝑄11.5

means the entire expression will be negative, hence good 1 displays

decreasing marginal utility.

𝜕(𝑀𝑈𝑄 ) 4 −1/3 4 𝑄10.5

iii) 2
= 9 𝑄10.5 𝑄2 =9 1/3 Given that 𝑄1 𝑎𝑛𝑑 𝑄2 are non-negative, it means the
𝜕𝑄2 𝑄2

entire expression will be positive, hence good 2 displays increasing marginal utility.

b) Cost function and Marginal cost function

We have been assuming that a firm produces 1 product only. But in some cases, a firm is likely to

produce several products. There is need to find out how the total cost of a firm changes when the

production one of the products change when the production of other products is held constant. The

total cost of the firm will be a function of production level of each of the products. If we have 𝑛

products then

𝑇𝐶 = 𝑓(𝑄1 , 𝑄2 , … … . 𝑄𝑛 )

Assuming we have 2 products only then

𝑇𝐶 = 𝑓(𝑄1 , 𝑄2 )

The marginal costs of good 1 is given as:

Page 44 of 167
and 2 are

𝑑𝑇𝐶 𝑑𝑇𝐶
𝑀𝐶1 = 𝑑𝑄 While marginal cost of good 2 is given as: 𝑀𝐶2 = 𝑑𝑄
1 2

Example

𝐶 = 8𝑄13 + 6𝑄22 + 4𝑄1 𝑄2 − 10

𝑑𝑇𝐶
𝑀𝐶1 = = 24𝑄1 2 + 4𝑄2
𝑑𝑄1

𝑑𝑇𝐶
𝑀𝐶2 = = 12𝑄2 + 4𝑄1
𝑑𝑄2

The firm is also interested in knowing if marginal costs are increasing or decreasing with output

change. This is computed by getting the derivatives of partial marginal cost functions with respect

to that good.

In our previous example:

𝑑𝑇𝐶
𝑀𝐶1 = = 24𝑄1 2 + 4𝑄2
𝑑𝑄1

To determine if MC1 increases or decreases with output we differentiate it with respect to Q1

𝑑𝑀𝐶1
= 48𝑄1 > 0 since Q1 is positive, the whole expression will be positive, hence good 1
𝑑𝑄1

displays increasing marginal cost.

Similarly, to determine if MC2 increases or decreases with output we differentiate it with respect

to Q2.

Page 45 of 167
 𝑑𝑀𝐶2
= 12 > 0, hence good 2displays increasing marginal cost.
𝑑𝑄2

c) Total Product and Marginal Product

In mathematics for economics 1, you learnt about the production function. The production function

describes the output level (total product) as a function of the level of inputs employed by the firm.

If we assume we have only two factors of production, capital and labour, we can denote the

production function as:

𝑄 = 𝑓(𝐿, 𝐾)

If we want to know how the level of output changes when one of the inputs is changed while the

other one is held constant, we use the partial derivatives. The marginal product of labour is the

partial derivative of the production function with respect to labour when capital is held constant.

i.e.

𝜕𝑄
𝑀𝑃𝐿 =
𝜕𝐿

On the other hand, the marginal product of capital is the partial derivative of the production

function with respect to capital when the labour is held constant

𝜕𝑄
𝑀𝑃𝐾 =
𝜕𝐾

Slope of an isoquant

It we get the total differential of a production function; it will be given as:

𝜕𝑄 𝜕𝑄
𝑑𝑄 = 𝑑𝐿 + 𝑑𝐾
𝜕𝐿 𝜕𝐾

Page 46 of 167
Along an isoquant the output level does not change, hence the total differential, dQ=0

If we equate the total differential to 0, we get:

𝜕𝑄 𝜕𝑄
𝑑𝑄 = 𝑑𝐿 + 𝑑𝐾 = 0
𝜕𝐿 𝜕𝐾

𝑀𝑃𝐿 𝑑𝐿 + 𝑀𝑃 𝐾 𝑑𝐾 = 0

𝑀𝑃𝐾 𝑑𝐾 = 𝑀𝑃𝐿 𝑑𝐿

Assuming that K is on the Y axis and L is on the X axis, the slope will be given as:

𝑑𝐾 𝑀𝑃𝐿
=−
𝑑𝐿 𝑀𝑃𝐾

This is the marginal rate of technical substitution (MRTS).

If the derivative of the marginal product of labour with respect to labour is positive, the marginal

product is said to be increasing as labour use increases. On the other hand, if it is negative, there

is diminishing marginal product. If it is 0, it has constant marginal product.

If the derivative of the marginal product of capital with respect to capital is positive, the marginal

product is said to be increasing as capital use increases. On the other hand, if it is negative, there

is diminishing marginal product. If it is 0, it has constant marginal product.

Example

A production function (total product) is given as:

𝑄 = 80𝐾 0.2 𝐿0.5

Page 47 of 167
Determine the marginal product of labour, marginal product of capital, the slope of the isoquant

and whether the marginal products are increasing or decreasing.

Solution

𝜕𝑄 40𝐾 0.2
𝑀𝑃𝐿 = = 40𝐾 0.2 𝐿−0.5 = 0.5
𝜕𝐿 𝐿

𝜕𝑄 −0.8 0.5
16𝐿0.5
𝑀𝑃𝐾 = = 16𝐾 𝐿 = 0.8
𝜕𝐾 𝐾

𝑑𝐾 𝑀𝑃𝐿 40𝐾 0.2 16𝐿0.5 2.5𝐾

𝑆𝑙𝑜𝑝𝑒 = =− = 0.5 ÷ 0.8 =
𝑑𝐿 𝑀𝑃𝐾 𝐿 𝐾 𝐿

Nature of the Marginal products:

𝜕𝑀𝑃𝐿 𝜕 2 𝑄 −20𝐾 0.2

= 2 = −20𝐾 0.2 𝐿−1.5 = <0
𝜕𝐿 𝜕𝐿 𝐿1.5

Since K and L are positive, the partial derivative of MPL with respect to L will be negative. Hence

marginal product of labour decreases as more labour is employed (Diminishing / Decreasing MPL).

𝜕𝑀𝑃𝑘 𝜕 2 𝑄 −1.8 0.5

−12.8𝐿0.5
= = −12.8𝐾 𝐿 = <0
𝜕𝐾 𝜕𝐾 2 𝐾 1.8

There is diminishing marginal product of capital as capital use increases.

Page 48 of 167
4.2.2 Partial Elasticity of Demand

In partial market equilibrium, the demand for a product is assumed to be a function of its own price

when other factors are held constant.

𝑄𝑑𝑥 = 𝑓(𝑃𝑥) ceteris paribus

𝑄𝑑𝑦 = 𝑓(𝑃𝑦)

This is a very simplified market model. In most cases, the demand of a product will be affected

by its own price, and also the prices of other related commodities such as its substitutes and

complements, the income among other variables.

Assuming we have a product A, its demand function can be given as:

𝑄𝐴 = 𝑓(𝑃𝐴 , 𝑃𝐵 , 𝑌)

Where:

𝑄𝐴 = 𝑞𝑢𝑎𝑛𝑡𝑖𝑡𝑦 𝑑𝑒𝑚𝑎𝑛𝑑𝑒𝑑 𝑜𝑓 𝑔𝑜𝑜𝑑 𝐴

𝑃𝐴 = 𝑃𝑟𝑖𝑐𝑒 𝑜𝑓 𝑔𝑜𝑜𝑑 𝐴

𝑃𝐵 = 𝑃𝑟𝑖𝑐𝑒 𝑜𝑓 𝑔𝑜𝑜𝑑 𝐵

𝑌 = 𝑐𝑜𝑛𝑠𝑢𝑚𝑒𝑟 𝑖𝑛𝑐𝑜𝑚𝑒

We will now look at how we compute the responsiveness of the quantity demanded of good A

with respect to price of good A, price of good B and the income.

Page 49 of 167
 i) Own Price elasticity of demand

It measures the percentage change in the quantity demanded of good A when its own price changes

keeping the price of good B (PB) and the income (Y) constant.

𝜕𝑄𝐴 . 𝑃𝐴
𝑃𝑟𝑖𝑐𝑒 𝑒𝑙𝑎𝑠𝑡𝑖𝑐𝑖𝑡𝑦 𝑜𝑓 𝑑𝑒𝑚𝑎𝑛𝑑 =
𝜕𝑃𝐴 . 𝑄𝐴

ii) Cross price elasticity of demand

It measures the percentage change in quantity demanded of good A when price of good B changes

when the price of good A (PA) and other factors such as the income (Y) are held constant.

𝜕𝑄 .𝑃
𝐶𝑟𝑜𝑠𝑠 𝑝𝑟𝑖𝑐𝑒 𝑒𝑙𝑎𝑠𝑡𝑖𝑐𝑖𝑡𝑦 𝑜𝑓 𝑑𝑒𝑚𝑎𝑛𝑑 = 𝜕𝑃 𝐴.𝑄𝐵
𝐵 𝐴

You can know whether two goods are substitutes or complements by looking at the sign of the

partial derivatives of their demand functions will respect to price of the other good e.g.

∂QA
> 0 the two goods are substitutes i.e. As price of good B increase, the demand of good A
∂PB

increase. If demand for a good goes down due to an increase in its own price and consumers

substitute it by buying more its substitute.

For complementary commodities

∂QA
<0
∂PB

As price of good B increase quantity of demanded of good A decrease. The decrease in demand

for good B due to an increase in its price causes the demand of good A to decrease since the two

goods are used together.

Page 50 of 167
 iii) Income elasticity of demand

It measures the percentage change in the quantity demanded of good A when income changes

when the price of good A and the price of good B are held constant.

𝐴 𝜕𝑄 .𝑌
𝐼𝑛𝑐𝑜𝑚𝑒 𝑒𝑙𝑎𝑠𝑡𝑖𝑐𝑖𝑡𝑦 𝑜𝑓 𝑑𝑒𝑚𝑎𝑛𝑑 = 𝜕𝑌.𝑄
𝐴

The income elasticity is positive for normal goods and negative for inferior and giffen goods.

Example

A demand function is given as:

𝑄𝐴 = 90 − 2𝑃𝐴 + 0.4𝑃𝐵 + 0.1𝑌

Find the own price, income and cross price elasticities of demand when:

𝑃𝐴 = 6 𝑃𝐵 = 10 𝑌 = 100

𝑄𝐴 = 90 − 2(6) + 0.4(10) + 0.1(100) = 90 − 12 + 4 + 10 = 92

Own price elasticity of demand

𝜕𝑄𝐴
= −2
𝜕𝑃𝐴

𝜕𝑄𝐴 . 𝑃𝐴 6
= = −2 ∗ = −0.13
𝜕𝑃𝐴 . 𝑄𝐴 92

A 1% change in the price of good A, holding PB and Y constant leads to a 0.13% decrease in the

quantity demanded of good A. It is inelastic to price change.

Page 51 of 167
Cross-Price elasticity of demand

𝜕𝑄𝐴
= 0.4
𝜕𝑃𝐵

𝜕𝑄𝐴 . 𝑃𝐵 10
= = 0.4 ∗ = 0.04
𝜕𝑃𝐵 . 𝑄𝐴 92

A 1% change in the price of good B, holding P A and Y constant leads to a 0.04% increase in the

quantity demanded of good A. It is inelastic to price change. Since the cross elasticity is positive

the two goods are substitutes.

Income elasticity of demand

𝜕𝑄𝐴
= 0.1
𝜕𝑌

𝜕𝑄𝐴 . 𝑌 100
= = 0.1 ∗ = 0.11
𝜕𝑌. 𝑄𝐴 92

A 1% change in the income holding PB and PA constant leads to a 0.11% increase in the quantity

demanded of good A. It is inelastic to income change. Since the income elasticity is positive, good

A is a normal good.

4.3 Self-Assessment Questions

a) Determine whether these 2 goods are substitutes or complements.

𝑄𝑑𝑥 = 20 − 3𝑃𝑥 + 4𝑃𝑦 + 0.2𝑌

𝑃𝑥 = 1 𝑃𝑦 = 2 𝑌 = 500

b) A total cost function is given by:

Page 52 of 167
 𝐶 = 𝑄23 + 𝑄1 𝑄2 − 6𝑄23 + 40

Find: MC1, MC2 and if function displays increasing or decreasing MC with respect to each of the

goods.

4.4 E-references

1. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical

Economics, 4th edition, McGraw Hill, Singapore
2. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
3. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau

Page 53 of 167
LECTURE FIVE: HOMOGENEITY OF A FUNCTION AND RETURNS TO SCALE

5.1 Introduction

In this lesson, we will discuss the concepts of homogeneity of a function, returns to scale and the

Euler’s Theorem. We will also perform some comparative statics.

5.2 Lesson Learning Outcomes

By the end of this lesson, the learner should be able to:

5.2.1 Understand how to compute the homogeneity of a function

5.2.2 Understand how to calculate the returns to scale of a function

5.2.3 Understand the Euler’s Theorem

5.2.4 Perform comparative statics

5.2.1 Homogeneity of a function

Given a function 𝑌 = 𝑓(𝑋1 , 𝑋2 )

We say it is homogenous of degree r if by multiplying all the independent variables by a constant

K, the dependent variable (Y) changes by a multiple 𝐾 𝑟

Example 1

𝑌 = 𝑓(𝑋1 , 𝑋2 ) = 4𝑋1 + 𝑋2

Page 54 of 167
New Y= 4𝑘𝑋1 + 𝑘𝑋2 = 𝑘(4𝑋1 + 𝑋2 ) = 𝑘1 𝑌 The function is homogenous of degree 1. Functions

which are homogenous of degree 1 are said to be linearly homogenous.

Example 2

Determine the homogeneity of the function given as:

𝑌 = 𝑋 2 + 5𝑋𝑊 + 𝑊 2

𝑌 = (𝑘𝑋)2 + 5(𝑘𝑋)(𝑘𝑊) + (𝑘𝑊)2

𝑁𝑒𝑤 𝑌 = 𝑘 2 (𝑋 2 + 5𝑋𝑊 + 𝑊 2 ) = 𝑘 2 (𝑌)

It’s homogenous of degree 2

Example 3

Determine the homogeneity of the function given as: 𝑄 = 80𝐾 0.2 𝐿0.5

𝑄 = 80(𝐶𝐾)0.2 (𝐶𝐿)0.5

𝑄 = 𝐶 0.2+0.5 80𝐾 0.2 𝐿0.5

𝑄 = 𝐶 0.7 80𝐾 0.2 𝐿0.5

5.2.2 Returns to scale

This refers to the proportion by which output changes when all inputs are multiplied by the same

multiple. For example, assume that a farmer has been using 2 employees and 1 tractor and the

resultant output produced is 1000bags of maize.

Page 55 of 167
If the farmer doubles the inputs/factors such that he employs 4 employees and 2 tractors, it can

result in three scenarios:

i) Using 4 employees and 2 tractor production is 2000 bags

This is a case of constant returns to scale. When inputs are doubled, output doubles. The

output changes by the same multiple by which inputs where multiplied by

ii) 4 employees 2 tractors produce 1500 bags

The output has increased by less than double i.e. By 1.5 times. This is a case of decreasing

returns to scale. The output increases by less than the multiple by which the inputs were

multiplied by.

iii) 4 Employees - 2 tractors - 3000 bags

The output has increased by more than double i.e. by three times. This is increasing returns to

scale. The output increases by more than the multiple by which the inputs were multiplied by.

A function is said to display:

i) Constant returns to scale if its degree of homogeneity is of degree 1.

ii) Decreasing returns to scale if its degree of homogeneity is less than 1.

iii) Increasing returns to scale if its degree of homogeneity is more than 1.

Example

Determine the nature of returns to scale for the following functions:

i) 𝑄 = 80𝐾 0.2 𝐿0.5

𝑄 = 80(𝐶𝐾)0.2 (𝐶𝐿)0.5

Page 56 of 167
 𝑄 = 𝐶 0.2+0.5 80𝐾 0.2 𝐿0.5

𝑄 = 𝐶 0.7 80𝐾 0.2 𝐿0.5

This function is homogenous of degree 0.7. Hence it displays decreasing returns to scale.

1 3
ii) 𝑄 = 𝐴𝐿2𝐾4

𝑁𝑒𝑤𝑄 = 80(𝐶𝐿)0.5 (𝐶𝐾)0.75

𝑛𝑒𝑤𝑄 = 𝐶 1.25 𝐴𝐾 0.75 𝐿0.5

The function is homogenous of degree 1.25. Hence displays increasing returns to scale.

1 1
iii) 𝑄 = 𝐴𝐿2𝐾2

𝑁𝑒𝑤𝑄 = 80(𝐶𝐿)0.5 (𝐶𝐾)0.5

𝑛𝑒𝑤𝑄 = 𝐶 1 𝐴𝐾 0.5 𝐿0.5

The function is homogenous of degree 1. Hence displays constant returns to scale.

5.2.3 Euler’s Theorem

According to this theorem, if a production function displays characteristics of constant returns to

scale, the sum of marginal product of labour multiplied by labour units (L) and marginal product

of capital multiplied by capital units (K) is equal to the total output.

i.e. if Q = A K L

Such that  +  = 1

Then: 𝑀𝑃𝐿 . 𝐿 + 𝑀𝑃𝐾 . 𝐾 = 𝑄 − − − − − − − − − − − − − −1

From the production function:

Page 57 of 167
 𝑄 = 𝐴𝐾 𝛼 𝐿𝛽

𝛽𝐴𝐾𝛼 𝐿 𝛽 𝛽𝑄
𝑀𝑃𝐿 = 𝛽𝐴𝐾 𝛼 𝐿𝛽−1 = = :
𝐿 𝐿

Similarly

𝛼𝐴𝐾𝛼 𝐿 𝛽 𝛼𝑄
𝑀𝑃𝐾 = 𝛼𝐴𝐾 𝛼−1 𝐿𝛽 = =
𝐾 𝐾

If we plug in the MPL and MPK

Into equation 1, we get:

𝛽𝑄 𝛼𝑄
.𝐿 + . 𝐾 = 𝛽𝑄 + 𝛼𝑄 = 𝑄(𝛽 + 𝛼)
𝐿 𝐾

Since 𝐵 +  = 1

𝑄(𝛽 + 𝛼) = 𝑄

This is a proof of the theorem.

If the production function is homogenous of degree 𝑛, then Euler’s theorem becomes

𝑀𝑃𝐿 . 𝐿 + 𝑀𝑃𝐾 . 𝐾 = 𝑛𝑄 − − − − − − − − − − − − − −2

Example

Proof Euler’s theorem for the production function.

Q = 3K3 – 5KL2 – L3

MPL = 10KL – 3L2

Page 58 of 167
MPK = 9K2 – 5L2

Degree of homogeneity

=3(Kd)3 – 5Kd (Ld)2 – (dL)3

= d3 (3K3 – 5KL2 – L3)

 Homogenous of degree 3

We want to proof the Euler’s theorem

MPL L + MPK K = nQ = 3Q

(-10KL – 3L2)L + (9K2 – 5L2)K = 3Q

-10KL2 – 3L3 + 9K3 – 5L2K = 3Q

Can be written as

3(3K3) -2(5KL2) -3(L3) – 5KL2 = 3Q

9K3 – 15KL2 – 3L3 = 3Q

3(3K3 - 5KL2 – L3) =3Q Q = (3K3 - 5KL2 – L3)

3Q = 3Q

5.2.4 Comparative statics

This section deals with how the equilibrium value of on endogenous variable changes when there

is a change in any of the exogenous variables or parameters.

Page 59 of 167
a) Market Model

Consider the simple one commodity market model given as:

𝑄 = 𝑎 − 𝑏𝑃 (𝑎, 𝑏) > 0

𝑄 = −𝑐 + 𝑑𝑃 (𝑐, 𝑑) > 0

𝑎 + 𝑐
𝑃∗ =
𝑏+𝑑

𝑎𝑑 – 𝑏𝑐
𝑄∗ =
𝑏+𝑑

P* and Q* are the endogenous variables and they have been purely expressed as a function of the

parameters a, b, c, and d.

We would want to analyze how P* changes due to change in any of the parameters. This is done

by partially differentiating P* with respect to any of the parameters, by making use of the quotient

rule: e.g

𝜕𝑃 ∗ 1 𝜕𝑃 ∗ 1
= >0 = >0
𝜕𝑎 𝑏+𝑑 𝜕𝑐 𝑏+𝑑

𝜕𝑃 ∗ −1 (𝑎 + 𝑐) 𝜕𝑃 ∗ −1 (𝑎 + 𝑐)
= <0 = <0
𝜕𝑏 (𝑏 + 𝑑)2 𝜕𝑑 (𝑏 + 𝑑)2

Since a, b, c and d are greater than 0, the expressions

𝜕𝑃 ∗ 𝜕𝑃 ∗ 1
𝑎𝑛𝑑 = >0
𝜕𝑎 𝜕𝑐 𝑏+𝑑

And

Page 60 of 167
𝜕𝑃 ∗ 𝜕𝑃 ∗ −1 (𝑎 + 𝑐)
𝑎𝑛𝑑 = <0
𝜕𝑏 𝜕𝑑 (𝑏 + 𝑑)2

How does the equilibrium Q* change as parameters change. This is done by getting the partial

derivatives of Q* with respect to a, b, c and d.

𝜕𝑄 ∗ 𝑑 𝜕𝑄 ∗ −𝑏
= >0 = <0
𝜕𝑎 𝑏+𝑑 𝜕𝑐 𝑏+𝑑

𝜕𝑄 ∗ −𝑑 (𝑎 + 𝑐) 𝜕𝑄 ∗ 𝑏(𝑎 + 𝑐)
= <0 = >0
𝜕𝑏 (𝑏 + 𝑑)2 𝜕𝑑 (𝑏 + 𝑑)2

b) National income model

Simple national income model (closed economy)

𝑌 = 𝐶 + 𝐼0 + 𝐺0

𝐶 = 𝑐0 + 𝑐1 (𝑌 𝑑 ) 𝑤ℎ𝑒𝑟𝑒 𝑐0 > 0 0 < 𝑐1 < 1

𝑇 = 𝑡0 + 𝑡1 𝑌 𝑡0 > 0 0 < 𝑡1 < 1

𝑌𝑑 = 𝑌 − 𝑇

𝑌 = 𝑐0 + 𝑐1 (𝑌 𝑑 ) + 𝐼0 + 𝐺0

𝑌 = 𝑐0 + 𝑐1 (𝑌 − 𝑇) + 𝐼0 + 𝐺0

𝑌 = 𝑐0 + 𝑐1 (𝑌 − 𝑡0 − 𝑡1 𝑌 ) + 𝐼0 + 𝐺0

𝑌 = 𝑐0 + 𝑐1 𝑌 − 𝑐1 𝑡0 − 𝑐1 𝑡1 𝑌 + 𝐼0 + 𝐺0

𝑌 − 𝑐1 𝑌 + 𝑐1 𝑡1 𝑌 = 𝑐0 − 𝑐1 𝑡0 + 𝐼0 + 𝐺0
Page 61 of 167
 𝑌(1 − 𝑐1 + 𝑐1 𝑡1 ) = 𝑐0 − 𝑐1 𝑡0 + 𝐼0 + 𝐺0

𝑐0 − 𝑐1 𝑡0 + 𝐼0 + 𝐺0
𝑌∗ =
(1 − 𝑐1 + 𝑐1 𝑡1 )

𝑌 ∗ is the equilibrium value of national income.

How does Y* change as exogenous variables change.

𝜕𝑌 ∗ 1
= 𝑇ℎ𝑖𝑠 𝑖𝑠 𝑡ℎ𝑒 𝑎𝑢𝑡𝑜𝑛𝑜𝑚𝑜𝑢𝑠 𝑖𝑛𝑣𝑒𝑠𝑡𝑚𝑒𝑛𝑡𝑠 𝑚𝑢𝑙𝑡𝑖𝑝𝑙𝑖𝑒𝑟.
𝜕𝐼0 (1 − 𝑐1 + 𝑐1 𝑡1 )

Given that

𝑐1 𝑎𝑛𝑑 𝑐1 𝑡1 are positive and less than 1, the autonomous investment multiplier is positive,

meaning an increase in the autonomous investments will lead to an increase in the equilibrium

value of the national income.

Using partial differentiation, we can also get the autonomous tax multiplier.

𝜕𝑌 ∗ −𝑐1
=
𝜕𝐼0 (1 − 𝑐1 + 𝑐1 𝑡1 )

The autonomous tax multiplier is negative. This means that an increase in taxes will lead to a

decrease in the national income.

Similarly, we can get the tax rate multiplier. It tells us how the national income changes due to a

unit change in the tax rate.

𝜕𝑌 ∗ −(𝑐1 )(𝑐0 − 𝑐1 𝑡0 + 𝐼0 + 𝐺0 )
=
𝜕𝑡1 (1 − 𝑐1 + 𝑐1 𝑡1 )2

Page 62 of 167
 −(𝑐1 )𝑌 ∗
=
(1 − 𝑐1 + 𝑐1 𝑡1 )

An increase in the tax rate reduces the value of the national income and vice versa.

National income model for an open economy

𝑌 = 𝐶 + 𝐼0 + 𝐺0 + 𝑋0 − 𝑀

𝐶 = 𝑐0 + 𝑐1 (𝑌 𝑑 ) 𝑤ℎ𝑒𝑟𝑒 𝑐0 > 0 0 < 𝑐1 < 1

𝑇 = 𝑡0 + 𝑡1 𝑌 𝑡0 > 0 0 < 𝑡1 < 1

𝑌𝑑 = 𝑌 − 𝑇

𝑀 = 𝑚0 + 𝑚1 𝑌

𝑌 = 𝑐0 + 𝑐1 (𝑌 𝑑 ) + 𝐼0 + 𝐺0 + 𝑋0 − (𝑚0 + 𝑚1 𝑌)

𝑌 = 𝑐0 + 𝑐1 (𝑌 − 𝑇) + 𝐼0 + 𝐺0 −𝑋0 − 𝑚0 − 𝑚1 𝑌

𝑌 = 𝑐0 + 𝑐1 (𝑌 − 𝑡0 − 𝑡1 𝑌 ) + 𝐼0 + 𝐺0 −𝑋0 − 𝑚0 − 𝑚1 𝑌

𝑌 = 𝑐0 + 𝑐1 𝑌 − 𝑐1 𝑡0 − 𝑐1 𝑡1 𝑌 + 𝐼0 + 𝐺0 −𝑋0 − 𝑚0 − 𝑚1 𝑌

𝑌 − 𝑐1 𝑌 + 𝑐1 𝑡1 𝑌 + 𝑚1 𝑌 = 𝑐0 − 𝑐1 𝑡0 + 𝐼0 + 𝐺0 −𝑋0 − 𝑚0

𝑌(1 − 𝑐1 + 𝑐1 𝑡1 + 𝑚1 ) = 𝑐0 − 𝑐1 𝑡0 + 𝐼0 + 𝐺0 − 𝑚0

𝑐0 − 𝑐1 𝑡0 + 𝐼0 + 𝐺0 − 𝑚0
𝑌∗ =
(1 − 𝑐1 + 𝑐1 𝑡1 + 𝑚1 )

We can get a number of multipliers.

Page 63 of 167
 𝜕𝑌 ∗ 1
=
𝜕𝐺0 (1 − 𝑐1 + 𝑐1 𝑡1 + 𝑚1 )

This is the government multiplier. An increase in government expenditure causes the national

income to increase.

𝜕𝑌 ∗ 1
=
𝜕𝐼0 (1 − 𝑐1 + 𝑐1 𝑡1 + 𝑚1 )

𝜕𝑌 ∗ −1
=
𝜕𝑀0 (1 − 𝑐1 + 𝑐1 𝑡1 + 𝑚1 )

5.3 Self-Assessment Questions

a) From the national income model for an open economy,

𝜕𝑌 ∗ 𝜕𝑌 ∗ 𝜕𝑌 ∗
Compute: 𝜕𝑐 , 𝜕𝑚 𝑎𝑛𝑑
1 1 𝜕𝑡1

b) Find the degree of homogeneity and nature of returns to scale for the following function. Prove

Euler’s Theorem using the same function

Q = 10 K¼ L¾

5.4 E-references

4. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical Economics,

4th edition, McGraw Hill, Singapore
5. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
6. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau

Page 64 of 167
 LECTURE SIX: CONSTRAINED OPTIMIZATION

6.1 Introduction

In this lesson we shall learn how we can maximize or minimize functions when we are faced with

a constraint.

6.2 Lesson Learning Outcomes

By the end of this lecture the learner should be able to:

6.2.1 Formulate optimization problems and the Lagrangian function.

6.2.2 Understand the first order conditions (f.o.c) for optimization.

6.2.3 Determine the optimal values of optimization problems.

6.2.4 Understand the second order condition (s.o.c) for maximum and minimum.

6.2.1 Formulation of optimization problems and the Lagrangian function

The structure of constrained optimization problem

In a constrained optimization problem, the decision maker would want to optimize but is faced

with a constraint. The problem is composed of two parts: -

a) Objective function – what is to be maximized or minimized.

b) Constraints or constraining equation – defines the limiting conditions

Page 65 of 167
Example

The Lagrangian function

Since the constrained optimization problem consists of two functions (the objective function and

the constraint), before the conducting the optimization problem, a Lagrangian function also known

as the augmented objective function is constructed from the two functions.

The construction of the Lagrangian function is illustrated using the following example:

Example

Optimize 𝑍 = 𝑓(𝑥, 𝑦)

Subject to 𝑎𝑥 + 𝑏𝑦 = 𝐵

Page 66 of 167
The Lagrangian function is constructed as follows:

Step 1:

Set the constraint equal to zero

𝐵 − 𝑎𝑥 − 𝑏𝑦 = 0

Step 2:

Multiply the resulting constraint by 𝜆, the Lagrangian multiplier.

𝜆(𝐵 − 𝑎𝑥 − 𝑏𝑦) = 0

Step 3:

Form the Lagrangian function L by adding the equation obtained in step 2 to the objective function

𝑓(𝑥, 𝑦). The Lagrangian function is given by:

𝐿 = 𝑓(𝑥, 𝑦) + 𝜆(𝐵 − 𝑎𝑥 − 𝑏𝑦)

The Lagrangian function has three choice variables, namely, 𝑥, 𝑦 and 𝜆. Hence we can rewrite the

Lagrangian function as follows:

𝐿 = 𝑓(𝑥, 𝑦, 𝜆 ) + 𝜆(𝐵 − 𝑎𝑥 − 𝑏𝑦)

After formulating the Lagrangian function, the next step is to obtain the first order conditions.

6.2.2 The first order conditions (f.o.c) for optimization

Consider the following constrained optimization problem in which the objective function and the

constraint are all presented in the general form:

Page 67 of 167
Optimize 𝑍 = 𝑓(𝑥, 𝑦)

Subject to 𝑔(𝑥, 𝑦) = 𝐵

In order to determine the critical values of 𝑥, 𝑦 and λ we apply the first order condition.

The first order condition involves the computation of the first partial derivatives of the Lagrangian

function with respect to 𝑥, 𝑦 and λ, setting them equal to zero and solving the resulting system of

simultaneous equations for 𝑥, 𝑦 and λ.

The process is presented here in three steps:

Step 1:

Construct the corresponding Lagrangian function from the two functions.

𝐿(𝑥, 𝑦, 𝜆 ) = 𝑓(𝑥, 𝑦) + 𝜆{𝐵 − 𝑔(𝑥, 𝑦)}

Step 2:

Find the first partial derivatives of 𝐿(𝑥, 𝑦, 𝜆 ) with respect to the three choice variables 𝑥, 𝑦 and λ

and equating them to zero.

𝐿𝑥 = 𝑓𝑥 − 𝜆𝑔𝑥 = 0

𝐿𝑦 = 𝑓𝑦 − 𝜆𝑔𝑦 = 0

𝐿𝜆 = 𝐵 − 𝑔(𝑥, 𝑦) = 0

Step 3:

Solve the resulting system of three simultaneous equations from step 2 for the critical values

𝑥̅ , 𝑦̅ 𝑎𝑛𝑑 𝜆̅.

Page 68 of 167
6.2.3 Optimal values of optimization problems

Let’s illustrate the process of obtaining the critical values of optimization problems using the

following numerical examples.

Example 1

Consider the following constrained maximization problem:

Maximize 𝑍 = 12𝑥 + 4𝑦 − 6𝑥 2

Subject to −𝑥 2 + 𝑦 = 4

i. Construct the corresponding Lagrangian function

ii. Determine the critical values of 𝑥, 𝑦 and λ

iii. What is the stationary value of Z?

Solution

i. The Lagrangian function:

𝐿(𝑥, 𝑦, 𝜆 ) = 12𝑥 + 4𝑦 − 6𝑥 2 + λ (4 + 𝑥 2 − 𝑦)

ii. The critical values of 𝑥, 𝑦 and λ are obtained by applying the first order condition

𝐿𝑥 = 12 − 12𝑥 + 2λ𝑥 = 0 ………………………………….1

𝐿𝑦 = 4 − λ = 0 …………………………………………………2

𝐿𝜆 = 4 + 𝑥 2 − 𝑦 = 0 ……………………………………………3

Solve the system of three simultaneous equations for 𝑥, 𝑦 and λ

From equation 2:

Page 69 of 167
λ=4

Substituting for λ into equation 1

12 − 12𝑥 + 2λ𝑥 = 0

12 − 12𝑥 + 2(4)𝑥 = 0

12 − 12𝑥 + 8𝑥 = 0

12 − 4𝑥 = 0

𝑥̅ = 3

From equation 3:

4 + 𝑥2 − 𝑦 = 0

Substituting the value of x we have;

4 + (3)2 − 𝑦 = 0

𝑦̅ = 13

From equation (2):

𝜆̅ = 4

The critical values are: 𝑥̅ = 3; 𝑦̅ = 13; λ̅ = 4

iii. The stationary value of z is obtaining by substituting the critical values of x and y into the

objective function:

𝑍̅ = 12(3) + 4(13) − 6(32 ) = 34

Page 70 of 167
Example 2

Given the following constrained minimization problem:

Minimize 𝑍 = 4𝑥 2 + 2𝑥𝑦 + 𝑦 2

Subject to 𝑥 + 𝑦 = 8

Required:

i. Construct the corresponding Lagrangian function

ii. Determine the critical values of 𝑥, 𝑦 and λ

iii. What is the stationary value of Z?

Solution

i. The lagrangian function

𝐿(𝑥, 𝑦, 𝜆 ) = 4𝑥 2 + 2𝑥𝑦 + 𝑦 2 + 𝜆(8 − 𝑥 − 𝑦)

ii. The critical values of 𝑥, 𝑦 and λ

𝐿𝑥 = 8𝑥 + 2𝑦 − λ = 0 ………………………………………….1

𝐿𝑦 = 2𝑥 + 2𝑦 − λ = 0 …………………………………………………2

𝐿𝜆 = 8 − 𝑥 − 𝑦 = 0 ……………………………………………………3

Solve the system of three simultaneous equations for 𝑥, 𝑦 and λ

From 1;

8𝑥 + 2𝑦 = λ …………………………………….(a)

Page 71 of 167
2𝑥 + 2𝑦 = λ………………………………………(b)

Since (a) and (b) give the same value λ, we equate them to obtain:

8𝑥 + 2𝑦 = 2𝑥 + 2𝑦

6𝑥 = 2𝑦 − 2𝑦

6𝑥 = 0

𝑥̅ = 0

From (3)

8 − 𝑥 − 𝑦 = 0 since 𝑥̅ = 0

8−0−𝑦 = 0

𝑦̅ = 8

From (a)

8𝑥 + 2𝑦 = λ

λ̅ = 8(0) + 2(8) = 16

The critical values are: 𝑥̅ = 0; 𝑦̅ = 8; λ̅ = 16

iii. The stationary value of Z is obtained by substituting the critical values of x and y into the

objective function

𝑍̅ = 4(0)2 + 2(0)(8) + 82 = 64

6.2.4 The second order condition (s.o.c) for maximum and minimum

Page 72 of 167
The second order condition (s.o.c) for constrained optimization makes use of the Bordered Hessian

matrix determinant.

As the name suggests, a Bordered Hessian is a Hessian that has borders. The border elements or

the elements on the borders of the Bordered Hessian are obtained from the constraint while the

rest of the elements are obtained from the Lagrangian function.

We illustrate the process using the following example.

Consider the following constrained optimization problem in which both the objective function and

the constraint are presented in the general form.

Optimize 𝑍 = 𝑓(𝑥, 𝑦)

Subject to 𝑔(𝑥, 𝑦) = 𝑘

The construction of the bordered Hessian is presented in the following steps:

Step 1: Construct the Lagrangian function

𝐿 = 𝑓(𝑥, 𝑦) + 𝜆{𝑘 − 𝑔(𝑥, 𝑦)}

Step 2: Determine the elements of the borders of the bordered Hessian

The elements of the borders of the bordered Hessian are made up of the first partial derivatives of

the constraint with respect to the arguments 𝑥 and 𝑦.

𝑔(𝑥, 𝑦) = 𝑘

The required partial derivatives are: 𝑔𝑥 𝑎𝑛𝑑 𝑔𝑦

Step 3: Determine the remaining elements of Bordered Hessian

Page 73 of 167
The rest of the elements of the Bordered Hessian are made up of the second partial derivatives of

the Lagrangian function with respect to 𝑥 and 𝑦.

𝐿 = 𝑓(𝑥, 𝑦) + 𝜆{𝑘 − 𝑔(𝑥, 𝑦)}

The required partial derivatives are:

𝐿𝑥𝑥 : The second partial derivative of the Lagrangian function with respect to 𝑥.

𝐿𝑦𝑦 : The second partial derivative of the Lagrangian function with respect to 𝑦.

𝐿𝑥𝑦 = 𝐿𝑦𝑥 : The cross partial derivative of the Lagrangian function with respect to 𝑥 and 𝑦.

Step 4: Form the required bordered Hessian

0 𝑔𝑥 𝑔𝑦
̅
𝐻 = (𝑔𝑥 𝐿𝑥𝑥 𝐿𝑥𝑦 )
𝑔𝑦 𝐿𝑦𝑥 𝐿𝑦𝑦

Step 5: Obtain the determinant of the bordered Hessian

0 𝑔𝑥 𝑔𝑦
̅
|𝐻 | = |𝑔𝑥 𝐿𝑥𝑥 𝐿𝑥𝑦 |
𝑔𝑦 𝐿𝑦𝑥 𝐿𝑦𝑦

̅ | > 0: The critical values yield a Maximum

If |𝐻

̅ | < 0: The critical values yield a Minimum

If |𝐻

Example
Page 74 of 167
Test if the critical values of the following optimization problem yield a maximum or minimum

Optimize 𝑍 = 𝑥𝑦

Subject to 2𝑥 + 𝑦 = 9

𝐿 = 𝑥𝑦 + 𝜆(9 − 2𝑥 − 𝑦)

𝐿𝑥 = 𝑦 − 2𝜆 = 0 ………………………………………………………….1

𝐿𝑦 = 𝑥 − 𝜆 = 0…………………………………………………………….2

𝐿𝜆 = 9 − 2𝑥 − 𝑦 = 0……………………………………………………...3

From 1:

0.5𝑦 = 𝜆………………………………………………………………….(a)

From 2:

𝑥 = 𝜆…………………………………………………………………….(b)

From (a) and (b);

0.5𝑦 = 𝑥

Substituting for 𝑥 into equation 3;

9 − 2(0.5𝑦) − 𝑦 = 0

9 = 2𝑦

𝑦̅ = 4.5

0.5𝑦 = 𝑥

Page 75 of 167
𝑥̅ = 0.5 ∗ 4.5 = 2.25

To test if the critical values yield a maximum or minimum, we obtain the determinant of the

bordered Hessian.

From the constraint:

𝑔(𝑥, 𝑦) = 9 − 2𝑥 − 𝑦

The first partial derivatives of the constraint will form the border elements:

𝑔𝑥 = −2

𝑔𝑦 = −1

From the Lagrangian function: We obtain the remaining elements of the Hessian

𝐿 = 𝑥𝑦 + 𝜆(9 − 2𝑥 − 𝑦)

𝐿𝑥 = 𝑦 − 2𝜆

𝐿𝑥𝑥 = 0

𝐿𝑦 = 𝑥 − 𝜆

𝐿𝑦𝑦 = 0

𝐿𝑥𝑦 = 𝐿𝑦𝑥 = 1

The bordered Hessian will therefore be expressed as follows:

0 −2 −1
̅ = (−2 0
𝐻 1)
−1 1 0
Page 76 of 167
We obtain the determinant of the bordered Hessian as follows:

0 −2 −1
̅ | = |−2
|𝐻 0 1|
−1 1 0

̅ | = 0 |0
|𝐻 1
| − (−2) |
−2 1
| + (−1) |
−2 0
|=4
1 0 −1 0 −1 1

̅ | = 4 > 0: The critical values yield a maximum.

|𝐻

6.3 Self-Assessment Questions

Given the following constrained optimization problem:

Optimize 𝑍 = 𝑥 2 + 𝑦 2

Subject to 5𝑥 + 2𝑦 = 25

Required:

i. The critical values of 𝑥, 𝑦 and 𝜆

ii. Test whether the critical values yield a maximum or minimum point

6.4 E-References

4. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical

Economics, 4th edition, McGraw Hill, Singapore
5. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
6. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau

Page 77 of 167
 LECTURE SEVEN: ECONOMIC APPLICATIONS OF CONSTRAINED
OPTIMIZATION

7.1: Introduction
Constrained optimization is very key in economic analysis. Most economic optimization problems

are subject to constraints.

In this Lesson we will discuss the economic applications of constrained optimization.

7.2 Lesson Learning Outcomes

By the end of this Lesson, the learner should be able to understand the following constrained

optimization problems in economics:

7.2.1 Utility Maximization

7.2.2 Cost Minimization

7.2.3 Output Maximization

7.2.1 Utility Maximization

Maximize

𝑈 = 𝑢(𝑞1 , 𝑞2 )

Subject to

𝑀 = 𝑝1 𝑞1 + 𝑝2 𝑞2

𝑞1 𝑞2 > 0

Where:

M is the level of income, 𝑞1 and 𝑞2 are the amounts of goods 1 and 2 respectively, and P1 and

P2 are the prices of good 1 and 2 respectively.

Page 78 of 167
 𝐿 = 𝑈(𝑞1 𝑞2 ) + 𝜆(𝑀 − 𝑝1 𝑞1 − 𝑝2 𝑞2 )

𝐿𝑞1 = 𝑈𝑞1 − 𝜆. 𝑝1 = 0

𝐿𝑞2 = 𝑈𝑞2 − 𝜆. 𝑝2 = 0 0

𝐿𝜆 = 𝑀 − 𝑝1 𝑞1 − 𝑝2 𝑞2 = 0

From the first equation,

𝑈𝑞1
𝑈𝑞1 = 𝜆. 𝑝1 → 𝜆 = 𝑝1

and from the second equation

𝑈𝑞2
𝑈𝑞2 = 𝜆𝑝2 𝜆=
𝑝2

𝑈𝑞1 𝑈𝑞2
𝜆= =
𝑃1 𝑃2

𝑆𝑖𝑛𝑐𝑒 𝑈𝑞1 = 𝑀𝑈𝑞1 𝑎𝑛𝑑 𝑈𝑞2 = 𝑀𝑈𝑞2

𝑀𝑈𝑞1 𝑀𝑈𝑞2
→ = This is the condition for utility maximization.
𝑃1 𝑃2

Ratio of 𝑀𝑈𝑞1 to price of 𝑃1 = Ratio of 𝑀𝑈𝑞2 to price of 𝑃2

If this is rearranged, it can be written as:

𝑀𝑈𝑞1 𝑃1
→ =
𝑀𝑈𝑞2 𝑃2

This condition states that, for utility maximization, the slope of the indifference curve should be

equal to the slope of the budget line.

Alternatively

If we get the total differential of the utility function

𝑈 = 𝑢(𝑞1 𝑞2 )

𝑑𝑢 = 𝑈𝑞1 . 𝜕𝑞1 + 𝑈𝑞2 . 𝜕𝑞2

Page 79 of 167
 Along an indifference curve the utility does not change, hence du=0

𝑈𝑞1 𝜕𝑞1 = −𝑈𝑞2 𝜕𝑞2

𝜕𝑞2 𝑈𝑞
= − 𝑈𝑞1 This is the slope of the indifference curve
𝜕𝑞1 2

Slope of Budget Line

𝑀 − 𝑝1 𝑞1 + 𝑝2 𝑞2

𝑀 𝑝1
𝑞2 = − 𝑞
𝑝2 𝑝2 1

𝜕𝑞2 −𝑝1
= This is the slope of the budget line
𝜕𝑞1 𝑝2

𝑈𝑞1 𝑝1
− =−
𝑈𝑞2 𝑝2
𝑈𝑞 𝑝1
→ 𝑈𝑞1 = ⁄𝑝2……………. This is the condition for utility maximization
2

We use the determinant of the bordered Hessian to check whether the critical values yield a

maximum of minimum i.e.

̅ | > 0: The critical values yield a Maximum

If |𝐻

This means that the critical values maximize utility.

Example

Given a utility function U=5xy, and a budget constraint given as 5x+y=30, determine the levels of

x and y that will maximize the utility of the consumer.

Page 80 of 167
Solution

𝐿 = 5𝑥𝑦 + 𝜆(30 − 5𝑥 − 𝑦)

𝐿𝑥 = 5𝑦 − 5𝜆 = 0 → 𝜆 = 𝑦

𝐿𝑦 = 5𝑥 − 𝜆 = 0 → 𝜆 = 5𝑥

𝐿𝜆 = 30 − 5𝑥 − 𝑦 = 0

From the first equation and second equations,

𝜆 = 𝑦 = 5𝑥

If we replace 𝑦 in the third equation we get:

𝐿𝜆 = 30 − 5𝑥 − 𝑦 = 0

5𝑥 + 𝑦 = 30

5𝑥 + 5𝑥 = 30

10𝑥 = 30

𝑥=3

𝑦 = 5𝑥 = 15

𝜆 = 𝑦 = 15

Maximum utility will be given as:

𝑈 = 5𝑥𝑦 = 5 ∗ 3 ∗ 15 = 225

Do the critical values yield a maximum or a minimum?

We construct the bordered Hessian as follows:

From the constraint:

𝑔(𝑥, 𝑦) = 30 − 5𝑥 − 𝑦

Page 81 of 167
𝑔𝑥 = −5

𝑔𝑦 = −1

From the lagrangian function:

𝐿 = 5𝑥𝑦 + 𝜆(30 − 5𝑥 − 𝑦)

𝐿𝑥 = 5𝑦 − 5𝜆 = 0

𝐿𝑥𝑥 = 0

𝐿𝑦 = 5𝑥 − 𝜆 = 0

𝐿𝑦𝑦 = 0

𝐿𝑥𝑦 = 𝐿𝑦𝑥 = 5

0 −5 −1
̅ = (−5 0
𝐻 5)
−1 5 0

Obtain the determinant of the bordered Hessian

0 −5 −1
̅ | = |−5
|𝐻 0 5|
−1 5 0

̅ | = 0 |0
|𝐻 5
| − (−5) |
−5 5
| + (−1) |
−5 0
| = 50
5 0 −1 0 −1 5

̅ | = 50 > 0: The critical values maximize utility.

|𝐻

Page 82 of 167
7.2.2 Cost Minimization

If the firm wants to minimize the cost subject to a given output level, the problem can be written

as:

Minimize 𝐶 = 𝑟𝑘 + 𝑤𝑙

𝑆. 𝑡

𝑄 = 𝑓(𝑘 , 𝑙)

Where r and w are input prices of capital and labour respectively while K and L are the units of

capital and labour used respectively.

If we form the lagrangian function:

𝐿 = 𝑟𝑘 + 𝑤𝑙 + 𝜆 (𝑄 − 𝑓(𝑘 , 𝑙))

𝐿𝑘 = 𝑟 − 𝜆𝑓𝐾′ (𝑘 , 𝑙) = 0

𝐿𝑙 = 𝑤 − 𝜆𝑓𝐿′ (𝑘 , 𝑙) = 0

𝐿𝜆 = 𝑄 − 𝑓 (𝑘 , 𝑙) = 0

𝑟 = 𝜆𝑓𝑘 ′ (𝑘 , 𝑙) 𝑤 = 𝜆𝑓𝑙′ (𝑘 , 𝑙)

𝑟 𝑟 𝑤
𝜆 = 𝑓′ (𝑘𝑙)
= 𝜆 = 𝑚𝑝
𝑘 𝑚𝑝𝑘 𝑙

𝑟 𝑤
=
𝑚𝑝𝑘 𝑚𝑝𝑙

𝑟 𝑚𝑝𝑘
𝑤
= 𝑚𝑝𝑙
→ 𝐂𝐨𝐧𝐝𝐢𝐭𝐢𝐨𝐧 𝐟𝐨𝐫 𝐜𝐨𝐬𝐭 𝐦𝐢𝐧𝐢𝐦𝐢𝐳𝐚𝐭𝐢𝐨𝐧

Page 83 of 167
This condition states that, for cost minimization, the slope of the Isocost line is equal to the slope

of the isoquant.

Proof

From the Isocost equation: 𝐶 = 𝑟𝑘 + 𝑤𝑙

Slope of the Isocost

𝑘 = 𝐶 − 𝑤𝑙

𝑐 𝑤
𝑘= − 𝐿
𝑟 𝑟

𝜕𝑘 −𝑤
=
𝜕𝐿 𝑟

Slope of Isoquant

From the production function:

𝑄 = 𝑓 (𝑘, 𝑙)

Total differential

𝜕𝑄 = 𝑓𝑘′ (𝑘, 𝑙). 𝜕𝑘 + 𝑓𝐿′ (𝑘, 𝑙)𝜕𝑙

Along an isoquant, the output does not change, hence 𝑑𝑄 = 0

𝑂 = 𝑚𝑝𝑘 . 𝜕𝑘 + 𝑚𝑝𝑙 . 𝑑𝑙

𝑚𝑝𝑘 . 𝜕𝑘 = −𝑚𝑝𝑙 𝑑𝑙

𝑑𝑘 𝑚𝑝 +𝑤
= + 𝑚𝑝 𝑙 = ……. The slope of the isoquant is equal to the slope of the isocost line.
𝜕𝑙 𝑘 𝑟

Page 84 of 167
 ̅ | < 0: Cost is minimized
If |𝐻

Example

The production function for a firm is given as 𝑄 = 12𝐿0.5 𝐾 0.5. If the firm wants to produce an

output of 240 units, find the optimal values of labour and capital that will minimize the total cost

of production given that labour cost per unit is 25 dollars and capital cost per unit is 50 dollars

Solution

Minimize Cost: 𝐶 = 𝑤𝐿 + 𝑟𝐾 Since 𝑤 = 25 and 𝑟 = 50

𝐶 = 25𝐿 + 50𝐾

Subject to: 𝑄 = 12𝐿0.5 𝐾 0.5

𝐿 = 25𝐿 + 50𝐾 + 𝜆 (240 − 12𝐿0.5 𝐾 0.5 )

𝐿𝑙 = 25 − 𝜆6𝐿−0.5 𝐾 0.5 = 0

𝐿𝑘 = 50 − 𝜆6𝐿0.5 𝐾 −0.5 = 0

𝐿𝜆 = 240 − 12𝐿0.5 𝐾 0.5 = 0

If we solve for the first order conditions:

𝜆6𝐿−0.5 𝐾 0.5 = 25

𝜆6𝐿0.5 𝐾 −0.5 = 50

Dividing the two equations we have:

Page 85 of 167
25 𝜆6𝐿−0.5 𝐾 0.5 𝐾
= =
50 𝜆6𝐿0.5 𝐾 −0.5 𝐿

1
→ 𝐾= 𝐿
2

If we substitute this in the 3rd equation, we get:

𝐿𝜆 = 240 − 12𝐿0.5 𝐾 0.5 = 0

12𝐿0.5 𝐾 0.5 = 240

12𝐿0.5 (0.5𝐿)0.5 = 240

240
𝐿0.5 (0.5𝐿)0.5 = = 20
12

𝐿(0.5)0.5 = 20

20
𝐿̅ = = 28.28
0.50.5

1
̅ = 𝐿 = 14.14
𝐾
2

𝐶̅ = 25(28.28) + 50(14.14) = 1414

Page 86 of 167
7.2.3 Output Maximization

The firm may want to maximize output subject to cost constraint as shown below.

Maximize output: 𝑄 = 𝑓(𝑘, 𝑙)

Subject to: 𝐶 = 𝑟𝑘 + 𝑤𝑙

𝐿 = 𝑓(𝑘, 𝑙) + 𝜆 (𝐶 − 𝑟𝑘 − 𝑤𝑙)

𝐿𝑘 = 𝑓𝐾′ (𝑘 , 𝑙) − 𝜆𝑟 = 0

𝐿𝑙 = 𝑓𝐿′ (𝑘 , 𝑙) − 𝜆𝑤 = 0

𝐿𝜆 = 𝐶 − 𝑟𝑘 − 𝑤𝑙 = 0

From 𝐿𝑘 and 𝐿𝑙

𝜆𝑟 = 𝑓𝑘 ′ (𝑘 , 𝑙) 𝜆𝑤 = 𝑓𝑙′ (𝑘 , 𝑙)

𝑓 ′ 𝑘 (𝑘,𝑙) 𝑚𝑝𝑘 𝑚𝑝𝑙

𝜆= = and 𝜆=
𝑟 𝑟 𝑤

𝑚𝑝𝑘 𝑚𝑝𝑙
=
𝑟 𝑤

𝑚𝑝𝑘 𝑟
= 𝑤 → 𝐂𝐨𝐧𝐝𝐢𝐭𝐢𝐨𝐧 𝐟𝐨𝐫 𝐨𝐮𝐭𝐩𝐮𝐭 𝐦𝐚𝐱𝐢𝐦𝐢𝐳𝐚𝐭𝐢𝐨𝐧
𝑚𝑝𝑙

Page 87 of 167
Example

The production function for garages which services cars and trucks is given as:

𝑄 = 15𝐿2/3 𝐾 1/3

If each unit of labour used costs $5 while each unit of capital cost $3, find the units of labour and

capital to be used so as to maximize the production given that the garage has only $450 dollars at

its disposal.

Solution

Set up the optimization problem:

Maximize output: 𝑄 = 15𝐿2/3 𝐾 1/3

Subject to: 𝑤𝐿 + 𝑟𝐾 = 𝐶

Since 𝑤 = 5; 𝑟 = 3 and 𝐶 = 450

Hence the isocost line is: 5𝐿 + 3𝐾 = 450

The augmented function will be given as:

𝐿 = 15𝐿2/3 𝐾 1/3 + 𝜆(450 − 5𝐿 − 3𝐾)

1
𝐿𝐿 = 10𝐿−3 𝐾 1/3 − 5𝜆 = 0
2 2
𝐿𝐾 = 5𝐿3 𝐾 −3 − 3𝜆 = 0

𝐿𝜆 = 450 − 5𝐿 − 3𝐾 = 0

If we solve the three first order conditions, we get:

Page 88 of 167
Dividing 𝐿𝐿 and 𝐿𝐾

2𝐾 5
=3
𝐿

6𝐾 = 5𝐿
5
𝐾 = 6𝐿

Substituting for K into 𝐿𝜆

𝐿𝜆 = 450 − 5𝐿 − 3𝐾 = 0

5𝐿 + 3𝐾 = 450

5
5𝐿 + 3 ∗ 𝐿 = 450
6

5𝐿 + 2.5𝐿 = 450

7.5𝐿 = 450

450
𝐿̅ = = 60 𝑻𝒉𝒊𝒔 𝒊𝒔 𝒕𝒉𝒆 𝒐𝒑𝒕𝒊𝒎𝒂𝒍 𝒗𝒂𝒍𝒖𝒆 𝒐𝒇 𝑳𝒂𝒃𝒐𝒖𝒓 𝒖𝒏𝒊𝒕
7.5

5 5
̅=
𝐾 ∗ 𝐿̅ = ∗ 60 = 50 𝑻𝒉𝒊𝒔 𝒊𝒔 𝒕𝒉𝒆 𝒐𝒑𝒕𝒊𝒎𝒂𝒍 𝒗𝒂𝒍𝒖𝒆 𝒐𝒇 𝒄𝒂𝒑𝒊𝒕𝒂𝒍 𝒖𝒏𝒊𝒕
6 6
2 1 2 1
𝑄̅ = 15𝐿̅3 𝐾
̅ 3 = 15(60)3 (50)3 = 15 ∗ 15.35 ∗ 3.68 = 847.32

7.3 Self-Assessment Questions

a) The production function for a firm is given as : 𝑄 = 10𝐿1/2 𝐾 1/2

If each unit of labour used costs $5 while each unit of capital cost $3, find the units of labour and

capital to be used so as to minimize the production cost given that the firm wants to produce 250

units of output.

Page 89 of 167
 b) A firm produces jam from two raw materials, strawberries and black berries. The

production function for the jam is given as: 𝑄 = 20𝑥 0.2 𝑦 0.8 where x is strawberries and y

is blackberries. If each unit of straw berries cost 20 shillings and each unit of blackberries

cost 2 shillings, and the firm has only 1450 shillings at its disposal:

i) Use the Lagrangian method to determine the optimal values of strawberries and

blackberries to be bought and the maximum amount of jam to be produced.

𝑀𝑃𝑥 𝑃𝑥
ii) Show that 𝑀𝑃𝑦 = 𝑃𝑦 at the maximum output

7.4 E-References

1. Allen R.G.D. Mathematical analysis for Economics

2. Bradley T. and Paul P. (1999) Essential Mathematics for Economics and Business, John
Willey and Sons Ltd, England.
3. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical
Economics, 4th edition, McGraw Hill, Singapore
4. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
5. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau
6. Monga G.S. (2005) Mathematics and Statistics for Economics, second revised edition,
Vikas publishing press

Page 90 of 167
 LECTURE EIGHT: MATRIX ALGEBRA I

8.1 Introduction

In this Lesson, we will discuss several concepts in matrices as well as matrix operations such as

addition, subtraction, multiplication and division.

8.2 Lesson Learning Outcomes

By the end of the lesson, the learner should be able to.

8.2.1 Understand the meaning of matrices

8.2.2 Understand different types of matrices

8.2.3 Compute the determinant of a matrix

8.2.4 Understand matrix operations

8.2.1 The meaning of Matrices

A simple economic model consists of only one equation. The equation gives the relationship

between the dependent variables and the independent. Such a relationship can be given as

e.g. 𝑦 = 𝑎0 + 𝑎1 𝑥

𝑤ℎ𝑒𝑟𝑒 𝑦 = 𝐷𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑡 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒

𝑥 = 𝐼𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑡 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒

𝑎0 𝑎𝑛𝑑 𝑎1 𝑎𝑟𝑒 𝑝𝑎𝑟𝑎𝑚𝑒𝑡𝑒𝑟𝑠 𝑜𝑓 𝑡ℎ𝑒 𝑚𝑜𝑑𝑒𝑙

It is easy to get the values of such parameters.

Page 91 of 167
The economic model can contain several equations, which can be linear or non-linear.

Each of the equations would represent one aspect of the model. For example, it can be a

constraint on one of the resources.

Matrices can give a way of writing linear simultaneous equations and helps us to have a clear

way of getting a solution for each of the variables.

A matrix can therefore be defined as a rectangular / square array of numbers, parameters or

variables. The members of this array are called the elements of the matrix. A matrix is an

ordered set of numbers, parameters or variables.

The position and the magnitude of each element in a matrix matters. The examples below

illustrate this:

No. of Students MALE FEMALE

Machakos University 4,000 6,000

Kenyatta University 12,000 20,000

This can be represented as a matrix in the form

4,000 6,000
[ ]
12,000 20,000

4,000 in the 1st row 1st column represents the number of male students in Machakos University.

6,000 in the 1st row 2nd column represents the number of female students in Machakos

University.

12,000 in the 2nd row 1st column represent the number of male students in Kenyatta University.

20,000 in the 2nd row 2nd column represent the number of female students in Kenyatta

University.

Page 92 of 167
A matrix is represented by a capital letter while its elements are represented by small letters

with 2 subscripts on them.

The 1st subscript represents the row where the element is found while the 2nd one represents

the column where the element is found.

The above matrix can be given as:

4,000 6,000 𝑎11 𝑎12

𝐴=[ ] = [𝑎 𝑎22 ]
12,000 20,000 21

𝐴 = [𝑎𝑖𝑗 ] 𝑖 − 𝑟𝑜𝑤𝑠 𝑖𝑛 𝐴

𝑗 − 𝑐𝑜𝑙𝑢𝑚𝑛𝑠 𝑖𝑛 𝐴

𝑎11 − 𝐸𝑙𝑒𝑚𝑒𝑛𝑡 𝑖𝑛 1𝑠𝑡 𝑟𝑜𝑤 1𝑠𝑡 𝑐𝑜𝑙𝑢𝑚𝑛

𝑎12 − 𝐸𝑙𝑒𝑚𝑒𝑛𝑡 𝑖𝑛 1𝑠𝑡 𝑟𝑜𝑤 2𝑛𝑑 𝑐𝑜𝑙𝑢𝑚𝑛

𝑎21 − 𝐸𝑙𝑒𝑚𝑒𝑛𝑡 𝑖𝑛 2𝑛𝑑 𝑟𝑜𝑤 1𝑠𝑡 𝑐𝑜𝑙𝑢𝑚𝑛

𝑎22 − 𝐸𝑙𝑒𝑚𝑒𝑛𝑡 𝑖𝑛 2𝑛𝑑 𝑟𝑜𝑤 2𝑛𝑑 𝑐𝑜𝑙𝑢𝑚𝑛

The dimensions of a matrix are given by the number of rows and number of columns in the

matrix, in that particular order. Our previous matrix is a 2 x 2 matrix.

A matrix with 𝑚 rows and 𝑛 columns is said to be a 𝑚 by 𝑛 matrix.

In general, the element

𝑎𝑖𝑗 belongs to 𝑖 𝑡ℎ 𝑟𝑜𝑤 and 𝑗 𝑡ℎ 𝑐𝑜𝑙𝑢𝑚𝑛

Page 93 of 167
 8.2.2 Types of Matrices

a) Column Vector – This is a matrix with only 1 column for example.

10
3
𝐴 = [15] A is 3 𝑥 1 𝐵 = [ ] B is a 2𝑥1 matrix
6
9

b) Row Vector

A matrix with only one row

𝐴 = [15 10 12] 𝐴 𝑖𝑠 𝑎 1 𝑥 3 𝑚𝑎𝑡𝑟𝑖𝑥

𝐵 = [7 15 20 9 11] 𝑇ℎ𝑖𝑠 𝑖𝑠 𝑎 1𝑥5 𝑚𝑎𝑡𝑟𝑖𝑥

c) Equal Matrices

Two matrices are equal if

i. They are of the same size.

ii. The elements in corresponding positions in the two matrices are the same.

For example

4 10 4 10
𝐴=[ ] 𝐵=[ ]
1 19 1 19

Matrix A’s elements can be represented by

𝑎11 𝑎12 𝑏11 𝑏12

𝐴 = [𝑎 𝑎22 ] 𝑤ℎ𝑖𝑙𝑒 𝐵 = [𝑏21 ]
21 𝑏22
Page 94 of 167
𝑎11 = 𝑏11 = 4 𝑒𝑙𝑒𝑚𝑒𝑛𝑡𝑠 𝑖𝑛 1𝑠𝑡 𝑟𝑜𝑤 1𝑠𝑡 𝑐𝑜𝑙𝑢𝑚𝑛

𝑎12 = 𝑏12 = 10 𝑒𝑙𝑒𝑚𝑒𝑛𝑡𝑠 𝑖𝑛 1𝑠𝑡 𝑟𝑜𝑤 2𝑛𝑑 𝑐𝑜𝑙𝑢𝑚𝑛

𝑎21 = 𝑏21 = 1 𝑒𝑙𝑒𝑚𝑒𝑛𝑡𝑠 𝑖𝑛 2𝑛𝑑 𝑟𝑜𝑤 1𝑠𝑡 𝑐𝑜𝑙𝑢𝑚𝑛

𝑎22 = 𝑏22 = 19 𝑒𝑙𝑒𝑚𝑒𝑛𝑡𝑠 𝑖𝑛 2𝑛𝑑 𝑟𝑜𝑤 2𝑛𝑑 𝑐𝑜𝑙𝑢𝑚𝑛

Hence Matrix A = Matrix B

Are the following two matrices C and D equal?

8 2 2 8
𝐷=[ ] 𝐶= [ ]
5 11 11 5

By rewrite

𝑑11 𝑑12 𝑐11 𝑐12

𝐷=[ ] 𝐶 = [𝑐 𝑐22 ]
𝑑21 𝑑22 21

𝑑11  𝑐11

𝑑21  𝑐21 Hence D and C are not equal

𝑑12  𝑐12

𝑑22  𝑐22

d) Transpose of a matrix

Page 95 of 167
This is a matrix formed when you interchange rows and columns so that a 𝑚𝑥𝑛 matrix become an

𝑛𝑥𝑚 matrix. For example

8 5 1
𝐴=[ ] 𝐴 𝑖𝑠 𝑎 2 𝑥 3 𝑚𝑎𝑡𝑟𝑖𝑥
2 3 9

8 2
𝐴𝑇 = [5 3] The transpose of A is a 3𝑥2 matrix
1 9

The transpose of a transpose is the original matrix i.e.

[𝐴𝑇 ]𝑇 = 𝐴

8 2
𝑇
𝐴 = [5 3]
1 9

[𝐴𝑇 ]𝑇 = [8 5 1
]
2 3 9

e) Square matrix

This is a matrix whose number of rows is equal to its number of columns.

E.g. 1 𝑥 1 2𝑥2 3𝑥3 4𝑥4 [𝑛 𝑥 𝑛]

Examples

2 8 10
10 20
𝐴=[ ] 𝐵 = [4 7 11]
15 40
9 5 6

Quick Test

Page 96 of 167
 Are these matrices square.

8 1
𝐴=[ ] 𝐵 = [5]
2 5

5 8 1
9 2 0
𝐶=[ ]
1 4 10
7 3 15

f) A diagonal matrix

This is a square matrix whose elements are all zero except the elements in its principal diagonal

(the diagonal from up left to right down). At least one of the elements is principal diagonal should

be non-zero.

For Example

9 0 0 0
8 0 0
0 0 0 0
𝐷 = [0 2 0] 𝐹=[ ]
0 0 0 0
0 0 0
0 0 0 11

1 0
𝐸=[ ]
0 1

g) An Identity / Unit matrix

A diagonal matrix, whose elements of the principal diagonal are all 1.

1 0 0
1 0
e.g. 𝐸=[ ] 𝐴 = [0 1 0]
0 1
0 0 1

h) A null / zero matrix

Page 97 of 167
A matrix whose elements are all zero.

0 0
0 0 0
0 0
𝐵=[ ] 4𝑥2 𝑛𝑢𝑙𝑙 𝑚𝑎𝑡𝑟𝑖𝑥 𝐶 = [0 0 0] 3𝑥3 𝑛𝑢𝑙𝑙 𝑚𝑎𝑡𝑟𝑖𝑥
0 0
0 0 0
0 0

𝐸 = [0] 1𝑥1 𝑛𝑢𝑙𝑙 𝑚𝑎𝑡𝑟𝑖𝑥

i) Sub- matrix

This a matrix formed from another matrix by deleting some specified row/s and or column/s of the

matrix.

For Example

5 7 0
𝐴 = [9 4 2]
3 1 8

9 4 2
𝐴1 = [ ] − 𝐹𝑜𝑟𝑚𝑒𝑑 𝑏𝑦 𝑑𝑒𝑙𝑒𝑡𝑖𝑛𝑔 𝑡ℎ𝑒 1𝑠𝑡 𝑟𝑜𝑤 𝑜𝑓 𝑚𝑎𝑡𝑟𝑖𝑥 𝐴.
3 1 8

5 7
𝐴2 = [ ]
9 4

− 𝐹𝑜𝑟𝑚𝑒𝑑 𝑏𝑦 𝑑𝑒𝑙𝑒𝑡𝑖𝑛𝑔 𝑡ℎ𝑒 3𝑟𝑑 𝑟𝑜𝑤 𝑎𝑛𝑑 3𝑟𝑑 𝑐𝑜𝑙𝑢𝑚𝑛 𝑜𝑓 𝑚𝑎𝑡𝑟𝑖𝑥 𝐴.

In any matrix, we can form several sub matrices from it.

j) Principal Sub-matrix

If B is a square matrix, then any square sub matrix formed from B whose principal diagonal is part

of the principal diagonal of B is a principal sub matrix of A.

Page 98 of 167
 5 10 2 4
9 1 6 9
e.g. 𝐵=[ ]
8 13 7 5
1 8 0 7

5 10 2 1 6 9
𝐵1 = [9 1 6] 𝐵3 = [13 7 5]
8 13 7 8 0 7

5 10 1 6
𝐵2 = [ ] 𝐵4 = [ ]
9 1 13 7

𝐵1 , 𝐵2 , 𝐵3 𝑎𝑛𝑑 𝐵4 𝑎𝑟𝑒 𝑎𝑙𝑙 𝑝𝑟𝑖𝑛𝑐𝑖𝑝𝑎𝑙 𝑠𝑢𝑏 𝑚𝑎𝑡𝑟𝑖𝑐𝑒𝑠 𝑜𝑓 𝐵.

8.2.3 Determinant of a matrix

Determinants are only defined for square matrices.

It is a uniquely defined scalar associated with the matrix.

The determinant of a matrix B is denoted as:

|𝑩|

i. Determinant of a 1 x 1 matrix

𝐵 = [𝑎11 ]

Its determinant is given as

|𝐵| = 𝑏11

𝐼𝑓 𝐵 = [10] 𝑖𝑡𝑠 𝑑𝑒𝑡𝑒𝑟𝑚𝑖𝑛𝑎𝑛𝑡 = 10

Page 99 of 167
ii. Determinant of a 2 x 2 matrix
𝑎11 𝑎12
𝐵 = [𝑎 𝑎22 ]
21

|𝐵| = 𝑎11 𝑎22 − 𝑎21 𝑎12

For example

8 6
𝐵=[ ]
9 1

|𝐵| = (8 𝑥 1) − (9 𝑥 6) = 8 − 54

= −48

Compute the determinant of matrix C below

8 0
𝐶=[ ]
4 1

iii. Determinant of a 3 x 3 matrix

𝑎11 𝑎12 𝑎13

𝐴 = [𝑎21 𝑎22 𝑎23 ]
𝑎31 𝑎32 𝑎33

This is easily done by rewriting the 1st 2 columns of matrix A next to the matrix on the right hand

side.

Page 100 of 167

The first 3 principal diagonals from left to right going down formed are added while other 3

diagonals (auxiliary diagonals) formed from left to right up are subtracted from the sum of the

previous 3.

For example

4 2 0
𝐴 = [8 9 2]
1 5 3

4 2 0 4 2
|𝐴| = |8 9 2| 8 9
1 5 3 1 5

4 2 0 4 2
|𝐴| = |8 9 2| 8 9
1 5 3 1 5

(𝐴) = {4 𝑥 9 𝑥 3 + 2 𝑥 2 𝑥 1 + 0 𝑥 8 𝑥 5} − {1 𝑥 9 𝑥 0 − 5 𝑥 2 𝑥 4 − 3 𝑥 8 𝑥 2}

= 108 + 4 + 0 − 0 − 40 − 48 = 24

Exercise

Compute the determinant of matrix B:

8 2 2
𝐵 = [6 5 10]
1 9 3

Solution: -518

Page 101 of 167

NOTE

In the course of this lecture we will learn about another method of getting the determinant of any

square matrix. The method is called Laplace Expansion.

Properties of determinants

i. If you interchange the rows and columns of a matrix it leaves the determinant unchanged.

Hence the determinant of the transpose is the same as the determinant of the original matrix.

For example:

5 8 (𝐵) = 5 − 72 = −67
𝐵=[ ]
9 1

5 9
𝐵𝑇 = [ ] = (𝐵𝑇 ) = 5 − 72 = −67
8 1

ii. If you interchange any 2 rows or any 2 columns, of a matrix it changes the sign of a matrix.

For example

15 1 (𝐶) = 45 − 2 = 43
𝐶=[ ]
2 3

12 3
𝐶1 = = |𝐶1 | = 12 − 45 = −43
15 1

Example 2

3 4 6 3 𝑥 5 𝑥0 + 4 𝑥 1 𝑥 2 + 6 𝑥 1 𝑥 1
𝐷 = [1 5 1] |𝐷| = −2 𝑥 5𝑥 6 − 1 𝑥 1 𝑥 3 − 0𝑥1 𝑥 4
2 1 0 = 0 + 8 + 6 − 60 − 3 − 0 = −49

Page 102 of 167

 If we interchange the 2nd and 3rd column we get another matrix given as:

3 6 4 3𝑥1𝑥1+6𝑥5𝑥2+4𝑥1𝑥0
𝐷1 = [1 1 5] |𝐷| = −2 𝑥 1 𝑥 4 − 0 𝑥 5 𝑥 3 − 1 𝑥 1 6
2 0 1 = 3 + 60 − 8 − 6 = 63 − 14 = 49

|𝐷| = −|𝐷1 |

iii. If each of the elements in a row or column is multiplied by a constant 𝑥, then the determinant

of the new matrix will be multiplied by 𝑥.

e.g.

8 2
𝐶=[ ] |𝐶| = 24 − 2 = 22
1 3

If x=2

16 4
𝐶 𝑛𝑒𝑤 = [ ]
1 3

|𝐶 𝑛𝑒𝑤| = |16 4
| = 48 − 4 = 44
1 3

Hence, |𝐶 𝑛𝑒𝑤| = 2|𝐶|

iv. Determinant of a matrix with 2 identical rows is 0.

8 8 2
𝐷 = [1 1 0]
3 3 7

8 8 2
|𝐷| = |1 1 0|
3 3 7

=8𝑥1𝑥7+8𝑥0𝑥3+2𝑥1𝑥3−3𝑥1𝑥2−3𝑥0𝑥8−7𝑥1𝑥8

=0

Page 103 of 167

v. If a matrix has 2 row or 2 columns which are multiples of one another, the determinant of the

matrix is zero.

5 10
e.g. 𝐸=[ ] 𝑖. 𝑒. 2𝑛𝑑 𝑐𝑜𝑙𝑢𝑚𝑛 𝑖𝑠 𝑡𝑤𝑖𝑐𝑒 𝑡ℎ𝑒 1𝑠𝑡 𝑐𝑜𝑙𝑢𝑚𝑛.
2 4

|𝐸| = 20 − 20 = 0

vi. If in any matrix, the elements of any of the row or column are all zero, the determinant of the

matrix is zero.

1 0 9
|𝑋| 1 𝑥 0 𝑥 18 + 0 𝑥 12 𝑥 5 + 9 𝑥 2 𝑥 0
𝑋 = [2 0 12]
−5 𝑥 0 𝑥 9 − 0 𝑥 12 𝑥 1 − 18 𝑥 2 𝑥 0 = 0
5 0 18

vii. If a matrix 𝐷∗ is formed from matrix 𝐷 by adding or subtracting a multiple of one row or

column to another row or column of then

|𝐷∗ | = |𝐷|

8 5 2 8 5 2
𝐷 = [1 6 4] 𝐷∗ = [17 16 8]
2 1 3 2 1 3

𝑤ℎ𝑒𝑟𝑒 𝑅2 𝑖𝑛 𝐷 ∗ 𝑖𝑠 𝑔𝑖𝑣𝑒𝑛 𝑎𝑠:

𝑁𝑒𝑤 𝑟𝑜𝑤2 = 𝑂𝑙𝑑 𝑅2 + 2 𝑜𝑙𝑑 𝑅1

|𝐷| = 115 |𝐷∗ | = 115

Page 104 of 167

8.2.4 Matrix Operations

a) Matrix addition and subtraction

We can only add/subtract matrices if they are of the same size (dimension). When

adding/subtracting matrices, elements in the same positions are added /subtracted. The

sum/difference of two matrices A and B will be a matrix of the same size as A and B.

Example:

4 9 10 1 15 4
𝐴 = [5 2 0] 𝐵 = [3 2 7]
6 1 6 8 3 5

𝐴+𝐵 =𝐶

5 24 14
𝐶=[8 4 7]
14 4 11

𝐴−𝐵 =𝐷

3 −6 6
𝐷=[ 2 0 −7]
−2 −2 1

b) Matrix multiplication

Matrix multiplication is a little more complex than addition and subtraction. Two matrices A and

B can only be multiplied in the order 𝐴𝑥𝐵 if they are conformable.

Page 105 of 167

Two matrices are conformable if the number of columns in the first matrix is the same as the

number of rows in the second matrix.

For example, if A is a 𝑚𝑥3 matrix, it can only be multiplied by another matrix of the order 3𝑥𝑛.

The reason for this requirement is the method used when multiplying the matrices.

Example

5 2 1 9 6
If: 𝐴 = [ ] 𝐵=[ ]
3 1 4 0 8

We can write this these matrices as:

𝑎11 𝑎12 𝑏11 𝑏12 𝑏13

𝐴 = [𝑎 𝑎22 ] 𝐵=[ ]
21 𝑏21 𝑏22 𝑏23

𝐴𝑥𝐵 𝑤𝑖𝑙𝑙 𝑏𝑒 𝑎𝑛𝑜𝑡ℎ𝑒𝑟 𝑚𝑎𝑡𝑟𝑖𝑥. 𝐿𝑒𝑡𝑠 𝑎𝑠𝑠𝑢𝑚𝑒 𝐴𝐵 = 𝐶

The dimensions of matrix C will be determined by the 2 matrices. It will have the same number of

rows as A and same number of columns as B i.e. C will be a 2x3 matrix.

𝑐11 𝑐12 𝑐13

𝐶 = [𝑐 𝑐22 𝑐23 ]
21

How do we get the elements of matrix C?

𝑐𝑖𝑗 = 𝐼𝑡 𝑖𝑠 𝑡ℎ𝑒 𝑝𝑟𝑜𝑑𝑢𝑐𝑡 𝑜𝑏𝑡𝑎𝑖𝑛𝑒𝑑 𝑏𝑦 𝑚𝑢𝑙𝑡𝑖𝑝𝑙𝑦𝑖𝑛𝑔 𝑡ℎ𝑒 𝑒𝑙𝑒𝑚𝑒𝑛𝑡𝑠 𝑖𝑛 𝑡ℎ𝑒 𝑖𝑡ℎ 𝑟𝑜𝑤

𝑜𝑓 𝑚𝑎𝑡𝑟𝑖𝑥 𝐴 𝑏𝑦 𝑡ℎ𝑒 𝑒𝑙𝑒𝑚𝑒𝑛𝑡𝑠 𝑖𝑛 𝑡ℎ𝑒 𝑗𝑡ℎ 𝑐𝑜𝑙𝑢𝑚𝑛 𝑜𝑓 𝑚𝑎𝑡𝑟𝑖𝑥 𝐵.

Page 106 of 167

 𝑐11 = 𝑎11 . 𝑏11 + 𝑎12 . 𝑏21 𝑐12 = 𝑎11 . 𝑏12 + 𝑎12 . 𝑏22 𝑐13 = 𝑎11 . 𝑏13 + 𝑎12 . 𝑏23

𝑐21 = 𝑎21 . 𝑏11 + 𝑎22 . 𝑏21 𝑐22 = 𝑎21 . 𝑏12 + 𝑎22 . 𝑏22 𝑐23 = 𝑎21 . 𝑏13 + 𝑎22 . 𝑏23

Using our example:

5 2 1 9 6
𝐴=[ ] 𝐵=[ ]
3 1 4 0 8

𝑐11 = 5𝑥1 + 2𝑥4 𝑐12 = 5𝑥9 + 2𝑥0 𝑐13 = 5𝑥6 + 2𝑥8

𝑐21 = 3𝑥1 + 1𝑥4 𝑐22 = 3𝑥9 + 1𝑥0 𝑐23 = 3𝑥6 + 1𝑥8

13 45 46
𝐶=[ ]
7 27 26

4 5
4 3 7
Two matrices A and B are given as: 𝐴 = [ ]. This is a 2𝑥3 matrix and 𝐵 = [6 2] 𝑡his
1 2 5
8 9

is a 3𝑥2 matrix. The resultant matrix will be a 2𝑥2 matrix.

4 5
4 3 7
𝐴𝐵 = 𝐶 𝐴=[ ] 𝐵 = [6 2]
1 2 5
8 9

𝑐11 𝑐12 16 + 18 + 56 20 + 6 + 63 90 89
.𝐶 = [𝑐 𝑐22 ] = [ 4 + 12 + 40 ]=[ ]
21 5 + 4 + 45 56 54

Page 107 of 167

Note: When multiplying matrices, 𝐴𝑥𝐵 ≠ 𝐵𝑥𝐴

It may not always be possible to multiply the matrices in the reverse order unless the two matrices

are square.

NB: Matrix division will be discussed in LECTURE NINE

8.3 Self-Assessment Questions

a) Given the following matrix:

4 9 10
𝐴 = [5 2 0]
6 1 6

i) What is the order of the matrix?

ii) Compute the transpose of the matrix.

iii) Compute the determinant of the matrix.

b) Compute the product AB and BA for the following matrices

4 9 10 1 15 4
𝐴 = [5 2 0] 𝐵 = [3 2 7]
6 1 6 8 3 5

8.4 E-References

1. Allen R.G.D. Mathematical analysis for Economics

2. Bradley T. and Paul P. (1999) Essential Mathematics for Economics and Business,
John Willey and Sons Ltd, England.
3. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical
Economics, 4th edition, McGraw Hill, Singapore
4. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists,
Macmillan press, London
5. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau
6. Monga G.S. (2005) Mathematics and Statistics for Economics, second revised edition,
Vikas publishing press

Page 108 of 167

 LECTURE NINE: MATRIX ALGEBRA II

9.1 Introduction

In this Lesson, we will discuss how to perform matrix division. We will also discuss how to

compute the rank of a matrix.

9.2 Lesson Learning Outcomes

By the end of the lesson, the learner should be able to.

9.2.1 Understand Matrix Inversion

9.2.2 Understand Cramer’s Rule

9.2.3 Compute the rank of a matrix

Before we discuss how to divide matrices, we discuss the following concepts.

i) Minor of a matrix

The minor of a square matrix C are determinants of square sub matrices of C.

𝑐11 𝑐12 𝑐13

𝑐
𝐴𝑠𝑠𝑢𝑚𝑒 𝐶 = [ 21 𝑐22 𝑐23 ]
𝑐31 𝑐32 𝑐33

To get the minor of the entire matrix we first get the minors of each of the elements.

The minor of any element is the determinant of the matrix formed by deleting the row and the

column where that element is found. By abbreviating the minor as 𝑀 then:

Page 109 of 167

 𝑐22 𝑐23 𝑐21 𝑐23 𝑐21 𝑐22
𝑀(𝐶11 ) = |𝑐 𝑐33 | 𝑀(𝐶12 ) = |𝑐 𝑐33 | 𝑀(𝐶13 ) = |𝑐 𝑐32 |
32 31 31

𝑐12 𝑐13 𝑐11 𝑐13 𝑐11 𝑐12

𝑀(𝐶21 ) = |𝑐 𝑐33 | 𝑀(𝐶22 ) = |𝑐 𝑐33 | 𝑀(𝐶23 ) = |𝑐 𝑐33 |
32 31 31

𝑐12 𝑐13 𝑐11 𝑐13 𝑐11 𝑐12

𝑀(𝐶31 ) = |𝑐 𝑐23 | 𝑀(𝐶32 ) = |𝑐 𝑐23 | 𝑀(𝐶33 ) = |𝑐 𝑐32 |
22 21 31

Example:

5 0 7
𝐶 = [2 9 3]
8 10 4

9 3 2 3 2 9
𝑀(5) = | |=6 𝑀(0) = | | = −16 𝑀(7) = | | = −52
10 4 8 4 8 10

0 7 5 7 5 0
𝑀(2) = | | = −70 𝑀(9) = | | = −36 𝑀(3) = | | = 50
10 4 8 4 8 10

0 7 5 7 5 0
𝑀(8) = | | = −63 𝑀(10) = | |=1 𝑀(4) = | | = 45
9 3 2 3 2 9

The minor of matrix C is therefore:

6 −16 −52
𝑀 (𝐶) = [−70 −36 50 ]
−63 1 45

Page 110 of 167

 ii) Cofactor of an element

This is given by the minor of an element multiplied by ( −1) raised to the power of the addition

of its subscripts.

e.g.

𝑐𝑜𝑓𝑎𝑐𝑡𝑜𝑟 𝑜𝑓 𝑎11 = 𝑀𝑖𝑛𝑜𝑟 (𝑎11 )𝑥(−1)1+1

𝑐𝑜𝑓𝑎𝑐𝑡𝑜𝑟 𝑜𝑓 𝑎12 = 𝑀𝑖𝑛𝑜𝑟 (𝑎12 )𝑥(−1)1+2

𝑐𝑜𝑓𝑎𝑐𝑡𝑜𝑟 𝑜𝑓 𝑎42 = 𝑀𝑖𝑛𝑜𝑟 (𝑎42 )𝑥(−1)4+2

Example

Find the cofactors of the elements in matrix C below:

5 2 6 𝐶11 𝐶12 𝐶13

𝐶 = [8 5 4] = [ 𝐶21 𝐶22 𝐶23 ]
1 1 7 𝐶31 𝐶32 𝐶33

5 4
𝑀(𝑐11 ) = | | = 35 − 4 = 31
1 7

8 4
𝑀(𝑐12 ) = | | = 56 − 4 = 52
1 7

8 5
𝑀(𝑐13 ) = | |=8−5=3
1 1

Page 111 of 167

𝐶𝑜𝑓(𝐶11 ) = 𝑀(𝐶11 ) 𝑥 (−1)1+1 = 31 𝑥 (−1)2 = 31

𝐶𝑜𝑓(𝐶12 ) = 𝑀(𝐶12 ) 𝑥 (−1)1+2 = 52 𝑥 (−1)3 = −52

𝐶𝑜𝑓(𝐶13 ) = 𝑀(𝐶13 ) 𝑥 (−1)1+3 = 3 𝑥 −14 = 3

Cofactor matrix

The cofactor matrix of C is the matrix formed by the cofactors corresponding to the elements of

matrix C.

e.g.

𝑐11 𝑐12 𝑐13

𝐶 = [𝑐21 𝑐22 𝑐23 ]
𝑐31 𝑐32 𝑐33

𝐶11 𝐶12 𝐶13

𝐶𝑜𝑓 (𝐶) = [𝐶21 𝐶22 𝐶23 ]
𝐶31 𝐶32 𝐶33

Example

Find the cofactor matrix of matrix C below:

5 2 6
𝐶 = [8 5 4]
1 1 7

Minors:

5 4 5 4 8 5
𝑀(5) = | | = 31 𝑀(2) = | | = 52 𝑀(6) = | |=3
1 7 1 7 1 1

Page 112 of 167

 2 6 5 6 5 2
𝑀(8) = | |=8 𝑀(5) = | | = 29 𝑀(4) = | |=3
1 7 1 7 1 1

2 6 5 6 5 2
𝑀(1) = | | = −22 𝑀(1) = | | = −28 𝑀(7) = | |=9
5 4 8 4 5 5

𝐶𝑜𝑓(5) = (−1)1+1 ∗ 𝑀(5) = 31

𝐶𝑜𝑓(5) = −11+1 (31) = 31 𝐶𝑜𝑓(2) = −11+2 (52) = −52 𝐶𝑜𝑓(6) = −11+3 (3) = 3

𝐶𝑜𝑓(8) = −12+1 (8) = −8 𝐶𝑜𝑓(5) = −12+2 (29) = 29 𝐶𝑜𝑓(4) = −12+3 (3) = 3

𝐶𝑜𝑓(1) = −13+1 (−22) = −22 𝐶𝑜𝑓(1) = −13+2 (−28) = 28 𝐶𝑜𝑓(7) = −13+3 (9) = 9

31 −52 3
𝐶𝑜𝑓𝑎𝑐𝑡𝑜𝑟 𝑜𝑓 𝐶 = [ −8 29 −3]
−22 28 9

iii) Adjoint matrix

This is the transpose of a cofactor matrix.

Example 1

In our example above, the adjoint of C will be given as the Transpose of the cofactor matrix of C.

31 −52 3
𝐶𝑜𝑓 𝐶 = [ −8 29 −3]
−22 28 9

Page 113 of 167

 31 −8 −22
𝐴𝑑𝑗𝑜𝑖𝑛𝑡 𝐶 = [−52 29 28 ]
3 −3 9

Example 2

Compute the adjoint of the following matrix

4 1
𝐴=[ ]
8 9

𝑀(4) = 9 𝑀(1) = 8 𝑀(8) = 1 𝑀(9) = 4

9 8
𝑀(𝐴) = [ ]
1 4

𝐶𝑜𝑓𝑎𝑐𝑡𝑜𝑟𝑠 𝑜𝑓 𝐴 = −11+1 (9) = 9 (−1)1+2 (8) = −8

−12+2 (1) = −1 (−1)2+2 (4) = 4

9 −8
𝐶𝑜𝑓 (𝐴) = [ ]
−1 4

The adjoint matrix is the transpose of the cofactor matrix

9 −1
𝐴𝑑𝑗𝑜𝑖𝑛𝑡 (𝐴) = [ ]
−8 4

iv) Inverse Matrix

The inverse of matrix C denoted as

𝐶 −1 𝑖𝑠 𝑑𝑒𝑓𝑖𝑛𝑒𝑑 𝑎𝑠:

Page 114 of 167

 𝐴𝑑𝑗𝑜𝑖𝑛𝑡 𝐶
𝐶 −1 =
|𝐶|

Example

10 9
𝐶=[ ]
2 4

Find the inverse of matrix C.

Steps

4 2
1. Get minor of 𝐶=[ ]
9 10

4 −2
2. Get cofactor of 𝐶 = [ ]
−9 10

4 −9
3. Get ad joint of 𝐶 = [ ]
−2 10

10 9
4. Get determinant of 𝐶 = | | = 40 − 18 = 22
2 4

𝐴𝑑 𝑗𝑜𝑖𝑛𝑡 𝐶
5. Get inverse of 𝐶 = 𝐶 −1 = |𝐶|

4 −9
[ ]
= [ −2 10 ]
22

Page 115 of 167

 4⁄ −9⁄ 2 −9⁄
𝐶 −1 = 1⁄22 [
4 −9
] = [ 22 22] = [ ⁄11 22]
−2 10 −2⁄ 10⁄ −1⁄ 5⁄
22 22 11 11

Example 2

Get the inverse of matrix B

6 2 1
𝐵 = [4 5 2]
1 3 1

Step 1

−1 2 7
1. Get minor 𝐵 = [−1 5 16]
−1 8 22

−1 −2 7
2. Get cofactor of 𝐵 = [ 1 5 −16]
−1 −8 22

−1 1 −1
3. Get adjoint 𝐵 = [−2 5 −8]
7 −16 22

4. Get determinant of B

|𝐵|

= 30 + 4 + 12 − 5 − 36 − 8

Page 116 of 167

 = 46 − 49 = −3

−1 1 −1 1⁄ −1⁄ 1⁄
[−2 5 −8] 3 3 3
𝐴𝑑𝑗 𝐵 7 −16 22 2 −5⁄ 8⁄
5. Get 𝐵 −1 = |𝐵|
= = ⁄3 3 3
−3
−7 16⁄ −22⁄
[ ⁄3 3 3]

The inverse of a matrix multiplied by the original matrix gives an identity matrix. 𝐵. 𝐵 −1 = 𝑰

Cofactor expansion of the determinant

We can get the determinant of a matrix by using the method of cofactor expansion. This is done

by choosing any row or column and using it for the expansion. For example:

𝑐11 𝑐12 𝑐13 5 2 6

𝐶 = [𝑐21 𝑐22 𝑐23 ] = [8 5 4]
𝑐31 𝑐32 𝑐33 1 1 7

If we use the first column, we can get the determinant of matrix C as follows:

|𝐶| = 𝑐11 ∗ 𝑐𝑜𝑓 𝑜𝑓 𝑐11 + 𝑐21 ∗ 𝑐𝑜𝑓 𝑜𝑓 𝑐21 + 𝑐31 𝑐𝑜𝑓 𝑜𝑓𝑐31

5 4 2 6 2 6
= 5 ∗ (−1)1+1 | | + 8(−1)2+1 | | + 1(−1)3+1 | |
1 7 1 7 5 4

= 5 ∗ (35 − 4) − 8(14 − 6) + 1(8 − 30)

Page 117 of 167

= 5 ∗ 31 − 8 ∗ 8 + 1 ∗ −22

= 69

Note: Check by the other method if the answer is the same.

Division of matrices

Now that we know how to compute the inverse of a matrix, it will be very easy to divide matrices.

To show how we divide matrices, we will use the case of a system of simultaneous equations. We

will learn the inverse method and also the Cramer’s rule of matrix division.

9.2.1 Matrix Inverse Method

If we have 2 simultaneous equations given as:

𝑎11 𝑋1 + 𝑎12 𝑋2 = 𝐷1

𝑎21 𝑋1 + 𝑎22 𝑋2 = 𝐷2

We can rewrite the equations in matrix form as:

𝑎11 𝑎12 𝑋1 𝐷1
[𝑎 𝑎22 ] ∗ [𝑋2 ] = [𝐷2 ]
21

𝑎11 𝑎12
If we represent the coefficients matrix [𝑎 𝑎22 ] 𝑎𝑠 𝐴
21

𝑋1 𝐷
The matrix of the variables [ ] 𝑎𝑠 𝑋 and the matrix of the constants [ 1 ] 𝑎𝑠 𝐷, we can rewrite
𝑋2 𝐷2

the system of the equations as:

𝐴𝑋 = 𝐷

Page 118 of 167

where A and D are known matrices and X is unknown, we can get matrix X by the following

operation:

𝐷
𝑋 = 𝐴 = 𝐴−1 . 𝐷

We pre-multiply matrix D with the inverse of matrix A.

Note: If we multiply 𝐷 by the inverse of 𝐴 it will not be possible because there is no conformity.

𝐴−1 = 2 ∗ 2 𝑚𝑎𝑡𝑟𝑖𝑥 𝐷 = 2 ∗ 1 ℎ𝑒𝑛𝑐𝑒, 2 ∗ 1 𝑎𝑛𝑑 2 ∗ 2 𝑖𝑠 𝑛𝑜𝑡 𝑝𝑜𝑠𝑠𝑖𝑏𝑙𝑒.

Hence to ensure we can get matrix 𝑋 which is a 2 ∗ 1 matrix, we have to ensure we mutiply the

inverse of 𝐴 by 𝐷

Example

If we have 2 simultaneous equations given as:

4𝑋1 + 6𝑋2 = 10

𝑋1 + 10𝑋2 = 30

4 6 𝑋 10
We can rewrite the matrices as:[ ] ∗ [ 1] = [ ]
1 10 𝑋2 30

10
[ ]
𝑋1 4 6 −1 10
[ ] = 30 = [ ] [ ]
𝑋2 4 6 1 10 30
[ ]
1 10

4 6
To get the inverse of the matrix [ ] we proceed as follows
1 10

Get the Determinant = 40-6=34

Get the minors

Page 119 of 167

 10 1
=[ ]
6 4

Get the cofactors

10 −1
=[ ]
−6 4

Get the adjoint

10 −6
=[ ]
−1 4
10 6
−1 1 10 −6 −
4 6
[ ] = [ ] = [ 34 34]
1 10 34 −1 4 1 4
−
34 34

10 6 10 6 80
𝑋1 − ∗ 10 − ∗ 30 −
[ ] = [ 34 34] [10] = [ 34 34 ] = [ 34]
𝑋2 1 4 30 1 4 110
− − ∗ 10 + ∗ 30
34 34 34 34 34

80 40
𝑋1 = − =−
34 17

𝑤ℎ𝑖𝑙𝑒

110 55
𝑋2 = =
34 17

9.2.2 Cramer’s Rule Method

Given a system of two equations as below:

𝑎1 𝑋 + 𝑏1 𝑌 = 𝐶1 − − − − − − − 𝑖

𝑎2 𝑋 + 𝑏2 𝑌 = 𝐶2 − − − − − − − ii

Page 120 of 167

 If we want to get the values of X and Y, we can use the substitution method to eliminate one

of them and finally get a solution as follows:

𝑎1 𝑋 = 𝐶1 − 𝑏1 𝑌

𝐶1 − 𝑏1 𝑌
𝑋=
𝑎1

If we substitute this in the second equation, we get:

𝐶1 −𝑏1 𝑌
𝑎2 + 𝑏2 𝑌 = 𝐶2
𝑎1

Solving for Y we get

𝑎2 𝐶1 − 𝑎2 𝑏1 𝑌+𝑎1 𝑏2 𝑌
= 𝐶2
𝑎1

𝑎2 𝐶1 − 𝑎2 𝑏1 𝑌+𝑎1 𝑏2 𝑌 = 𝑎1 𝐶2

𝑌(𝑎1 𝑏2 − 𝑎2 𝑏1 ) = 𝑎1 𝐶2 − 𝑎2 𝐶1

𝑎1 𝐶2 − 𝑎2 𝐶1
𝑌=
𝑎1 𝑏2 − 𝑎2 𝑏1

By replacing these values in the expression of X, we get:

𝐶1 𝑏2 − 𝐶2 𝑏1
𝑋=
𝑎1 𝑏2 − 𝑎2 𝑏1

Notice that the numerator and the denominator are a product of two values, just like a determinant

of a 2𝑥2 matrix. We can therefore write the expressions of Y and X using determinants as:

Page 121 of 167

 𝑎1 𝐶1
| |
𝑎 𝐶2
𝑌= 2
𝑎 𝑏1
| 1 |
𝑎2 𝑏2

and

𝐶 𝑏1
| 1 |
𝐶2 𝑏2
𝑋=
𝑎 𝑏1
| 1 |
𝑎2 𝑏2

In general, Cramer’s rule is written as:

𝑎1 𝑋 + 𝑏1 𝑌 = 𝐶1

𝑎2 𝑋 + 𝑏2 𝑌 = 𝐶2

𝑎1 𝑏1 𝑋 𝐶
[ ] ∗ [ ] = [ 1]
𝑎2 𝑏2 𝑌 𝐶2

𝐶 𝑏1
| 1 |
𝐶 𝑏2
𝑋= 2
𝑎 𝑏1
| 1 |
𝑎2 𝑏2

X can be obtained by getting the determinant of the matrix formed by replacing the 1st column

of the coefficients matrix by the matrix on the right hand side of the equation and dividing it

by the determinant of the coefficients matrix.

Y on the other hand is given as:

𝑎1 𝐶1
| |
𝑎 𝐶2
𝑌= 2
𝑎 𝑏1
| 1 |
𝑎2 𝑏2

Page 122 of 167

Y can be obtained by getting the determinant of the matrix formed by replacing the 2nd column

of the coefficients matrix by the matrix on the right hand side of the equation and dividing it

by the determinant of the coefficients matrix.

Example

A system of two simultaneous equations is given as:

𝑋 + 2𝑌 = 10

5𝑋 + 8𝑌 = 40

Determine the values of X and Y using the Cramer’s rule

Solution

1 2 𝑋 10
[ ]∗[ ]=[ ]
5 8 𝑌 40

10 2
| | 80 − 80 0
𝑋 = 40 8 = =
1 2 8 − 10 −2
| |
5 8

𝑋=0

1 10
| |
𝑌= 5 40 = 40 − 50 = − 10
1 2 8 − 10 −2
| |
5 8

𝑌=5

If we have a system of three simultaneous equations given as:

Page 123 of 167

 𝑎11 𝑋1 + 𝑎12 𝑋2 + 𝑎13 𝑋3 = 𝐶1
𝑎21 𝑋1 + 𝑎22 𝑋2 + 𝑎23 𝑋3 = 𝐶2
𝑎31 𝑋1 + 𝑎32 𝑋2 + 𝑎33 𝑋3 = 𝐶3

We can rewrite the equations as:

𝑎11 𝑎12 𝑎13 𝑋1 𝐶1

[𝑎21 𝑎22 𝑎23 ] ∗ [𝑋2 ] = [𝐶2 ]
𝑎31 𝑎32 𝑎33 𝑋3 𝐶3

The values of the unknown 𝑋1 , 𝑋2 , 𝑎𝑛𝑑 𝑋3 𝑤𝑖𝑙𝑙 𝑏𝑒 𝑔𝑖𝑣𝑒𝑛 𝑎𝑠:

𝐶1 𝑎12 𝑎13 𝑎11 𝐶1 𝑎13 𝑎11 𝑎12 𝐶1

|𝐶2 𝑎22 𝑎23 | |𝑎21 𝐶2 𝑎23 | |𝑎21 𝑎22 𝐶2 |
𝐶 𝑎 𝑎 𝑎 𝐶 𝑎 𝑎 𝑎 𝐶
𝑋1 = 𝑎 3 𝑎32 𝑎33 𝑋2 = 𝑎 31 𝑎 3 𝑎33 𝑋3 = 𝑎 31 𝑎 32 𝑎 3
11 12 13 11 12 13 11 12 13
𝑎
| 21 𝑎 22 𝑎 23 | 𝑎
| 21 𝑎 22 𝑎 23 | 𝑎
| 21 𝑎 22 𝑎 23 |
𝑎31 𝑎32 𝑎33 𝑎31 𝑎32 𝑎33 𝑎31 𝑎32 𝑎33

For example:

𝑋1 + 𝑋2 + = 12
2𝑋1 + 5𝑋2 + 2𝑋3 = 20
6𝑋1 + 3𝑋2 + 6𝑋3 = 0

1 1 0 𝑋1 12
[2 5 2] ∗ [𝑋2 ] = [20]
6 3 6 𝑋3 0

12 1 0
|20 5 2|
168
𝑋1 = 0 3 6 = =7
1 1 0 24
|2 5 2|
6 3 6

1 12 0
|2 20 2|
𝑋2 = 6 0 6 = 120 = 5
1 1 0 24
|2 5 2|
6 3 6

Page 124 of 167

 1 1 12
|2 5 20|
𝑋3 = 6 3 0 = −228 = −9.5
1 1 0 24
|2 5 2|
6 3 6

9.2.3 The Rank of a matrix

The rank of a matrix is the maximum number of linearly independent rows in a matrix.

The rank of an 𝑚 𝑥 𝑛 matrix is the maximum order of a non-vanishing determinant that can be

constructed from the rows and columns of that matrix.

The rank can be at most 𝑚 𝑜𝑟 𝑛 whichever is smaller, because the determinant is defined for only

square matrices.

e.g. 𝑓𝑜𝑟 𝑎 4 𝑥 2 𝑚𝑎𝑡𝑟𝑖𝑥, 𝑟𝑎𝑛𝑘 𝑐𝑎𝑛 𝑏𝑒 𝑎𝑡 𝑚𝑜𝑠𝑡 2.

𝑖𝑓 𝑤𝑒 ℎ𝑎𝑣𝑒 𝑎 𝑚𝑎𝑡𝑟𝑖𝑥 𝐴 𝑡ℎ𝑒𝑛 𝑟𝑎𝑛𝑘 (𝐴) ≤ min(𝑚, 𝑛)

How to find the rank of a matrix of order 𝒎 𝒙 𝒏.

1. Can be done by inspection for smaller dimension matrix.

2. Transform the matrix into an Echelon matrix

An echelon matrix is a matrix such that all elements in principal diagonal are one and all

elements below the principal diagonal are zeros.

1 6 10
0 1 5
0 0 1

This can be done by at least 3 operations.

Page 125 of 167

 i. Interchange any 2 rows in matrix.

ii. Multiply / divide row by any scalar (𝐾) ≠ 0

iii. Add 𝑘 times any row to another row.

None of these operations alters the rank of the matrix.

Example 1

Compute the rank of the following matrix

0 −1 −4
𝐵 = [3 1 2]
6 1 0

Interchange row 1 and row 3

6 1 0
𝐵1 = 3 1 2
0 −1 −4

Change 6 in the1st row 1st column into a 1 by dividing the entire row 1 by 6.

1 1⁄6 0
3 1 2
0 −1 −4

Change the 3 in row 2, 1st column into a 0. This will be done by the following operation:

New𝑅2 = 𝑂𝑙𝑑𝑅𝑜𝑤2 − 3𝑛𝑒𝑤𝑅𝑜𝑤1

1 1⁄6 0
0 1⁄2 2
0 −1 −4

Page 126 of 167

Next step:

Change ½ in row 2, 2nd column into a 1 by multiplying the entire row 2 by 2

1 1⁄6 0
𝐵4 = 0 1 4
0 −1 −4

Change −1 in 𝑟𝑜𝑤 3 𝑖𝑛𝑡𝑜 𝑎 0 by the following operation:

𝑁𝑒𝑤𝑅3 = 𝑂𝑙𝑑𝑅3 + 𝑁𝑒𝑤𝑅2

1 1⁄6 0
0 1 4
0 0 0

This is an Echelon matrix

→ 2 non zero rows

Hence the matrix has a rank of 2.

9.3 Self-Assessment Questions

4 1 2
a) Given a matrix A =5 8 11
3 0 9

Determine the inverse of the matrix.

b) Compute the rank of the following matrix

7 6 3 3
𝐵=0 1 2 1
8 0 0 8

Page 127 of 167

c) Given the following system of equations, Compute the equilibrium values of the unknowns
using matrix inversion and Cramer’s Rule.
2𝑋1 + 2𝑋2 + = 24
4𝑋1 + 10𝑋2 + 4𝑋3 = 40
12𝑋1 + 6𝑋2 + 12𝑋3 = 0

9.4 Self-Assessment Questions

7. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical Economics,
4th edition, McGraw Hill, Singapore
8. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
9. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau

Page 128 of 167

 LECTURE TEN: ECONOMIC APPLICATIONS OF MATRIX ALGEBRA I

10.1 Introduction

In this Lesson, we will discuss how matrix algebra is applied in economics.

10.2 Lesson Learning Outcomes

By the end of the lesson, the learner should be able to.

10.2.1 Understand the input-output model

10.2.2 Understand market equilibrium analysis using matrix algebra

10.2.1 Input-Output Model

In any economy, there are several industries that are interdependent. The output of an industry is

likely to be sold to other industries as an input, can be bought by the industry itself, or can be used

for final demand. For example, the output of a poultry farmer, who is in agriculture industry, can

be eggs among other products. The eggs can be bought by other farmers for hatching, can be

bought by bakers in the service industry or can be bought by consumers for final consumption. An

industry also requires inputs from other industries.

An input-output model helps to know the level of output required for each industry to satisfy the

inter-industry requirements and final demand. The correct level of output depends on total

Page 129 of 167

requirements by all industries. Due to inter industry dependence correct output levels are

determined by input requirements in whole economy.

Assumptions

1. Each industry produce 1 homogenous product.

2. Each industry uses a fixed input-output ratio (factor combination) for production of a unit of

output.

3. Production is subject to constant returns to scale. i.e. if we multiply inputs by a constant 𝑘

output changes by a multiple of K.

Consider the following case of an economy with three sectors Agriculture, manufacturing and

service sectors. All figures are given in million Kshs.

Input to

Agriculture Manufacturing Services Final Total

demand Output

Output Agriculture 150 225 125 100 600

From Manufacturing 210 250 140 300 900

Service 170 0 30 100 300

Other inputs 70 425 5

Total inputs 600 900 300

The total output from each of the sectors must satisfy the final demand and the inter-industry

requirements.

Page 130 of 167

 For example, 150 million worth of output from agriculture is sold to agriculture as input, 225

to manufacturing, 125 is sold to service and 100 for final demand.

The output from an industry can also be demanded for final consumption. This can be by the

local consumers and also for exports. The final demand for each sector is represented in the

column labeled Final demand.

An industry requires output from other industries as inputs (raw materials) in its production

activities. For example, the services sector requires 125 million worth of inputs from

agriculture, 140 million from manufacturing and 30 million from itself (service). In addition,

an industry also requires other inputs such as labour and imported raw materials. This is

represented in the row labeled other inputs. The total requirements of the service sector for

other inputs is 5 million.

From your knowledge in macroeconomics, the total product should be used to purchase inputs

in the firm. Hence total output for each sector/industry should be equal to the total inputs for

each sector/industry.

To produce output in 𝑗 𝑡ℎ sector, the input required from 𝑖 𝑡ℎ sector is a fixed amount denoted as

𝑎𝑖𝑗 .

1𝑠𝑡 𝑠𝑢𝑏𝑠𝑐𝑟𝑖𝑝𝑡 𝑑𝑒𝑛𝑜𝑡𝑒𝑠 𝑡ℎ𝑒 𝑖𝑛𝑝𝑢𝑡 𝑤ℎ𝑖𝑙𝑒 𝑡ℎ𝑒 2𝑛𝑑 𝑠𝑢𝑏𝑠𝑐𝑟𝑖𝑝𝑡 𝑑𝑒𝑛𝑜𝑡𝑒𝑠 𝑡ℎ𝑒 𝑜𝑢𝑡𝑝𝑢𝑡

Considering a case of 3 sectors, the fixed input-output ratios can be represented in matrix form

as:

Input – output Coefficient Matrix

Page 131 of 167

 𝑎11 𝑎12 𝑎13
𝑎21 𝑎22 𝑎23
𝑎31 𝑎32 𝑎33

Output

𝐼𝑛𝑝𝑢𝑡
𝐼 𝐼𝐼 𝐼𝐼𝐼
𝑂𝑢𝑡𝑝𝑢𝑡 𝐼 𝑎11 𝑎12 𝑎13
𝐼𝐼 𝑎21 𝑎22 𝑎23
[𝐼𝐼𝐼 𝑎31 𝑎32 𝑎33 ]

𝑎12 → How much of input from sector 1 is required to produce 1 unit of output in sector 2.

In order to calculate the fixed input-output ratios, divide the input to the sector from each sector

by the total output in the sector. This can be done as shown in the table below:

Input

Agriculture Manufacturing Services

Output Agriculture 150/600 225/900 125/300

Manufacturing 210/600 250/900 140/300

Service 170/600 0/900 30/300

The input-output coefficient matrix will be given as:

Page 132 of 167

 𝐼𝑛𝑝𝑢𝑡
𝐼 𝐼𝐼 𝐼𝐼𝐼
𝑂𝑢𝑡𝑝𝑢𝑡 𝐼 0.25 0.25 0.42
𝐼𝐼 0.35 0.28 0.47
[𝐼𝐼𝐼 0.28 0 0.10]

The matrix is also known as the input-output matrix or matrix of technical coefficients.

The proportion of primary inputs which are the other inputs required by an industry apart from

what it buys from other sectors e.g. labour, imported raw materials e.t.c. can be obtained by the

following:

i) Add the elements in each column in the input output table i.e.
𝑛

∑ 𝑎𝑖𝑗 =< 1 𝐹𝑜𝑟 𝑠𝑒𝑐𝑡𝑜𝑟 1, 𝑡ℎ𝑖𝑠 𝑐𝑎𝑛 𝑏𝑒 𝑔𝑖𝑣𝑒𝑛 𝑎𝑠:

𝑖=1

𝑎11 + 𝑎21 + 𝑎31 = 0.25 + 0.35 + 0.28 = 0.88

iv) Deduct the total sum from step (i) from 1.

𝑝𝑟𝑜𝑝𝑜𝑟𝑡𝑖𝑜𝑛 𝑜𝑓 𝑃𝑟𝑖𝑚𝑎𝑟𝑦 𝑖𝑛𝑝𝑢𝑡𝑠 = 1 − ∑ 𝑎𝑖𝑗

𝑖=1

𝐼𝑛 𝑜𝑢𝑟 𝑐𝑎𝑠𝑒 𝑡ℎ𝑖𝑠 𝑤𝑖𝑙𝑙 𝑏𝑒 𝑔𝑖𝑣𝑒𝑛 𝑎𝑠:

𝑃𝑟𝑜𝑝𝑜𝑟𝑡𝑖𝑜𝑛 𝑜𝑓 𝑝𝑟𝑖𝑚𝑎𝑟𝑦 𝑖𝑛𝑝𝑢𝑡𝑠 𝑖𝑛 𝑠𝑒𝑐𝑡𝑜𝑟 1 = 1 − 0.88 = 0.12

Page 133 of 167

How can we represent the total output requirements for each sector/industry?

As noted earlier, the output from a sector is required for inter-industry requirements and also for

final demand.

For an industry to produce an output that will meet input demands and the final demand

𝑡ℎ𝑒𝑛.

𝑋1 = 𝑎11 𝑋1 + 𝑎12 𝑋2 + 𝑎13 𝑋3 + 𝐷1

𝑋2 = 𝑎21 𝑋1 + 𝑎22 𝑋2 + 𝑎23 𝑋3 + 𝐷2

𝑋3 = 𝑎31 𝑋1 + 𝑎32 𝑋2 + 𝑎33 𝑋3 + 𝐷2

Putting the like terms together.

𝑋1 − 𝑎11 𝑋1 − 𝑎12 𝑋2 − 𝑎13 𝑋3 = 𝐷1

𝑋2 − 𝑎21 𝑋1 − 𝑎22 𝑋2 − 𝑎23 𝑋3 = 𝐷2

𝑋3 − 𝑎31 𝑋1 − 𝑎32 𝑋2 − 𝑎33 𝑋3 = 𝐷2

(1 − 𝑎11 )𝑋1 − 𝑎12 𝑋2 − 𝑎13 𝑋3 = 𝐷1

−𝑎21 𝑋1 + (1 − 𝑎22 )𝑋2 − 𝑎23 𝑋3 = 𝐷2

−𝑎31 𝑋1 − 𝑎32 𝑋2 + (1 − 𝑎33 ) 𝑋3 = 𝐷2

In matrix form the three equations can be written as:

(1 − 𝑎11 ) 𝑎12 𝑎13 𝑋1 𝐷1

[ 𝑎21 (1 − 𝑎22 ) 𝑎23 ] [𝑋2 ] = [𝐷2 ]
𝑎31 𝑎32 (1 − 𝑎33 ) 𝑋3 𝐷3

There are three matrices in this formulation:

Page 134 of 167

 (1 − 𝑎11 ) 𝑎12 𝑎13
1. [ 𝑎21 (1 − 𝑎22 ) 𝑎23 ] which can also be denoted as 𝐼 − 𝐴, that is, the
𝑎31 𝑎32 (1 − 𝑎33 )

identity matrix minus the input-output coefficient matrix. This matrix ( 𝐼 − 𝐴) is known as

the Leontief matrix.

𝑋1
𝑋
2. [ 2 ] is the matrix of outputs from the three sectors. It can be denoted as 𝑋.
𝑋3

𝐷1
3. [𝐷2 ] is the final demand matrix denoted as 𝐷.
𝐷3

The three matrices can be represented as:

[1 − 𝐴 ] [ 𝑋 ] = 𝐷

If we know the fixed input-output ratios, the final output that can be produced to satisfy

the inter-industry requirements and the final demand is given as:

𝐷
𝑋= = (𝐼 − 𝐴)−1 𝐷
(𝐼 − 𝐴)

To get the total output requirements we pre-multiply the final demand matrix by the inverse

of the Leontief matrix.

Example

The input-output ratio coefficients for a three sector economy are given as:
𝑎11 𝑎12 𝑎13 0.2 0.3 0.2
𝑎
𝐴 = [ 21 𝑎22 𝑎23 ] = [0.4 0.1 0.2] while the final demand matrix is given as:
𝑎23 𝑎32 𝑎33 0.1 0.3 0.2

10
𝐷=[5]
6

Determine the total output required in each of the three sectors to meet the inter-industry input

requirements and the final demand.

Page 135 of 167

Solution

The Leontief matrix [1 − 𝐴 ] will be given as:

1 0 0 0.2 0.3 0.2

𝐼 − 𝐴 = [0 1 0] − [0.4 0.1 0.2]
0 0 1 0.1 0.3 0.2

0.8 −0.3 −0.2

1 − 𝐴 = [−0.4 0.9 −0.2]
−0.1 −0.3 0.8

Determinant of matrix (𝐼 − 𝐴) = 0.384

Inverse of matrix 𝐼−𝐴 is given as

1 1 0.66 0.30 0.24

[1 − 𝐴 ]−1 = [𝐴𝑑𝑗 𝐴 ] = [0.34 0.62 0.24]
|1 − 𝐴 | 0.384
0.21 0.27 0.60

The output for the three sectors is given as: 𝑋 = [1 − 𝐴 ]−1 𝐷

1 0.66 0.30 0.24 10

= [0.34 0.62 0.24] [ 5 ]
0.384
0.21 0.27 0.60 6

24.84
𝑋 = [20.68]
18.36

X represents the output required in the three sectors to meet the inter-industry input

requirements and the final demand.

Page 136 of 167

10.2.2 Application of matrix algebra in the market equilibrium models

The equilibrium prices and quantities for a market with 3 commodities can be determined using

the matrix method. If the demand and supply functions for the three commodities is given as:

A 3 product market model is given as:

𝑄𝑑1 = 50 − 2𝑃1 + 5𝑃2 − 3𝑃3 𝑄𝑠1 = 8𝑃1 − 5

𝑄𝑑2 = 22 + 7𝑃1 − 2𝑃2 + 5𝑃3 𝑄𝑠2 = 12𝑃2 − 5

𝑄𝑑3 = 17 + 𝑃1 + 5𝑃2 − 3𝑃3 𝑄𝑠3 = 4𝑃3 − 1

Determine the equilibrium prices and quantities for the three products.

At equilibrium:

𝑄𝑑1 = 𝑄𝑠1

𝑄𝑑2 = 𝑄𝑠2

𝑄𝑑3 = 𝑄𝑠3

If we equate the demand and the supply for each of the three products, we get:

50 − 2𝑃1 + 5𝑃2 − 3𝑃3 = 8𝑃1 − 5

22 + 7𝑃1 − 2𝑃2 + 5𝑃3 = 12𝑃2 − 5

17 + 𝑃1 + 5𝑃2 − 3𝑃3 = 4𝑃3 − 1

This gives us 3 simultaneous equations:

Page 137 of 167

 10𝑃1 − 5𝑃2 + 3𝑃3 = 55

−7𝑃1 + 14𝑃2 − 5𝑃3 = 27

−𝑃1 − 5𝑃2 + 7𝑃3 = 18

We can rewrite the equations as:

10 −5 3 𝑃1 55
𝑃
[−7 14 −5] ∗ [ 2 ] = [27]
−1 −5 7 𝑃3 18

55 −5 3
|27 14 −5|
𝑃1 = 18 −5 7 = 7
10 −5 3
|−7 14 −5|
−1 −5 7

10 55 3
|−7 27 −5|
𝑃2 = −1 18 7 = 9
10 −5 3
|−7 14 −5|
−1 −5 7

10 −5 55
|−7 14 27|
𝑃3 = −1 −5 18 = 10
10 −5 3
|−7 14 −5|
−1 −5 7

TASK: Solve for the equilibrium quantities by substituting for the values of the

equilibrium prices into the demand or supply functions

𝑄1 = 51 𝑄2 = 103 𝑄3 = 39

Page 138 of 167

 10.3 Self-Assessment Questions

Given the following information of a 3-sector economy

Input to

Agriculture Manufacturing Services Final Total

demand Output

Output Agriculture 300 450 250 200 1200

From Manufacturing 420 500 280 600 1800

Service 340 0 60 200 600

Other inputs 140 850 10

Total inputs 1200 1800 600

Required:

i) The input-output matrix

ii) The Leontief Matrix

400
iii) The output requirement for each sector given that the final demand changes to 𝐷 = (1200)
400

10.4 E-References
10. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical Economics,
4th edition, McGraw Hill, Singapore
11. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
12. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau

Page 139 of 167

 LECTURE ELEVEN: ECONOMIC APPLICATIONS OF MATRIX ALGEBRA II

11.1 Introduction

In this Lesson, we will discuss how matrix algebra is applied in economics.

11.2 Lesson Learning Outcomes

By the end of the lesson, the learner should be able to.

11.2.1 Compute the equilibrium national income using matrix algebra

11.2.2 Compute general equilibrium in the IS-LM framework using matrix algebra

11.2.1 Application of matrix algebra in the National Income Model

A Simple national income model is given as:

𝑌 = 𝐶 + 𝐼0 + 𝐺0

𝐶 = 𝑐0 + 𝑐1 (𝑌 𝑑 )

𝑌𝑑 = 𝑌 − 𝑇

𝐶 = 𝑐0 + 𝑐1 (𝑌 − 𝑇)

𝑇 = 𝑡0 + 𝑡1 𝑌

𝑌 = 𝑛𝑎𝑡𝑖𝑜𝑛𝑎𝑙 𝑖𝑛𝑐𝑜𝑚𝑒, 𝐶 = 𝑐𝑜𝑛𝑠𝑢𝑚𝑝𝑡𝑖𝑜𝑛 𝑒𝑥𝑝𝑒𝑛𝑑𝑖𝑡𝑢𝑟𝑒,

𝑌 𝑑 = 𝑑𝑖𝑠𝑝𝑜𝑠𝑎𝑏𝑙𝑒 𝑖𝑛𝑐𝑜𝑚𝑒, 𝑇 = 𝑡𝑎𝑥

𝑤ℎ𝑒𝑟𝑒 𝑐0 > 0 0 < 𝑐1 < 1

Page 140 of 167

 𝑡0 > 0 0 < 𝑡1 < 1

i) Write this model in three simultaneous equations representing the three endogenous

variables.

𝑌 − 𝐶 = 𝐼0 + 𝐺0

−𝑐1 𝑌 + 𝐶 + 𝑐1 𝑇 = 𝑐0

−𝑡1 𝑌 + 𝑇 = 𝑡0

This can also be written as:

1 −1 0 𝑌 𝐼0 + 𝐺0
[−𝑐1 1 𝑐1 ] ∗ [𝐶 ] = [ 𝑐0 ]
−𝑡1 0 0 𝑇 𝑡0

ii) Use Cramer’s rule to determine the expressions for the equilibrium values of the

endogenous variables (𝑌, 𝐶 and 𝑇)..

𝐼0 + 𝐺0 −1 0
| 𝑐0 1 𝑐1 |
𝑡0 0 0
𝑌=
1 −1 0
|−𝑐1 1 𝑐1 |
−𝑡1 0 0

𝑐0 + 𝐼0 + 𝐺0 − 𝑐1 𝑡0
𝑌=
(1 − 𝑐1 + 𝑐1 𝑡1 )

1 𝐼0 + 𝐺0 0
|−𝑐1 𝑐0 𝑐1 |
−𝑡1 𝑡0 0
𝐶=
1 −1 0
| 1 1 𝑐1 |
−𝑐
−𝑡1 0 0

Page 141 of 167

 𝑐0 + 𝑐1 (1 − 𝑡1 )( 𝐼0 + 𝐺0 ) − 𝑐1 𝑡0
𝐶=
(1 − 𝑐1 + 𝑐1 𝑡1 )

1 −1 𝐼0 + 𝐺0
|−𝑐1 1 𝑐0 |
−𝑡1 0 𝑡0
𝑇=
1 −1 0
| 1 1
−𝑐 𝑐1 |
−𝑡1 0 0

𝑡0 (1 − 𝑐1 )𝑐0 + 𝑡1 (𝑐0 + 𝐼0 + 𝐺0 )
𝑇=
(1 − 𝑐1 + 𝑐1 𝑡1 )

Example 1
Find the equilibrium income, consumption and tax given the following national income model:
𝑌 =𝐶+𝐼+𝐺

𝐶 = 100 + 0.8𝑌 𝑑
𝑇 = 10 + 0.1𝑌
𝐼 = 50
𝐺 = 30
Step 1: Rewrite the equations putting the endogenous variables on the L.H.S and the exogenous variables
on the R.H.S
𝑌 =𝐶+𝐼+𝐺
𝑌 − 𝐶 = 𝐼 + 𝐺 (Substitute for I and G)

𝑌 − 𝐶 = 50 + 30
𝑌 − 𝐶 = 80…………………………………………………………………. (1)

𝐶 = 100 + 0.8𝑌 𝑑

𝑌𝑑 = 𝑌 − 𝑇
𝐶 = 100 + 0.8(𝑌 − 𝑇)
𝐶 = 100 + 0.8𝑌 − 0.8𝑇)
−0.8𝑌 + 𝐶 + 0.8𝑇 = 100……………………………………………………. (2)

𝑇 = 10 + 0.1𝑌

Page 142 of 167

−0.1𝑌 + 𝑇 = 10……………………………………………………………… (3)

Presenting (1), (2) and (3) in matrix form we have:

𝑌 − 𝐶 = 80
−0.8𝑌 + 𝐶 + 0.8𝑇 = 100
−0.1𝑌 + 𝑇 = 10
1 −1 0 𝑌 80
⌊−0.8 1 0.8⌋ ⌊𝐶 ⌋ = ⌊100⌋
−0.1 0 1 𝑇 10
Using Cramer’s Rule:
80 −1 0
|100 1 0.8|
𝑌 = 10 0 1 = 172 = 614.29
1 −1 0 0.28
|−0.8 1 0.8|
−0.1 0 1
1 80 0
|−0.8 100 0.8|
𝐶 = −0.1 10 1 = 149.6 = 534.29
1 −1 0 0.28
|−0.8 1 0.8|
−0.1 0 1

1 −1 80
|−0.8 1 100|
𝑇 = −0.1 0 10 = 20 = 71.43
1 −1 0 0.28
|−0.8 1 0.8|
−0.1 0 1

Example 2
You are given the following national income model:
𝑌 =𝐶+𝐼+𝐺
𝐶 = 120 + 0.8𝑌
𝐼 = 100 + 0.1𝑌

Page 143 of 167

 𝐺 = 300
Required:
i) Present this information in matrix format:
ii) Using Cramer’s rule, find the equilibrium values of the endogenous variables

Solution:
𝑌 − 𝐶 − 𝐼 = 300………………………………………………………………… (1)
−0.8𝑌 − 𝐶 = 120……………………………………………………………….. (2)
0.1𝑌 + 𝐼 = 100………………………………………………………………….. (3)

1, 2 and 3 in matrix form:

1 −1 −1 𝑌 300
⌊−0.8 1 0 ⌋ ⌊𝐶 ⌋ = ⌊120⌋
−0.1 0 1 𝐼 100

Using Cramer’s Rule:

300 −1 −1
|120 1 0|
𝑌= 100 0 1 = 520 = 5,200
1 −1 −1 0.1
|−0.8 1 0|
−0.1 0 1

1 300 −1
|−0.8 120 0 |
𝐶 = −0.1 100 1 = 428 = 4,280
1 −1 −1 0.1
|−0.8 1 0|
−0.1 0 1

1 −1 300
|−0.8 1 120|
62
𝐼 = −0.1 0 100 = = 620
1 −1 −1 0.1
|−0.8 1 0|
−0.1 0 1

Page 144 of 167

11.2.2 Application of matrix algebra in computing equilibrium in the IS-LM framework

Equilibrium in the goods market: Represented by the IS Curve / Equation

Given:
𝑌 =𝐶+𝐼
𝐶 = 250 + 0.2𝑌
𝐼 = 10 − 0.4𝑟
𝑌 = 250 + 0.2𝑌 + 10 − 0.4𝑟
0.8𝑌 + 0.4𝑟 = 250…………………………………………….. IS Equation

Equilibrium in the money market: Represented by the LM Curve / Equation

𝑀𝑑 = 120 + 0.1𝑌 − 0.3𝑟

𝑀 𝑠 = 140
At equilibrium: 𝑀𝑑 = 𝑀 𝑠

120 + 0.1𝑌 − 0.3𝑟 = 140

0.1𝑌 − 0.3𝑟 = 20………………………………………………………. LM Equation

The general equilibrium

It requires simultaneous equilibrium in both the goods and money markets. From the equations derived
above:

0.8𝑌 + 0.4𝑟 = 250

0.1𝑌 − 0.3𝑟 = 20
Presenting the two equations in matrix format:
0.8 0.4 𝑌 250
[ ]⌈ ⌉ = ⌈ ⌉
0.1 −0.3 𝑟 20

Using Cramer’s Rule:

250 0.4
| | −83
𝑌 = 20 −0.3 = = 296.43
0.8 0.4
| | −0.28
0.1 −0.3

Page 145 of 167

 0.8 250
| | −9
𝑟 = 0.1 20 = = 32.14
0.8 0.4
| | −0.28
0.1 −0.3

11.3 Self-Assessment Questions

a) Given the following national income model, compute the equilibrium values of the endogenous
variables using matrix inversion and Cramer’s rule:
𝑌 =𝐶+𝐼+𝐺

𝐶 = 180 + 0.7𝑌 𝑑
𝑇 = 70
𝐼 = 60 + 0.1𝑌
𝐺 = 80
b) Determine how matrix algebra can be applied in determining the equilibrium values of interest

rates and national income in the macro economy using the following IS and LM functions.

IS equation: 𝑟 = 8 − 0.006𝑦

LM equation: 𝑟 = −3 + 0.003𝑦

11.4 E-References
13. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical Economics,
4th edition, McGraw Hill, Singapore
14. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
15. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau

Page 146 of 167

 LECTURE TWELVE: INTEGRATION

12.1 Introduction

In this lesson we shall learn the different rules that are used to integrate functions

12.2 Lesson Learning Outcomes

By the end of the lesson, the learner should be able to.

12.2.1 Understand the meaning of integration

12.2.2 Understand the rules of integration

12.2.3 Understand the meaning and computation of definite integrals

12.2.1 The meaning of integration

Integration is simply the reverse of differentiation. If we differentiate a function, we get the slope

of the function. On the other hand, if we have the slope, and we want to get back to the original

function, we use the process of integration.

For example:

If 𝑌 = 𝑋 𝑛

𝜕𝑌
= 𝑛𝑋 𝑛−1
𝑑𝑋

To get back to the original function, we do the reverse of what we have done.

i.e. add 1 to the power and divide by the new power.

Page 147 of 167

 𝑛𝑋 𝑛−1+1
𝑌=
𝑛

Integration can also be interpreted as the area below a curve between specified limits.

Consider the following two functions: 𝐹(𝑥), a primary function and 𝑓(𝑥), the derived

function i.e. the derivative of 𝐹(𝑥).

In order to obtain the primary function 𝐹(𝑥)from the derived function 𝑓(𝑥), we have to

integrate the derived function 𝑓(𝑥).

𝑑
∫ 𝑓(𝑥)𝑑𝑥 = ∫ 𝐹(𝑥)𝑑𝑥 = 𝐹(𝑥)
𝑑𝑥

We will therefore define the term integration as the process of obtaining a primary function from

the corresponding derived function.

Constant of integration

Consider the primary function:

𝐹(𝑥) = 𝑥 2

The derivative of this function with respect to x is:

𝑓(𝑥) = 2𝑥

If we integrate the derived function, we obtain the primary function:

∫ 𝑓(𝑥)𝑑𝑥 = ∫ 2𝑥𝑑𝑥 = 𝑥 2

It is important to note that, we have other functions whose derivative is 2𝑥 as shown below.

𝐹(𝑥) = 𝑥 2 + 13

𝐹(𝑥) = 𝑥 2 + 9

𝐹(𝑥) = 𝑥 2 + 40

among others.

Page 148 of 167

Due to this fact, whenever integration is performed, it is important to add an arbitrary constant to

the solution known as the constant of integration.

∫ 𝑓(𝑥)𝑑𝑥 = 𝐹(𝑥) + 𝑐

Where: 𝑐 is the constant of integration.

Examples

Compute the following:

i) ∫ 2𝑥 𝑑𝑥

ii) ∫ 3𝑥 2 𝑑𝑥

Solution

i) ∫ 2𝑥 𝑑𝑥 = 𝑥 2 + 𝑐

ii) ∫ 3𝑥 2 𝑑𝑥 = 𝑥 3 + 𝑐

Initial and boundary conditions

In some instances, it may be necessary to determine the specific numerical values of the constant

of integration. This is possible when the problem being discussed provides initial conditions e.g.

𝒚 = 𝟕𝟎, 𝑤ℎ𝑒𝑛 𝒙 = 𝟎 or boundary conditions e.g. 𝒚 = 𝟔𝟗, 𝑤ℎ𝑒𝑛 𝒙 = 𝟓.

Example 1

Compute the following given the boundary condition stated in the problem:

𝑦 = ∫ 2𝑥 𝑑𝑥 𝑦 = 125 𝑤ℎ𝑒𝑛 𝑥 = 5

Solution

𝑦 = ∫ 2𝑥 𝑑𝑥 = 𝑥 2 + 𝑐

Page 149 of 167

 𝑦 = 125 𝑤ℎ𝑒𝑛 𝑥 = 5

125 = 52 + 𝑐

𝑐 = 125 − 25 = 100

𝑦 = 𝑥 2 + 100

Example 2

Compute the following given the initial condition stated in the problem:

𝑦 = ∫ 3𝑥 2 𝑥 = 0 𝑤ℎ𝑒𝑛 𝑦 = 77

Solution

𝑦 = ∫ 3𝑥 2 = 𝑥 3 + 𝑐

77 = 03 + 𝑐

𝑐 = 77

𝑦 = 𝑥 3 + 77

12.2.2 Rules of integration

a) Constant function Rule

∫ 𝐾. 𝑑𝑥 = 𝐾𝑥 + 𝐶

The constant rule can be understood further by using the power multiplied by a constant rule.

The rule can be written as:

𝐾𝑥 0+1
∫ 𝐾𝑥 0 . 𝑑𝑥 = 1
+ 𝐶 = 𝐾𝑥 + 𝐶

The constant 𝐶 is added because it is usually lost in the process of differentiating. It can take

many values.

Examples

i) ∫ 10. 𝑑𝑥 = 10𝑥 + 𝑐

Page 150 of 167

 ii) ∫ 99. 𝑑𝑥 = 99𝑥 + 𝑐

b) Power Rule

𝑛
𝑥 𝑛+1
∫ 𝑥 . 𝑑𝑥 = +𝑐
𝑛+1

Examples

𝑥 4+1 𝑥5
i) ∫ 𝑥 4 . 𝑑𝑥 = 5
+𝑐 =5 +𝑐

3 7
4
ii) ∫ 𝑥 4 . 𝑑𝑥 = 7 𝑥 4 + 𝑐

c) Power multiplied by constant rule

𝑛
𝐴𝑥 𝑛+1
∫ 𝐴𝑥 . 𝑑𝑥 = +𝑐
𝑛+1

Examples

5𝑥 4
i) ∫ 5𝑥 3 . 𝑑𝑥 = 5 ∫ 𝑥 3 . 𝑑𝑥 = 4
+𝑐

ii) ∫ 10𝑥 4 . 𝑑𝑥 = 10 ∫ 𝑥 4 . 𝑑𝑥 = 2𝑥 5 + 𝑐

d) Sum/difference rule

The integral of a sum/difference of two functions

∫(𝑔(𝑥) ± ℎ(𝑥)) 𝑑𝑥 = ∫(𝑔(𝑥))𝑑𝑥 ± ∫ ℎ(𝑥)𝑑𝑥 + 𝑐

Example

20𝑥 2 3𝑥 3
∫(20𝑥 + 3𝑥 2 ) 𝑑𝑥 = ∫(20𝑥)𝑑𝑥 + ∫ 3𝑥 2 𝑑𝑥 + 𝑐 = + + 𝑐 = 10𝑥 2 + 𝑥 3 + 𝑐
2 3

e) Integration of linear function raised to a power

1 (𝑚𝑥+𝑘)𝑛+1
∫(𝑚𝑥 + 𝑘)𝑛 . 𝑑𝑥 = 𝑚 𝑛+1
+𝐶

Example

1 (2𝑥+10)4
∫(2𝑥 + 10)3 . 𝑑𝑥 = 2 4
+𝐶

Page 151 of 167

f) Exponential function Rule

i) 𝑦 = 𝑒𝑥

𝑑𝑦
= 𝑒𝑥
𝑑𝑥

∫ 𝑒 𝑥 . 𝑑𝑥 = 𝑒 𝑥 + 𝑐

ii) 𝑦 = 𝑒 𝑘𝑥

𝑑𝑦
= 𝑘𝑒 𝑘𝑥
𝑑𝑥

1
∫ 𝑒 𝑘𝑥 . 𝑑𝑥 = 𝑘 𝑒 𝑘𝑥 + 𝑐

In general:

∫ 𝑓 ′ (𝑥)𝑒 𝑓(𝑥) 𝑑𝑥 = 𝑒 𝑓(𝑥) + 𝑐

Examples

Evaluate:

i) ∫ 2𝑒 𝑥 . 𝑑𝑥

∫ 2𝑒 𝑥 . 𝑑𝑥 = 2 ∫ 𝑒 𝑥 . 𝑑𝑥 = 2𝑒 𝑥 + 𝑐

ii) ∫ 𝑒 7𝑥 . 𝑑𝑥
1
∫ 𝑒 7𝑥 . 𝑑𝑥 = 7 𝑒 7𝑥 + 𝑐

1
𝑥
iii) ∫ 𝑒 2 . 𝑑𝑥
1 1 1
𝑥 1
∫ 𝑒 2 . 𝑑𝑥 = 1 𝑒 2𝑥 + 𝑐 = 2𝑒 2𝑥 + 𝑐
2

Page 152 of 167

g) Logarithm rule
𝑑 1
If 𝑦 = 𝐿𝑛𝑥 𝑡ℎ𝑒𝑛: 𝐿𝑛(𝑥) = 𝑥
𝑑𝑥

This means:
1
∫ 𝑥 𝑑𝑥 = 𝐿𝑛𝑥 + 𝑐

𝑓 ′ (𝑥)
∫ 𝑑𝑥 = 𝐿𝑛 𝑓(𝑥) + 𝑐
𝑓(𝑥)

Examples
4
i) ∫ 𝑥 . 𝑑𝑥

4 1
∫ . 𝑑𝑥 = 4 ∫ . 𝑑𝑥 = 4 𝑙𝑛𝑥 + 𝑐
𝑥 𝑥

1
ii) ∫ 3𝑥 . 𝑑𝑥

1 1 1 1
∫ . 𝑑𝑥 = ∫ . 𝑑𝑥 = 𝑙𝑛𝑥 + 𝑐
3𝑥 3 𝑥 3

3𝑥 2 +3
iii) ∫ 𝑥 3 +3𝑥+4 . 𝑑𝑥

3𝑥 2 + 3
∫ . 𝑑𝑥 = ln(𝑥 3 + 3𝑥 + 4) + 𝑐
𝑥 3 + 3𝑥 + 4

h) Integration by substitution

Used if appropriate

𝑑𝑣
∫ 𝑓(𝑢). . 𝑑𝑥 = ∫ 𝑓(𝑢). 𝑑𝑣 = 𝑓(𝑢) + 𝑐
𝑑𝑥

Page 153 of 167

Comes from chain rule

𝑑 𝑓(𝑢) 𝑑 𝑑𝑣 𝑑𝑣 𝑑𝑣
= 𝑓(𝑢). = 𝑓 ′ (𝑢). = 𝑓𝑢
𝑑𝑥 𝑑𝑣 𝑑𝑥 𝑑𝑥 𝑑𝑥

Example

i. ∫ 2𝑥. (𝑥 2 + 1)𝑑𝑥 𝑙𝑒𝑡 𝑢 = 𝑥 2 + 1

𝑑𝑢 𝑑𝑢
= 2𝑥 𝑑𝑥 =
𝑑𝑥 2𝑥

𝑑𝑢
∫ 2𝑥. 𝑢 = ∫ 𝑢𝑑𝑢
2𝑥

(𝑥 2 + 1)2
= +𝑐
2

3𝑥 2
ii. ∫ (𝑥 3 +18)7 . 𝑑𝑥

𝑑𝑢
Let 𝑢 = 𝑥 3 + 18 = 3𝑥 2
𝑑𝑥

1 −7
−𝑢−6
∫ 𝑑𝑢 = ∫ 𝑢 𝑑𝑢 = +𝑐
𝑢7 6

Substituting for u

−(𝑥 3 + 18)−6 −1
+𝑐 = +𝑐
6 6(𝑥 + 18)6
3

12.2.3 Definite integral

The integrals we have looked at are called indefinite integrals in that we do not have a definite

numerical value for the integral. i.e.

Page 154 of 167

 ∫ 𝑓(𝑥)𝑑𝑥 = 𝐹(𝑥) + 𝐶

As noted earlier, integration can be defined as the area below a curve between two specified values.

Y=f(x)

0 b X
a

If we choose two values of the independent variable (X), e.g. 𝑎 and 𝑏 where 𝑎 < 𝑏, we can get a

definite value for the integral by substituting the values of 𝑥 in the integral. i.e.

𝐹(𝑥) + 𝐶 = [𝐹(𝑏) + 𝐶] − [𝐹(𝑎) + 𝐶] = 𝐹(𝑏) − 𝐹(𝑎)

This would give us a definite numerical value for the integral by ensuring that the variable 𝑥 and

the constant 𝐶 𝑎𝑟𝑒 not in the final solution.

In the definite integral, 𝑎 is the lower limit and 𝑏 is the upper limit.

If we are interested in getting the definite integral, we can rewrite our integral as:

Page 155 of 167

 𝑏
∫ 𝑓(𝑥)𝑑𝑥 = 𝐹(𝑥)]𝑏𝑎 = 𝐹(𝑏) − 𝐹(𝑎)
𝑎

Example

3
i. ∫1 𝑥 2 . 𝑑𝑥

3 3
2
𝑥3 33 13 26
∫ 𝑥 . 𝑑𝑥 = + 𝑐] = − =
1 3 1
3 3 3

3
ii. ∫1 4𝑋 2 . 𝑑𝑥

3 3
4𝑋 3
2
4(3)3 4(1)3
∫ 4𝑋 𝑑𝑥 = ] =[ − ]=
1 3 1 3 3

4 104
= 36 − =
3 3

3
3𝑥 2
∫ . 𝑑𝑥 = [ln(𝑥 3 + 3) + 𝑐]32 = ln(33 + 3 + 𝑐) − ln(23 + 3 + 𝑐)
2 𝑥3 + 3

ln(30) − ln(11)

Simplify

12.3 Self-Assessment Questions

Evaluate the following:

i) ∫ 8𝑒 2𝑥+3 . dx

5𝑥 4 +3𝑥 2
ii) ∫ 𝑥 5 +𝑥 3 +7 . 𝑑𝑥

1
iii) ∫−1(3𝑥 2 + 4𝑥 + 5). 𝑑𝑥

12.4 E-References

Page 156 of 167

16. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical Economics,
4th edition, McGraw Hill, Singapore
17. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
18. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau

Page 157 of 167

 LECTURE THIRTEEN: ECONOMIC APPLICATIONS OF INTEGRATION

13.1 Introduction

In this lesson we will learn how the concept of integration is used in economics.

13.2 Lesson Learning Outcomes

By the end of the lesson, the learner should be able to.

13.2.1 Understand the computation of the consumer surplus using the concept of integration

13.2.2 Understand the computation of the consumer surplus using the concept of integration

13.2.3 Obtain the cost function from the marginal cost function

13.2.4 Obtain the total revenue function from the marginal revenue function

13.2.5 Obtain the consumption function from the marginal propensity consume

13.2.6 Capital Formation

13.2.1 Consumer surplus

Consumer surplus is the benefit to the consumer because of being able to buy goods at a lower

price than what he/she was willing to pay. It is the difference between what a consumer pays and

what she/he was willing to pay. It is given as the area below the demand curve but above the market

price. If we have a linear demand function, it can be represented as:

Page 158 of 167

 Price

Market price/
equilibrium
price

P=f(Q)

0
Equilibrium Quantity demanded
quantity

The shaded area is the consumer surplus.

Using the concept of integration, consumer surplus is given as:

𝑄𝑒
𝐶𝑜𝑛𝑠𝑢𝑚𝑒𝑟 𝑠𝑢𝑟𝑝𝑙𝑢𝑠 = ∫ 𝑝. 𝑑𝑄 − 𝑃𝑒 𝑄𝑒
0

Example

Assume that a demand function is given as:

𝑃 = 40 − 2𝑄

Determine the consumer surplus when 𝑚𝑎𝑟𝑘𝑒𝑡 𝑝𝑟𝑖𝑐𝑒 = 10

𝑊ℎ𝑒𝑛 𝑃 = 10

10 = 40 − 2𝑄

Page 159 of 167

 2𝑄 = 40 − 10 = 30

𝑄 = 15

15 15
2𝑄 2
𝐶𝑜𝑛𝑠𝑢𝑚𝑒𝑟 𝑠𝑢𝑟𝑝𝑙𝑢𝑠 = ∫ (40 − 2𝑄). 𝑑𝑄 = 40𝑄 − | − 10 ∗ 15
0 2 0

= 40(15) − 152 − 0 − 150

= 600 − 225 − 150

𝐶𝑜𝑛𝑠𝑢𝑚𝑒𝑟 𝑠𝑢𝑟𝑝𝑙𝑢𝑠 = 225

13.2.2 Producer surplus

This is the benefit to the producer because of being able to sell goods at a higher price than what

he/she was willing to sell. It is the difference between what a producer receives and what she/he

was willing to receive. It is given as the area above the supply curve but below the market price.

If we have a supply function, it can be represented graphically as:

Page 160 of 167

 Price

Market price
=equilibrium price
P=f(Qs)

0 Equilibrium Quantity supplied

quantity

The shaded area is the producer surplus. Using the concept of integration, it is calculated

algebraically using the following formula:

𝑄𝑒
𝑝𝑟𝑜𝑑𝑢𝑐𝑒𝑟 𝑠𝑢𝑟𝑝𝑙𝑢𝑠 = 𝑃𝑒 𝑄𝑒 − ∫ 𝑝. 𝑑𝑄
0

Example

A supply function for fruits is given as:

𝑃 = −10 + 𝑄

Determine the producer surplus when the Price =30

Solution

When 𝑃 = 30, 𝑄𝑠 = 30 + 10 = 40

40
𝑝𝑟𝑜𝑑𝑢𝑐𝑒𝑟 𝑠𝑢𝑟𝑝𝑙𝑢𝑠 = 30 ∗ 40 − ∫ (−10 + 𝑄). 𝑑𝑄
0

Page 161 of 167

 40
𝑄2
= (1200) − [−10𝑄 + ]
2 0

40 ∗ 40
= (1200) − [−10(40) + ]−0
2

= 1200 + 400 − 800

𝑃𝑟𝑜𝑑𝑢𝑐𝑒𝑟 𝑠𝑢𝑟𝑝𝑙𝑢𝑠 = 800

13.2.3 Marginal cost and Total cost function

If we have the marginal cost function for a firm, the total cost can be computed by integrating the

marginal cost function.

Example

𝑀𝐶 = 40𝑄 − 6

𝑇𝐶 = ∫ 𝑀𝐶. 𝑑𝑄

𝑇𝐶 = ∫ 40𝑄 − 6. 𝑑𝑄

40𝑄 2
= − 6𝑄 + 𝐶
2

= 20𝑄 2 − 6𝑄 + 𝐶

The arbitrary constant 𝐶 in the total cost function is the fixed cost.

Assume that when 𝑇𝐶 = 5000, 𝑄 = 10

5000 = 20(10 ∗ 10) − 6 ∗ 10 + 𝐶

Page 162 of 167

 𝐶 = 5000 − 2000 + 60 = 3060

𝑇𝐶 = 20𝑄 2 − 6𝑄 + 3060

13.2.4 Marginal revenue and Total revenue function

Given the marginal revenue function as 𝑀𝑅 = 𝑓(𝑄)

The total revenue function can be obtained by integrating the marginal revenue function with

respect to output (Q).

TR=∫ 𝑀𝑅. 𝑑𝑄

Example

The marginal revenue function for a firm is given as:

𝑀𝑅 = 20 − 50𝑄

Determine the total revenue function for the firm

50𝑄 2
TR=∫ 𝑀𝑅. 𝑑𝑄=∫ 20 − 50𝑄𝑑𝑄 = 20𝑄 − = 20𝑄 − 25𝑄 2
2

Note

In a total revenue function, we will not need the integration constant because when output is zero,

there will be no revenue hence the value of the constant will be zero.

Page 163 of 167

13.2.5 Consumption function from the marginal propensity to consume.

Assume that the 𝑀𝑃𝐶 = 0.8

Determine the consumption function

𝑑𝐶
𝑀𝑃𝐶 = = 0.8
𝑑𝑌

𝐶 = ∫ 0.8𝑌 𝑑𝑌 = 0.8𝑌 + 𝐾

In the consumption function, K represents the autonomous consumption. We can get its value by

getting a pair of values for C and Y

For example, assume in our above case that when Y=5 billion, C=4.2 billion

𝐶 = 0.8𝑌 + 𝐾

4.2 = 0.8(5) + 𝐾

𝐾 = 4.2 − 4.0

= 0.2

Hence:

𝐶 = 0.2 + 0.8𝑌

13.2.6 Capital Formation

Capital formation refers to the process of adding to a given capital stock. If we denote capital

stock with 𝑲(𝒕), the rate of capital formation / net investment flow, 𝑰(𝒕) is given by:

𝑑𝐾
= 𝐾 = ̇ 𝐼(𝑡)
𝑑𝑡
Page 164 of 167
Given the rate of capital formation, capital stock or the time path of capital is given by:

𝐾(𝑡) = ∫ 𝐼(𝑡) 𝑑𝑡

Once we know the rate of net investment flow function 𝑰(𝒕), we can apply definite integration to

obtain the amount of capital formation or total net investment over some specified period of

time.

𝑡
𝑪𝒂𝒑𝒊𝒕𝒂𝒍 𝑭𝒐𝒓𝒎𝒂𝒕𝒊𝒐𝒏 = ∫ 𝐼(𝑡)𝑑𝑡 = [𝐾(𝑡)]𝑡0 = 𝐾(𝑡) − 𝐾(0)
0

Example

Find the time path of capital 𝐾(𝑡), given the following rates of net investment flow functions

2
i) 𝐼(𝑡) = 15𝑡 3

2 3 5 5
𝐾(𝑡) = ∫ 𝐼(𝑡) 𝑑𝑡 = ∫ 15𝑡 3 𝑑𝑡 = ∗ 15 𝑡 3 + 𝑐 = 9𝑡 3 + 𝑐
5

Find the amount of capital formation (total net investment) over the period [2,6] given the

following rates of net investment flows:

1
i) 𝐼(𝑡) = 60𝑡 3
1 4
𝐶𝑎𝑝𝑖𝑡𝑎𝑙 𝑓𝑜𝑟𝑚𝑎𝑡𝑖𝑜𝑛 = ∫ 𝐼(𝑡)𝑑𝑡 = ∫ 60𝑡 3 𝑑𝑡 = [45𝑡 3 + 𝑐]62 = 490.61 − 113.39

= 377.22

Find the time path of capital 𝐾(𝑡), given the rate of capital formation and initial

conditions

Page 165 of 167

 1
𝐼(𝑡) = 9𝑡 2 𝑐𝑎𝑝𝑖𝑡𝑎𝑙 𝑠𝑡𝑜𝑐𝑘 𝑎𝑡 𝑡 = 0 𝑖𝑠 𝐾(0) = 20

1 3
𝐾(𝑡) = ∫ 9𝑡 2 𝑑𝑡 = 6𝑡 2 + 𝑐

𝑐𝑎𝑝𝑖𝑡𝑎𝑙 𝑠𝑡𝑜𝑐𝑘 𝑎𝑡 𝑡 = 0 𝑖𝑠 𝐾(0) = 20

3
20 = 6(0)2 + 𝑐

Thus 𝑐 = 20
3
Hence 𝐾(𝑡) = 6𝑡 2 + 20

13.3 Self-Assessment Questions

a) Find the producer’s surplus given the following supply functions:

i. 𝑃 = 3 + 2𝑄; 𝑃𝑒 = 19

ii. 𝑃 = 3 + 𝑄2; 𝑃𝑒 = 19

b) Find the consumer’s surplus given the following demand functions:

𝑃 = 50 − 0.5𝑄; 𝑃𝑒 = 30

𝑃 = 13 − 𝑄 2; 𝑃𝑒 = 4

c) Assuming that the rate of net investment flow is given by the following function and that

𝐾(0) = 30:

2
𝐼(𝑡) = 24𝑡 5

i) Find the time path of capital stock 𝐾(𝑡)

ii) Find the amount of capital formation over the interval [1,4]

Page 166 of 167

 13.4 E-References

19. Chiang Alpha, Wainwright K. (2005), Fundamental methods of Mathematical Economics,

4th edition, McGraw Hill, Singapore
20. Holden K., Pearson A.W. (1985), Introductory Mathematics for Economists, Macmillan
press, London
21. Mukras M.S. (1986), Elements of Mathematical Economics, Kenya Literature bureau

THE END

Page 167 of 167