ADVANCED

MATHEMATICAL
METHODS
FOR ECONOMICS
B.A. (HONS.) ECONOMICS
SEMESTER-III
DSC-9

DEPARTMENT OF DISTANCE AND CONTINUING EDUCATION

UNIVERSITY OF DELHI
Advanced Mathematical Methods for Economics

SYLLABUS
Advanced Mathematical Methods for Economics
Syllabus Mapping

Unit I: Multivariate Optimization with Constraints

Constrained optimisation with equality and inequality constraints: geometric Lesson 1: Constrained
characterisation, Lagrange characterisation using calculus and applications; Optimisation (Pages 3–35)
properties of value function: envelope theorem, applications.

Unit II: Linear Programming

Introduction, graphical solution, matrix formulation, duality, economic Lesson 2: Linear Programming
interpretation. (Pages 39-72)

Unit III: Integration, Differential Equations, and Difference Equations

Definite integrals, indefinite integrals and economic applications; first order Lesson 3: Definite Integral
and second order difference (Pages 75–99)
Lesson 4: Integration
(Pages 101–127)
Lesson 5: Differential Equations
(Pages 129–154)
Lesson 6: Difference Equation
(Pages 155–181)

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Advanced Mathematical Methods for Economics

CONTENTS
UNIT I: MULTIVARIATE OPTIMIZATION
WITH CONSTRAINTS

Lesson 1 Constrained Optimisation 3–35

1.1 Introduction
1.2 Two Variables, One Equality Constraint
1.3 The Lagrangian Multiplier Method
1.3.1 Economic Interpretations of the Lagrange Multiplier
1.3.2 A Case of Cost Minimization
1.4 Sufficient Conditions
1.4.1 Global Sufficiency Theorem
1.4.2 Local Sufficient Conditions
1.5 More General Lagrangian Problem
1.5.1 A Case of Utility Maximization
1.5.2 A General Case
1.6 Envelope Theorem
1.7 Inequality Constrained Optimization
1.8 Summary
1.9 Glossary
1.10 Answer to In-Text Questions
1.11 Self-Assessment Questions
1.12 References

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

UNIT II: LINEAR PROGRAMMING

Lesson 2 Linear Programming 39–72

2.1 Learning Objective
2.2 Introduction
2.3 Basic Terminology
2.4 Assumptions of the LPP model
2.5 Formulation of LP Models
2.6 Graphical Solutions under Linear Programming
2.7 Duality
2.8 Summary
2.9 Glossary
2.10 Answer to In-Text Questions
2.11 Self-Assessment Questions
2.12 Suggested Readings

UNIT III: INTEGRATION, DIFFERENTIAL EQUATIONS,

AND DIFFERENCE EQUATIONS

Lesson 3 Definite Integral 75–99

3.1 Introduction
3.2 The Fundamental Theorem of Calculus
3.3 Properties of Definite Integral
3.4 Definite integral
3.5 Economic Application
3.6 Marginal Function to Total Function
3.7 Consumer Surplus and Producer Surplus

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Advanced Mathematical Methods for Economics

3.8 Summary
3.9 Glossary
3.10 Answer to In-Text Questions
3.11 Self-Assessment Questions
3.12 Suggested Readings
Lesson 4 Integration 101–127
4.1 Learning Objectives
4.2 Introduction
4.3 Area under Curves
4.4 Integration as an Inverse Process of Differentiation
4.5 Integration by substitution
4.6 Integration by partial fraction
4.7 Marginal Propensity to Save
4.8 Summary
4.9 Glossary
4.10 Answer to In-Text Questions
4.11 Self-Assessment Questions
4.12 Suggested Readings
Lesson 5 Differential Equations 129–154
5.1 Learning Objectives
5.2 Introduction
5.3 Order and Degree of a Differential Equation
5.4 Linear and Non-Linear Differential Equations
5.5 Homogeneous differential equations
5.6 Non -Homogeneous differential equations
5.7 Variable Separation Method
5.8 General and Particular Solutions

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

5.9 Stability and Phase

5.10 Second order differential equation
5.11 Summary
5.12 Glossary
5.13 Answers to In-Text Questions
5.14 Self-Assessment Questions
5.15 Suggested Readings
Lesson 6 Difference Equation 155–181
6.1 Learning Objectives
6.2 Introduction
6.3 Degree and Order of the difference equation
6.4 Solving First Order Difference Equations
6.5 Equilibrium or Stationary Value
6.6 Behaviour of Solutions of First Order Equations
6.7 Economic application of first Order Difference Equations
6.8 Second Order Difference Equations
6.9 Summary
6.10 Glossary
6.11 Answer of In-Text Questions
6.12 Self-Assessment Questions
6.13 Suggested Readings

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
UNIT I: MULTIVARIATE OPTIMIZATION WITH
CONSTRAINTS

LESSON 1 CONSTRAINED OPTIMISATION

Constrained Optimisation

LESSON 1 NOTES

CONSTRAINED OPTIMISATION

Swarup Santra
Assistant Professor,
Department of Economics,
Satyawati College, University of Delhi)

Structure
1.1 Introduction
1.2 Two Variables, One Equality Constraint
1.3 The Lagrangian Multiplier Method
1.3.1 Economic Interpretations of the Lagrange Multiplier
1.3.2 A Case of Cost Minimization
1.4 Sufficient Conditions
1.4.1 Global Sufficiency Theorem
1.4.2 Local Sufficient Conditions
1.5 More General Lagrangian Problem
1.5.1 A Case of Utility Maximization
1.5.2 A General Case
1.6 Envelope Theorem
1.7 Inequality Constrained Optimization
1.8 Summary
1.9 Glossary
1.10 Answer to In-Text Questions
1.11 Self-Assessment Questions
1.12 References

1.1 INTRODUCTION

When maximizing or minimizing a function, we are free to consider any values of an

x-variable and y-variable on the real line as a possible solution then the problem is said
to be unconstrained. However, in economic optimization problems, the variables
involved are almost always required to satisfy certain constraints that restrict the values
of x and y variables to find the possible solutions. For instance, prices and quantities Self-Instructional
Material 3
are often non-negative by definition, and scarcities dictate that consumption quantities

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES are usually bounded from above. In addition, production quotas, budget limitations
and other constraints might restrict the range of choice.
This lesson starts by considering the problem of maximizing or minimizing a
function whose variables are restricted to satisfy one or more equality constraints. A
typical economic example concerns a consumer who chooses how much of the available
income m to spend on a good x whose price is p and how much income to leave over
for expenditure y on other goods. Note that the consumer then faces the budget
constraint px+ y = m. Suppose that preferences are represented by the utility function
u(x. y). In mathematical terms therefore, the consumer faces the problem of choosing
(x, y) in order to maximize u(x, y) subject to px+y=m, a typical constrained maximization
problem. In this case, because y = m – px, the same problem can be expressed as the
unconstrained maximization of the function f(x) = u (x, (m – px)) with respect to the
single variable x. Indeed, this method reduces a constrained optimization problem to
an unconstrained one.
Consider more generally a consumer who is faced with the problem of deciding
how much to buy of n different commodities in a certain period. Denote the utility
function by U (x1, x2, ... ..., xn). Let the price per unit of commodity i be fixed and
equal to pi so that (p1x1 + ... ... ... + pnxn), is the amount required to buy the commodity
vector (x1, x2, ... ..., xn). Assume moreover that the consumer intends to spend an
amount m on the ‘n’ number of commodities. Then it is possible to purchase any
commodity vector (x1, x2, ... ..., xn) that satisfies the budget constraint, p1x1 + ... ... ...
+ pnxn= m. The problem facing the consumer is among all commodity vectors (x1, x2,
... ..., xn) that satisfy the budget constraint, find one which maximizes utility.
Briefly formulated, the problem is
max 𝑈 (𝑥1 , 𝑥2 , … … , 𝑥𝑛 ) … … … … … (1.1)
𝑥 1 ….𝑥 𝑛

Subject to p1x1 + ... ... ... + pnxn= m

It is also tacitly assumed that x1 > 0...., xn > 0. Again, because we can solve the
budget constraint for xn, say, in terms of , x1, ... , xn–1, problem (1) can also be
expressed as an unconstrained maximization problem.
When the constraint is a complicated function, or when there are several equality
constraints to consider, this substitution method might be difficult or impossible to
carry out in practice. In such cases, other techniques should be used. In particular,
Self-Instructional economists make much use of the method of Lagranian multipliers. Actually, the same
4 Material
method is sometimes used by economists even for problems that are quite easy to

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

express as unconstrained problems. The reason is that Lagrange multipliers have NOTES
important economic interpretations.

1.2 TWO VARIABLES, ONE EQUALITY CONSTRAINT

In this section, we will introduce the two variables optimization problem with one
equality constraint. Suppose that we want to maximize (or minimize) a function,
f = (x, y), where the function is strictly concave function and x and y are the two
variables. In absence of any constraint (i.e., unconstraint), the maximization or
minimization problem, the solution (x*, y*) satisfies the conditions, f x( x, y )  0 and
f y ( x, y )  0.
However, in a constraint optimization problem, the solution set is also needed
to satisfy a constraint. Consider an equality constraint, g(x, y) = c. In case we want to
maximize the function f (x, y), the problem can be represented as:
𝑀𝑎𝑥 𝑓(𝑥, 𝑦)
… … … … … … … … … … (1.2)
𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝑔(𝑥, 𝑦) = 𝑐
The above problem can be shown geometrically as in Figure 1.1 below. The
function f (x, y) is illustrated in three dimensions. The graph of the function f is a
surface like a bowl. The equation, g(x, y) = c represents a curve in the xy-plane.

Self-Instructional
Figure 1.1 Material 5

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES The K-curve on the bowl in the above figure 1.1 lies directly above the constraint
curve, g(x, y) = c. At the point A, the function f is maximum. However, this is the
maximum without considering any constraint. If we consider the constraint maximization,
point B is the maximum point on the K-curve. Therefore, point B is the solution of the
maximization problem with constraint. Analytically, now the problem is to find the
coordinates (x, y) of point B.
The same problem can be shown with the level curves for the function, f. The
Figure 1.2 below, shows the level curves of the function f. The point is the maximum
point for the unconstraint maximization.

Figure 1.2

However, in the case of constraint optimization problem, the constraint curve

g(x, y) = c passes through the level curves. Therefore, we can proceed along the
constraint curve and attain higher values of the function f. In such case, point is the
maximum problem. The points Q or P are also on the constraints curve satisfying the
equation g(x, y) = c. However, points Q or P are not the maximum points as the point
lies also on the constraint curve, but, it attains the higher level curve than point Q or P
attain. If one move along the constraint curve g(x, y) = c, then the point B is the only
point where it attains the highest possible level curve. This observation implies that the
slope of the tangent to the curve g(x, y) = c at (x, y) is equal to the slope of the tangent
Self-Instructional to the level curve of f(x, y) at that point.
6 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

The slope of the level curve of f(x, y) is given by: NOTES

𝑑𝑦 𝑓1′ (𝑥, 𝑦)
= − ′ … … … … … … … … . (1.3)
𝑑𝑥 𝑓2 (𝑥, 𝑦)

And the slope of the constraint curve, g(x, y) = c is given by:

𝑑𝑦 𝑔1′ (𝑥, 𝑦)
= − ′ … … … … … … … … . (1.4)
𝑑𝑥 𝑔2 (𝑥, 𝑦)

Since, at the point of tangency, the slope of level curve and constraint curve
must be equal, hence we have:
𝑓1′ (𝑥, 𝑦) 𝑔1′ (𝑥, 𝑦)
= … … … … … … … … . (1.5)
𝑓2′ (𝑥, 𝑦) 𝑔2′ (𝑥, 𝑦)

If the coordinates of the solution point is (x*, y*), then this point satisfies the
condition for maximization as:
𝑓1′ (𝑥 ∗ , 𝑦 ∗ ) 𝑔1′ (𝑥 ∗ , 𝑦 ∗ )
′ (𝑥 ∗ ∗ ) =
𝑓2 ,𝑦 𝑔2′ (𝑥 ∗ , 𝑦 ∗ )

And this solution point (x*, y*) also satisfies the constraint curve as
g(x*, y*) = c. These, two equations together solve the problem and give the solution of
two unknown variables, x and y.
In case of minimization problem,
𝑀𝑖𝑛 𝑓(𝑥, 𝑦)
𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝑔(𝑥, 𝑦) = 𝑐

We can get the same conditions.

These two conditions together provide the necessary condition for (x, y) to
solve the optimization problem.
Example 1.1:Find the only possible solution to the consumer demand problem
max 𝑥 𝛼 𝑦 𝛽 , 𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 (𝑝𝑥 + 𝑦 = 𝑚)

Where α and β are positive constants.

Self-Instructional
Material 7

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Solution:
With 𝑓(𝑥, 𝑦) = 𝑥 𝛼 𝑦 𝛽
And 𝑔(𝑥, 𝑦) = 𝑝𝑥 + 𝑦

We have, 𝑓1′ (𝑥, 𝑦) = 𝛼𝑥 𝛼 −1 𝑦 𝛽

𝑓2′ (𝑥, 𝑦) = 𝛽𝑥 𝛼 𝑦 𝛽 −1
𝑔1′ (𝑥, 𝑦) = 𝑝
𝑔2′ (𝑥, 𝑦) = 1

Therefore, the slope of objective function = slope of constraint.

𝛼𝑥 𝛼 −1 𝑦 𝛽 𝑝
𝛼 𝛽 −1
=
𝛽𝑥 𝑦 1
𝛼𝑦
Or, =𝑝
𝛽𝑥

Solving the latter equation for y gives, 𝑦 = 𝛽 𝛼 𝑝𝑥, which inserted into the
budget constraint, we can get;
𝛼
𝑝𝑥 = 𝑚
𝛼+𝛽

𝛽
𝑦= 𝑚
𝛼+𝛽
This solution is very sensible. It says that the consumer should spend the fraction
𝛼 𝛽
of income on the first good and the fraction on everything else.
𝛼+𝛽 𝛼 +𝛽

IN-TEXT QUESTIONS
Find the possible solutions to the following constrained optimization problems:
1. max 𝑓(𝑥, 𝑦) = 𝑥 + 𝑦; 𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝑔(𝑥, 𝑦) = 𝑥 2 + 𝑦 = 1
2 2
2. min 𝑓(𝑥, 𝑦) = 𝑥 + 𝑦 ; 𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝑥 + 2𝑦 = 4
1 1
3. max 𝑓(𝑥, 𝑦) = 10 𝑥 2 𝑦 3 ; 𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝑔(𝑥, 𝑦) = 2𝑥 + 4𝑦 = 9
Self-Instructional
8 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

NOTES
1.3 THE LAGRANGIAN MULTIPLIER METHOD

We can use the Lagrange Multiplier Method for solving constrained optimization
problems. Let us consider the previous constrained maximization problem as:
𝑀𝑎𝑥 𝑓(𝑥, 𝑦)
𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝑔(𝑥, 𝑦) = 𝑐
The first-order condition (equation 1.5) is:
𝑓1′ (𝑥, 𝑦) 𝑔1′ (𝑥, 𝑦)
′ (𝑥, = ′
𝑓2 𝑦) 𝑔2 (𝑥, 𝑦)

And after rearranging the first-order condition, we get:

𝑓1′ (𝑥, 𝑦) 𝑓2′ (𝑥, 𝑦)
= … … … … … … … . . (1.6)
𝑔1′ (𝑥, 𝑦) 𝑔2′ (𝑥, 𝑦)

If (x0, y0) solves the problem, the left-hand side and the right-hand side are
equal at the point (x0, y0). Let the ratio of the equation 1.6, be , such that;
𝑓1′ (𝑥, 𝑦) 𝑓2′ (𝑥, 𝑦)
= = 𝜆 … … … … … … … … . (1.7)
𝑔1′ (𝑥, 𝑦) 𝑔2′ (𝑥, 𝑦)

The common value  is called a Lagrange Multiplier.

From the equation 1.7, we can get:
𝑓1′ (𝑥, 𝑦) − 𝜆 𝑔1′ (𝑥, 𝑦) = 0
……………………… (1.8)
𝑓2′ (𝑥, 𝑦) − 𝜆 𝑔2′ (𝑥, 𝑦) = 0

The same conditions can be obtained by forming the Lagrangian function ℒ by;
ℒ (𝑥, 𝑦) = 𝑓 (𝑥, 𝑦) − 𝜆 (𝑔(𝑥, 𝑦) − 𝑐) … … … … … … … . (1.9)
Where,  is called the Lagrange multiplier.
The partial derivatives of the Lagrangian function, ℒ (x, y), with respect to x, y,
and  are;
𝜕ℒ(𝑥, 𝑦)
= 𝑓1′ (𝑥, 𝑦) − 𝜆 𝑔1′ (𝑥, 𝑦)
𝜕𝑥
𝜕ℒ(𝑥, 𝑦) Self-Instructional
= 𝑓2′ (𝑥, 𝑦) − 𝜆 𝑔2′ (𝑥, 𝑦) … … … … … … … … … (1.10) Material 9
𝜕𝑦
𝜕ℒ(𝑥, 𝑦)
= 𝑔(𝑥, 𝑦) − 𝑐
𝜕𝜆
© Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics
(𝑥, 𝑦)
= 𝑓2′ (𝑥, 𝑦) − 𝜆 𝑔2′
𝜕𝑦
NOTES
𝜕ℒ(𝑥, 𝑦)
= 𝑔(𝑥, 𝑦) − 𝑐
𝜕𝜆
By equating the first-order partial derivatives equal to zero, we can first-order
conditions are;

𝑓1′ (𝑥, 𝑦) − 𝜆 𝑔1′ (𝑥, 𝑦) = 0

𝑓2′ (𝑥, 𝑦) − 𝜆 𝑔2′ (𝑥, 𝑦) = 0 … … … … … … … . . (1.11)
𝑔(𝑥, 𝑦) − 𝑐 = 0

By rearranging the equations (11), we can get;

𝑓1′ (𝑥, 𝑦) = 𝜆 𝑔1′ (𝑥, 𝑦) … … … … … … … … . (1.12𝑎)

𝑓2′ (𝑥, 𝑦) = 𝜆 𝑔2′ (𝑥, 𝑦) … … … … … … … … … (1.12𝑏)
𝑔(𝑥, 𝑦) = 𝑐 … … … … … … … … … … … … . … . (1.12𝑐)

By rearranging and combining equations (1.12a) and (1.12b), we can get;

𝑓1′ (𝑥, 𝑦) 𝑓2′ (𝑥, 𝑦)
= = 𝜆
𝑔1′ (𝑥, 𝑦) 𝑔2′ (𝑥, 𝑦)

which gives the same result as before; the slope of the objective function and the slope
of constraint function are equal. Also the equation (1.12c) is itself the constraint. One
way of looking the Lagrangian multiplier method is as a way of delivering the tangency
conditions for an optimal Solution.
By using the three equations, 1.12a, 1.12b, 1.12c, we can get the solution for
the three unknowns, x, y, and .
Example 1.2:
Find the solution of the maximization problem, using the Lagrangian method,
max 𝑥𝑦, 𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 (2𝑥 + 𝑦) = 𝑚
Solution: The Lagrangian function is:
ℒ(𝑥, 𝑦) = 𝑥𝑦 − 𝜆 (2𝑥 + 𝑦 − 𝑚)

So, the necessary conditions for the solution of the problem are:
ℒ1′ (𝑥, 𝑦) = 𝑦 − 2𝜆 = 0
Self-Instructional ℒ2′ (𝑥, 𝑦) = 𝑥 − 𝜆 = 0
10 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

And, 2x + y = m NOTES
The first two equations imply that y = 2 and x = ,
Therefore, y = 2x
Now, insert the value of y into the constraint. Then,
2x + 2x = m,
hence, 𝑥 = 𝑚 4 and 𝑦 = 𝑚 2

And,  = 𝑥 = 𝑚 4

1.3.1 Economic Interpretations of the Lagrange Multiplier

Let us consider the maximization problem, again as;

𝑀𝑎𝑥 𝑓(𝑥, 𝑦)
𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝑔(𝑥, 𝑦) = 𝑐

Suppose, x* and y* are the values of x and y that solve this problem. In general,
x* and y* depend on c.
So, we assume that;
𝑥 ∗ = 𝑥 ∗ (𝑐)
… … … … … … … … . . (1.13)
𝑦 ∗ = 𝑦 ∗ (𝑐)

We also assume that these two functions of x* and y* are differentiable of c.

Now, the objective function can be transformed as;
𝑓 ∗ (𝑐) = 𝑓 𝑥 ∗ (𝑐), 𝑦 ∗ (𝑐) … … … … … … … … . (1.14)

Here, f *(c) is called the (optimal) value function for the problem.
When using the Lagrangian multiplier method, the associated value of the
Lagrange multiplier, , also depends on c, such that  = (c); provided that certain
regularity conditions are satisfied.
After differentiating the optimal value function, f*(c) with respect to c, we can have;
𝑑𝑓 ∗ (𝑐)
= 𝜆(𝑐) … … … … … … … … . (1.15)
𝑑𝑐 Self-Instructional
Material 11

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Thus, the Lagrange multiplier that  = (c) is the rate at which the optimal value
of the objective function changes with respect to changes in the constraint constant, c.
Proof of (1.15):
Assume that the function 𝑓 ∗ (𝑐) = 𝑓 𝑥 ∗ (𝑐), 𝑦 ∗ (𝑐) is differentiable.
After taking the differentials of the function;
𝑑 𝑓 ∗ (𝑐) = 𝑑 𝑓(𝑥 ∗ , 𝑦 ∗ ) = 𝑓1′ (𝑥 ∗ , 𝑦 ∗ ) 𝑑𝑥 ∗ + 𝑓2′ (𝑥 ∗ , 𝑦 ∗ ) 𝑑𝑦 ∗ … … … … … … (1.16)

But from the first-order conditions, equations 1.12a and 1.12b; 𝑓1′ (𝑥, 𝑦) = 𝜆 𝑔1′ (𝑥, 𝑦)
and 𝑓2′ (𝑥, 𝑦) = 𝜆 𝑔2′ (𝑥, 𝑦); we can rewrite equation (1.16) as;
𝑑 𝑓 ∗ (𝑐) = 𝜆 𝑔1′ (𝑥 ∗ , 𝑦 ∗ ) 𝑑𝑥 ∗ + 𝜆 𝑔2′ (𝑥 ∗ , 𝑦 ∗ ) 𝑑𝑦 ∗ … … … … … … . (1.17)

Now, take the constraint at the solution as;

𝑔(𝑥 ∗ , 𝑦 ∗ ) = 𝑔 𝑥 ∗ (𝑐), 𝑦 ∗ (𝑐) = 𝑐 … … … … … … (1.18)

By taking the differential of the equation (18), we can get

𝑔1′ (𝑥 ∗ , 𝑦 ∗ ) 𝑑𝑥 ∗ + 𝑔2′ (𝑥 ∗ , 𝑦 ∗ )𝑑𝑦 ∗ = 𝑑𝑐 … … … … … … . . (1.19)

After multiplying both sides of equation (1.19) by , we obtain;

𝜆 𝑔1′ (𝑥 ∗ , 𝑦 ∗ ) 𝑑𝑥 ∗ + 𝜆 𝑔2′ (𝑥 ∗ , 𝑦 ∗ ) 𝑑𝑦 ∗ = 𝜆 𝑑𝑐 … … … … … . . (1.20)

By comparing equations (1.17) and (1.20), we can get;

𝑑 𝑓 ∗ (𝑐) = 𝜆 𝑑𝑐
In particular, if dc is a small change in c, then we can have;
𝑑𝑓 ∗ (𝑐)
= 𝜆
𝑑𝑐

Shadow Price: In economic applications, c often denotes the available stock of some
resource, and f (x, y) denotes the utility function or profit function. Then, (c)dc, for
dc > 0, measures approximately the increase in utility or profit that can be obtained
from dc units more of resources. In economics,  is called a shadow price of the
resource.

Self-Instructional
12 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

1.3.2 A Case of Cost Minimization NOTES

Consider a firm uses inputs K and L of capital and labor, respectively, to produce a
single output Q. Consider the production function as;
1 1
𝑄 = 𝐹(𝐾, 𝐿) = 𝐾 2 𝐿 4 … … … … … … . . (1.21)
Assume that there are perfect completive input markets; so, the prices of capital
(r) and labor (w) are exogenously given.
Hence, the total cost of the firm is;
𝐶 = 𝑟 𝐾 + 𝑤 𝐿 … … … … … … (1.22)
As input prices, r and w are given, then the total cost depends on K and L, only.
Therefore, the cost function will be as:
𝐶(𝐾, 𝐿) = 𝑟 𝐾 + 𝑤 𝐿

Now considering that the firm wants to minimize cost, for given amount of
output, Q .
Hence, we can now present the cost minimization problem of the firm as:
𝑀𝑖𝑛 𝐶(𝐾, 𝐿) = 𝑟 𝐾 + 𝑤 𝐿
1 1 … … … … … … … … … … . . (1.23)
𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝐾 2 𝐿 4 =𝑄

Let’s form the Lagrangian function as;

1 1
ℒ (𝐾, 𝐿) = (𝑟 𝐾 + 𝑤 𝐿) − 𝜆 𝐾 2 𝐿 4 − 𝑄 … … … … . (1.24)

Calculating the partial derivatives and then making them equal to zero, we get;

𝜕ℒ 1 1 1
= 𝑟 − 𝜆 𝐾− 2𝐿 4 = 0
𝜕𝐾 2
𝜕ℒ 1 1 3
= 𝑤 − 𝜆 𝐾 2 𝐿− 4 = 0 … … … … … … … … . . (1.25)
𝜕𝐿 4
𝜕ℒ 1 1
= 𝐾 2 𝐿 4− 𝑄=0
𝜕𝜆

From the first-order conditions we can have;

1 1 1 Self-Instructional
𝑟= 𝜆 𝐾 − 2 𝐿 4 … … … … … … (1.26𝑎) Material 13
2

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 1 1 3
𝑤= 𝜆 𝐾 2 𝐿− 4 … … … … (1.26𝑏)
4
1 1
𝐾 2 𝐿 4 = 𝑄 … … … … … … . . (1.26𝑐)
The equation (1.26c) is the constraint itself.
From the equations (1.26a) and (1.26b), we can obtain;
1 1 1 3
𝜆 = 2𝑟𝐾 2 𝐿− 4 = 4𝑤𝐾 − 2 𝐿 4 … … … … … . . (1.27)

1 1 1 3
Or, 2𝑟𝐾 2 𝐿− 4 = 4𝑤𝐾 − 2𝐿 4

1 1
Now, multiplying both sides by 𝐾 2 𝐿 4, we can get;

2𝑟𝐾 = 4𝑤𝐿
So, 𝐿 = 𝑟 2𝑤 𝐾

Substituting L into the constraints, we can obtain;

1 1 1
𝐾 2 𝑟 4
𝐾 4 = 𝑄
2𝑤
3 1 1 1
𝐾 4 = 2 4 𝑟− 4 𝑤 4 𝑄
1 1 1 4
Hence, 𝐾 ∗ = 2 3 𝑟− 3 𝑤 3 𝑄 3 … … … … … … … . . (1.28)

And so, 𝐿∗ = 𝑟 2𝑤 𝐾 ∗ = 2−2 3 𝑟 2 3 𝑤 −2 3 𝑄4 3 … … … … … … … (1.29)

Therefore, the minimum cost;
2 2 1 4
𝐶 ∗ = 𝑟 𝐾 ∗ + 𝑤 𝐿∗ = 3 2− 3 𝑟 3 𝑤 3 𝑄 3 … … … … … … . (1.30)

Again, by inserting the values of K * and L* into equation (1.27), we can get;
4 2 1 1
𝜆= 2 3 𝑟 3 𝑤 3 𝑄 3

𝜕 𝐶∗
Now, taking partial derivative of C * with respect to r, we get, = 𝐾 ∗;
𝜕𝑟

𝜕 𝐶∗
Then, taking partial derivative of with respect to w, we get, = 𝐿∗;
𝜕𝑤

Self-Instructional
14 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

𝜕 𝐶∗ NOTES
And, taking partial derivative of with respect to Q, we get, 𝜕𝑄
= 𝜆.

IN-TEXT QUESTIONS
4. Solve the following problems by the Lagrangian method.
max(𝑥 2 + 3𝑥𝑦 + 𝑦 2 ), 𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝑥 + 𝑦 = 100
5. Solve the following problem by the Lagrangian method.
max 12𝑥 𝑦 , 𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 3𝑥 + 4𝑦 = 12
1 1
6. Consider the problem max 10 𝑥 2 𝑦 3 ; 𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 (2𝑥 + 4𝑦 = 𝑚). Write down
the necessary conditions and solve it for x, y, and \lambda as function of m.

1.4 SUFFICIENT CONDITIONS

We saw in the previous section that the first-order conditions for a maximum and a
minimum of a constrained problem are identical, as in the unconstrained case, and so
it again becomes necessary to look at the second-order conditions, sufficient conditions.
The Lagrangian multiplier method provides the necessary conditions (first-order) for
the solution of the optimization problem;
𝑀𝑎𝑥 (𝑀𝑖𝑛)𝑓(𝑥, 𝑦)
… … … … … … … . . (1.31)
𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝑔(𝑥, 𝑦) = 𝑐

In order to ascertain that we have really found the solution, more arguments are
needed. Sometimes we can rely on the extreme-value theorem.
If (x0, y0) does solve the optimization problem (1.31), then the Lagrangian multiplier
method implies that the Lagrangian function;
ℒ (𝑥, 𝑦) = 𝑓 (𝑥, 𝑦) − 𝜆 (𝑔(𝑥, 𝑦) − 𝑐) is stationary at (𝑥0 , 𝑦0 ).

But, the Lagrangian function, ℒ (x0, y0), does not necessarily have a maximum (minimum)
at (x0, y0). Suppose, however, that (x0, y0) happens to maximize ℒ (x, y) among all
(x, y0). Then,
Self-Instructional
Material 15

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES ℒ (𝑥0 , 𝑦0 ) = 𝑓 (𝑥0 , 𝑦0 ) − 𝜆 (𝑔(𝑥0 , 𝑦0 ) − 𝑐)

And, hence, ℒ (x0, y0)  ℒ (x, y), for all (x. y).
If (x0, y0) also satisfies the constraint g(x0, y0) = c, then we can write;
f (x0, y0)  f (x, y), for all (x, y), such that g(x, y) = c.
Hence, (x0, y0) really does solve the maximization problem (1.31).
A corresponding result is obtained for the minimization problem in (1.31).
provided that (x0, y0) minimizes ℒ (x, y), among all pairs of (x, y). By combining these
two observations, we can now introduce the global sufficiency theorem.

1.4.1 Global Sufficiency Theorem

One approach to these is global: assumptions are built into economic model to ensure
that the objective function and the constraint function(s) have the right general shape.
It is sufficient for maximum (minimum) that the objective function be quasi-concave
(-convex) and that the constraint functions(s) defines a convex set.
Suppose that the objective function f(x, y) and the constraint function g(x, y) in
optimization problem (1.31) are continuously differentiable functions on an open convex
set A in 2, and let (x0, y0)  A be an interior stationary point for the Lagrangian
function:
ℒ (𝑥, 𝑦) = 𝑓 (𝑥, 𝑦) − 𝜆 (𝑔(𝑥, 𝑦) − 𝑐)

Suppose further that 𝑔(𝑥0 , 𝑦0 ) = 𝑐, Then,

ℒ (𝑥, 𝑦) 𝐶𝑜𝑛𝑐𝑎𝑣𝑒 ⟹ {(𝑥0 , 𝑦0 ) 𝑠𝑜𝑙𝑣𝑒𝑠 𝑡ℎ𝑒 𝑚𝑎𝑥𝑖𝑚𝑖𝑠𝑎𝑡𝑖𝑜𝑛 𝑝𝑟𝑜𝑏𝑙𝑒𝑚;

ℒ (𝑥, 𝑦) 𝐶𝑜𝑛𝑣𝑒𝑥 ⟹ {(𝑥0 , 𝑦0 ) 𝑠𝑜𝑙𝑣𝑒𝑠 𝑡ℎ𝑒 𝑚𝑖𝑛𝑖𝑚𝑖𝑠𝑎𝑡𝑖𝑜𝑛 𝑝𝑟𝑜𝑏𝑙𝑒𝑚.

Note that the Lagrangian function ℒ (x, y) = f (x, y) –  (g(x, y) – c) is concave

if f(x, y) is concave and l (g(x, y)) is convex. The sum of (concave) and (- convex) is
same as sum of concave and concave. Therefore, the sum is the concave function.
For example, the linear function, (rK + wL) is a convex function and the
Cobb-Douglas production function, Q = K1/2 L1/4 is a concave function. Now, as
  0, then,
Self-Instructional
16 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

(–) (K1/2 L1/4 – Q) is a convex function. NOTES

As, the sum of two convex function is also a convex function.
Therefore, [(rK + wL) + (–) (K1/2 L1/4 – Q)] is the sum of two convex functions,
then the Lagrangian function, ℒ (x, y) , is also a convex function.

1.4.2 Local Sufficient Conditions

However, for some purposes, particularly comparative statics, it is useful to have the
second-order conditions in local form in terms of small deviations around the optimal
point. The local second-order conditions in the unconstrained case could be expressed
in terms of the signs of leading principal minors of the Hessian determinant of the
objective function. In case of constrained optimization problem, the local second-
order conditions would be based on the bordered Hessian determinants.
Consider the local constrained maximization (or minimization) problem;
𝑙𝑜𝑐𝑎𝑙 max(𝑚𝑖𝑛) 𝑓(𝑥, 𝑦)
… … … … … … … . . (1.32)
𝑠𝑢𝑏𝑗𝑒𝑐𝑡 𝑡𝑜 𝑔(𝑥, 𝑦) = 𝑐

The first-order condition is essentially as follows:

Consider the objective function, z = f(x, y) and g 2 ( x , y )  0, then the objective
function z = f (x, y) is a function of x, only, because of the presence of the constraint.
By computing dz/dx and taking into account how y depends on x, we obtain a necessary
condition for local extreme points. To find a sufficient condition for local extreme
points, we consider the sign of d2z/dx2.
𝑑𝑧 𝑑𝑦
= 𝑓1′ (𝑥, 𝑦) + 𝑓2′ (𝑥, 𝑦). … … … … … . . (1.33)
𝑑𝑥 𝑑𝑥

If the constraint is g(x, y) = c and, y is a function of x; y = h(x), then

𝑔 𝑥, ℎ (𝑥) = 𝑐

The total differential;

𝑔1′ (𝑥, 𝑦) + 𝑔2′ (𝑥, 𝑦) ℎ′ = 0

Self-Instructional
Material 17

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 𝑔 1′ (𝑥,𝑦 )
Hence, ℎ′ = − 𝑔 2′ (𝑥,𝑦 )
;

Then, we can rewrite the equation (1.33), as;

𝑑𝑧 𝑔1′ (𝑥, 𝑦)
= 𝑓1′ (𝑥, 𝑦) − 𝑓2′ (𝑥, 𝑦) ′ … … … … … . . (1.34)
𝑑𝑥 𝑔2 (𝑥, 𝑦)

2 𝑑𝑧
The derivative 𝑑𝑑 𝑥𝑧2 is just the total derivative of 𝑑𝑥 with respect to x. Assuming
that the functions f and g are continuous functions and twice differentiable; and y is a
function of x.
Then,
𝑑2 𝑧 ′′ ′′ ′ ′′
𝑔′
′′ ′ ) 1 ′
′′
(𝑔11 ′′ ′
+ 𝑔12 𝑦 )𝑔2′ − (𝑔21′′ ′′ ′ )𝑔 ′
+ 𝑔22 𝑦 1
2
= (𝑓11 + 𝑓12 𝑦 ) − (𝑓21 + 𝑓22 𝑦 ′ − 𝑓2 ′ 2
𝑑𝑥 𝑔2 (𝑔2 )

Because the functions f and g are twice continuously differentiable functions,

then
′′ ′′ ′′ ′′
𝑓12 = 𝑓21 and 𝑔12 = 𝑔21 .

′
Moreover, 𝑦 ′ = − 𝑔1 𝑔′
2

And, from the first-order conditions; 𝑓1′ = 𝜆 𝑔1′ and 𝑓2′ = 𝜆 𝑔2′

We have to use these relationships to substitute y and f 2 , and then we obtain,

𝑑2 𝑧 1 ′′ ′′ )(𝑔 ′ )2 ′′ ′′ )𝑔 ′ ′ ′′ ′′ ′ 2
= [(𝑓11 − 𝜆 𝑔11 2 − 2(𝑓12 − 𝜆 𝑔21 1 𝑔2 + (𝑓22 − 𝜆𝑔22 )(𝑔1 ) ] … (1.35)
𝑑 𝑥2 (𝑔2′ )2

However, for the right-hand side of the equation (1.35), to make it easy to remember,
we can write the expression with the square brackets in a symmetric form. Now, let us
form the bordered Hessian determinant as;

0 𝑔1′ (𝑥, 𝑦) 𝑔2′ (𝑥, 𝑦)

′ (𝑥, ′′ ′′ ′′
(𝑓12 ′′ )
𝐷(𝑥, 𝑦) = 𝑔1 𝑦) (𝑓11 − 𝜆 𝑔11 ) − 𝜆𝑔12 … … … … … … . (1.36)
𝑔2′ (𝑥, 𝑦) ′′
(𝑓21 − 𝜆𝑔21 ) ′′ ′′
(𝑓22 − 𝜆𝑔22 ′′
)

Self-Instructional
18 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

Then, NOTES

𝑑2 𝑧 1
= 𝐷(𝑥, 𝑦) … … … … … … … (1.37)
𝑑𝑥 2 (𝑔2′ )2

We have arrived at the following result:

A sufficient condition for (x0, y0) to solve the optimization problem (1.32), is
that (x0, y0) satisfies the first-order conditions and, moreover, that the bordered Hessian,
D(x0, y0) given by equation (1.36) is greater than zero in the maximization case, and
is less than zero in the minimization case.
If D(x0, y0) > 0; then objective function is maximized.
If D(x0, y0) < 0; then objective function is minimized.
The conditions on the sign of the determinant D(x, y) are called the (local)
second-order conditions for problem (1.32).

1.5 MORE GENERAL LAGRANGIAN PROBLEMS

Constrained optimization problems in economics usually involve more than just two
variables, ( x1 ,., xn ) . We begin by considering the problem;

max  min  f  x1 , ., xn 

.. 1.38 
subject to g  x1 ,., xn   c

We can generalize the Lagrangian multiplier method from the previous sections. Let 
be the Lagrangian multiplier. Then we can write the Lagrangian function as;

( x1 , ., xn )  f  x1 , ., xn     g  x1 , ., xn   c  .. 1.39 

Next, compute the partial derivatives of ( x1 ,., xn ) and equate them to
zero, so that

Self-Instructional
Material 19

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES
f1'  x1 ,., xn     g1'  x1 , ., xn   c   0
 .. 1.40 
f n'  x1 , ., xn     g n'  x1 , ., xn   c   0

These ‘n’ number of equations of partial derivatives together with the constraint
form (n + 1) equations. These (n+1) number of equations can determine the (n + 1)
number of unknowns  x1 ,., xn  and .

The relevant bordered Hessian is now the (n+1) X (n+1) determinant,

11 12  1n 1n

21 22  2 n g2
    
D  x1 ,., xn   n1 n 2  nn gn
g1 g 2  g n 0

a) The objective function f  x1 ,., xn  will be maximized subject to

g  x1 ,., xn   c ; is successive principal minors of D  x1 ,., xn 
alternate in sign, with  1 .
n

b) The objective function f  x1 ,., xn  will be minimized subject to

g  x1 ,., xn   c ; if all the principal minors of D  x1 ,., xn  are strictly
negative.

1.5.1 A case of Utility Maximization

Consider the utility maximization problem with budget constraint. The utility maximization
problem can be written as:

max U  x1 , ., xn 
Self-Instructional 1.41
20 Material subject to  p1 x1 ..  pn xn   m

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

Then the Lagrangian function is as: NOTES

  x1 , ., xn   U  x1 , ., xn     p1 x1  ..  pn xn   m  .. 1.42 

Taking partial derivatives of the Lagrangian function with rest to , (i = 1, …, n),

and then equating it with zero, we can obtain;

i'  x1 ,., xn  U i'  x1 ,., xn    pi  i  1, 2,., n 

Writing x   x1 ,., xn  , we can have

U1'  x  U 2'  x  U '  x

    n   1.43
p1 p2 pn

Apart from the last equation, which serves only to determine the Lagrangian
multiplier  , we have (n – 1) equations. (For n = 2, there is one equation; for n = 3.
there are two equations; and so on.) In addition, the constraint must hold. Thus, we
have n equations to determine the values of  x1 ,., xn  .

U
The partial derivative, U i'  x   is called the marginal utility of the ith
xi
commodity. The equation (1.43) say that if x   x1 , ., xn  maximizes the utility
subject to the budget constraint, then the ratio between the marginal utility of a com-
modity and its price per unit must be the same for all the commodities.
Assume that the system consisting of the equations in (1.43) plus the budget
constraint is solved for  x1 ,., xn  as functions of p1 , p2 ,., pn and m. It gives,
The equation (1.44) gives the amount of the ith commodity demanded by the
individual when facing prices ( p1 , p2 ,., pn ) and income m. Therefore, D1, D2, .... ,
Dn, are called the (individual) demand functions.

1.5.2 The General Case

Sometimes, economists need to consider optimization problems with more than one
equality constraint. The corresponding general Lagrangian problem is; Self-Instructional
Material 21

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES max  min  f  x1 ,., xn 

 g1  x1 , ., xn   c1

subject to 
 g  x , ., x   c
 m 1 n m

The Method of Lagrange multipliers can be extended to treat the above

optimization problem. Associate Lagrange multipliers with each of the m constraints,
and define the new Lagrangian function by;
m
 x1 , ., xn   f  x1 , ., xn    j  g j  x1 , ., xn   c j . 1.45 
j 1

The necessary first-order conditions for an optimum are that the partial derivatives
of the Lagrangian function with rest to each vanish, so that,

  f  x1 ,., xn  m  g j  x1 ,., xn 
   j  0 .. 1.46 
xi xi j 1 xi

For all j=1,2,…..,n.

Together with the m equality constraints, these n equations form a total of
(n + m) equations in the (n + m) unknowns  x1 ,., xn  and  1 ,., m  .
Example 1.3:
Solve the problem

 x  2y  z 1
 
max  min  x 2  y 2  z 2 , subject to 
2 x  y  3z  4
Solution:
The Lagrangian function is

 
  x, y , z   x 2  y 2  z 2  1  x  2 y  z  1  2  2 x  y  3 z  4 

The first-order conditions:

Self-Instructional
22 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

 NOTES
 2 x  1  22  0
x


 2 y  21  2  0
y


 2 z  1  32  0
z


 x  2 y  z 1  0
1


 2 x  y  3z  4  0
2

So there are five equations to determine the five unknowns x, y, z, 1 and 2 .

Solving the first three equations in first-order conditions, simultaneously for 1

and 2 gives

2 4
1  x  y,
3 5

4 2
2  x y,
5 5
Inserting these expressions in the third equation, and rearranging them, we can get:
x y z 0
The above equation with the last two equations in first-order conditions form a
system of equation with three unknowns. By using Cramer’s rule, we can solve it as:
16 1 11
x , y  , z 
15 3 15
And the corresponding values of Lagrangian multiplier as:
52 54
1  , 2  Self-Instructional
75 75 Material 23

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES There is a geometric argument that might be introduced here. Each of the two
constraints represents a plane in R3, and the points satisfying both constraints conse-
quently lie on the straight line where the two planes intersect. Now,  x 2  y 2  z 2  is
the square of the distance from the origin to the point (x, y, z). Therefore, our problem
is to find the minimum and the maximum distances from the origin to the points on a
straight line. No maximum distance can possibly exist, but it is geometrically obvious
that there is a minimum distance, and that it must be attained at the point we have
found.
Example 1.4:
Consider the utility maximization problem;

max U  x1 ,., xn 
subject to  p1 x1 ..  pn xn   m

Solution:
Let, U *  p1 , p2 ,., pn , m  denote the maximum utility obtainable when prices are
( p1 , p2 ,., pn ) and the income is m.

This U *  p1 , p2 ,., pn , m  is called the indirect utility function.

U *
We see that,  
m
Thus,  is approximately the increase in maximum utility from increasing income
by one unit. Therefore,  is generally called the marginal utility of income.

Self-Instructional
24 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

NOTES
IN-TEXT QUESTIONS
7. Solve the following problem, assuming that it has a solution:
 x  y  z 1
min( x 2  2 x  2 y 2  z 2  z ), subject to 
2 x  y  z  5

8. Solve the problem:

 x2  2 y 2  z 2  1
max  min  x  y, subject to 
 x  y  z 1
9. Solve the minimization problem as:

 x2  y  z  2
min f  x, y, z    y  z  3 , subject to 
2

x  y  2z  2
2

1.6 ENVELOPE THEOREM

Optimization problems in economics usually involve functions that depend on a number

of parameters, like prices, tax rates, income levels, and so on. Although these parameters
are held constant during optimization, they vary according to the economic situation.
For example, we may calculate a firm’s profit maximum treating the prices it faces as
parameters. But then we want to know how the profit maximum responds to changes
in those prices. So it is important to know what happens to the optimal solution when
the situation changes- that is, if the parameters change. The Lagrange multipliers
considered in the previous section gave some information of this kind.
More generally, consider the problem;

max f  x , r 
x
. 1.47 
subject to g j  x , r   0
Self-Instructional
For all j = 1, ……, m Material 25

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Where, r   r1 ,., rk  is a vector of parameters. Here r is kept constant

during the maximization with respect to x   x1 ,, xn  .
Note that the parameters may appear in the objective as well as in the
constraint functions. The maximum value of f  x, r  obtained in (1.47) will depend
on r and we denote it by f*(r). Assuming that the maximum value exists as;

f *  r   max  f  x, r  : g j  x , r   0, j  1, .., m .. 1.48 

Thus, f*(r) is the maximum value of all numbers f  x, r  as x runs through all
x, where, g j  x, r   0, j  1,.., m . The function f*(r) is called the value function
for problem (1.47). If we let x1*  r  ,.., xn*  r  denote the values of

 x1 ,, xn  for which the maximum value in (1.48) is obtained, then,

   
f *  r   f x*  r  , r  f x1*  r  , x2*  r  ,.., xn*  r  , r1 , r2 , ., rn . 1.49 

First, consider the case in which there are no constraints. When n = l, construction
of the function f*(r) as the ‘envelope’ of all the different f(x, r) functions is indicated in
Figure 1.3, below.

Figure 1.3

Self-Instructional
26 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

After taking partial derivatives of equation (1.49), with respect to rj, we obtain: NOTES

f *  r 


f x*  r  , r  for all  j  1, 2, ., k .. 1.50 
r j rj
Economists call this an envelope theorem. It is a very useful result that should be
studied carefully. Note that if r, is changed, then f*(r) changes for two reasons:

a) First, a change in r j changes the vector r and thus it changes f(x, r) directly..

b) Second, a change in r, changes all the functions, x1*  r  , x2*  r  ,.., xn*  r  ,

and hence, f  x*  r  , r  is changed indirectly..

The result in (1.50) shows that the total effect on the value function of a small
change in r, is found by simply computing the partial derivative of f  x*  r  , r  with
respect to r j , ignoring the indirect effect of the dependence of x*  r  x’ on r altogether..
At first sight, this seems very surprising. On further reflection, however, you
may realize that the first-order conditions for x*(r) to maximize f  x, r  with respect
to x imply that small changes in x induced by a small change in r have negligible effects
on the value of f(x*, r).
Lagrangian Method:
Let form the Lagrangian function as:
m
  x, r   f  x , r   i gi  x, r .. 1.51
j 1

Where, Lagrangian multiplier, i  i  r 

Therefore, the first-order conditions for Lagrangian function as:

f *  r 


 x*  r  , r  for all  j  1, 2, ., k .. 1.52 
r j r j

According to this result, the total effect on the value of f*(r) of a small change in
r j , is found by simply differentiating the Lagrangian   x, r  partially with respect to
Self-Instructional
Material 27

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES r j , treating the x’s as well as the i ‘s as constants. This is the (general) envelope
theorem.
Example 1.5:
Consider the problem
1 1
min C  rk  wL; subject to K L  Q
2 4

Let C* = C (r, w, Q) denote the value function for the problem.

And let Lagrangian function be:

 1 1 
  rK  wL    K 2 L4  Q 
 
The partial derivatives of Lagrangian function,  , with respect to r, w, Q are:

  
K,  L , and 
r w Q
And,

C 8 C 8 C 8
K,  L , and 
r w Q

1.7 INEQUALITY CONSTRAINED OPTIMISATION

So far this lesson has considered how to maximize or minimize a function subject to
equality constraints. The final sections concern nonlinear programming problems, which
involve inequality constraints. Some particularly simple inequality constraints are those
requiring certain variables to be non-negative. These often have to be imposed in
order that the solution should make economic sense. In addition, bounds on the
availability of resources are often expressed as inequalities rather than equalities.
A fairly general nonlinear programming problem is the following:
Self-Instructional
28 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

NOTES
max f  x1 ,.., xn 
 g1  x1 ,.., xn   C1
 1.53
subject to  ..
 g  x ,.., x   C
 m 1 n m

The set of vectors x   x1 ,.., xn  that satisfies all the constraints is called
the constraint set, the admissible set, or the feasible set.
Note that minimizing f  x1 ,.., xn  and maximizing – f  x1 ,.., xn  are
equivalent. Also an inequality constraint of the form g j  x1 ,.., xn   C j can be
rewritten as  g j  x1 ,.., xn   C j .
In principle, such problems can be solved by the classical methods. These involve
examining the stationary pints of f  x1 ,.., xn  in the interior of the feasible set S,
and the behavior of f  x1 ,.., xn  on the boundary of S.
However, since the 1950s, economists have generally tackled such problems
by using an extension of the Lagrangian multiplier method due originally to H. W,
Kuhn and A W. Tucker.
Kuhn-Tucker Conditions:
Consider first the simple nonlinear programming problem:

max f  x, y  ; subject to g  x, y   c . 1.54 

Associate a constant Lagrange multiplier  with the constraint g  x, y   c ,

and define the Lagrangian function as:
  x, y   f  x, y     g  x, y   c . 1.55 

Now, take the partial derivatives with respect to x, y, and equate the partials of   x, y 
to zero:
1'  x, y   f1'  x, y    g1'  x, y   0.. 1.56a 

'2  x, y   f 2'  x, y    g 2'  x, y   0. 1.56b 

Self-Instructional
Material 29

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Now, we introduce the complementary slackness condition;

  0   0 if g  x, y   c . 1.57 

It is also required that the solution (x, y) is needed to satisfy the inequality constraint;

g  x, y   c . 1.58

If we find all the pairs (x, y) (together with suitable values of  that satisfy all
these conditions), then we have all the candidates for the solution of problem (1.54).
Note that the conditions (1.56a) and (1.56b) are exactly the conditions used in
the Lagrangian method. And the condition (1.58) obviously has to be satisfied. So, the
only new feature is condition (1.57).
In fact, condition   0 is rather tricky. It requires that  is non-negative.

a)  = 0 if g  x, y   c .

b) if  > 0, we must have g  x, y   c .

An alternative formulation of this condition is that:

  0,  .  g  x, y   c   0 . 1.59 

The equation (1.59) says that the two inequalities

i)   0 and

ii) g  x, y   c

are complementary inequalities in the sense that at most one can be ‘slack’ that is,
at most one can hold with inequality. Equivalently, at least one must be an equality.
The General Case of Kuhn-Tucker Condition:
Consider the nonlinear programming problem

Max f  x 
. 1.60 
subject to g j  x   C j

Self-Instructional
30 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

NOTES
Where, f  x  and g1 , g 2 ,.., g m are continuously differentiable with f  x 
concave and ( g1 , g 2 ,.., g m ) all convex. Suppose that there exist numbers,
1 , 2 ,, m and a feasible vector x0 such that

 
 f x0 m
   0 i  1, 2,, n 1.61
 g j x0
xi

j 1
j
xi

 j  0 ( 0 if g j  x 0   C j  j  1, 2,.., m . 1.62 

Then, x 0 solves the problem:

max f  x, y   2   x  1  e y
2 2

1.8 SUMMARY

In this lesson, we have introduced the constrained optimization. The economic behavior
of an individual or any economy is restricted or limited by some factors. In such cases,
we have to apply constrained optimization. In this lesson, we have learned equality
type of constraints. The problem can be solved by substituting the value of any variable
from the constraints into the objective function. There is very famous method for this,
the Lagrangian multiplier method. We have form a Lagrangian function by combing the
objective function and the constraint(s). The result is very similar for both the methods.
However, the first-order conditions can say about the extreme point only. But
the solution cannot ensure any maximum or minimum, because the first-order or
necessary conditions for maximization or minimization are same. So, we need second
order conditions. Any concave objective function and convex constraint(s) gave the
maximization problem. Similarly, any convex objective function and concave
constraint(s) gave the minimization problem. For more general rule, we have to calculate
the bordered Hessian determinant from the second-order derivatives of the objective
function. The signs of the principal minors can ensure the maximization of the minimization
Self-Instructional
problem. Material 31

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Next, we have studied the envelope theorem. Then, we have studied the
inequality constrained optimization by introducing the Kuhn-Tucker conditions.

1.9 GLOSSARY

 Constrained optimization: Constrained optimization is the process of optimizing

an objective function with respect to some variables in the presence of constraints
on those variables.
 Lagrangian function: The relationship between the gradient of the function
and gradients of the constraints rather naturally leads to a reformulation of the
original problem, known as the Lagrangian function.
 Lagrangian method: Lagrangian method is a strategy for finding the local
maxima and minima of a function subject to equality constraints (i.e., subject to
the condition that one or more equations have to be satisfied exactly by the
chosen values of the variables). This basic idea is to convert a constrained
problem into a form such that the derivative test of an unconstrained problem
can still be applied.
 Shadow Price: The shadow price is the change per infinitesimal unit of the
constraint. In the optimal value of the objective function of an optimization
problem, it is obtained by relaxing the constraint. If the objective function is
utility, it is the marginal utility of relaxing the constraint. If the objective function
is cost, it is the marginal cost of strengthening the constraint.

1.10 ANSWER TO IN-TEXT QUESTIONS

1. x=1/2 and y=3/4

2. x=4/5 and y=8/5
3. x=27/10 and y=9/10
Self-Instructional
32 Material
4. x=50, y=50, and  =250

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

5. x = 8/3, y =1, and  = 4 NOTES

6. x = 3m/10, y = m/10, and =  = 2.5(10/27m)1/6
7. x = 2, y = -1/6, z = -5/6
8. x = (0 or 4/5), y = (0 or 2/5), and z = (1 or -1/5); (x,y,z) = (0,0,1) or (x,y,z)
= (4/5, 2/5, -1/5)
9. (x,y,z) = (1,1,0) or (x,y,z) = (-1/2, 1, ¾)

1.11 SELF-ASSESSMENT QUESTIONS

1. Find the possible solution of constrained optimization, without using Lagrangian:

max  min  3xy, subject to x 2  y 2  8

max  min  x  y, subject to x 2  3xy  3 y 2  3

2. Consider the problem:

min f  x, y   x 2  y 2 , subject to x  2 y  a
a. Solve the problem by using the constraint to eliminate y.
b. Write down the Lagrangian function for the problem and solve the necessary
conditions.
3. Consider the problem:
max xy , subject to x  y  2
Using the Lagrangian method, prove that (x, y) = (1,1) solves the problem with
 =1.
4. Consider the problem:
1 1
max10 x 4 y 2 , subject to 4 x  2 y  m.

Write down the necessary conditions, and solve them for x, y, and  as function
Self-Instructional
of m. Material 33

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 5. Consider the problem:

max U  x, y   100  e x e  y , subject to px  qy  m.

Write down the necessary conditions for the solution of the problem, and solve
them for x, y, and as function of p, q, and m.
6. Consider the problem:

 
Min x 2  y 2 , subject to  x  2 y  a  .

Now, compute bordered Hessian determinant, D(x,y).

7. What interpretation can be placed on a Lagrangian multiplier at the optimal
solution?
8. Form the bordered Hessian determinant for a problem with two variables and
constraints. What sign of the determinant is sufficient to ensure a maximum? A
minimum? Now generalize for n > 2 variables.
9. What conditions on the objective function and the feasible set are sufficient to
ensure existence of a maximum and minimum solution to a constrained
optimization problem?
10. What condition is sufficient to ensure the existence of the Lagrangian multipliers
for a given constrained optimization problem?
11. What does the envelope theorem tell us about the interpretation of a Lagrangian
multiplier in a constrained optimization problem? Give some examples of such
Lagrange multipliers and their interpretations.
12. State the Kuhn-Tucker condition for both a maximization problem and
minimization problem.
13. Consider the problem:

max f  x, y   2   x  1  e y
2 2

subject to x 2  y 2  a
Where, a is a positive constant.
Self-Instructional Write down the Kuhn-Tucker conditions for the solution of the problem.
34 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Constrained Optimisation

NOTES
1.12 REFERENCES

 Sydsaeter, Knut and Hammond, Peter J. (2002). Mathematics for Economic

Analysis, Pearson Educational.
 Hoy, M., J. Livernois, C. McKenna, R. Rees and T. Stengos (2001).
Mathematics for Economics, Prentice-Hall India.

Self-Instructional
Material 35

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
UNIT II: LINEAR PROGRAMMING

LESSON 2 LINEAR PROGRAMMING

Linear Programming

LESSON 2 NOTES

LINEAR PROGRAMMING

Sarabjeet Kaur
Assistant Professor,
Department of Economics,
Zakir Husain College (E), University of Delhi)
Structure
2.1 Learning Objective
2.2 Introduction
2.3 Basic Terminology
2.4 Assumptions of the LPP model
2.5 Formulation of LP Models
2.6 Graphical Solutions under Linear Programming
2.7 Duality
2.8 Summary
2.9 Glossary
2.10 Answer to In-Text Questions
2.11 Self-Assessment Questions
2.12 Suggested Readings

2.1 LEARNING OBJECTIVE

After studying this lesson, you will be able to:

 Understand what is linear programming
 Locate areas of application with its scope
 Know how to formulate LP models and use graphical procedures to solve them
 Analyze the importance of duality in linear programming
 Understand dual formation and solution

Self-Instructional
Material 39

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES
2.2 INTRODUCTION

To accomplish particular goals, all organizations—large or small—need to make the

best use of their limited or scarce resources. Money, personnel, materials, machine
capacity, technology, time, etc. are all examples of scarce resources. The decision-
maker must be aware of all relevant information on the organization’s activities as well
as the relationships that regulate the chosen activities and their results in order to
produce the best results with the available resources. The desired result might be
measured in terms of revenue, time, ROI, costs, etc.
The most popular and frequently used method of mathematical programming,
amongst all the well-known optimal models, is linear programming. Essentially, linear
programming is a deterministic mathematical technique that includes allocating limited
or scarce resources optimally in accordance with the established norm of optimality.
Frequently, the criterion of optimality is profits, costs, return on investment, time, distance,
etc.
The linear programming method was created by George B. Dantzig in 1947
while he was a member of the US Air Force during World War II. As a method for
selecting the best plan out of several plans for attaining the objective through diverse
activities like procurement, recruitment, maintenance, etc., linear programming was
created. Decisions in linear programming are made with certainty, meaning that the
resource availability and the relationships between the variables are known. As a result,
the chosen course of action will almost always provide ideal outcomes. Thus, linear
programming offers a quantitative basis for assisting a decision maker identify the most
favourable and feasible course of action from a set of accessible options in attain the
goal in the most optimal manner.

2.3 BASIC TERMINOLOGY

 Linear:
When two or more variables are related in a direct or precisely proportional
Self-Instructional
40 Material way, the relationship is said to be linear.

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

 Programming: NOTES
The term “programming” refers to the systematic adoption of numerous alternative
courses of action while making decisions.
 Optimal:
The term “optimal” refers to a program that maximizes or minimizes some criterion
or measure of effectiveness, such as profit, cost, or sales. The availability of
resources during the planning horizon is referred to as “limited.”
 Feasible solution:
A feasible solution is one that satisfies all of the constraints.
 Infeasible solution:
A solution that violates at least one constraint is said to be infeasible.
 Feasible region:
The set of all feasible solutions makes up the feasible region.
 Optimal solution:
The feasible solution having the highest favourable value of the objective is an
optimal solution
 The most favourable value:
If the objective function is to be maximised, it is the greatest value; if the objective
function is to be minimised, it is the lowest value.
 Corner point feasible (CPF):
A solution that is at the corner of the feasible region is known as a corner point
feasible (CPF).
 Optimal solution v/s CPF solutions:
Consider any linear programming problem with feasible solutions and a bounded
feasible region to examine the relationship between the optimal solution and
CPF solutions. The problem must have at least one optimum solution and a
CPF solution. The most favourable solution is the optimal solution. Therefore, a
CPF solution must exist if a problem only has one perfect solution. If there are
numerous optimal solutions to the problem, at least two of them must be CPF
Self-Instructional
solutions. Material 41

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Basic requirements

(a) Decision variables and their relationship
(b) Objective function
(c) Constraints
(d) Alternative courses of action
(e) Non-negative restriction
(f) Linearity

2.4 ASSUMPTIONS OF THE LPP MODEL

The model formulation contains implicitly all of the linear programming assumptions.
The model must particularly have a linear goal function that is subject to linear restrictions
from a mathematical perspective. However, from the viewpoint of a model, these
mathematical features of a linear programming model need to make specific assumptions
about the activities and data of the issue that are depicted, especially those on the
impact of changing the activity levels.
(a) Proportionality: Each activity’s contribution to the value of the objective function
varies in direct proportion to its activity (xj). Similar to this, the level of each
activity’s contribution to the left side of each functional constraint is proportional
to the activity xj. Consequently, this assumption rules out any exponent other
than 1 for any variable in any term of any function in linear programming model.
If this presumption is violated, non-linear programming will be used. In addition,
a mixed integer programming extension of linear programming can be used if
the assumption is fails.
(b) Divisibility (or continuity): The LP model’s divisibility assumption states that
decision variables may have both integer and non-integer values. The fundamental
premise is that operations can be carried out at fractional levels. These numbers
may seem reasonable only as integers; in these cases, we require an extension
of LP called integer programming.
(c) Additivity: In a linear programming model, each function—whether it is the
objective function or the function on the left of a functional constraint—represents
Self-Instructional
42 Material the sum of the individual contribution of the respective activities.

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

Cross product terms (terms involving the product of two or more variables) are NOTES
not prohibited by the proportionality assumption, even though exponents other
than one are not allowed. This latter scenario isn’t permitted by the additivity
assumption.
(d) Certainty: Every parameter holds a known and constant value that can be
determined with absolute certainty, according to the certainty assumption. To
put it another way, the model is predicated on the idea that the reactions to the
values of the variables are a perfect match to the responses denoted by the
coefficients. The certainty assumption, however, does not always hold true in
real-world circumstances since model parameters take into account both the
present and the future, adding an element of uncertainty.
(e) Data: In order to create a linear program model to solve a problem, data must
be available to define the problem.

2.5 FORMULATION OF LP MODELS

Let’s consider a LPP having ‘n’ decision variables x1, x2……… xn subject to set of
‘m’ restrictions or constraints which are linear function to the decision variables. The
standard format of a general LPP may be written in the following format:

Self-Instructional
Material 43

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES where:

 x1, x2, …………………, xn, are decision variables.

 C1, C2, …………………, Cn, are called profit or cost coefficients.
 aij are called substitution coefficients.
 b1, b2, …………………, bm, represents minimum requirement or maximum
availability of m constraints.
 The non-negativity constraint represents that all xj’s must be positive (non-
negative).

Mathematical Formulation of LPP

Mathematical Formulation is the most important part of the Linear Programming

Problem. Mathematical formulation of LPP means converting descriptive given problem
into mathematical form. Generally the following steps are followed:

Step 1: Identify Decision Variables

Identify the decision variables whose value is supposed to be determined by following

one basic principle into the mind, i.e., if the problem is of maximization then variables
denoting profit/revenue/customer reach, etc. should be assumed as decision variable
and if the problem is of minimization, then variables denoting cost/loss, etc. should be
taken as decision variable.

Step 2: Write Down the Objective Function

Identify the type of optimization i.e. Max or Min. Linearity assumptions should be
invoked to write the objective function. While writing the objective function, we are
also assuming CRS.

Step 3: Write Down the Constraints

Identify the constraints expressed in terms of decision variables with appropriate equality
or inequality. In diet problems, if the problem was simply to minimize costs, then the
answer would be very easy. Buy no food. After all, that costs you nothing. The problem
Self-Instructional with that is that the people in your institution will die.
44 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

You want to minimize expenditures, but only after you have met the nutritional NOTES
requirements. You need a way to decide whether the food you buy actually satisfies
nutritional requirements.

Step 4: Write the Non-Negativity Restriction

Include this restriction of the decision variables. Non negative restrictions mean that all
decision variables must be zero or more than zero.

IN-TEXT QUESTIONS
Multiple Choice Questions:
1. Linear Programming is a
(a) Constrained optimization technique
(b) Technique for economic allocation of limited resource
(c) Mathematical technique
(d) All of the above
2. A constraint in a LP model restricts
(a) Value of an objective function
(b) Value of a decision variable
(c) Use of the available resources
(d) All of the above
3. Constraints of an LP model represents
(a) Limitations
(b) Requirements
(c) Balancing limitations and requirements
(d) All of the above

Example 2.1:

Two Mines Example

The Two Mines Company own two different mines that produce an ore which, after
Self-Instructional
being crushed, is graded into three classes: high, medium and low-grade. The company Material 45

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES has contracted to provide a smelting plant with 12 tons of high-grade, 8 tons of medium-
grade, and 24 tons of low-grade ore per week. The two mines have different operating
characteristics as detailed below.
Mine Cost per day (£'000) Production (tons/day)
High Medium Low
X 180 6 3 4
Y 160 1 1 6

Consider that mines cannot be operated in the weekend. How many days per
week should each mine be operated to fulfil the smelting plant contract?

Solution

What we have is a verbal description of the Two Mines problem. What we need to do
is to translate that verbal description into an equivalent mathematical description.
In dealing with problems of this kind we often do best to consider them in the order:

 Variables
 Constraints
 Objective
This process is often called formulating the problem (or more strictly formulating
a mathematical representation of the problem).

Variables

These represent the “decisions that have to be made” or the “unknowns”.

We have two decision variables in this problem:
x = number of days per week mine X is operated
y = number of days per week mine Y is operated
Note here that x  0 and y  0.

Constraints

It is best to first put each constraint into words and then express it in a mathematical
Self-Instructional
46 Material
form.

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

Ore production constraints - Balance the amount produced with the quantity NOTES
required under the smelting plant contract
Ore
High 6x + 1y  12
Medium 3x + 1y  8
Low 4x + 6y  24
Days per week constraint - We cannot work more than a certain maximum
number of days a week e.g. for a 5 day week we have
x5
y5

Inequality constraints

Note we have an inequality here rather than an equality. This implies that we may
produce more of some grade of ore than we need. In fact we have the general rule:
given a choice between an equality and an inequality choose the inequality
For example - if we choose an equality for the ore production constraints we
have the three equations 6x + y = 12, 3x + y = 8 and 4x + 6y = 24 and there are no
values of x and y which satisfy all three equations (the problem is therefore said to be
“over-constrained”). For example the values of x and y which satisfy 6x + y = 12 and
3x + y = 8 are x = 4/3 and y = 4, but these values do not satisfy 4x + 6y = 24.
The reason for this general rule is that choosing an inequality rather than an
equality gives us more flexibility in optimizing (maximizing or minimizing) the objective
(deciding values for the decision variables that optimize the objective).

Implicit constraints

Constraints such as days per week constraint are often called implicit constraints because
they are implicit in the definition of the variables.

Objective

Again in words our objective is (presumably) to minimize cost which is given by

Self-Instructional
180x + 160y Material 47

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Hence we have the complete mathematical representation of the problem:

Minimize 180x + 160y
subject to
6x + y  12
3x + y  8
4x + 6y  24
x5
y5
x, y  0

Notes

The mathematical problem given above has the form

 All variables continuous (i.e. can take fractional values)
 A single objective (maximize or minimize)
 The objective and constraints are linear i.e. any term is either a constant or a
constant multiplied by an unknown (e.g. 24, 4x, 6y are linear terms but xy or x2
is a non-linear term)
Any formulation which satisfies these three conditions is called a linear program
(LP).

Example 2.2:

Giapetto Example:
Giapetto’s wooden soldiers and trains. Each soldier sells for $27, uses $10 of raw
materials and takes $14 of labor & overhead costs. Each train sells for $21, uses $9
of raw materials, and takes $10 of overhead costs. Each soldier needs 2 hours finishing
and 1 hour carpentry; each train needs 1 hour finishing and 1 hour carpentry. Raw
materials are unlimited, but only 100 hours of finishing and 80 hours of carpentry are
available each week. Demand for trains is unlimited; but at most 40 soldiers can be
sold each week. How many of each toy should be made each week to maximize
Self-Instructional
48 Material profits?

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

Solution NOTES
Decision variables completely describe the decisions to be made (in this case, by
Giapetto). Giapetto must decide how many soldiers and trains should be manufactured
each week. With this in mind, we define:

x1 = the number of soldiers produced per week

x2 = the number of trains produced per week

Objective function is the function of the decision variables that the decision
maker wants to maximize (revenue or profit) or minimize (costs). Giapetto can
concentrate on maximizing the total weekly profit (z).
Here profit equals to (weekly revenues) – (raw material purchase cost) – (other
variable costs). Hence Giapetto’s objective function is:
Maximize z = 3 x1 + 2 x2
Constraints show the restrictions on the values of the decision variables. Without
constraints Giapetto could make a large profit by choosing decision variables to be
very large. Here there are three constraints:
Finishing time per week
Carpentry time per week
Weekly demand for soldiers
Sign restrictions are added if the decision variables can only assume non-
negative values (Giapetto can not manufacture negative number of soldiers or trains!)
All these characteristics explored above give the following Linear Programming
(LP) model

max z = 3 x1 + 2 x2 (The Objective function)

Subject to
2 x1 +  100 (Finishing constraint)
x1 + x2  80 (Carpentry constraint)
Self-Instructional
Material 49

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES x1  40 (Constraint on demand for soldiers)

x1, x2 > 0 (Sign restrictions)
Example 2.3:
A company produces 2 types of cowboy hats. Each hat of the first type requires twice
as much labor time as the second type. The company can produce a total of 500 hats
a day. The market limits the daily sales of first and second types to 150 and 250 hats.
Assuming that the profits per hat are $8 per type A and $5 per type B, formulate the
problem as Linear Programming model in order to determine the number of hats to be
produced of each type so as to maximize the profit.

Solution:

Let x1 be the no. of hats of type A.

x2 be the no. of hats of type B.
8x1 is the total profit for hats of type A.
5x2 is the total profit for hats of type B.
Hence, objective function will be equal to
Max Z = 8x1 + 5x2 (Subject to constraints)
2 x1 + x2  500 (Labor time for total production)
x1 150 (No. of hats of type A to be sold)
x2  250 (No. of hats of type B to be sold)
x1 , x2  0 (Non-negativity constraints)

Example 2.4:

A firm can produce 3 types of cloth say A, B and C. Three kinds of wool are required
for it say red, green and blue. One unit length of type A cloth needs 2 metres of red
wool and 3 metres of blue wool. One unit length of B type cloth needs 3 metres of red
wool, 2 metres of green wool and 2 metres of blue wool; and 1 unit length of type C
cloth needs 5 metres of green wool and 4 metres of blue wool. The firm has a stock of
Self-Instructional 8 metres of red wool, 10 metres of green wool and 15 metres of blue wool, it is
50 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

assumed that the income obtained from one unit length of type A cloth is 3, of B 5 NOTES
and of C 4.
Determine how the firm should use the available material so as to maximize the
income from the finished cloth. Formulate the above problem as LPP.
Solution:
Let, x1 be the type of cloth A
x2 be the type of cloth B
x3 be the type of cloth C
therefore,
3 x1 is the profit for type A cloth
5 x2 is the profit for type B cloth
4 x3 is the profit for type C cloth.

Clot
Materials h Max. Material
A B C Available
Red 2 3 -- 8
Blue 3 2 4 15
Green -- 2 5 10
Profit per unit ( ) 3 5 4

The objective function is given by

Maximize Z = 3 x1 + 5 x2 + 4 x3 (Subject to constraints)

2 x1 + 3  8 (Material Constraint)

3 x1 + 2 x2 + 4 x3  15 (Material Constraint)
2 x2 + 5 x3  10 (Material Constraint)
x1 , 5 x2 , 4 x3  0 (Non-negativity constraints)

Self-Instructional
Material 51

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Solving LPP

LP Solutions: Four Cases

When an LP is solved, one of the following four cases will occur:
1. The LP has a unique optimal solution.
2. The LP has alternative (multiple) optimal solutions. It has more than one (actually
an infinite number of) optimal solutions
3. The LP is infeasible. It has no feasible solutions (The feasible region contains no
points).
4. The LP is unbounded. In the feasible region there are points with arbitrarily
large (in a max problem) objective function values.

2.6 GRAPHICAL SOLUTIONS UNDER LINEAR

PROGRAMMING

Any LP with only two variables can be solved graphically

If we can find the values of the decision variables x1 , x2 , x3 ,.... xn , which can optimize
(maximize or minimize) the objective function Z, then we say that these values of xi
are the optimal solution of the Linear Program (LP). The graphical method is applicable
to solve the LPP involving two decision variables x1 , and x2 , we usually take these
decision variables as x, y instead of x1 , x2 . To solve an LP, the graphical method
includes two major steps.
(a) The determination of the solution space that defines the feasible solution. Note
that the set of values of the variable x1 , x2 , x3 ,.... xn which satisfy all the
constraints and also the non-negative conditions is called the feasible solution of
the LP.
(b) The determination of the optimal solution from the feasible region.

Self-Instructional
52 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

To determine the feasible solution of an LP, we have the following steps. NOTES
Step 1 : Find the feasible region of the LLP.
Step 2 : Find the co-ordinates of each vertex of the feasible region. These co-
ordinates can be obtained from the graph or by solving the equation of the lines.
Step 3 : At each vertex (corner point) compute the value of the objective
function.
Step 4 : Identify the corner point at which the value of the objective function is
maximum (or minimum depending on the LP). The coordinates of this vertex is the
optimal solution and the value of Z is the optimal value.
If an LPP has many constraints, then it may be long and tedious to find all the
corners of the feasible region.

IN-TEXT QUESTIONS
Fill in the blanks:
4. Graphical method can be used only if there are …….. decision variables.
5. While solving a LP graphically, the area bounded by the constraints is called
…………….
6. If the given problem is maximization, Zmax then locates the solution point at
the ……………point of the feasible zone from the origin

Example 2.5:

Production of wooden tables and chairs

In this example, we will look at the profit maximization problem that a furniture
manufacturer faces. The manufacturer uses wood and labor (inputs) to produce tables
and chairs (output). We consider that the manufacturer earns 60 and 80 for each
unit of table and chair respectively. Also the availability of wood is 3000 units and that
of labor is 1100 units. And it takes 300 units of wood and 50 units of labour to make
a table and 200 units of wood and 100 units of labour to make a chair. The following
table depicts the relevant information for this LP problem.
Self-Instructional
Material 53

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Table

Input Table (X1) Chair (X2) Availability

Wood 300 200 3000
Labor 50 100 1100
Unit Profit 60 80

Step-1: Formation of the LP problem

With help of the table we will formulate the objective function. An objective
function is developed in such a way that it is either maximized or minimized. In our
example the objective would be to maximize unit profit. Wood and labor row in the
table is used to formulate the constraint line. We state the non-negativity conditions.
Maximize: Z = 60X1 + 80X2 (objective function, Z = profit)
Subject to: 300X1 + 200X2  3000 (wood constraint)
50X1 + 100X2  1100 (labor constraint)
X1, X2  0 (non-negativity conditions)
Step-2: Plotting the constraint lines on the graph
We plot the number of tables on x-axis and number of chairs on y-axis. Next
we plot the constraints lines by finding the x and y intercepts for the two constraints
equation as follows:

Wood

300X1 + 200X2  3000

Set X2 = 0 and solve for X1
300X1 = 3000
X1 = 10 tables (All wood is used to make tables)
300X1 + 200X2  3000
Set X1 = 0 and solve for X2
200X2 = 3000
Self-Instructional X2 = 15 chairs (All wood is used to make chairs)
54 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

Labor NOTES

50X1 + 100X2  1100

Set X2 = 0 and solve for X1
50X1 = 1100
X1 = 22 tables (All wood is used to make tables)
50X1 + 100X2  1100
Set X1 = 0 and solve for X2
100X2 = 1100
X2 = 11 chairs (All wood is used to make chairs)
Now we plot the wood constraint line using the intercept X1 = 10 and X2 = 15
chairs. And the labor constraint line by using the intercept X1 = 22 and X2 = 11 chairs.
(See figure 1)

Figure 1: Constraint Lines

Step-3: Determining the valid side of the constraint line

Origin (0, 0) is used to check the valid side for both constraint lines. 300(0) + 200(0) Self-Instructional
 3000 is valid, therefore the side of the origin is the valid side of the wood constraint Material 55

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES line. 50(0) + 100(0)  1100 is valid, so the side of the origin is the valid side of the
labor constraint line. Arrows are marked in the figure indicating the valid side of each
constraint line. (Figure 2)

Figure 2: Identification of Feasible Region

Step-4: Identification of the feasible solution region

The feasible region is the area on the valid side of both the constraint lines. Any point
located on the invalid side of the constraint line is infeasible. And because of the non-
negativity conditions, the feasible region is restricted to first (positive) quadrant.

Step-5: Plotting the objective function to determine the direction of

improvement

Now, we will arbitrarily set profit, Z = 480 and Z = 720 and then find the x and y
intercept and plot the two lines.
Setting Z = 480
and X2 = 0 and solving for X1
 60(X1) = 480
 X1 = 8
Self-Instructional
56 Material Now, setting X1 = 0 and solving for X2

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

 80(X2) = 480 NOTES

 X2 = 6
Next, Setting Z = 720 and X2 = 0 and solve for X1
 60(X1) = 720
 X1 = 12
Now, setting X1 = 0 and solving for X2
 80(X2) = 720
 X2 = 9
Now we plot the objective function lines when Z = 480 and Z = 720 as shown
in figure 3. We can see from the figure that as we move away from the origin Z rises.

Figure 3: Determining the Direction of Increasing Value

Step-6: Find the appropriate corner

Since we want to maximize Z, we will draw a line parallel to the objective function
lines that touches the last point in the feasible region while moving away from the
origin. This identifies the most attractive corner, which gives us the amounts of wood
and labor that will result in the maximization of profits (maximize Z). Thus, it represents
Self-Instructional
the optimal solution to the LP problem. (Figure 4) Material 57

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES

Figure 4: Finding the Most Appropriate Corner

The most appropriate corner lies at the intersection of the wood and labor
constraint lines. The intersection point of the two constraint is (4, 9). When we solve
for Z with this solution we get Z = 60*4 + 80*9 = 960.
Thus, the maximum profit of 960 can be obtained by producing 4 tables and
9 chairs.
Example 2.6:
Giapetto
Since the Giapetto LP has two variables, it may be solved graphically.
Answer
The feasible region is the set of all points satisfying the constraints.
Max z = 3 x1 + 2 x2
subject to
2 x1 + x2  100 (Finishing constraint)
x1 + x2  80 (Carpentry constraint)
x1  40 (Demand constraint)
Self-Instructional
58 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

x1 , x2  0 (Sign restrictions) NOTES

The set of points satisfying the LP is bounded by the five sided polygon DGFEH.
Any point on or in the interior of this polygon (the shaded area) is in the feasible
region.
Having identified the feasible region for the LP, a search can begin for the optimal
solution which will be the point in the feasible region with the largest z-value
(maximization problem).
To find the optimal solution, a line on which the points have the same z-value is
graphed. In a max problem, such a line is called an iso-profit line while in a min
problem, this is called the iso-cost line. (The figure shows the iso-profit lines for
z = 60, z = 100, and z = 180).
In the unique optimal solution case, iso-profit line last hits a point (vertex -
corner) before leaving the feasible region.

The optimal solution of this LP is point G where ( x1 , x2 ) = (20, 60) giving

z = 180.

Self-Instructional
Material 59

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES A constraint is binding (active, tight) if the left-hand and right-hand side of the
constraint are equal when the optimal values of the decision variables are substituted
into the constraint.
A constraint is nonbinding (inactive) if the left-hand side and the right-hand
side of the constraint are unequal when the optimal values of the decision variables are
substituted into the constraint.
In Giapetto LP, the finishing and carpentry constraints are binding. On the other
hand, the demand constraint for wooden soldiers is nonbinding since at the optimal
solution x1 < 40 (= 20).
However, there are limitations and they are as discussed below.
1. Linearity of relations: A primary requirement of LP is that the objective function
and every other constraint must be linear. This means that inputs and outputs
can be added, multiplied and divided. However, in real life situations, several
business and industrial problems are non-linear in nature.
2. Single objective: LP takes in account a single objective only, i.e. profit
maximization or cost minimization. However, in present dynamic environment,
there is no single universal objective.
3. It is complex to determine the particular objective function. Even if a particular
objective function is laid down, it may not be so easy to find out various
technological, financial and other constraints which may be operative in pursuing
the given objective.
4. Given a specified objective and a set of constraints it is feasible that the
constraints may not be directly expressible as linear inequalities.
5. Even if the above problems are surmounted, a major problem is one of estimating
relevant values of the various constant co-efficient that enter into a linear
programming mode, i.e. prices etc.
6. This technique presumes perfect competition in product and factor markets.
But perfect competition is not a reality.
7. Certainty: LP assumes that the values of the coefficient of decision variables
are known with certainty. Due to this, LP cannot be applied to a wide variety of
Self-Instructional
60 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

problems where the values of the coefficients are probabilistic. For example, NOTES
the LP technique is based on the hypothesis of constant returns. In reality, there
are either diminishing or increasing returns which a firm experiences in production.
8. Constant parameters: Parameters appearing in LP are assumed to be constant,
but in practical situations it is not so.
9. It is a highly mathematical and complicated technique. The solution of a problem
with linear programming requires the maximization or minimization of a clearly
specified variable. The solution of a linear programming problem is also arrived
at with such complicated method as the simplex method which comprises of a
huge number of mathematical calculations.
10. Largely, LP models reflects trial & error solutions and it’s hard to gauge solutions
which are optimal to the variety complexities of economics.

2.7 DUALITY

For every linear programming problem, there is an associated linear programming

problem. The former problem is called primal and the latter is called its dual and vice
versa. The two problems may appear to have superficial relationship between each
other but they possess very intimately related properties and useful one, so that the
optimal solution of one problem gives complete information about the optimal solution
to the other. In other words, the optimal solutions for both the problems are same.
One part of a Linear Programming Problem (LPP) is called the Primal and the
other part is called the Dual. In other words, each maximization problem in LP has its
corresponding problem, called the dual, which is a minimization problem. Similarly,
each minimization problem has its corresponding dual, a maximization problem. For
example, if the primal is concerned with maximizing the contribution from the three
products A, B, and C and from the three departments X, Y, and Z, then the dual will
be concerned with minimizing the costs associated with the time used in the three
departments to produce those three products. An optimal solution from the primal and
the dual problem would be same as they both originate from the same set of data.
Self-Instructional
Material 61

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Formulation of a Dual Problem

In an economy, the resources are scarce therefore they need to be valued. So when
we solve a linear programming model, we focus upon two problems:
 The primal resource allocation problem
 The dual resource valuation problem
If the primal problem has n variables and m constraints, then the dual problem
will have m variables and n constraints.
Standard form of the Primal Problem is:

Now, we formulate the Standard form of the Dual Problem from the above primal:

Subject to:

Self-Instructional
62 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

In condensed form this can be written as: NOTES

Thus, the dual problem uses the same parameters as the primal problem, but in
different locations.
To highlight the comparison, now look at these same two problems in matrix
notation, where c and y = [y1, y2, . . . , ym] are row vectors but b and x are column
vectors.

The primal-dual table for linear programming given below helps to highlight
the correspondence between the two problems. It shows all the linear programming
parameters (the aij, bi, and cj) and how they are used to construct the two problems.
All the headings for the primal problem are horizontal, whereas the headings for the
dual problem are read by turning the book sideways. For the primal problem, each
column (except the right side column) gives the coefficients of a single variable in
the respective constraints and then in the objective function, whereas each row
(except the bottom one) gives the parameters for a single constraint. For the dual

Self-Instructional
Material 63

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES problem, each row (except the Right Side row) gives the coefficients of a single
variable in the respective constraints and then in the objective function, whereas
each column (except the rightmost one) gives the parameters for a single constraint.
In addition, the Right Side column gives the right-hand sides for the primal problem
and the objective function coefficients for the dual problem, whereas the bottom
row gives the objective function coefficients for the primal problem and the right-
hand sides for the dual problem.

Steps for Dual Formulation

The following procedure is adopted to convert primal problem into its dual. Simplex
method is applied to obtain the optimal solution for both primal and dual form.
 Step 1 : For each constraint, in primal problem there is an associated variable
in the dual problem.
 Step 2 : The elements of right-hand side of the constraints will be taken as the
coefficients of the objective function in the dual problem.
 Step 3 : If the primal problem is maximization, then its dual problem will be
minimization and vice versa.

Self-Instructional
64 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

 Step 4 : The inequalities of the constraints should be interchanged from  to  NOTES

and vice versa and the variables in both the problems are non-negative.
 Step 5: The rows of primal problem are changed to columns in the dual problem.
In other words, the matrix A of the primal problem will be changed to its transpose
(A) for the dual problem.
 Step 6: The coefficients of the objective function will be taken to the right hand
side of the constraints of the dual problem.
Example 2.7:
Max Z = 4 x1  10 x2  9 x3
Subject to:
5 x1 14 x2  5 x3  15

8 x1 12 x2  8 x3  8

12 x1  4 x2  8 x3  10

2 x1  5 x3  5
x1 , x2 , x3  0
Constructing the primal dual table:

Therefore the dual can be expressed as:

Min Y = 15 y1 + 8 y2 + 10 y3 + 5 y4

Self-Instructional
Material 65

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Sub to

Thus, the dual problem may be viewed as a restatement in linear programming terms
of the goal of the simplex method; namely, to reach a solution for the primal problem
that satisfies the optimality test. We can form the dual of a linear program in
minimization normal form in the same way in which we formed the dual in the
maximization case:
 Switch the type of optimization,
 Introduce as many dual variables as the number of primal constraints (not
counting the non-negativity constraints),
 Define as many dual constraints (not counting the non-negativity constraints) as
the number of primal variables.
 Take the transpose of the matrix of coefficients on the left-hand side of the
inequality,
 Switch the roles of the vector of coefficients in the objective function and the
vector of right-hand sides in the inequalities.
Example 2.8:
Write the dual of the following problem.
Maximize: Z = –6 x1 + 7 x2

Subject to, – x1 + 2 x2 – 5

3 x1 + 4 x2  7

x1 , x2  0
Solution:
The given problem is considered as primal linear programming problem. To convert it
Self-Instructional into dual, the following procedure is adopted.
66 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

Step 1: There are 2 constraints and hence the dual problem will have 2 variables. NOTES
Let us denote them as y1 and y2 .
Step 2: The right hand side of the constraints are –5 and 7 which are taken as
the coefficients of the variables y1 and in the objective function.
Step 3: The primal objective problem is maximization and hence the dual seeks
minimization for the objective function. Hence, the objective functions for the dual
problem is given by
Minimize Z = –5 y1 + 7 y2
Step 4: The inequalities of the constraints for the primal problem are of the type
(). Hence, the inequalities for the dual constraints will be of the type ().
Step 5: The coefficient matrix for the primal problem is

 1 2 
A=  
 3 4
The transpose of this matrix which serves as the coefficient matrix for the dual
problem is given by.
 1 3 
A=  
 2 4
Step 6: The coefficients of the objective function for the given primal are –6
and 7. They are taken on the right hand side of the constraints for the dual problem.
Hence, the constraints for the dual problem are represented as
– y1 + 3 y2 –6
2 y1 + 4 y2 7
y1 , y2  0

Notes The study of duality is very important in LP. Knowledge of duality allows one
to develop increased insight into LP solution interpretation. Also, when solving the
dual of any problem, one simultaneously solves the primal. Thus, duality is an
alternative way of solving LP problems. However, given today’s computer
capabilities, this is an infrequently used aspect of duality. Therefore, we concentrate
on the study of duality as a means of gaining insight into the LP solution. Self-Instructional
Material 67

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Interpretation of the result

For any primal-dual pair of linear programs, if either has an optimal solution, then both
have optimal solutions, and their optimal objective function values are equal. This is
called a weak duality theorem.

Weak Duality Theorem

If x  ( x1 ,....., xn ) is a feasible solution to the primal with objective function value z ,

and y  ( y1 ,...., yn ) is a feasible solution to dual with objective function value w then
z  w, and if z  w then x and y are both optimal to their respective LPs.
In mathematical optimization theory, duality is the principle that states that
optimization problems may be viewed from either of two perspectives, the primal
problem or the dual problem. As per the weak duality theorem, the optimal vales of
the primal and dual are the same. However, in general the optimal values of the
primal and dual problems need not be equal. Their difference is called the duality
gap.

Economic interpretation of Duality

If we interpret our primal LP problem as a classical “Resource Allocation” problem,

its dual can be interpreted as a “Resource Valuation” problem. The dual variables give
the shadow prices for the primal constraints. Suppose we have a profit maximization
problem with a resource constraint. Then the value yi of the corresponding dual variable
in the optimal solution tells us that we get an increase of yi in the maximum profit for
each unit increase in the amount of resource.
Sometimes, finding an initial feasible solution to the dual is much easier than
finding one for the primal. Also, it is much easier to solve the dual as compared to the
primal at times. A primal with many constraints and few variables can be converted
into a dual with few constraints and many variables. Thus duality provides a lot of
computational advantages in a problem with lesser number of variables and a multitude
of constraints.

Self-Instructional
68 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

NOTES
IN-TEXT QUESTIONS
Answer the following questions.
7. One part of a Linear Programming Problem (LPP) is called the Primal and the
other part is called the Dual. (True / False)
8. In the primal problem, the objective function is an exponential combination of
n variables. (True / False)
9. In the dual problem, the dual vector multiplies the constants that determine the
positions of the constraints in the primal. (True / False)
10. Duality is quite useful when investigating changes in the parameters of an LPP
( the technique known as sensitivity analysis). (True / False)
11. It is advantageous to solve the dual of a primal having less number of constraints
because the number of constraints usually equals the number of iterations
required to solve the problem. (True / False)

2.8 SUMMARY

 Linear programming determines the way to achieve the best outcome (such as
maximum profit or lowest cost) in a given mathematical model and some list of
requirements represented as linear equations.
 It is a technique to ensure the optimum allocation of scarce resources in order
to deliver for the fulfilment of ever increasing demands in the market.
 Linear Programming is used as a helping tool in nearly all functional areas of
management.
 The graphical method to solve linear programming problem helps to visualize
the procedure explicitly.
 It also helps to understand the different terminologies associated with the solution
of LPP.
 In mathematical optimization theory, duality is the principle that states that
optimization problems may be viewed from either of two perspectives, the primal Self-Instructional
Material 69

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES problem or the dual problem. The dual problem may be viewed as a restatement
in linear programming of a primal.
 One part of a Linear Programming Problem (LPP) is called the Primal and the
other part is called the Dual. In other words, each maximization problem in LP
has its corresponding problem, called the dual, which is a minimization problem.
Similarly, each minimization problem has its corresponding dual, a maximization
problem.

2.9 GLOSSARY

 Constraints: It is a condition that a solution to an optimization problem must

satisfy.
 Feasible Region: It is the region containing solution.
 Feasible Solution: If a solution satisfies all the constraints, it is called a feasible
solution.
 Shadow Price: The amount that the objective function value changes per unit
change in the constraint is known as shadow price.
 Dual Problem: In the dual problem, the dual vector multiplies the constants
that determine the positions of the constraints in the primal.
 Duality principle: In optimization theory, the duality principle states that
optimization problems may be viewed from either of the two perspectives, the
primal problem or the dual problem.
 Primal Problems: In the primal problem, the objective function is a linear
combination of n variables.

2.10 ANSWER TO IN-TEXT QUESTIONS

1. (d) 2. (d)
Self-Instructional 3. (d) 4. Two
70 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Linear Programming

5. Feasible region 6. far most NOTES

7. True 8. False
9. True 10. True
11. True

2.11 SELF-ASSESSMENT QUESTIONS

1. Explain the linear programming problem giving two examples.

2. What are the essential characteristics of a linear programming model?
3. What do you understand by ‘Graphical Method’? Give its limitations.
4. Explain the graphical method of solving a Linear programming Model involving
two variables.
5. Define and explain the following:
(i) Optimum Solution
(ii) Feasible Solution
(iii) Unrestricted Variables
6. A firm manufactures headache pills in two sizes A and B. Size A contains I
grains of aspirin, 5 grains of bicarbonate and 1 grain of codeine. It is found by
uses that it requires at least 12 grains of aspirin, 74 grains of bicarbonate and 24
grains of codeine for providing immediate effect. It is required to determine the
least number of pills a patient should take to get immediate relief.
Formulate the problem as a standard LPP.
7. Consider a small plant which makes 2 types of automobile parts say A and B. It
buys castings that are machined, bored, and polished. The capacity of machining
is 25 per hour for A and 40 hours for B, capacity of boring is 28 per hour for A
and 35 per hour for B, and the capacity of polishing is 35 per hour A and 25
hour of B. Casting for port A costs 2 each and for part B they cost 3 each.
They sell for 5 and 6 respectively. The three machines have running costs of
20, 14 and 17.50 per hour. Assuming that any combination of parts A and Self-Instructional
B can be sold, what product mix maximizes profit? Material 71

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 8. A ship is to carry 3 types of liquid cargo – X, Y and Z. There are 3000 litres of
X available, 2000 litres of Y available and 1500 litres of Z available. Each litre
of X, Y and Z sold fetches a profit of 30, 35 and 40 rupees respectively. The
ship has 3 cargo holds-A, B and C of capacities 2000, 2500 and 3000 litres
respectively. From stability considerations, it is required that each hold be filled
in the some proportion. Formulate the problem of loading the ship as a linear
programming problem. State clearly what are the decision variables and
constraints.

2.12 SUGGESTED READINGS

 J.K. Sharma, Operations Research, Theory and Applications, MacMillan India

Ltd.
 Kanti Swarup, P.K Gupta & Manmohan, Operations Research, Sultan Chand
 Publications, New Delhi
 Michael W. Carter, Camille C. Price, Operations Research: A Practical
Introduction,
 CRC Press, 2001
 Paul A. Jensen, Jonathan F. Bard, Operations Research Models and Methods,
John
 Wiley and Sons, 2003
 Richard Bronson, Govindasami Naadimuthu, Schaum’s Outline of Theory and
 Problems of Operations Research, McGraw-Hill Professional; 1997
 Sydsaeter, K. Hammond, P. (2002), Mathematics for economic analysis. Person
Edition
Online links
 http://web.mit.edu/15.053/
 http://agecon2.tamu.edu/
Self-Instructional  http://lucatrevisan.wordpress.com/
72 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
UNIT III: INTEGRATION, DIFFERENTIAL
EQUATIONS, AND DIFFERENCE
EQUATIONS

LESSON 3 DEFINITE INTEGRAL

LESSON 4 INTEGRATION

LESSON 5 DIFFERENTIAL EQUATIONS

LESSON 6 DIFFERENCE EQUATION

Definite Integral

LESSON 3 NOTES

DEFINITE INTEGRAL

Sarabjeet Kaur
Assistant Professor,
Department of Economics,
Zakir Husain College (E), University of Delhi)
Structure
3.1 Introduction
3.2 The Fundamental Theorem of Calculus
3.3 Properties of Definite Integral
3.4 Definite integral
3.5 Economic Application
3.6 Marginal Function to Total Function
3.7 Consumer Surplus and Producer Surplus
3.8 Summary
3.9 Glossary
3.10 Answer to In-Text Questions
3.11 Self-Assessment Questions
3.12 Suggested Readings

3.1 INTRODUCTION

The Definite Integral inputs a function and outputs a number, which gives the area
between the graph of the input and the x-axis. It is represented by the following notation:
b
a
f ( x)dx
This is nothing but the area under the curve as shown below:

Self-Instructional
Material 75

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES It takes a specific value and free from the variable, x and an arbitrary constant,
C. Here, ‘a’ is called the lower limit of the integration while ‘b’ is known as the upper
limit.

3.2 THE FUNDAMENTAL THEOREM OF CALCULUS

The Fundamental Theorem of Calculus identifies the relationship among the processes
of differentiation and integration. That relationship says that differentiation and integration
are inverse processes.

The Fundamental Theorem of Calculus: Part 1

If f is a continuous function on [a,b], then the function indicated by

g(x) =  f (u)du for x a,b]
is continuous on [a,b], differentiable on (a,b) and g’(x) = f(x)

Figure: 3.1

If f(t) is continuous on [a,b], the function g(x) which is equal to the area enclosed
by the u-axis and the function f(u) and the lines u = a and u = x will be continuous on
[a,b] and differentiable on (a,b). Most prominently, while we differentiate the function
g(x), we will discover that it is equal to f(x). The above graph demonstrates the function
f(u) and the area g(x).
Self-Instructional
76 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

The Fundamental Theorem of Calculus: Part 2 NOTES

If f is a continuous function on [a,b], then

b

f  x  dx  F  b   F  a 
a

where F is any antiderivative of f.

If f is continuous on [a,b], the definite integral with integrand f(x) and limits a and
b is just equal to the value of the antiderivative F(x) at b minus the value of F at a. This
property permits us to simply resolve definite integrals, if we can locate the antiderivative
function of the integrand.
Part one and part two of the Fundamental Theorem of Calculus can be combined
as below.
Combining the Fundamental Theorem of Calculus Part 1 and Part 2
Let f be a continuous function on [a,b].
b

1. If g(x) = f  u  du, g’(x) = f(x)

a

2. f  x  dx  F  b   F  a  where F is any antiderivative of f

a

3.3 PROPERTIES OF DEFINITE INTEGRAL

If f(x) is continuous and has an integral over the (closed) interval [a,b] then we can
write:
b b
(1) f  x  𝑑𝑥 = − f  x  𝑑𝑥
a a

(2) f  x  𝑑𝑥 = 0
a
Self-Instructional
Material 77

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES b c b

(3) f  x  𝑑𝑥 =  f  x  𝑑𝑥 + f  x  𝑑𝑥 (for a<c<b)

a a c

b b

(4) k.f  x  𝑑𝑥 = k . f  x  𝑑𝑥 (for any scalar k)

a a

(i) If f(x) is continuous and even function over [a,-a] then,

a a

 f  x  dx = 2 f  x  dx
a 0

(ii) If f(x) is continuous and an odd function over [a,-a] then,

a

 f  x  dx = 0
a

(iii)

Steps for Evaluating Definite Integral

Let I = f  x  dxx
a

 First, find the indefinite integral,  f(x) dx = F(x) + C

 Substitute, x = b upper limit in this integral, i.e. F(b) + C
 Substitute, x = a lower limit in this integral i.e. F(a) + C
 Subtract, second {F(b)+c} from third {F(a)+C}

Example 3.1: Evaluate x dx

a
Self-Instructional
78 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

Solution: NOTES
b

I = x dx
a

Example 3.2: Evaluate:  x2

1
dx

Solution:

Example 3.3: Find the area between the regions of parabola y = x2 and straight line
y = |x| over the interval [-1,1] or {(x,y) x2  y  |x|}

Self-Instructional
Material 79

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Solution:
Given, y= x2 and y = |x| i.e, y=x or y = –x

Figure: 3.2

The area under consideration is

= Area OAB + Area OCD
= 2 *Area OAB (Because curve is symmetrical about the y axis)

1 1

= 2  xdx   x 2dx 
0 0 

Self-Instructional
80 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

NOTES
IN-TEXT QUESTIONS

1. 0  2 x2 dx =
a. 2
b. 2/3
c. 8/3
d. None of these
2. 0  2 (x2 + 3) dx equals.
a. 24/3
b. 25/3
c. 26/3
d. None of the above.
3. If  2x dx = f(x) + C, then f(x) is
a. 2x
b. 2x loge2
c. 2x / loge2
d. 2x+1/x+1

4. 1  2 dx/x2 equals
a. 1
b. -1
c. 2
d. ½

3.5 ECONOMIC APPLICATION

Integral Calculus can be applied in the various fields of economics. We primarily use
three major Concepts of Economics, such as Lorenz Curve and Gini Coefficient, Self-Instructional
Material 81

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Social Welfare in the economy and transformation of marginal functions into total
functions.

Rate of growth or sale

If the rate of growth or sale of a function is a known function of t say f(t) where t is a
time measure, then total growth or sale of a product over a time period t is given by,

Total sale =

Example 3.4: The rate of new product is given by f (x) = 100 – 90e–x where x is the
number of days the product is on the market. Find the total sales during the first four
days. (e–4=0.018)
Solution:

3.6 MARGINAL FUNCTION TO TOTAL FUNCTION

1. Cost Function:

It represents the relation between the input prices & the output quantity. Marginal
cost is the additional cost spent on producing one more unit of output.

dC  x 
MC = is the formula for marginal cost, where C(x) is the cost function.
Self-Instructional dx
82 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

C = C(x) =  MC dx + K where, K is the constant of integration, being the NOTES

constant it is the fixed cost.

2. Total Revenue:

It denotes the sales of a firm based on a quantity of goods. It is the total income of the
company and is calculated by multiplying the quantity of goods sold by the price of the
goods.

d
R(x) represents the total revenue and MR is the marginal revenue MR = R x
dx
R(x) =  MR dx + K
Where, K is the constant of integral
Demand function = P = R(x)/x
Example 3.5: If marginal revenue (MR) = 16 – q2 find the maximum total revenue,
also find the total, average revenue demand.
Solution:
When TR is maximum, then MR= 0

Therefore, Demand (AR) = 16 – q2/3

Example 3.6: The marginal revenue function of a commodity is stated as MR =
12 – 3x2 + 4x. Find the total revenue and the demand function.

Self-Instructional
Material 83

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Solution:
MR = 12 – 3x2 + 4x
MR = (12 – 3x2 + 4x) + K
MR = (12x – x3 + 2x2) [The constant of integral is 0 in this case.]
The above function is the revenue function.
Demand function:

IN-TEXT QUESTIONS
5. If the marginal revenue function of a firm is MR= e–x/10, then revenue is
(a) –10e-x/10
(b) 1 – e-x/10
(c) 10 (1 – e-x/10)
(d) e-x/10 + 10
6. The area bounded by the parabola y2 = 4x bounded by its latus rectum is
(a) 16/3 sq.units
(b) 8/3 sq.units
(c) 72/3 sq.units
(d) 1/3 sq.units

3.7 CONSUMER SURPLUS AND PRODUCER

SURPLUS

Consumer surplus is the difference between the price consumers are willing to pay for
a good service and the actual price. It is measured as the area between the demand
curve and the equilibrium price. It basically represents the benefit consumers get for
Self-Instructional purchasing goods at a price lower than the maximum they are willing to pay. Producer’s
84 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

surplus is the difference between what producers are willing & able to supply a good NOTES
for & the price they actually get. It is measured as the area between the Equilibrium
Price & the Supply Curve. It is the extra benefits producers get from selling a product
at a price that is higher than their minimum accepted price, as shown by the supply
curve.

Figure 3.3: Consumer and Producer Surplus

Since Social Welfare in the Economy is calculated as the sum of both Consumer
Surplus and Producer Surplus, we can find Society’s state of development using Integral
Calculus.
Consumer Surplus and Producer Surplus are calculated using the supply and
demand curves. Assuming that the equilibrium price is Pe and Equilibrium quantity be
Qe then, we use the demand and supply functions to calculate Consumer and Producer
Surplus, respectively.
Let,
Demand Function = D(q) and
Supply Function = S(q)

Self-Instructional
Material 85

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Then,

Example 3.7: If the demand function is P(x) = 30-20-x2, what will be the consumer
surplus (CS)?
Solution:

Figure 3.4:

Self-Instructional
86 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

NOTES
IN-TEXT QUESTIONS
7. When x0 = 5 and p0 = 3 the consumer surpluses for the demand function
pd = 28 – x2 is
(a) 250 units
(b) 250/3 units
(c) 251/2 units
(d) 251/3 units
8. When x0 = 2 and P0 = 12 the producer surplus for the supply function Ps =
2x2 + 4 is
(a) 31/5 units
(b) 31/2 units
(c) 32/3 units
(d) 30/7 units
9. Area bounded by y = x between the lines y = 1, y = 2 with y = axis is
(a) 1/2 sq.units
(b) 5/2 sq.units
(c) 3/2 sq.units
(d) 1 sq.unit
10. The marginal cost function is MC = 100x. find AC given that TC = 0 when
the output is zero

Example 3.8: Obtain the producer surplus, when the demand and supply function is
given;
Self-Instructional
D = 20 – 4x and S = 4+ 4x Material 87

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES

Figure 3.5:

20
Example 3.9: For a certain good the demand curve is p = D(q) = and the
q 1
supply curve is p = S(q) = q + 2. Find the equilibrium price & equilibrium quantity.
Then compute the consumer & producer surplus.

Self-Instructional
88 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

Solution: NOTES
To find the equilibrium quantity, we let D(q) = S(q) to obtain

20
 q+2
q 1

20 = (q + 1)(q + 2)
0 = q2 + 3q - 18
The positive solution gives the equilibrium quantity Qe = 3, and the equilibrium
price is Pe = 5.
We compute consumer and producer surplus using the above formulae:

Social Welfare in the Economy = Consumer Surplus + Producer Surplus =

12.73 + 4.50 = 17.23

Income Inequality and Lorenz Curve

The Lorenz Curve is a graph of income inequality that shows what percentage of a
country’s income is earned by a percentage of the country’s households. The X-axis
of a Lorenz Curve represents cumulative share of people from lower to higher income
groups or the percentage of households and the Y-axis represents the cumulative share Self-Instructional
of Income earned. Material 89

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES

Figure 3.6: Lorenz curve

The Line of equality shows a perfectly equal income distribution, i.e., X% of

households earn X% of total Income. This is depicted by the straight line BD. The
closer the Lorenz curve is to the Line of Equality, the more evenly distributed is the
income. The more it sags away, greater is the inequality and uneven distribution of
income.
The Gini Coefficient is a measure of Statistical dispersion for Income
Inequality. It is defined as a ratio wherein the numerator is the area between the
Lorenz curve and the Line of equality and the denominator is the area beneath the
Line of Equality. Its value ranges between 0 and 1. Here, 0 corresponds to perfect
income equality (i.e. everyone has the same income) and 1 corresponds to perfect
income inequality (i.e. one person has all the income, while everyone else has zero
income).
Example 3.10:
If the area between the line of perfect equality and Lorenz curve is A, and the area
under the Lorenz curve is E, then the Gini coefficient is A/(A+E). Since A+E =
0.5, which implies 2A = 1-2E = G (Gini Coefficient). If the Lorenz curve is
represented by the function Y = L(X), showing the proportion of total income
Self-Instructional
90 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

held by the poorest x% of households, the value of E can be found with NOTES
integration,

Therefore Gini Coefficient will be,

The line of inequality is given by y = x

Example 3.11:
Given L1(X) = x2 and L2(X) = (x + 2x3)/3. Find the Gini Coefficient of both the
Income Distributions.
Solution:

Self-Instructional
Material 91

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Example 3.12:

A company produces 50,000 units per week with 200 workers. The rate of change of
productions with respect to the change in the number of additional labor x is represented
as 300 – 5x2/3. If 64 additional labors are employed, find out the additional number of
units, the company can produce.
Solution:
Let p be the additional product produced for additional of x labor,

 The number of additional units produced are 16128.

Total number of units produced by 264 workers are
50000 + 16128 = 66128 units.
Hence, if 64 additional labors are employed, the company can produce 66128
units.
Example 3.13:
The price of a machine is 6,40,000 if the rate of cost saving is represented by the
function f(t) = 20,000t. Find out the number of years required to recoup the cost of
the function.
Solution:
Saving Cost:
t

S(t) =  20000t dt
0
Self-Instructional
92 Material = 10000 t2

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

To recoup the total price, NOTES

10000 t2 = 640000
t2 = 64
t =8
When t = 8 years, one can recoup the price.
Example 3.14:
The marginal revenue function (in thousands of rupees) of a commodity is
10 + e–0.05x Where x is the number of units sold. Find the total revenue from the sale of
100 units (e–5 = 0.0067)
Solution:
Given, Marginal revenue R(x) = 10 + e–0.05x
Total revenue from sale of 100 units is

Total revenue = 1019.87 × 1000

= 1019870
The total revenue from the sale of 100 units is 1019870
Example 3.15:
In year 2000 world gold production was 2547 metric tons and it was growing
exponentially at the rate of 0.6% per year. If the growth continues at this rate, how
many tons of gold will be produced from 2000 to 2013? (e0.078 = 1.0811)

Self-Instructional
Material 93

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Solution:
Annual consumption at time t = 0 (In the year 2000) is p0 = 2547 metric ton.
Total production of Gold from 2000 to 2013 =

Therefore, 34426.95 m tons of gold can be produced.

Example 3.16:
The rate of change of sales of a company after an advertisement campaign is represented
as, f (t) = 3000e–0.3t where t represents the number of months after the advertisement.
Find out the total cumulative sales after 4 months and the sales during the fifth month.
Also find out the total sales due to the advertisement campaign. (e–1.2 = 0.3012,
e–1.5 = 0.2231)
Solution:

Self-Instructional
94 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

NOTES

Example 3.17:
Mr. Rohit invests 10,000 in ABC Bank each year, which pays an interest of 10%
per annum compounded continuously for 5 years. How much amount will there be
after 5 years. (e0.5 = 1.6487)
Solution:
p = 10000, r = 0.1, N = 5

Self-Instructional
Material 95

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES

Rohit will get 64,870 after 5 years.

3.8 SUMMARY

 Integral Calculus is motivated by the problem of defining and calculating the

area of the region bounded by the graph of the functions.
 The functions that could possibly have given function as a derivative are called
anti derivatives (or primitive) of the function.
 The formula that gives all the anti derivatives is called the indefinite integral
of the function and such process of finding anti derivatives is called
integration.
 There is a connection, known as the Fundamental Theorem of Calculus, between
indefinite integral and definite integral which makes the definite integral as a
practical tool for science and engineering.
 The Fundamental Theorem of Calculus defines the relationship among the
processes of differentiation and integration.
 The inverse relationship between differentiation and integration means that, for
each statement about differentiation, we can write down a matching statement
regarding integration.
 In the method of Integration by substitution, we reduce an integral in non-standard
form into an integral in standard form by altering the variable into a new variable
with appropriate substitution.
Self-Instructional
96 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

 Integration by parts depends on the product rule for differentiation, for articulating NOTES
one integral in provisions of another. Integration by parts is used when we observe
two dissimilar functions that don’t appear to be associated to each other via a
substitution.

3.9 GLOSSARY

 Integration: The formula that gives all the anti derivatives is called the indefinite
integral of the function and such process of finding anti derivatives is called
integration.
 Integration by Parts: It is a method depending on the product rule for
differentiation, for articulating one integral in provisions of another.
 Producer surplus: Producer surplus is the difference between what the producer
is willing to receive to supply goods and the actual amount received.

3.10 ANSWER TO IN-TEXT QUESTIONS

1. (c) 8/3
2. (c) 26/3
3. (c) 2x / loge2
4. (d) ½
5. (c)10 (1 – e-x/10)
6. b
7. B
8. C
9. C
10. A
Self-Instructional
Material 97

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES
3.11 SELF-ASSESSMENT QUESTIONS

1. Illustrate how integration can be defined as an Inverse Process of

Differentiation.
2. Depict the concept of the Fundamental Theorem of Calculus with examples.
3. Evaluate the following integrals:

4. Find the consumer and producer’s surplus under pure competition, for demand
8 1
function p = – 2, and the supply function p = (x + 3) , where p is the
x 1 2
price and x is the quantity.
5. The demand function of a monopolist is x = 210 – 3p and his average cost
function is AC = x + 6 + 10/x, where p and x refers to the price and quantity of
the commodity, respectively. Determine consumer surplus at the price and quantity
which the monopoly would like to fix.
6. After producing 35 units of a product, the production manager determines
that the production facility is following the learning curve of the form fx =
300 – 190e–2, where f(x) is the rate of labor hours required to produce xth
unit. How many labor hours would be required to produce additional 25
units?
7. The supply function of a producer is given by: p = 0.4e2, where x denotes
thousand units. Find the producer surplus when sales are 2000 units.
8. After tuning up 50 cars, a company determines that its tuning facility follows a
learning curve of the form f(x) = 1000x–1, where f(x) is the rate of labor-hours
required to tune up the xth unit. How many total labor-hours should they estimate
are required to tune up an additional 50 cars?
Self-Instructional
98 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Definite Integral

9. After an advertisement campaign, the rate of ales of a product is given by NOTES

S(t) = 1000e–0.5, where t is the time in months. Find:
(i) Total cumulative sales after 2 months.
(ii) Sales during the 3rd month; and
(iii) Total sales as a result of the campaign
10. At t = 0, the annual world use of natural gas was 50 trillion cubic feet. The
annual gas consumption is increasing at the rate of 10% continuously. Find the
total consumption for the first 10 years and the consumption of 10th year.
11. The demand function p = 85 – 5x and supply function p = 3x – 35. Calculate
the equilibrium price and quantity demanded. Also calculate consumer surplus.
12. The demand and supply functions under perfect competition are
pd = 1600 – x2 and ps = 2x2 + 400 respectively. Find the producer surplus.
13. Under perfect competition for a commodity, the demand and supply laws are
pd = (8 / [x + 1]) – 2 and ps = [x + 3] / 2 respectively. Find the consumer and
producer surplus.
14. The rate of new product is given by f(x) = 100 – 90 e–x where x is the number
of days the product is on the market. Find the total sales during the first four
days. (e–4 = 0.018)

3.12 SUGGESTED READINGS

 Sydsaeter, K. Hammond, P. (2002), Mathematics for economic analysis. Person

Edition

Self-Instructional
Material 99

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

LESSON 4 NOTES

INTEGRATION

Sarabjeet Kaur
Assistant Professor,
Department of Economics,
Zakir Husain College (E), University of Delhi)
Structure
4.1 Learning Objectives
4.2 Introduction
4.3 Area under Curves
4.4 Integration as an Inverse Process of Differentiation
4.5 Integration by substitution
4.6 Integration by partial fraction
4.7 Marginal Propensity to Save
4.8 Summary
4.9 Glossary
4.10 Answer to In-Text Questions
4.11 Self-Assessment Questions
4.12 Suggested Readings

4.1 LEARNING OBJECTIVES

After studying this unit, you will be able to:
• Understand the Integration as inverse process of differentiation
• Illustrate the process of Integration by Substitution
• Understand the concept of Integration by parts

4.2 INTRODUCTION
Differential Calculus is centred on the concept of the derivative. Differentiation
is the process used to calculate the instantaneous rate of change of a function
whereas integration is the reverse operation of differentiation. Together,
differentiation and integration make up the essential operations of calculus and Self-Instructional
Material 101
are related by the fundamental theorems of calculus.

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES As differentiation of a function F(x) gives another function f(x) or F′(x)

= f (x). Reverse operation, if f (x) is given, we get back the original function
F(x). This reverse operation to the derivation is called anti-derivation or
integration. Historically, the concept of integration was developed before that of
differentiation. Much later, it was pointed out that differentiation and integration
are reverse of one another.

4.3 AREA UNDER CURVES

There are several ways to calculate the area under the curve, but the antiderivative
approach is the most widely used. Knowing the curve’s equation, its bounds,
and its enclosing axis will allow you to determine the area under the curve. In
general, there are formulas for calculating the areas of geometrical forms like
squares, rectangles, quadrilaterals, polygons, and circles, but there are no formulas
specifically designated for calculating the area of curves. The integration method
helps in equation solution and area determination.
In simple terms, integration is applied to determine the area under a curve.
It can also be found out geometrically.  However, the idea of integration and
differentiation does not depend on geometry as analytically; rather, a mathematical
interpretation is needed solely to comprehend intuitively.
Suppose y as a function of x (i.e. y = f (x)) be a continuous and positive
function on the closed interval [a, b] as shown in the figure (4.1) and we have to
compute area (A) under the given graph.

Self-Instructional
102 Material
Figure: 4.1

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

Let A(x) be the area under curve y = f (x) on the closed interval [a, x]. NOTES
Since, A(a) = 0; as, there is no area starting from ‘a’ till ‘a’ and the total
area can be defined as A = A(b).
Now, we suppose that ‘x’ increases by small amount; say ∆x. Then, A(x +
∆x) is the area under curve y = f (x) over the closed interval [a, x + ∆x]. Hence,
the required area is given by A(x + ∆x) − A(x). It is the area {∆A} under the
curve y = f (x) over the closed interval [x, x + ∆x]. Let ∆A be very small i.e.
magnified and this area can not be exceed the area of rectangle with edges ∆x
and f(x + ∆x) and cannot be lesser than area of the rectangle with edges ∆x and
f(x). Hence, ∀ x>0;

If we take ∆x →0 in the above equation, then the interval [x, x + ∆x]

shrinks to the single point ‘x’ and the value f(x + ∆x) approaches f(x). So, the
function A(x) is differentiable and it measures the area under the curve y = f(x)
over the closed interval [a, x]. Then, the derivative of the function is given by;

Therefore, we can conclude that the derivatives of the area function A(x)
is a curve height function {i.e. y = f (x)}.
Now, suppose F(x) is another continuous function with the function y = f(x)
as its derivative;

Then, F’(x) = A’(x) = f(x); ∀x € (a, b)

Because,

d
 A ( x ) − F ( x )  = A′ ( x ) − F′ ( x ) = 0
dx 
It must also be true that,
A(x) = F(x) +C {C is some constant}
Self-Instructional
Material 103

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES If A(a) = 0, then

A (a) = F(a) + C = 0
Or C = -F(a), put this value in above equation

In short, the method for finding the area under the curve y = f(x) and its
domain (a,b) or above the x–axis from x = a to x = b has following steps:
▪ Find an arbitrary function F(x), that is continuous over the interval (a, b)
such that
F’(x) = f(x) ∀ x∈(a,b)  ...(4.1)
▪ Then the required area of the function is given by
A(x) = F(b) –F(a)  ...(4.2)
Further, if the value of the function has negative value in the closed interval
[a, b]. then, the required area is A(x) = -(F(b) –F(a)) or take the modulus of the
area. The area of the curve y = f(x) below the x-axis and bounded by the x-axis
is obtained by taking the limits a and b. A = |∫f(x) dx|. As the area of a region is
always positive, hence, A(x) is also positive.

Area of the Curve below the axis

Figure: 4.2

Self-Instructional
104 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

The area of the curve that is partially above and partially below the axis is NOTES
split into two portions and calculated separately. Since the area below the axis
is negative, the modulus is used. As a result, the total area equals the sum of the
two areas (A = |A1| + A2).

Figure: 4.3

Example 4.1:

Find the area under the straight line y = x over the interval [0,1]

Figure: 4.4
Self-Instructional
Material 105

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Solution:

To find out the area (A) in the figure given, we will use above equations (4.1)
and (4.2), we must find a function that has as its derivative x;

x2
F(x) =
2
d 1
[ as ax n anx
= = n −1
x, here,
= n 2 and
= a ]
dx 2
x
F’(x) = 2 =x
2
Thus, the required area is given by;

1 1
A= F(1) – F(0) = −0 =
2 2
Example 4.2:
Find out the area under the parabola; y= (1/3)x2 over the interval [a, b].

Figure: 4.5

Solution:
To find out the shaded area A in the given figure (4.5), we will use equations
(4.1) and (4.2) given above as we have to find a function, that has x as its
Self-Instructional derivative.
106 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

Let, NOTES

1 2
F(x) = x
3
1 2 2
Then, F’(x) =3* x =x
3
Thus, the required area is given by

1 2 1 2 1 2 2
A = F(b) – F(a) = b - a = (b – a )
3 3 3

IN-TEXT QUESTIONS
1. Find the area under the graph of polynomial y = x3 over the interval [0,1].
2. Find the bounded area of the graph of function y = (ex + e-x) over the
interval (-1,1).
3. Find the area under straight line, y = cx + d over the interval [0,1].
4. Compute the area under the parabola y = 4x2 over the interval [0,1].

4.4 INTEGRATION AS AN INVERSE PROCESS OF

DIFFERENTIATION

Integration is defined as the inverse process of differentiation. Rather than

differentiating a function, we are provided with the derivative of a function, and
requested to locate its primitive, i.e., the original function. This type of process
is known as Integration or anti differentiation.
When you integrate some function f(x), you find its antiderivative function,
which is often denoted F(x). This function can ‌calculate the area underneath the
curve of f(x).

Self-Instructional
Material 107

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Notation for integrating

The symbol for ‘Integral’ is a stylish “S”.

Interpreting notation

• The symbol ∫ indicates that we are calculating the antiderivative function

of f(x); known as the integral sign. It is a distorted form of ‘S’ meaning
“sum”. This is because originally integration was defined as the sum of a
certain infinite series.
• The function f(x) is called the integrand. It is a function we want to integrate.
• The differential dx indicates that integration f(x) is being done with respect
to the variable x.
• f(x)dx means integral of the function f(x) i.e. integral of the integrand.
• The process of finding integral of a function is called integration.
• F(x) is the antiderivative function which provides back f(x) when
differentiated.
• The letter C represents a constant value and is known as the constant of
integration.
• An indefinite integral is a family of functions of type y = F(x) + C. From
geometrical point of view, an indefinite integral is a collection of curves,
each of which is obtained by translating one of the curves parallel to itself,
upwards or downwards (along y-axis).
Self-Instructional
108 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

F(x) as the Antiderivative function NOTES

When you take the derivative of F(x), you get back f(x) again. That is, the
derivative/ differential coefficient of integration of a function is the function
itself. Incidentally, this is the first rule of integration. Their relationship can be
thought of like this:
If ∫f(x)dx = F(x)+C, then
F’(x) = f(x)
Or, it might be easier to think about it like this:
∫f’(x)dx = f(x)+C
Example 4.3: Find anti-derivative of the function f(x) = x3

Solution:
d 3
Since,
dx
( )
x = 3x 2 , therefore, 3x2 dx = x3 +C

Figure: 4.6

Standard Integration Rules

Assuming that f and g are continuous functions, here is a list of the most
important integration rules and properties that you should know:

1. Constant Rule
∫adx= ax + C for some constant a
2. Constant Multiplier Rule
∫kf(x)dx = k ∫f(x)dx for some constant k
Self-Instructional
3. Power Rule Material 109

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES
x n +1
∫xndx = ​+ C for some real number n where n≠1
n +1
4. Sum Rule
∫[f(x)+g(x)]dx = ∫f(x)dx + ∫g(x)dx
In general,

5. Difference Rule
∫[f(x) − g(x)]dx = ∫f(x)dx − ∫g(x)dx
In general,

6. Reciprocal Rules

1
∫ dx = ∫ (x – 1) dx = ln∣x∣ + C
x
1 1
∫ dx = ​ln(ax + b) + C
ax + b a
7. Exponential and Logarithmic Function Rules
∫ex dx = ex + C
∫ax dx = ln(a)ax​+ C, for any positive real number a
∫ln(x)dx = x ln(x) − x + C
8. Integral of Multiplication/integration by part.

Self-Instructional
110 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

9. Other results NOTES

Example 4.4: f(x) = x

Solution:
x2
∫f(x) dx = ∫x dx = +C
2
Example 4.5: Find the integral ∫ (5x4 + 3x2 + 2x - 1) dx
Solution:
∫ (5x4 + 3x2 + 2x - 1) dx

= ∫ 5x4 dx + ∫ 3x2 dx + ∫2x dx - ∫1dx

= 5∫ x4 dx + 3∫ x2 dx + 2∫x dx - ∫dx

x5 x3 x2
=5 +C1 + 3 + C2 + 2 + C3 – x + C4
5 3 2
= x5 + x3 + x2 – x + C1 + C2 + C3 + C4
= x5 + x3 + x2 – x + C [C = C1 + C2 + C3 + C4]

Self-Instructional
Material 111

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Example 4.6: Evaluate

Solution:

Example 4.7: Integrate

Solution

This method is used when basic rules of integration, given above are not directly
applicable. In such cases, the integrand is transformed (by suitable substitution
of a given variable by a new variable) into a form in which basic rules are
applicable. The basic idea underlying this method is the use of a function of a
function (or chain rule).
This method is explained through the following example. This method
is used when basic rule of integration, given above are not directly applicable.
In such cases, the integrand is transformed (by suitable substitution of a given
variable by a new variable) into a form in which basic rules are applicable. The basic
idea underlying this method is the use of a function of a function (or chain rule).
This method is explained through the following example. This method
Self-Instructional is used when basic rule of integration, given above are not directly applicable.
112 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

In such cases, the integrand is transformed (by suitable substitution of a given NOTES
variable by a new variable) into a form in which basic rules are applicable. The
basic idea underlying this method is the use of a function of a function (or chain rule).

Example 4.8: Evaluate

Solution

Self-Instructional
Material 113

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES IN-TEXT QUESTION

5.

6.

7. Derivative of a function is unique but a function can have infinite
antiderivatives. State true or false.
8. Find ∫(ax + b)3dx = ___________

4.5 INTEGRATION BY SUBSTITUTION

Self-Instructional This method is used when basic rules of integration are not applicable. Here,
114 Material the integrand is transformed (by suitable substitution of a given variable by a

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

new variable) into a form in which basic rules are applicable. The basic idea NOTES
underlying this method is the use of a function of a function (or chain rule).
This method is explained through the following example:
Example 4.9: Evaluate ∫(2x2 + 7)5 .4x.dx
Solution:

I = ∫ (2x2 + 7)5 .4x.dx

Put 2x2 + 7 = u, then 4x.dx =du
Therefore I = ∫u5du = (u5 + 1)/5+1 + c = (2x2 +7)6/6 + c
Example 4.10: ∫ x(x 2 + 1)15 dx

Solution:

Put x2 + 1 = t
2x dx = dt
Therefore, I = 1/2∫t 15 dt = (1/ 2) t 16/16 + C
= t 16/32 + C

Integration by parts

The method of integration by parts is based on the product rule of differentiation.

We know that if u and v are two functions of x, then
d(uv)/dx = u.dv/dx + v.du/dx
Integrating both sides, w.r.t. x we get
uv = ∫u(dv/dx).dx + ∫v(du/dx).dx
Or ∫u.dv = uv - ∫v.du  ...(4.3)
Let u = f(x) and dv = g(x).dx
Therefore, du = f′(x) dx, v = ∫g(x) dx
Substituting these values in equation (4.3), we have
∫f(x).g(x).dx = f(x) ∫g(x).dx - ∫ [∫g(x) dx] f′(x) dx
Self-Instructional
Or ∫f(x) g(x)dx = f(x) ∫g(x)dx - ∫f′(x) [∫g(x)dx] dx  ...(4.4) Material 115

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Taking f(x) as the first and g(x) as the second function the above equation
can be stated in words as:
The integral of the product of two functions = first function ×integral of
second - integral of [differential coefficient of first ×integral of second]
Note: To apply this formula the integrand should be expressible as a product of
two functions such that one of them can be easily integrated. This function is
taken as the second function.

Example 4.11: Evaluate

Solution:

Self-Instructional
116 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

NOTES

4.6 INTEGRATION BY PARTIAL FRACTIONS

This method is used for the integration of rational functions. A rational

function f(x) is defined as the ratio of two polynomials say p(x)/q(x), q(x)
≠ 0. If the degree of p(x) is less than the degree of q(x) then f(x) is called
proper rational function otherwise it is called an improper rational function.

We shall state without proof that every proper rational function can be
expressed as a finite sum of partial fractions.
If a rational function is improper it can be expressed as a sum of polynomial
(which can be easily integrated) and a proper rational function by division of
polynomial in numerator by the polynomial in denominator.
The method of integration by partial fractions is explained through the
following example:
Example 4.12: Evaluate

Self-Instructional
Material 117

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Solution:

Now compare both sides of the equation; 2x (A+B) + 3 (A-B) =1

Hence 2 (A+B) = 0 or A= -B and 3(A-B) = 1 or B = -1/6 and A=1/6, now
by equation (i)

Example 4.13: Evaluate ∫x -1 / (x + 1) (x – 2) dx

Solution:

A B
x - 1 / (x + 1) (x – 2) = +  ...(1)
x +1 x − 2
Therefore, x-1 = A (x-2) + B(x+1)  ...(2)

Self-Instructional
118 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

Putting, x-2 = 0 or x = 2 in (2) we get, B = 1/3 NOTES

Putting, x+1 = 0 or x = -1 in (2) we get, A= 2/3
Substituting A and B in (1) we get,
x −1 2 1 1 1
= + .
( x + 1)( x − 2 ) 3 x + 1 3 x − 2
Thus,

x −1 2 1 1 1
= dx dx + . dx
( x + 1)( x − 2 ) 3 x +1 3 x−2
2 1
= log ( x + 1) + log ( x − 2 ) + C
3 3

Economic Application

Cost functions
TC = (FC+VC) is the total cost of producing an output x,
dTC
marginal cost function MC =
dx
Using integration, as the reverse process of differentiation, we obtain,
Cost function TC = ∫ (MC) dx + k
Where k is the constant of integration which is to be evaluated.
Average cost function AC = C/X, x ≠ 0

Example 4.14:

The marginal cost function of manufacturing x shoes is 6 + 10x − 6x2. The cost
of producing a pair of shoes is ₹12. Find the total and average cost function.

Self-Instructional
Material 119

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Solution:

Given,
Marginal cost MC = 6 + 10x − 6x2
C = ∫ MC dx + k
= ∫ (6 + 10x − 6x2) dx + k
= 6x + 5x2 − 2x 3 + k
When x = 2, C = 12 (given)
12 = 12 + 20 −16 + k
k = -4
C = 6x + 5x2 − 2x3 – 4

Example 4.15: If marginal propensity to consume (MPC) function is given as

follows;

Dc
= 0.5 – 0.001y ; Then find total consumption function. Given at
dy
income zero, c is 0.02.
Solution:

Self-Instructional
120 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

NOTES
IN-TEXT QUESTIONS
9. The demand and supply functions are given by D (x) = 16 − x2 and
S (x) = 2x2 + 4 are under perfect competition, then the equilibrium
price x is
(a) 2
(b) 3
(c) 4
(d) 5
10. If MR and MC denotes the marginal revenue and marginal cost
functions, then the profit functions is
(a) P = ∫ (MR − MC) dx + k
(b) P = ∫ (MR + MC) dx + k
(c) P = ∫ (MR) (MC) dx + k
(d) P = ∫ (R – C) dx + k
11. The marginal revenue and marginal cost functions of a company are
MR = 30 − 6x and MC = −24 + 3x where x is the product, then the
profit function is
(a) 9x2 + 54x
(b) 9x2 − 54x
(c) 54x − 9x2/2
(d) 54x – [9x2/2] + k
12. If the marginal revenue MR = 35 + 7x − 3x2, then the average revenue
AR is
(a) 35x + 7x2/2 − x3
(b) 35 + 7x/2 − x2
(c) 35 + 7x/2 + x2
(d) 35 + 7x + x2

Self-Instructional
Material 121

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 4.7 MARGINAL PROPENSITY TO SAVE

It is the fraction of a rise in income that is not spent on an increase in consumption.

Integration when used upon marginal propensity to save (MPS) helps us to find
the savings function. S=∫𝑀𝑃𝑆 𝑑𝑌+𝐶

Example 4.16:

𝑑𝑆/𝑑𝑌=0.3−0.1𝑌-1/2, if the aggregate savings S is nil when Y=81, find S(Y).

Solution:

(𝑌) = ∫ (0.3−0.1𝑌−1/2) 𝑑𝑌
= 0.3𝑌−0.2𝑌1/2+𝐶
Taking S = 0, Y = 81
0 = 0.3(81) − 0.2√81+𝐶
C = -22.5
Hence, the desired function is (𝑌) = 0.3𝑌 − 0.2𝑌1/2 − 22.5

Example 4.17:

The marginal cost and marginal revenue with respect to commodity of a firm are
given by C′(x) = 8 + 6x and R′(x) = 24. Find the total profit given that the total
cost at zero output is zero.

Solution:

Self-Instructional
122 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

NOTES

4.8 SUMMARY

• Integral Calculus is motivated by the problem of defining and calculating

the area of the region bounded by the graph of the functions.
• The functions that could possibly have given function as a derivative are
called anti derivatives (or primitive) of the function.
• The formula that gives all the anti derivatives is called the indefinite
integral of the function and such process of finding anti derivatives is
called integration.
• There is a connection, known as the Fundamental Theorem of Calculus,
between indefinite integral and definite integral which makes the definite
integral as a practical tool for science and engineering.
Self-Instructional
Material 123

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES • The Fundamental Theorem of Calculus defines the relationship among the
processes of differentiation and integration.
• The inverse relationship between differentiation and integration means that,
for each statement about differentiation, we can write down a matching
statement regarding integration.
• In the method of Integration by substitution, we reduce an integral in non-
standard form into an integral in standard form by altering the variable
into a new variable with appropriate substitution.
• Integration by parts is a method depending on the product rule for
differentiation, for articulating one integral in provisions of another.
Integration by parts is used when we observe two dissimilar functions that
don’t appear to be associated to each other via a substitution.

4.9 GLOSSARY

• Anti-derivatives: The functions that could possibly have given function

as a derivative are called anti derivatives (or primitive) of the function.
• Integration: The formula that gives all the anti-derivatives is called
the indefinite integral of the function and such process of finding anti
derivatives is called integration.
• Integration by Parts: It is a method depending on the product rule for
differentiation, for articulating one integral in provisions of another.
• Indefinite Integral: The Indefinite Integral is the Anti-derivative, the
inverse operation to the derivative. F is an indefinite integral of f when f
is a derivative of F.
• Consumer’s Surplus: This notion was introduced by Alfred Marshall to
measure the net benefit that a consumer enjoys from his act of purchasing
a particular commodity in the market. It is defined in terms of the excess
of the consumer’s total willingness to pay in units of money over his actual
expenditure.
Self-Instructional
124 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

• Producer surplus: Producer surplus is the difference between what the NOTES
producer is willing to receive to supply goods and the actual amount
received.
• Equilibrium: If an initial value has a solution that is a constant function
(i.e., independent of t), then the value of the constant is called an equilibrium
state or stationary state of the equation.

4.10 ANSWER TO INTEXT QUESTIONS

1. ¼
2. (e-1/e)
1. ½ (a+b)
2. 4/3
3. b
4. d
5. true
6.
7. A
8. A
9. D
10. B

4.11 SELF-ASSESSMENT QUESTIONS

1. Illustrate how integration can be defined as an Inverse Process of

Differentiation.
2. Depict the concept of the Fundamental Theorem of Calculus with examples.

Self-Instructional
Material 125

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 3. Evaluate the following integrals:

4. Calculate consumer surplus if the demand function p = 50 − 2x and x = 20.
5. Calculate consumer surplus if the demand function p = 122 − 5x − 2x2 and
x = 6.
6. The demand function p = 85 − 5x and supply function p = 3x − 35. Calculate
the equilibrium price and quantity demanded. Also calculate consumer’s
surplus.
7. Calculate the producer surplus at x = 5 for the supply function p = 7 + x.
8. If the supply function for a product is p = 3x + 5x2 .Find the producer’s
surplus when x = 4.
9. The marginal revenue of a company is given by MR=100+20Q+3Q2,
where Q is amount of units sold for a period. Find the total revenue function
if at Q = 2 it is equal to 260.
10. The marginal cost function of firm is MC = (log x)2. Find the total cost of
100 units if the cost of producing one unit is `22.
11. The additional cost (in lakhs of rupees) of producing a motor car is given
by 6+4x2+1.5e- x, where x is the quantity produced. Determine the total
cost of producing 5 motor cars if the fixed cost is ` 7 lakhs. It is given that
e-5 = 0.006.
4
12. If the Marginal revenue function for output x is given by MR = 2 −1,
( 2 x + 3)
find the total revenue function and the demand function.

Self-Instructional
126 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Integration

NOTES
3p
13. The price elasticity of demand for a commodity is 𝜋𝑑 = ( p − 1)( p + 2 ) .
Find the corresponding demand function if the quantity demanded is 8
units when p = `2.
14. The marginal cost function of manufacturing x shoes is 6 + 10x − 6x2. The
cost of producing a pair of shoes is ₹12. Find the total and average cost
function.
15. A company has determined that the marginal cost function for a product
of a particular commodity is given by MC = 125 + 10x − x2/9 where C
rupees is the cost of 9 producing x units of the commodity. If the fixed cost
is ₹250 what is the cost of producing 15 units.
16. The marginal cost function MC = 2 + 5ex (i) Find C if C (0) =100 (ii)
Find AC.

4.12 SUGGESTED READINGS

• Sydsaeter, K. Hammond, P. (2002), Mathematics for economic analysis.

Person Edition

Self-Instructional
Material 127

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

LESSON 5 NOTES

DIFFERENTIAL EQUATIONS

Sarabjeet Kaur
Assistant Professor,
Department of Economics,
Zakir Husain College (E), University of Delhi)
Structure
5.1 Learning Objectives
5.2 Introduction
5.3 Order and Degree of a Differential Equation
5.4 Linear and Non-Linear Differential Equations
5.5 Homogeneous differential equations
5.6 Non -Homogeneous differential equations
5.7 Variable Separation Method
5.8 General and Particular Solutions
5.9 Stability and Phase
5.10 Second order differential equation
5.11 Summary
5.12 Glossary
5.13 Answers to In-Text Questions
5.14 Self-Assessment Questions
5.15 Suggested Readings

5.1 LEARNING OBJECTIVES

After studying this lesson, you will be able to:

• Define a differential equation, its order and degree.
• Determine the order and degree of a differential equation.
• Form differential equation from a given situation.
• Illustrate the terms “general solution” and “particular solution” of a
differential equation through examples. Self-Instructional
Material 129
• Solve the first order and second order differential equations.

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 5.2 INTRODUCTION

A differential equation is an equation that relates a function and its derivatives.

dy
Equations which involve one or more differential coefficients such as ,
2 dx
d y
(or differentials) etc. and independent and dependent variables are called
dx 2
differential equations. A first order differential equation is an equation that
involves only the first derivative of an unknown function. The general form
of a first order differential equation is:
dy
= f(x,y)
dx
where y is the function and f(x,y) is some function of x and y.
A differential equation involving derivatives of the dependent variable
with respect to only one independent variable is called an ordinary differential
equation and a differential equation involving derivatives with respect to more
than one independent variables is called a partial differential equation.
Note: The differential equations are solved and the solutions thus obtained
are interpreted in the context of the problem.

5.3 ORDER AND DEGREE OF A DIFFERENTIAL

EQUATION

The order and degree of a differential equation are important concepts in

differential equations. They help to classify and determine the complexity of
a differential equation.
The order of a differential equation refers to the highest order derivative
that appears in the equation. The degree of a differential equation is the power
to which the highest derivative is raised after the equation has been put into
standard form. The standard form is obtained by expressing the equation so
that the highest derivative appears alone on one side of the equation, and the
other terms are on the other side. In other words, the degree of a differential
equation refers to the highest power (positive integer only) of the highest
130
Self-Instructional
Material
order derivative that appears in the equation.

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

NOTES
2
d y dy
For example, the differential equation 2
–3 + 2 y = 0 is a second
dx dx
order differential equation because the highest order derivative that appears
d2 y
is , and the degree of the equation is 2 because the highest power of
dx 2
d2 y
that appears is 1.
dx 2
Here are some examples of differential equations with their orders and
degrees:

The order and degree of a differential equation are important because they
help us to determine the type of solution that we should expect. Generally, higher
order and degree differential equations are more difficult to solve and may require
advanced techniques. In addition, some types of differential equations may not
have solutions at all.
Example 5.1: Find the order and degree of the differential equation:

Self-Instructional
Material 131

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Solution:

The given differential equation is

Or

The above equation has fractional power. Therefore, we first square both
sides to remove fractional index. Squaring both sides, we have:

Hence each of the order of the differential equation is 2 and the degree
of the differential equation is also 2.
Note: Before finding the degree of a differential equation, it should be free from
radicals and fractions as far as derivatives are concerned.

Classification of Differential Equations

A differential equation can be classified as ordinary or partial differential equation.

A differential equation involving a single independent variable and the derivatives
with respect to it, is called an ordinary differential equation. An ordinary
differential equation contains total derivatives or total differentials. For example,

Self-Instructional
132 Material
A differential equation involving two or more independent variables and
© Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Differential Equations

the partial derivatives with respect to them, is called a partial differential NOTES
equation. For example,

5.4 LINEAR AND NON-LINEAR DIFFERENTIAL

EQUATIONS

A differential equation in which the dependent variable and all of its derivatives
occur only in the first degree and are not multiplied together is called a linear
differential equation. A differential equation which is not linear is called non-
linear differential equation. For example, the differential equation:

d2 y dy
2
–3 +2y=0
dx dx
2
 d2 y  dy
The differential equation  2  – 3 + 2 y = 0 is non-linear as it has
 dx  dx
d2 y
degree of two. Further, the differential equation y + 2 = x is non-linear
dx 2
d2 y
because the dependent variable y and its derivative are multiplied together.
dx 2

Self-Instructional
Material 133

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES IN-TEXT QUESTIONS

Fill in the blanks:
1. A ................................ involves independent variables, dependent
variables, their derivatives and constants.
2. Ordinary differential equations are those, which involve only one ......
.......................... variable.
3. The process of ................................ formulations of natural phenomena
leads to a differential equation.
State whether the following statements are true or false:
4. Ordinary differential equations involve more than one independent
variable.

Formation of a Differential Equation

Consider the family of all straight lines passing through the origin (see Fig. 5.1).
This family of lines can be represented by
y = mx  ...(5.1)
Differentiating both sides, we get

dy
= m ...(5.2)
dx
Substituting (5.2) in (5.1), we get

dy
y= x  ...(5.3)
dx
dy
So y = mx and y = x represent the same family.
dx
Equation 5.3 is the differential equation.

Self-Instructional
134 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

NOTES

Figure 5.1

Working Rule:

To form the differential equation corresponding to an equation involving two

variables, say x and y and some arbitrary constants, say, a, b, c, etc.
(i) Differentiate the equation as many times as the number of arbitrary
constants in the equation.
(ii) Eliminate the arbitrary constants from these equations.

IN-TEXT QUESTIONS
Fill in the blanks:
5. A differential equation can be classified as an ordinary or ...................
............. differential equation.
6. A differential equation involving a single independent variable and the
derivatives with respect to it, is called an ................................ differential
equation.
7. A differential equation involving ................................ independent
variables and the partial derivatives with respect to them, is called a
partial differential equation.

Self-Instructional
Material 135

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Example 5.2: Form the differential equation representing the family of curves.

y = ax2 + bx  ...(1)
Differentiating both sides, we get

dy
= 2 ax + b  ...(2)
dx
Differentiating again, we get
d2 y
= 2 a ...(3)
dx 2
1 d2 y
a=  ...(4)
2 dx 2
(The equation (1) contains two arbitrary constants. Therefore, we
differentiate this equation two times and eliminate ‘a’ and ‘b’).
On putting the value of ‘a’ in equation (2), we get
dy d2 y
= x 2 + b ...(5)
dx dx
dy d2 y
b
= − x 2  ...(6)
dx dx
Substituting the values of ‘a’ and ‘b’ given in (4) and (5) above in
equation (1), we get

Self-Instructional
136 Material which is the required differential equation.

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

Note: If an equation contains one arbitrary constant then the corresponding NOTES
differential equation is of first order and if an equation contains two arbitrary
constants then the corresponding differential equation is of second order and so on.

5.5 HOMOGENEOUS DIFFERENTIAL EQUATIONS

A differential equation is said to be homogeneous differential equation if it

is of the form

where P(x, y) and Q(x, y) are homogeneous functions of equal degree in

variables x and y.
In order to solve homogeneous differential equations we need to follow
mainly three following steps.
1. Put y = vx in the given differential equation and evaluate dy/dx.
2. Substitute the values of y and dy/dx in the main equation and bring the
equation in the form of separable variable.
3. Solve by the method of separable variable
Note: Homogeneous equations do not have constant terms.

5.6 NON-HOMOGENEOUS DIFFERENTIAL

EQUATION

A non-homogeneous differential equation is a differential equation that

involves both a function and its derivatives, as well as a non-zero function of
the independent variable. The general form of a non-homogeneous differential
equation of order n can be written as:

dy
+ ay = b, b ≠ 0
dx Self-Instructional
Material 137

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Solving a non-homogeneous differential equation involves finding a

particular solution to the equation that satisfies the non-homogeneous part
f(x), as well as finding the general solution to the corresponding homogeneous
differential equation (i.e., the differential equation obtained by setting f(x) equal
to zero). The general solution to the non-homogeneous differential equation is
then given by the sum of the particular solution and the general solution to the
homogeneous equation.

5.7 VARIABLE SEPARATION METHOD FOR

SOLVING DIFFERENTIAL EQUATION

The separation of variables method involves rearranging the equation so that all
the y terms are on one side and all the x terms are on the other side, and then
integrating both sides with respect to x and y, respectively. The general solution
can then be found by solving for y.
Steps to solve a differential equation using the variable separation method:

1. Rewrite the differential equation in the form dy/dx = f(x) g(y), where y is
the unknown function and f(x) and g(y) are known functions.
2. Separate the variables by dividing both sides by g(y) and dx, then integrate
both sides with respect to x and y, respectively, to obtain ∫(1/g(y)) dy =
∫f(x) dx + C, where C is the constant of integration.
3. Solve for y by evaluating the integral on the left-hand side, which may
require using integration techniques such as substitution or partial fractions.
4. Obtain the general solution by adding the constant of integration C to the
expression for y.
Also note that while the variable separation method can be used to solve
only homogenous differential equations, it may not work for more complex
or nonlinear differential equations. In such cases, other techniques such as the
method of integrating factors or numerical methods may be required

Self-Instructional
138 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

Example 5.3: (x2 +y2)dx – 2xydy = 0 NOTES

Solution:
Put y = vx we get dy/dx = v+ x dv/dx
Substitute these values in equation above equation, we get

Which is now in the separable variable form. So, the solution can be
obtained by direct integration. Integrating both side we get,

-log (1 – v2) = log x + log c where c is an arbitrary constant

log (1 – v2) + log x = log c′, where c′ = c-1
log (x (1-v2) = c′
Now substitute the value of v in above equation, we get
x2 - y2 = c′
Which is the general solution of the given differential equation.

Self-Instructional
Material 139

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 5.8 GENERAL AND PARTICULAR SOLUTIONS

Finding the solutions of a differential equation is a reverse process. Here we try

to find an equation which gives rise to the given differential equation through the
process of differentiations and elimination of constants. The equation so found
is called the primitive or the solution of the differential equation.
In differential equations, a general solution is a solution that contains a set
of arbitrary constants that are determined by initial or boundary conditions. A
particular solution is a solution that is obtained by setting the arbitrary constants
in the general solution to specific values that satisfy the given initial or boundary
conditions.

Note:

(1) If we differentiate the primitive, we get the differential equation and if we

integrate the differential equation, we get the primitive.
(2) Solution of a differential equation is the one which satisfies the differential
equation.

General Solution

To find the general solution of a differential equation, we need to integrate the

equation with respect to the independent variable. For example, consider the
first-order differential equation:
dy/dx = f(x)
To find the general solution of this equation, we integrate both sides
with respect to x:
∫dy = ∫f(x)dx
The indefinite integral on the left-hand side gives us the general solution:
y = F(x) + C
where F(x) is the antiderivative of f(x), and C is the arbitrary constant
of integration. The general solution contains an arbitrary constant, which can
take any value.
Self-Instructional
140 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

Particular Solution NOTES

To find the particular solution of a differential equation, we need to use the initial
or boundary conditions to determine the value of the arbitrary constant in the
general solution. For example, consider the same first-order differential equation:
dy/dx = f(x)
with the initial condition y(0) = y0. To find the particular solution that
satisfies this initial condition, we first find the general solution:
y = F(x) + C
Next, we use the initial condition to determine the value of the arbitrary
constant C:
y(0) = F(0) + C = y0
Therefore, the particular solution that satisfies the initial condition is:
y = F(x) + y0
where y0 is the value of y at x = 0.
Similarly, for a second-order differential equation, we need two initial
or boundary conditions to determine the values of the two arbitrary constants
in the general solution.
Example 5.4: Find the particular solution of

when y (0) = 3 (i.e. when x = 0, y = 3).

Solution:
The given differential equation is

(3y2 +1) dy = 2xdx (variable separated) …..(1)

integrate both sides of (1), we get
∫ (3y2 +1) dy = ∫ 2xdx
Self-Instructional
where C is an arbitrary constant. Material 141

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES y3 + y = x2 + C  ....(2)
It is given that, y(0) = 3.
On substituting y = 3 and x = 0 in (2), we get
27 + 3 = C
C = 30
Thus, the required particular solution is
y3 + y = x2 +30

Solution for non-homogenous differential equations

A non-homogeneous differential equation is a differential equation that includes

a forcing function or source term, which depends on the independent variable. To
solve a non-homogeneous differential equation, we need to find both the general
solution and the particular solution.

General Solution:

The general solution of a non-homogeneous differential equation consists of two

parts: the complementary function (CF) and the particular integral (PI).
The CF is the solution of the corresponding homogeneous differential
equation, obtained by setting the forcing function to zero. It represents the
solution of the differential equation in the absence of any external influences.
The CF is determined by solving the homogeneous differential equation using
standard techniques such as separation of variables, substitution, or the method
of integrating factors.
The PI is a particular solution of the non-homogeneous differential equation,
which represents the effect of the forcing function. It is a function that satisfies
the non-homogeneous differential equation, but not necessarily the homogeneous
differential equation. The PI is usually found using the method of undetermined
coefficients or the variation of parameters.
Once we have the CF and the PI, we can combine them to obtain the general
solution of the differential equation:
Self-Instructional y = yCF + yPI
142 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

Example 5.5: Consider the following equation NOTES

dy
+ α y b where, b ≠ 0
=
dt
The above equation is non homogeneous differential equation as R.H.S.
of the equation is non zero. It is also known as the complete equation.
The solution of the non-homogeneous/complete differential equation
consists two solutions
(a) Complementary solution
(b) Particular solution.
For solution of non-homogeneous differential equation we 1st find out
particular solution of the complete differential equation.
For this, we assume there is no change in y. If this is true then dy/dt should
be equal to zero, hence the above equation becomes:
0 +α y =
b
Yp = b/ α
Where yp is the particular solution and α ≠ 0
For complementary solution we take ‘b’ of equation) = 0

dy
+α y =
0
dt

dy
= −α y
dt

1
dy = −α .dt
y
Integrating both sides

1
dy = −α .dt
y
log y = - α t Self-Instructional
Material 143

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Taking anti log of the above equation

y = e −α t + c

y = e −α t e c

y = A.e −α t

Where A = ec
The general solution of the equation is given by the sum of particular and
complementary solution of non-homogeneous differential equation.
General solution of equation:

b
=y A.e −α t +  ---(1)
α
As the above equation contains constant term, we have to find out A at
y(0), i.e,

b
y ( t ) A.e −α t +
=
α
b
y ( 0 ) A.e −α .0 +
=
α
b
=A y (0) −  ---(2)
α
Substituting (2) in (1)
 b b
y = y ( 0 ) −  .e −α +
 α α

Example 5.6: Solve dy/dt + 4y = 10

Solution:

For particular solution we assume y is constant i.e. 𝑑𝑦/𝑑𝑡 = 0, then equation

becomes
Self-Instructional
144 Material 0 + 4y = 10

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

yp = 10/4 NOTES
yp = 2.5
For complementary solution we take R.H.S. of equation = 0
i.e.
dy/dt+ 4y = 0
dy/dt = - 4y
1/y . dy = -4 dt
Integrating both side
∫ 1/y . dy =∫ -4 dt
ln y = -4t + c [Where ‘c’ is constant]
Taking antilog
y=e -4t+c

y=e -4t
e c

y = A. e -4t
where A = e c

General solution consists particular solution and complementary solution

therefore
yt = yp +yc
yt = 2.5 + A. e -4t

For finding the value of A; put t = 0

y0 = 2.5 + A e0
A = y0 - 2.5
Hence, the required solution is
y(t) = 2.5 + (y0 - 2.5). e -4t

Self-Instructional
Material 145

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 5.9 STABILITY AND PHASE

Differential equation phase diagrams provide a graphical representation of the

behavior of solutions to a given differential equation. These diagrams can provide
insights into the long-term behavior of solutions and can be used to analyze the
stability of solutions.
To create a phase diagram, we plot the slope field of the differential
equation, which is a graphical representation of the slopes of the tangent lines
to the solution curves at each point in the plane. This is done by evaluating the
derivative of the solution function at many points in the plane and plotting a short
line segment with that slope at each point. These line segments can be connected
to form a direction field that shows the overall behavior of the solution curves.
The phase diagram is then created by plotting a collection of sample solution
curves that show the behavior of the solutions at different initial conditions. The
curves are plotted in such a way that they never intersect, and their shape and
direction reflect the behaviour of solutions in different regions of the plane.

Stability Analysis:

One important use of phase diagrams is to analyze the stability of solutions. A

solution is said to be stable if small perturbations to the initial conditions result
in solutions that approach the same equilibrium solution. In other words, if the
solution curve approaches an equilibrium point as time goes to infinity, the
equilibrium point is stable.
To analyze stability, we can use the concept of a critical point, which is a
point where the derivative of the solution function is zero. A critical point is also
known as an equilibrium point or a stationary point.
There are two types of critical points: stable and unstable. A stable critical
point is one where small perturbations to the initial conditions result in solutions
that converge towards the critical point as time goes to infinity. An unstable
critical point is one where small perturbations to the initial conditions result in
solutions that diverge away from the critical point.
Self-Instructional
146 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

To determine the stability of a critical point, we can examine the behaviour NOTES
of the solution curves near the critical point. If the solution curves approach the
critical point from all directions, the critical point is stable. If the solution curves
approach the critical point from some directions but diverge away from it in other
directions, the critical point is unstable.
Example 5.7: Find the critical points then sketch the phase diagram for the
following differential equation:     
            dy/dt = (1/2) y ( y ˗ 2 )2 ( y ˗ 4 )
Step 1:  Find the critical points by setting f(y) = 0.
Critical Points:   f(y) = 0 at y = 0, y = 2, and y = 4
Step 2:  Identify each region where slopes may change from positive to negative,
stay the same, or change from negative to positive.  Evaluate the slopes on each
side of the critical points by taking values within each region.
The four regions in this example include:  y < 0, 0 < y < 2,   2 < y < 4, y > 4
i.e.   For y = ˗ 1, dy/dt = 45/2 > 0,          
For y = 1, dy/dt = ˗ 3/2 < 0
        For y = 3,     dy/dt = ˗ 3/2 < 0, and
For y = 5, dy/dt = 45/2 > 0
Step 3: Plot Phase Diagram:

        

Note:  A right directed arrow indicates that the dependent variable, y, is increasing.
Self-Instructional
A left directed arrow indicates that the dependent variable, y, is decreasing. Material 147

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Results: At the critical point y = 0, the response is stable.

           At the critical point y = 2, the response is semi-stable.
           At the critical point y = 4, the response is unstable.

i.e. In the neighborhood of y = 0, the value of y(t) tends toward a stable
value.
In the neighborhood of y = 2, the value of y(t) hovers around a stable value.
In the neighborhood of y = 4, the value of y(t) departs from a stable value.
Example 5.8: In the following equation dy/dt = f(y), sketch the graph of f(y)
versus y, determine the critical (equilibrium) points, and classify each one as
asymptotically stable or unstable. Draw the phase line, and sketch several graphs
of solutions in the ty-plane.
dy/dt = e y − 1, −∞ < y0 < ∞.

Solution:
Here is a graph of f(y) = e y
− 1:

Self-Instructional
148 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

The only zero is at y = 0. The function is positive for positive y, and negative NOTES
for negative y. This gives the following phase line:

There is only one equilibrium solution, namely y = 0. Since nearby solutions

are pushed away from it – y′ is positive for y > 0 and negative for y < 0 – it is
an unstable equilibrium.

IN-TEXT QUESTIONS
Fill in the blanks:
8. By allocating different values for c, we obtain a family of curves where
c is known as the ................................ of the family.
9. Differential equations are formed by ................................ of arbitrary
constants.
10. To eliminate ................................ arbitrary constants, we require two
more equations besides the given relation.
11. The elimination of two arbitrary constants lead us to ...........................
..... order derivatives.
12. Elimination of n arbitrary constants leads us to nth order derivatives and
hence a differential equation of the ................................ order.
13. By eliminating the arbitrary constants from the specified equation and
the equations attained by the ................................, we obtain the requisite
differential equations.
State whether the following statements are true or false:
14. The specified equation is differentiated as many times as there are
arbitrary constants.
15. Elimination of n arbitrary constants leads to a differential equation of
the (n+1)th order.

Self-Instructional
Material 149

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 5.10 SECOND ORDER DIFFERENTIAL EQUATIONS

A second order differential equation is an equation that involves the second

derivative of a function. Solving a second order differential equation involves
finding a function that satisfies the equation.
Consider the differential equation

D2 y
= f ( x)
dx 2
It may be noted that it is a differential equation of second order. So its
general solution will contain two arbitrary constants.

d2 y
Now we have, = f ( x)
dx 2

d  dy 
or   = f ( x)
dx  dx 
Integrating both sides, we get

dy
= f ( x ) dx + C1 where C1 is arbitrary constant
dx
f ( x ) dx = ϕ ( x )
Then

dy
= ϕ ( x ) + C1
dx
Again on integrating both sides

y ϕ ( x ) dx + C1x + C 2
=
Where C2 is another arbitrary constant. Therefore in order to find the
particular solution, we need two conditions.

Self-Instructional
150 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

NOTES
2
d y
Example 5.9: Solve 2
= xe x
dx
Solution:

Now integrating both sides of given equation, we have

dy
= xe x dx + C1
dx
where C1 is an arbitrary constant

dy
=xe x − e x dx + C1
dx
dy
= xe x − e x + C1  -- (1)
dx
Again on integrating both sides of (1), we get

y= ( xe x − e x + C1)dx + C 2
where C2 is another arbitrary constant.

y xe x − ∫ e x dx − e x + C1x + C 2
=

y = xe x − 2 e x + C1x + C 2
Which is the required general solution.

5.11 SUMMARY

• Differential equation represents the relationship between function of one

variable or more than one variable and its various derivative.
• Types of differential equation is broadly, classified into order and degree
of differential equation.
• Order of a differential equation is the highest derivative of differential
equation.
Self-Instructional
Material 151

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES • Degree of a differential equation is the highest power of highest order

derivative.
• Solution to the differential equation is the relation between variables which
is free from derivatives and satisfies the equation. In simple word solution
to the differential equation is the integral of the function.
• Solution of the differential equation is divided into two part.
(i) Particular solution: When a solution of differential equation does not
contain constant term, it is called particular solution.
(ii) General solution: When solution to the differential equation includes
constant term, it is called general solution.

5.12 GLOSSARY

• Arbitrary Constant: By allocating different values for c, we obtain a

family of curves where c is known as the parameter or arbitrary constant
of the family.
• Differential Equation: It is a differential equation that involves independent
variables, dependent variables, their derivatives and constants.
• Order: It is the highest derivative that appears in a differential equation.
• Degree: It is the highest power of the highest order derivative that appears
in a differential equation.
• Homogeneous equation: It is a differential equation in which the non-
constant terms are zero.
• Non-homogeneous equation: It is a differential equation in which the
non-constant terms are not zero.
• Linear equation: It is a differential equation in which the dependent
variable and its derivatives appear only linearly.
• Non-linear equation: It is a differential equation in which the dependent
variable and its derivatives appear in a non-linear fashion.

Self-Instructional
• Ordinary differential equation: It is a differential equation that involves
152 Material only ordinary derivatives.

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Differential Equations

• Solution: It is a function that satisfies a differential equation. NOTES

• General solution: It is the family of all possible solutions to a differential
equation that involves one or more arbitrary constants.
• Particular solution: It is a specific solution to a differential equation that
satisfies the initial or boundary conditions.

5.13 ANSWERS TO INTEXT-QUESTIONS

1. differential equation 2. independent

3. mathematical 4. False
5. partial 6. ordinary
7. two or more 8. parameter or arbitrary constant
9. elimination 10. two
11. second 12. nth
13. differentiation 14. True
15. False

5.14 SELF-ASSESSMENT QUESTIONS

1. Form the differential equation of all circles with their centres on the line
y = 2x.
2. Solve the differential equation (1 + x)ydx + (1 + y)xdy = 0
3. Solve

4. Solve the differential equation (2x – y + 1)dx + (2y – x – 1)dy = 0.
5. Solve the differential equation (hx + by + f ) dy + (ax + hy + g)dx = 0.
Self-Instructional
Material 153

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 6. Solve

7. (x2 + y2)dx = 2xydy

5.15 SUGGESTED READINGS

• Sydsaeter, K. Hammond, P. (2002), Mathematics for economic analysis.

Person Edition
• D.A. Murray, Introductory Course in Differential Equations, Orient
Blackswan.

Online links

• http://web.mit.edu/15.053/
• http://agecon2.tamu.edu/
• http://lucatrevisan.wordpress.com/

Self-Instructional
154 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

LESSON 6 NOTES

DIFFERENCE EQUATION

Sarabjeet Kaur
Assistant Professor,
Department of Economics,
Zakir Husain College (E), University of Delhi)
Structure
6.1 Learning Objectives
6.2 Introduction
6.3 Degree and Order of the difference equation
6.4 Solving First Order Difference Equations
6.5 Equilibrium or Stationary Value
6.6 Behaviour of Solutions of First Order Equations
6.7 Economic application of first Order Difference Equations
6.8 Second Order Difference Equations
6.9 Summary
6.10 Glossary
6.11 Answer of In-Text Questions
6.12 Self-Assessment Questions
6.13 Suggested Readings

6.1 LEARNING OBJECTIVES

After studying this unit, you will be able to:

 Understand the concept of linear equations.
 Discuss the equations reducible to linear form.
 Second Order Linear Difference Equation.

6.2 INTRODUCTION

A difference equation is a mathematical equation that describes how a variable changes

over time. Unlike differential equations, which describe continuous changes in a variable, Self-Instructional
Material 155
difference equations describe changes that occur at discrete time intervals.

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES The general form of a difference equation is:

X(t) = f(X(t-1), X(t-2), ..., X(t-k))

where X(t) represents the value of a variable at time t, and f is a function that
describes how the variable’s value changes based on its past values at time periods
t-1, t-2, ..., t-k. The value of k represents the number of past time periods that are
considered relevant to the current value of the variable.
For example, a simple difference equation that models population growth can
be:
P(t) = P(t-1) + r * P(t-1)
where P(t) represents the population at time t, P(t-1) represents the population
at the previous time period (t-1), and r represents the growth rate of the population.
This equation states that the population at time t is equal to the population at
time t-1 plus the product of the growth rate and the population at time t-1. Using
this equation, we can predict the population at any future time period based on its
past values and the growth rate.
Difference equations are important in Economics because they allow us to
model and analyze dynamic systems, where variables change over time in response
to various factors. In particular, difference equations are used to model a wide range
of economic phenomena, including economic growth, investment, consumption,
inflation, and asset prices.
Difference equations are also used extensively in economic forecasting, as
they allow economists to predict future trends based on historical data and current
conditions. Moreover, they provide a way to understand the long-term behaviour
of economic systems and to test the effects of different policy interventions on these
systems.

6.3 DEGREE AND ORDER OF THE DIFFERENCE

EQUATION

A difference equation is an equation that relates the values of a sequence at different

times. The degree of a difference equation is the highest power of the dependent
Self-Instructional
156 Material
variable (the sequence) that appears in the equation. The order of a difference

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

equation is the highest order of the difference operator (Δ or E) that appears in the NOTES
equation.
To determine the degree and order of a difference equation, we first need to
identify the dependent variable and the difference operator. The dependent variable
is the sequence that we are interested in, and the difference operator is the operator
that relates the sequence values at different times.
For example, consider the following difference equation:
yn - 3yn-1 + 2yn-2 = xn
In this equation, the dependent variable is yn, and the difference operator is
Δ2, which is the second-order forward difference operator (Δyn = yn - yn-1, Δ2yn
= Δ(Δyn) = Δ(yn - yn-1) = yn - 2yn-1 + yn-2).
The degree of the difference equation is determined by looking at the highest
power of the dependent variable that appears in the equation. In this case, the
highest power of yn is 1, so the degree of the equation is 1.
The order of the difference equation is determined by looking at the highest
order of the difference operator that appears in the equation. In this case, the
highest order of Δ is 2, so the order of the equation is 2.
Therefore, the degree and order of the difference equation yn – 3yn-1
+ 2yn-2 = xn are 1 and 2, respectively.

Linear Difference Equation:

A difference equation is said to be linear when each term of the equation is defined
as a linear function of the preceding terms.
For example: The equation yn = 3yn-1 + 2yn-2 is linear.
whereas the equation yn - 3y2n-1 + 2yn-2 is not linear as it contains the non-
linear term y2n-1.
The equation of the following form is called an nth-order linear constant
coefficient difference equation:
y = a1 yt–1 + a2 yt–2 + …. + an yt–n + b
where a1, a2, …, an and b are constants.
Self-Instructional
Material 157

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES (assuming an  0, otherwise the order will be less than n). It is linear because
the dependent variable y is not raised to any power and there are no product terms
or, constant coefficients because a1…, an are constants and do not change with t.
This equation will be homogeneous if b = 0. If b  0, then it is nonhomogenous.
In this lesson, we shall work only with difference equations of this special type of
orders one and two (n = 1, 2).

IN-TEXT QUESTIONS
1. What is the order of a first-order difference equation?
a) 0
b) 1
c) 2
d) It depends on the equation.
2. First-order difference equations are commonly used to model:
a) Population growth
b) Economic inflation
c) Stock market prices
d) All of the above

6.4 SOLVING FIRST ORDER DIFFERENCE

EQUATIONS

In solving a difference equation, we find a time path y(t) from a given initial condition.
A first order difference equation takes the form:
Yt = f(t, yt-1) for all t.
We can solve such an equation by successive calculation, also called the
recursive method, taking the initial value of y (say y0) as given. Thus,
y1 = F(1, y0)
Self-Instructional y2 = F(2, y1) = f(2, f(1, y0)) and so on.
158 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

Note that, given any value y0, there exists a unique solution path y1, y2, ..... However, NOTES
resorting to calculation of the solution through such a method doesn’t tell us much about
the properties of the solution. We should have a general formula, which exists if the form
of f is simple.

Let us start with a first-order homogeneous linear difference equation with

constant coefficient (b) equals to zero. It takes the form
xn = axn-1 (n = 1, 2, ...)
Starting with a given x0 it is possible to calculate xn for small n
Note that x1=a x0
x2 = ax1 = a(ax0 ) = a2x0
Similarly, x3= a3x0
x4=a4x0
Hence, In general
For the equation xn = anxn-1
The solution is given by xn=anx0
Non homogeneous first order linear difference equation (i.e. b  0)
It takes the form:
xn = axn-1 + b
xn = a (axn-2 + b) +b
= a2xn-2 + ab + b
= a2(axn-3 + b) + ab + b
= a3xn-3 + a2b + ab + b
Continuing in the same manner, we get:
xn = an-1x1 + an-2b + an-3b + .....+ ab + b
= anx0 + b(1 + a + a2 +... + an-1)
Now, 1 + a + a2 +... + an-1 is a geometric progression of (n-1) terms with
first term as 1 and a common ratio of a. Clearly it can be summed up as


b 1 an  provided a  1
Self-Instructional
1 a Material 159

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES So for the difference equation of the type:

xn = axn-1 + b
where a and b are constants and a 1
The solution is given by:

xn  a x0 
n 
b 1 an 
1 a
or
b b
x n  a n (x 0  ) for a  1
1 a 1 a
For a = 1, 1 + a + ... + an-1 = n
So xn = x0+nb
Also if a = 1 the difference equation is of the type
xn = xn-1 + b n2
which can be seen as
xn = (xn-2 + b) + b
xn = xn-1 +2 b
= (xn-3 + b) + b + b
= xn-3 + 3b
….
xn = x1 + (n-1)b
Again we observe the pattern here. In each case, formula for xn begins with
the term x0 then adds the terms an–1b1, an–2b2 ,……….…………….,abn–1, bn.
n

Hence the overall sum can be expressed as, anx0+ a

k 1
nk
bk

i.e. xn = a x0 +
n
a
k 1
nk
bk ....... (6.2)

So if initial value x0 is known, xn can be determined for all n.

Self-Instructional
160 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

NOTES
6.5 EQUILIBRIUM OR STATIONARY VALUE

For a given value y0, the value of yt changes with t. But there may be some value
of y0 for which yt doesn’t change. Such a solution exists if
y* = b/(1 – a)
and yt is constant, equal to b/(1 – a).
We call y* the equilibrium value of y and rewrite the solution as
yt = at(y0 – y*) + y*.

Example 6.1: Solve yt+1 =α yt + β ...(6.3)

where α and β are constants.
Look for a stationary or equilibrium value of yt over time which can be
repeated for any t consistently satisfying the above equation. May be you consider
y as an equilibrium value of yt such that

To understand the above example, we need to remember the dynamic

multiplier.
Write
C = α yt-1 +β …..(6.4)
Let the investment be fixed at I for every t so that we have

Self-Instructional
Material 161

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Use the above relation (6.4) we have

If an equilibrium income Y is found, the solution can be written as

1
is the Keynesian multiplier
1 α
It is important to remember that we have solved equation (6.1) for the
stationary level of yt i.e., y . There is no guarantee that the actual path of y converges
to y . In case yt approaches y , then
If these values of yt and yt+1 hold, we can write
gt = yt – y ……..(6.5)
Since, yt and satisfy (6.3), we have
yt+1 = αyt+ βand
= α+ β
Thus,

From equation (6.5),

Self-Instructional
162 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

Since, gt+1 = α gt NOTES

gt = α gt-1
…..

g1 = α g0
Substituting backward,
gt+1 = α gt = α2 gt-1 = α3 gt-2…….
gt+1 = αt-1 g0
or gt = αt g0 for t = 0,1,2……
Thus, any difference equation of the form yt = αyt-1 has a solution yt = αt y0,
where y0 is the value of y at some chosen initial point.

General Solution

Suppose we intend to solve the equation

yt+1 + ayt = c …(6.7)
Its general solution will consist of particular solution yp and complementary
function yc i.e., yp + yc= yg. In this approach, the yp component represents the inter-
temporal equilibrium level of y while that of yc gives the deviations if the time path
from that equilibrium. The solution is called general solution due to an arbitrary
constant. In order to get a definite solution, we need an initial condition.
Let us work with complementary function. From (6.7), we get it’s reduced
form as
yt+1 + ayt = 0 …….(6.8)
It is seen above that yt = αt y0 is a solution to the difference equation. In that
case we have yt+1 = αt+1 y0 as well.
We modify this and rewrite:
yt = Abt and yt+1 = Abt+1
Substitution of these into (6.8) gives:
Abt + aAbt+1 = 0
A bt (b+a) = 0 Self-Instructional
Material 163

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES Or b+a = 0
 b = –a
We must have b = –a in the trial solution such that the complementary
solution can be written as:
yc = Abt = A(-a)t
Particular solution needs to be recited such that it is in agreement with the
general solution. Consider the simplest value of y. If yt has an equilibrium value k
such that it remains constant over time, we have yt = k as well as yt+1 = k.
Substitution of these values to the trial solution gives:
k + ak = c
c
or k =
1 a
Since the value k satisfies the equation, the particular solution can be written as:
c
yp = k = for a  0
1 a
In case a = –1, however, the particular solution is not defined and some other
solution of (6.7) must be searched for.
Substituting k into (6.7), we get
k (t+1) + akt =c

c
or k = t  1  a = C and Ct = yp
t

The general solution can now be written in one of the following forms:

Self-Instructional
164 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

Notice that the solution above still remains indeterminate. This is due to the NOTES
presence of arbitrary constant A. We have to take the help of initial condition
(yt = y0 ) for eliminating it. Thus, taking t = 0, we have:

The definite solution therefore, becomes

IN-TEXT QUESTIONS
3. The solution to a first-order difference equation involves:
a) Finding the first difference of the sequence.
b) Taking the integral of the sequence.
c) Solving a characteristic equation.
d) Using the forward difference operator.
4. Which of the following best describes a nonhomogeneous difference equation?
a) It includes a forcing term.
b) It involves the difference between consecutive terms.
c) It has a constant term in the equation.
d) It doesn't have any external influence or forcing term

Self-Instructional
Material 165

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES
6.6 BEHAVIOUR OF SOLUTIONS OF FIRST ORDER
EQUATIONS

The solution of a difference equation gives an expression for the relevant variable
as an explicit function of time. In other words, a time path of the variable is
obtained. To investigate the nature of this time path of a solution of the first order
equation, we write the solution for a  1. The behaviour of the solution path
depends on the value of a.
1. Monotonic convergence, when 0 < a < 1. If the value is 0<a<1, then yt
converges to y* and the solution is stable and Monotonic convergence.

2. Damped Oscillatory convergence and stable: When ”1 < a < 0.

Self-Instructional
166 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

3. Divergence and Explosion: If a >1 NOTES

4. Explosive oscillations: If a < - 1

Self-Instructional
Material 167

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 5. If a = –1: Then y(t) oscillates between the two values y(0) and b–y(0)

6. If a = 1: Then y(t) diverges to infinity

Self-Instructional 7. If a = 0: Then y(t) = B

168 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

NOTES

In short,
a > 1 time path explodes (diverges)
a < 1 time path converges
a > 0 time path non-oscillating
a < 0 time path oscillating.
Thus, the condition for stability is |a| < 1.

6.7 ECONOMIC APPLICATION OF FIRST ORDER

DIFFERENCE EQUATIONS

Example 6.2: Suppose a corporate employee has availed a home loan of 1500000/
- at an interest rate of 12% per annum over five years. Find his monthly repayment.
Solution:
Let us denote xn as the amount owing after n months. Then x0 = 1500000/- as
for the base month (zero) whole amount is left to be paid. Let b denote the amount
repaid every month i.e. his equated monthly instalment (EMI). Now if xn–1 is the Self-Instructional
Material 169

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES outstanding loan then interest is to be computed on this capital. Adding the outstanding
loan and the interest computed and subtracting the EMI from it will be then equal
to the amount left to be paid after n months. Formulating it as a difference equation,
we get:
x n = xn–1 + (1% of xn–1)–b
 xn = xn–1+ 0.01xn–1 –b
 xn = 1.01 xn–1 –b …(1)
For this problem, we know x0 = 1500000
Clearly (1) is a first order linear non homogeneous difference equation with
constant coefficients.
As seen in our previous sections, the solution to such difference equation is
given by:

1.01  1
n

xn = (1.01)nx0–b
1.01  1
= (1.01)n x0– 100b((1.01)n–1)
As the loan has been availed for a period of 5 years or equivalently 60
months, so at the end of the term, n = 60, we must have x60 = 0
Also, x18 = (1.01)60x0 – 100b((1.01)60 – 1)
 0 = 1500000 (1.01)60 – 100b((1.01)60 – 1)

1500000*1.0160
b = 100 1.0160  1  33366.7

 
So repayments of 33366.70 are to be made every month.
Example 6.3: Let us investigate the behaviour of price in a market with the demand
and supply functions:
Dt = 86 – 0.8 Pt
St = – 10 + 0.2 Pt – 1
Solution:

Self-Instructional
Assuming market clearing in each period (Dt = St), we have:
170 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

(–0.8) Pt = 0.2 Pt – 1 – 96 NOTES

or, Pt = (–0.25)Pt – 1 + 120
The solution is:

Since -0.25 = 0.25 < 1, the time path of P is oscillating but converges. The
market is stable and with the passage of time price approaches the equilibrium value
96.

Growth Model

Consider a national economy for which xn denotes the national income in the nth period,
an is the total investment and bn is the total savings in nth period. Suppose the economy
grows in such a way that the savings is proportional to national income and investment
depends directly on the change in income. Mathematically, this can be written by of
the following equations.
bn = αxn ..... (6.9)
an = β(xn– xn–1) ..... (6.10)
where α, β are positive constants.
By the equilibrium condition, we know,
an = bn .... (6.11)
as per difference equation,
β(xn– xn–1) = α xn
 (β–α)xn = βxn–1

β
 xn = x (n = 1, 2, 3... ) ....(6.12)
β  α n–1
Self-Instructional
Material 171

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES This is a homogeneous first order linear difference equation with constant
coefficients whose solution is given by:
n
 β 
xn =   x0 .....(6.13)
βα
where x0 is the initial national income.
Also equation (c) can be written as
 β 
xn =  1   x
 β  α  n–1
or xn=(1+g)xn–1
β X  X n 1
where g = = n which is constant proportional growth.
βα X n 1
And so the solution can be rewritten as xn = (1+g)nx0

6.8 SECOND ORDER DIFFERENCE EQUATIONS

A second-order difference equation is a mathematical equation that relates the values

of a sequence to the values of previous terms in the sequence. It involves the second
difference of the sequence, which is the difference between consecutive terms.
A general second-order difference equation can be written as:
where:
 yt+2 is the value of the sequence at the (t+2)th term.
 a and b are constants that determine the relationship between current and
previous terms.
Here b  0 and n  2.

Solution
To solve a second-order difference equation, there are different methods that can
Self-Instructional
172 Material be employed depending on the nature of the equation:

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

1. Homogeneous Linear Difference Equations: NOTES

Consider the following second order constant coefficient equation

Yt+2 = a.yt+1 + b.yt

We need to find two solutions of the above equation.

If we guess that the solution takes the form ut = mt
In order for ut to be a solution, we must have
mt(m2 + am + b) = 0
or, if m  0,
m2 + am + b = 0.
This is called the characteristic (or auxiliary) equation of the difference
equation and its solutions are

 1  2 
–(1/2)a ±  a – b .
 4  

Behaviour of Solutions of Homogeneous Equations

 1  2 
Looking at the component    a – b  ., we distinguish three cases:
 4  

i) Distinct real roots:

If a2 > 4b, the characteristic equation has distinct real roots, and the general
solution of the homogeneous equation is:
Amt1 + Bmt2,
where m1 and m2 are the two roots.
ii) Repeated real root
If a2 = 4b, then the characteristic equation has a single root, and the
General solution of the homogeneous equation is
Self-Instructional
(A + Bt)mt, Material 173

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES where m = –(1/2)a is the root.

iii) Complex roots

If a2 < 4b, then the characteristic equation has complex roots, and the
general solution of the homogeneous equation is
Art cos(θt + ω),
Where A and ω are constants, r = b , and cos θ = –a/(2 b ), or,,
alternatively,
C1rt cos(θt) + C2rt sin(θt),
where C1 = A cos ω and C2 = –A sin ω (using the formula that cos(x+y)
= (cos x)(cos y) – (sin x)(sin y)).
When the characteristic equation has complex root, the solution oscillates.
Art is the amplitude (which depends on the initial conditions) at time t, and
r is growth factor.
θ/2π is the frequency of the oscillations and ω is the phase (which depends
on the initial conditions).
If |r| < 1 then the oscillations are damped; if |r| > 1 then they are explosive.

IN-TEXT QUESTIONS
5. The solution to a second-order difference equation involves:
a) Finding the first difference of the sequence.
b) Taking the integral of the sequence.
c) Solving a characteristic equation.
d) Differentiating the sequence twice.
6. Which of the following best describes a homogeneous difference equation?
a) It includes a forcing term.
b) It involves the difference between consecutive terms.
c) It has a constant term in the equation.
d) It doesn't have any external influence or forcing term.
Self-Instructional
174 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

Fill in the blanks NOTES

7. The behavior of solutions of a homogeneous difference equation is influenced
by the magnitudes of the characteristic roots. If the absolute value of a
characteristic root is less than 1, the solution is ___________.
8. If the absolute value of a characteristic root is greater than 1, the solution is
___________.
9. If the absolute value of a characteristic root is equal to 1, the solution is
___________.
10. The behavior of solutions of a homogeneous difference equation can also be
affected by the presence of _______________.

Stability

We say that a system of differential equations is stable if its long-run behaviour is not
sensitive to the initial conditions.
Consider the second-order equation:

Write the general solution as

where A and B are determined by the initial conditions. This solution is stable
if the first two terms approach 0 as t , for all values of A and B. In this case,
for any initial conditions, the solution of the equation approaches the particular
solution ut*. If the first two terms approach zero for all A and B, then ut and vt must
approach zero. You can take A = 1 and B = 0 to see that ut approaches zero. On
the other hand, take A = 0 and B = 1 to see that vt approaches 0. A necessary
and sufficient condition for this to be so is that the moduli of the roots of the
characteristic equation be both less than 1. Note that the modulus of a complex
number:

α+βi α 2  β 2 is the absolute value of number if the number is real.

Self-Instructional
Material 175

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES There are two cases:

• If the characteristic equation has complex roots then the modulus of each

 1 2
root is b (the roots are α ± βi, where α = –a/2 and β = b –  a  .
 4 
So for stability we need b < 1.
• If the characteristic equation has real roots then the modulus of each
root is its absolute value. So for stability, we need the absolute values
of each root to be less than 1, or |–a/2 + (a2/4 –b)| < 1 and |–a/2 –
(a2/4 – b)| < 1.

Example 6.5: Find the general solution of the difference equation xn-6xn-1+9xn-2=0
Solution:
Let the trial solution be as xn = c kn and proceeding on the lines as discussed in
the article above, we get the equation as k2 - 6k + 9 = 0 or (k-3)2 = 0 which has
equal roots k = 3, 3
So the general solution is given by xn = (Dn+c)3n
Now suppose we are given initial conditions,
x1 = 3 and x2 = -9
Then from the general solution evaluated, we get
x1 = (D+C)3 = 3 ie. D +C = 1
x2 = (2+C)9 = -9 ie. 2D+C = -1
Solving these equations D
simultaneously we get D = -2 and C = 3
So the particular solution in this case is xn = (-2n+3)3n
Example 6.6: Find the general solution of the difference equation xn-xn-1-2xn-2=0.
Find the particular solution that satisfies the given initial conditions x1=1 and x2=5.
Hence evaluate x6.
Solution:

Using the trial solution xn = c kn , we get the auxiliary equation as k2-k-2=0.

Self-Instructional
176 Material Or (k+1)(k-2) = 0 or k = -1, 2 : real and distinct

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

Hence the general solution is given by xn=A2n +B (-1)n where A and B are NOTES
arbitrary constants.
Now using the given conditions:
2A - B = 1
4A + B = 5
which gives A = 1 and B = 1
Hence the particular solution is xn = 2n + (-1)n
Now x6 = 26 + (-1)6 = 64+1 = 65
Economic applications of second-order difference equations:
Second-order difference equations have various economic applications, particularly
in modelling dynamic systems and analyzing economic behavior over time. Here are
a few examples:
1. Economic Growth Models: Difference equations can be used to model
economic growth and development. One of the most well-known examples
is the Solow-Swan growth model. It utilizes a second-order difference equation
to represent the accumulation of capital and its effect on economic output
over time. The equation relates the current level of capital to its previous
levels and incorporates parameters such as savings rates, depreciation, and
technological progress.
2. Investment Decision Making: Difference equations can be employed
to study investment decisions and the dynamics of investment behaviour.
For instance, consider a company deciding how much to invest in new
capital each period. The investment decision might depend on the current
and lagged levels of output or profit, incorporating a second-order
difference equation. Analyzing this equation can provide insights into
investment patterns and the impact of different economic factors on
investment decisions.
3. Consumption and Savings Behaviour: Difference equations can also be
utilized to model consumption and savings behavior of individuals or
households. For example, the Euler equation, a second-order difference
equation, describes how individuals make consumption choices over time by
considering their current and future levels of income, interest rates, and
preferences. By solving this equation, economists can analyze the intertemporal Self-Instructional
allocation of consumption and savings. Material 177

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES 4. Macroeconomic Fluctuations: Difference equations can capture the

dynamics of macroeconomic variables and the business cycle. For instance,
the Phillips curve, which describes the relationship between inflation and
unemployment, can be represented using a second-order difference
equation. This equation links the current and lagged levels of inflation and
unemployment, offering insights into the short-run dynamics of these
variables.
5. Asset Pricing Models: Difference equations are also employed in asset
pricing models in finance and economics. For example, the autoregressive
conditional heteroscedasticity (ARCH) model uses a second-order difference
equation to model the volatility of asset returns. This equation relates the
conditional variance of returns to its lagged values and provides a framework
for studying the dynamics of volatility in financial markets.
By solving these equations, economists can make predictions, assess policy
implications, and gain insights into the dynamics of economic systems.

6.9 SUMMARY

1. General difference equation is written as:

xn = f (xn-1,xn-2,….x0)
expressing a functional relationship in terms of successive differences.
2. Linear difference equation consists of linear functions of preceding terms.
Order of a linear difference equation is the largest difference between indices
of terms in the equations.
General first order linear difference equation is:
xn = anxn-1+bn
Where an, bn are functions of n.
3. If bn = 0 equation is a linear homogeneous first order difference equation.
If bn  0, equation is a linear non homogenous first order difference equation.
If an is constant, equation is a first order linear difference equation with
constant coefficients otherwise with variable coefficients.
Self-Instructional
178 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

4. For a homogenous equation with constant coefficients NOTES

xn = axn-1,
Solution is xn = anx0 where x0 is the given initial value.
For a non- homogeneous equation xn = axn-1 + b.
Solution is:

and xn = x0 + nb

6.10 GLOSSARY

 Constant Coefficient Difference Equation: A difference equation has

constant coefficient if the coefficients ai’s associated with the y values are
constant and do not change over time.
 Difference Equation: A difference equation is an equation involving the
values of an unknown function y(x) for different values of x. The independent
variable – time in problems of economic dynamics – takes only discrete
values. The form of the equation is,

Where a1, a2, …., an and b are constants, is an example of an nth order linear,
constant coefficient, difference equation.
 Homogeneous Difference Equation: A difference equation is homogeneous
if the constant term b is zero.
 Linear Difference Equation: A difference equation is linear if (i) the
dependent variable y is not raised to any power and (ii) there are no product
terms.
 Non-homogeneous Difference Equation: A difference equation is
nonhomogeneous if the constant term, b, is non-zero.

Self-Instructional
Material 179

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Advanced Mathematical Methods for Economics

NOTES  Order of a Difference Equation: It is determined by the maximum number

of periods lagged.

6.11 ANSWERS OF IN-TEXT QUESTIONS

1. b) 1
2. d) All of the above
3. d
4. a
5. c
6. d
7. stable/convergent
8. unstable/divergent
9. marginally stable
10. repeated roots or higher-order roots

6.12 SELF-ASSESSMENT QUESTIONS

1. Investigate the behaviour of price in a market, i.e., the stability of a system with
demand and supply functions:
a) Dt = 86 – 0.8 Pt
St = –10 + 0.8 Pt –1
b) Dt = 86 – 0.8 Pt
2. Find the time path represented by the equation

Self-Instructional
180 Material

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
Difference Equation

3. Find the solution of the equation NOTES

4. The demand and supply for the model are Cob-web given as
Qdt = 19 – 6 Pt and Qst =6 Pt-1 – 5. Find the intertemporal equilibrium price and
comment on the stability of the equilibrium.

6.13 SUGGESTED READINGSSS

 Allen, R.G.D., 1959, Mathematical Economics (Second Edition) St.

Martin’s
 Press Inc., New York.
 Baumol, W.J., 1974, Economic Dynamics (Second Edition) Macmillan,
New
 York. Chapters 9, 10 and 11.
 Chiang, Alpha C. 1984, Fundamental Methods of Mathematical
Economics
 (Third Edition): Mc-Graw Hill International Edition, New Delhi.

Self-Instructional
Material 181

© Department of Distance & Continuing Education, Campus of Open Learning,

School of Open Learning, University of Delhi
 Notes
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
 Notes
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
 Notes
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
___________________________________________________________________________
DEPARTMENT OF DISTANCE AND CONTINUING EDUCATION
UNIVERSITY OF DELHI