CALCULUS FOR ECONOMISTS

Module II

Course Team: Ato Fredu Nega

Ato Torben Kenea

Mekelle University
Faculty of Business and Economics
Department of Economics
Mekelle
April 2004
III Differential Calculus: Fundamental Techniques 38
3.1. INTRODUCTION .................................................................................... 38
3.2. THE CONCEPT OF THE DERIVATIVE ............................................................ 39
3.3 THE RULES OR THEOREMS OF DERIVATES ..................................................... 41
3.3.1 The constant rule......................................................................... 41
3.3.2 The Simple Power rule ................................................................. 41
3.3.3 The coefficient rule ...................................................................... 42
3.3.4 The sum or difference rule............................................................ 42
3.3.5 The product rule .......................................................................... 43
3.3.6 The quotient rule ......................................................................... 44
3.3.7 Differentiation of a function .......................................................... 45
3.3.8 Derivatives of implicit functions ..................................................... 46
3.3.9 Derivatives of logarithmic and exponential functions ....................... 48
3.3.9.1 Derivatives of logarithmic functions ......................................... 48
3.3.9.2 Derivatives of exponential functions ........................................ 51
3.4. DIFFERENTIALS AND HIGHER ORDER DERIVATIVES ......................................... 54
3.4.1 Differentials................................................................................. 54
3.4.2 Second and higher order derivatives.............................................. 56
3.4 PARTIAL DERIVATIVES AND TOTAL DERIVATIVES .............................................. 57
3.4.1 Partial derivatives ........................................................................ 59
3.4.2 Total derivatives (differentials)...................................................... 60
3.5 SOME APPLICATIONS OF DIFFERENTIAL CALCULUS ........................................... 61
3.5.1. L Hôpital’s Rule.......................................................................... 61
3.5.2.Homogeneous Functions and Euler’s Theorem................................ 62
3.5.2.1 Linear Homogeneity ............................................................... 63
3.5.2.2.Properties of linearly homogeneous production function............ 63
3.5.2.3 Euler theorem ....................................................................... 64
IV. APPLICATIONS OF DIFFERENTIAL CALCULUS .......................................... 65
4.1. INTRODUCTION .................................................................................... 65
4.1.1. Increasing, and Decreasing Functions........................................... 65
4.1.2. Testing For Concavity.................................................................. 67
4.2. FREE (UNCONSTRAINED) OPTIMIZATION ...................................................... 70
4.2.1. The concept of optimum (extreme) value...................................... 70
4.2.1.1. Relative Extreme Value.......................................................... 70
4.2.1.2. Absolute Extrema.................................................................. 74
4.3. CONSTRAINED OPTIMIZATION ................................................................... 76
4.4. TESTING FOR THE NATURE OF OPTIMA ....................................................... 80

37
 III Differential Calculus: Fundamental Techniques
3.1. Introduction
This module deals with the techniques and applications of differential
calculus. It deals with concepts and techniques of mathematical
economics, which is commonly used in comparative statics analyses of
economic variables. It mainly deals with the concept of instantaneous
rate of change-namely the derivative. In economics, we are interested
in instantaneous rates of changes of various economic variables. For
instance, we deal with marginal utility, marginal product, marginal
revenue, value of marginal product, and other marginal values, which
are the instantaneous rate of change of their respective total
functions.

This module is divided in to two main parts where the first part deals
with the basic techniques of differential calculus; the second part deals
with the applications of differential calculus. The first part specifically
deals with concepts such as the derivative, rules/theorems of
derivative, differentials and higher order derivative, total and partial
derivatives, and finally about some applications of differential calculus
such as L Hôpital’s rule, which is used to deal with indeterminate
limits, homogeneous functions and Euler’s theorem. On the other
hand, the second theorem discusses about application concepts such
as increasing, decreasing, concave up, concave down, and
optimizations.

Let’s begin with the basic techniques and finally we will see its
applications in the area of economic reasoning. The concepts are
simple and valuable toolkits for an enterprising economist and all it
needs is curiosity, closeness and above all hard work. Enjoy, your
reading!

38
3.2. The Concept of the Derivative

If f is a function defined by y=f (x), then the derivative of f (x) for any
dy
value of x, is denoted by , y’ or f ’(x) is the instantaneous rate of
dx
change of f (X) which is given by:

dy f ( x + σx) − f ( x)
= Lim
dx σx →0 σx
As you might remember from the previous discussion, the derivative is
also the slope of the tangent line to f (x) at a point. Before we directly
go to the discussion of the rules of derivatives, let’s see some
examples.

Examples:
dy
1. Given f (x)=3x+7, find
dx
dy
2. Given f (x)=x2, find
dx
dy
3. Given f (x)=Xn , find
dx

Solutions:

1. For f (x)=3X+7,
dy f ( x + σx) − f ( x)
= Lim
dx σx →0 σx
[[3( x + σx) + 7] − (3x + 7)]
= Lim
σx → 0 σx
3x + 3σx + 7 − 3 x − 7
= Lim
σx → 0 σx
3σx
= Lim
σx → 0 σx

= Lim 3
σx → 0

=3

39
2. For f (x)=X2
dy f ( x + σx) − f ( x)
= Lim
dx σx →0 σx
( x + σx) 2 − x 2
= Lim
σx → 0 σx
( x 2 + 2σx.x + σx 2 − x 2 )
= Lim
σx → 0 σx
= Lim(2 x + σx)
σx → 0

= 2x

3. f ( x) = x n
dy f ( x + σx) − f ( x)
= Lim
dx σx →0 σx
( x + σx) − x n
n
= Lim
σx → 0 σx
n(n − 1) n − 2 2
[ x n + nx n −1σx + x σx + ...... + σx n ] − [ x n ]
= Lim 2
σx → 0 σx

n ( n − 2) n − 2
= Lim{nx n −1 + x σx + ...... + σx n −1}
σx → 0 2

= nx n −1

d n n −1
Thus, { X } = n X
dx

Directly evaluating f (x) and f (x+σx) to come up with the derivative is

some how tiresome if not difficult. However, using the result of
example-3 above along with other rules will simplify our calculations.

40
Note that,
i. The derivative, which is one of the most fundamental concepts in
calculus, is the same as the following two concepts.
a. Slope of a line tangent to a curve at x
b. Instantaneous rate of change of f(x) at x
ii. The process of obtaining f’ (x) from f (x) is known as
differentiation and if the derivative exists, then the function is
differentiable at a point or over an interval. In the following
section, we will present the rules of derivatives, which will
largely simplify our calculation of the derivative.

3.3 The Rules or Theorems of Derivates

In this section, we formally state the different rules of derivatives that
are important in several problems of getting the derivative. These are
the constant rule, the simple power rule, the coefficient rule,
sum/difference rule, product rule, quotient rule, chain rule, and others.

3.3.1 The constant rule

If c is constant and if f(x)=c, then f’(x)=0

Examples
a. If f (x)=2, then f ‘(x)=0
b. If f (x)=-12, then f ‘(x)=0

3.3.2 The Simple Power rule

If f ( x) = x n , where n ∈ ℜ, the f ' ( x) =nxn-1

41
.

Examples

Find f ' ( x) of functions of

i. ⇒ Y = 2 x 5 ,
ii. ⇒ Y = x ,
iii. ⇒ Y = x 2 x

Solution
i. Y = 2x 5 1 1
ii. Y = x = x 2
iii. Y = x 2
x=x x 2 2
dy
= 2.5 x 5−1 dy 1 2 −1
1 5
dx = x Y=X 2

= 10 x 4 dx 2 5
dy 5 2 −1
1 −2
1
= x
= x dx 2
2 3
dy 1 dy 5 2
= = x
dx 2 x dx 2

3.3.3 The coefficient rule

For any constant C and function m(x),
If f ( x) = C m( x), then f ' ( x) = C m' ( x)

Example

Y = 3x 2
dy dx 2
=3
dx dx
= 3[2 x ]
= 6x
3.3.4 The sum or difference rule

If f ( x) = u ( x) ± v( x), ~ thenf ' ( x) = u ' ( x) ± v' ( x)

Example

1. If y = x − 3x + 10 , then the derivative

3 2

42
 dy 3
( x − 3 x 2 + 10)
dx
d 3 d d
= x − 3 x 2 + 10
dx dx dx
= 3x − 6 x
2

2. y = x 4 − 2 x 2 + 1
dy 4
( x − 2 x 2 + 1)
dx
dy 4 dy dy
= x − 2 x2 + 1
dx dx dx
= 4X − 4X
3

Exercise
1. The total revenue for a commodity is described by
R( x) = 300 x − 0.02 x 2 . What is the marginal revenue when 50 units
are sold and interpret your result?

2. Suppose the total cost function for the production of x units of a

product is given by: Tc=2X+120
a. Find the instantaneous rate of change of average cost with
respect to the number of units produced.
b. Find the level of production at which this rate of change equals
zero.

3. The profit from the sale of x stereos per month for a certain co. is
given by:
π ( x) = 800 − x 2 + 0.2 x 2
Find the rate of change in the monthly profit for 100 stereos.

3.3.5 The product rule

Although most of the time a function is given as a sum or difference of
terms, there are times when functions are presented in products.
If f ( x) = u ( x).v( x), then f ' ( x) = u ' ( x).v( x) + v' ( x).u ( x)
Example
1. Y = (3x − 2 x 2 )(5 + 4 x)
dy
= (5 + 4 x)(3 − 4 x) + (3 x − 2 x 2 )(4)
dx
= 15 − 20 x + 12 x − 16 x 2 + 12 x − 8 x 2
= − 24 x 2 + 4 x + 15

43
 1  dy x2 +1
2. If Y = ( x − 1) + 1 , find . The answer is
x  dx x2

3.3.6 The quotient rule

If u (x) and V (x) are two functions and if f(x) is the quotient,
u ( x) v( x)u ' ( x) − u ( x)v' ( x)
If f ( x) = , then f ' ( x) =
v( x) [v( x)] 2

Example
(1 − 2 x)(3 x + 2) dy
1. Y = find ,
5x − 4 dx
Solution

 d (1 − 2 x)(3 X + 2) d (5 x − 4) 
 ((5 x − 4 ) ) − ((1 − 2 x)(3 x + 2) 
dY  dx dx 
=
dx (5 x − 4) 2

(5 x − 4)[(1 − 2 x)(3) + (3 x + 2)(−2)] − [5(1 − 2 x)(3x + 2)]

=
(5 x − 4) 2
(5 x − 4)(−1 − 12 x) − [(5 − 10 x)(3x + 2)]
=
(5 x − 4) 2
− 30 X 2 + 48 X − 6
=
(5 X − 4) 2

x −1 dy
2. Y = , find ,
x +1 dx
Solution
d ( x − 1) d ( x + 1)
( x + 1) − ( x − 1)
dx dx
( x + 1) 2
( x + 1)(1) − ( x − 1)(1)
=
( x + 1) 2
x +1− x +1
=
( x + 1) 2
2
=
( x + 1) 2

44
3.3.7 Differentiation of a composite function
df
If the function f is a function of g and g is a function of (x), then is:
dx

df df dg
= . Y = f ( g ), Z = g ( x)
dx dg dx

Example
dy
1. Y = I 4 + 3I 3 and I = x 2 , find
dx
Solution
dy dy dI
= . = [4 I 3 + 9 I 2 ][2 x]
dx dI dx
= 8 xI 3 + 18 xI 2
= 8 x( x 2 ) 3 + 18 x( x 2 ) 2
= 8 x 7 + 18 x 5

dy
2. Y = x 2 − 1 find
dx
Solution

Let x 2 − 1 = u
1
Y =u 2
dy dy du 1 x x
= . = .2 x = =
dx du dx 2 u u x2 −1

dy
3. Y = ( x 2 − 4 x) 6 , find
dx
Solution
Let x2 –4X =U,
So that Y= U6
dy dy du
= * = 6u 5 (2 x − 4)
dx du dx
dy
= 6( x 2 − 4 x) 5 .(2 x − 4)
dx
= (12 x − 24)( x 2 − 4 x) 5

4. The total physical output P of workers is a function of the number of

workers “x”. The function f(x)=p is called physical productivity

45
 function and is given by P = 10(3x + 1) 3 − 10, find the marginal physical
productivity.

5. Suppose the revenue function for a certain product is given by

R( x) = 15(2 x + 1) −1 + 30 x − 5, where x is in thousands units and R is in
thousands.
a. Find the MR when 200 units are sold
b. How is revenue changing when 2000 units are sold

3.3.8 Derivatives of implicit functions

Implicit functions are those functions whose dependent and
independent variables are not explicitly stated. That is, we don’t know
whether a given variable is a cause or an effect variable without any a
priori information when we encounter implicit functions. When the
dependent and independent variables are not explicitly stated, one
dy
may not easily come up with . But using implicit differentiation one
dx
dy
can calculate .
dx

Example

1. 3xy − y = 2
d d (2)
[(3xy − y )] = ...differentiating ~ both ~ sides ~ w.r.t. X
dx dx
d 3xy dy
= − =0
dx dx
dy dxy
=3
dx dx
dy dx
= 3.x + y … Applying the product rule to (xy)
dx dx
dx dy
= 3x +y
dy dx
dy dy
− 3x =y
dx dx

46
 y
[ ]
dy Y 1
= =
dx 1 − 3x 1 − 3x
[ ]
1
2 −1 2
=[ ][ ]... sin ce ~ 3 xy − y = 2, y (3x − 1) = 2, and ~ y =
3x − 1 1 − 3x 3x − 1
−2
=
(3x − 1) 2

2. 3xy − x = 2
3xy = x + 2
d (3xy ) d ( x + 2)
=
dx dx
 dy dx 
3 x + y  =1
 dx dx 
dy 1
x + y=
dx 3
dy 1
x = −Y
dx 3
dy 1 y
= −
dx 3x x
1 − 3y
=
3x
 + 2
x
1 − 3 
 3x  x+2
= Since Y =
3x 3x
x+2
1−
= x
3x
1
x−x−2 1
= .
x 3x
−2
= 2
3x
dy
3. Find of x 2 + y 3 = 2xy 2
dx
d ( x 2 + y 2 ) d 2 xy 2
= … Differentiating both sides w.r.t X
dx dx

47
 dy 2  dx dy 2  2
2x + = 2 y 2 +x  … Applying he product rule to xy
dx  dx dx 
2
dy dy  2  dy 2 dy   dy 2
2x + . = 2 y + x  .   … Applying the chain rule to
dy dx   dy dx   dx
dy  dy 
2 x + 2 y. = 2 y 2 + x 2 y 
dx  dx 
dy dy
2x + 2 y = 2 y 2 + 4 xy
dx dx
dy dy
2y − 4 xy = 2 y 2 − 2x
dx dx
dy 2 y 2 − 2 x
=
dx 2 y − 4 xy

Note:
You can use the following generalized steps
1. Treat y as a function of x and differentiate each term of the
equation with respect to x.
dy
2. Put all terms containing the factor on one side of the
dx
equation and the rest of the terms on the other side.
dy
3. Factor out from all terms that contain it
dx
dy
4. Solve the equation for
dx
3.3.9 Derivatives of logarithmic and exponential functions
The concept and rules governing logarithmic and exponential functions
were described previously in the first chapter. In this section, we will
review point out how to find the derivatives of logarithmic and
exponential functions.

3.3.9.1 Derivatives of logarithmic functions

dy
If Y = log ex = ln x, find
dx
y + σy = ln( x + σx)

48
 σy = ln(x +σx) − y
= ln(x +σx) − ln x
 σx 
= ln1+ 
 x
2 3 4
 σx   σx   σx 
     
σx  x   x   x 
σy = − + − + ......See expansion from the previous discussion
x 2 3 4
σy 1 σx σx (σx)
2 3
= − + − + .......
σx x 2x 3x3 4x 4
σy 1
Lim =
σx→0 σx x
d 1
Thus, ln x =
dx x
If the function is stated to any base than e, it is possible to transform
it to the natural logarithm form and apply the above formula.
For example, if you are given:
y = log ax
ln x
=
ln a
 ln x 
d 
dy  ln a 
= 1
dx dx … Since ln a and are constants.
ln a
1 d ln x
=( )( )
ln a dx
1 1
= .
x ln a
1
= log ea
x

d log ax 1
∴ =
dx x ln a

If U=f (x), using the chain rule,

d log ua 1
= .u ' ( x)
dx u ln a

49
Examples:
du
Find for each of the functions below
dx

1. y = x 3 + 3 ln x
2. y = log 4 ( x 3 + 1)
3. y = ln x 4
4. y = ln 3 x
5. x ln y = 4

Solution
dy d ln x 3
1. = 3x 2 + 3 = 3x 2 +
dx dx x

2. Let x 3 + 1 = u
ln u
So y = log u4 =
ln 4
dy dy du 1  d ln u du 
= . = . . 
dx du dx ln 4  du dx 
1 1
= . .3x 2
ln 4 u
3 x 2 log e 4
=
x3 + 1

3. y = ln x 4 let u = x 4
dy dy du
= .
dx du dx
d ln u du
= .
du dx
1
= .4 x 3
u
4x 3
= 4
x
4
=
x

50
4. Let 3 x = u
y = ln u
dy dy du 1 1
= . = . 3
dx du dx u 2 x
1 3
= .
3. x 2 x
1
=
2x
4
5. ln y =
x
d ln y dy − 4
. = Differentiating both sides w.r.t. X and applying the chain rule.
dy dx x 2
1 dy − 4
= =
y dx x 2
dy − 4 y
= = 2
dx x
4
− 4e x
= 2
... sin ce ~ y = e 4 / x
x

3.3.9.2 Derivatives of exponential functions

dy
If y = a x , a > 0& ≠ 1, then can easily be found by changing it to the
dx
logarithmic function.

ln y = x ln a
d ln y dy dx
= . ln a
dy dx dx
1 dy
= ln a
y dx
dy
= y ln a
dx
dy
= a x ln a
dx
da x
Thus, = a x ln a
dx

51
 da u ( x )
If Y=aU(x), Using chain rule = u ' ( x )a u ( x ) ln a . As a result, we have
dx
da u ( x )
= u ' ( x ).a u ( x ) ln a
dx
The exponential function ex is a unique function with special behaviour.
That is.
de x
= ex
dx

However, in case we have ef(x), we can use the chain rule to evalute its
derivative. That is let f (x)=u and we will have eu.
If Y=eu and U= f(x)
dy dy du
=
dx du dx
= e u * u ' ( x)

de f ( x )
Thus, = f ' ( x )e f ( x )
dx

dy
Example: Find in each of the functions given below
dx
1. y = 4 x
2
2. y = e 3 x
3. lny=x
4. y = e ex
x
5. y = 10 x
3
6. y = e 4 x
2
7. y = e ln x

Solution
1. y = 4 x

Solution
dy
Directly inserting in the formula, we get = 4 x ln 4
dx
2
2. y = e 3 x

52
 Solution

dy 2
= 6 xe 3 x
dx
3. ln y = x
Solution
y = ex
dy
∴ = ex
dx
x
4. y = e e
Solution
dy de x e x
= .e
dx dx
x
= e x .e e
x
+x
= ee
2
5. y = 10 x
Solution
let x 2 = u, then ~ y = 10 u
dy dy du
= .
dx du dx
= (10 u ln 10.2 x)
2
= 2 x(ln 10)10 x

3
6. Y = e 4 x let u = 4 x 3 ∴ y = e u

dy dy du
= .
dx du dx
3
= 12 x 2 .e 4 x

2 dy d ln x 2 ln x 2
7. y = e ln x = .e
dx dx
2 2
= e ln x .
x
ln x 2
2e
=
x

53
3.4. Differentials and Higher Order Derivatives

3.4.1 Differentials
dy
Until now we have used the symbol to represent the derivative of
dx
f(x) with respect to x. Now let’s assign different meanings to dy and
dx. Let f(x) =y and dx be a real variable thus, differential of y (dy) is
given by:
dy = f ' ( x)dx

Note:
1. dy and dx are variables the former being dependent and the
latter independent variable.
dy
2. If dx ≠ 0, one can have = f ' ( x)
dx
3. The main use of differentials is to find out approximations of
changes. In suing them for approximations, the value of dx
should be very small.

Examples

Find the differential of y (dy) for questions number 1,2, and 3.

1. y = x 3 − 4 x 2 + 5
x−2
2. y = 2
x
3. y = x , find dy when x=1 and dx=0.01
2

4. Using differentials, approximate the following number

a. 16.5 b. 3
65 c. 65
Solution

1. dy = (3x 2 − 8 x)dx
dy d ( x − 2) dx 2 x 2 − 2 x( x − 2)
2. = x2 − ( x − 2) =
dx dx dx x4

54
  x 2 − 2x 2 + 4x 
dy =  dx
 x4 
−x+4
= dx
x3

3. dy = 2 xdx
dy =2(1)(0.01)
dy =0.02
4. a. Let y = x
dy 1
=
dx 2 x
Taking x=16 and dx=0.5
1
dy = (0.5)
2 16
= 0.0625

y1 = y 0 + σy = x + σx = 16 + 0.0625 = 4.0625
∴ 16.5 ≅ 4.0625

b. Let y = 3 x
dy 1 13 −1 1
= x = 2
dx 3 3x 3
1
∴ dy = dx
3 x2
3

Taking x=64 and dx=1

1 1
∴ dy = .(1) = 16 = 0.021
33 64 2 3x
y1 = y 0 + dy = f (64) + dy = 4.021
∴ 3 65 ≅ 4.021
c. Let y = x
1
dy = dx
2 x
Taking x = 64 and , dx = 1
1
dy = = 0.0625
16
y + σy = f (64) + σy = 8.0625
65 ≅ 8.0625

55
3.4.2 Second and higher order derivatives

In some cases, the derivative of a function turns to be a function. In

this case, we can take the derivative of f ' ( x) to determine the
df ' ( x) d2y
instantaneous rate of change of f ' ( x). = f ' ' ( x) = 2 is known as
dx dx
the second order derivative of f(x). One can find the second derivative
of f(x) by differentiating it twice. Similarly, the third derivative of
 d3y
f ( x)  f " ( x) or 3  is found by differentiating f(x) three times.
 dx 

Example:
Find the first five derivatives of the following functions.

1. y = 2 x 4 − 3x 2 2. y = 3x 3 − 4 x 2 + 5

3. y = 2 x − 1 4. y = 4 x 3 + 5 x 2 + 3

5. y = 3x 4 + 6 x 3 − 3x 2 + 4 1
6. y =
x2

Solution:
1. f ' ( x) = 3x 3 − 6 x
f ' ' ( x) = 24 x 2 − 6
f ' ' ' ( x) = 48 x
f ( 4 ) ( x) = 48
f ( 5) ( x) = 0
dy
2. = 9 x 2 − 8x
dx
d2y
= 18 x − 8
dx
d3y
= 18
dx 3
d4y
=0
dx 4
d5y
=0
dx 5
Why don’t you try for the rest? Do them, they are simple!

56
Exercises
1. Given c( x) = x 3 − 9 x 2 + 33 x + 30 , determine the slope of marginal
cost at x=7.
2. If the production function is given by Y=20L3/4K5/4, what is
a. The range over which there is increasing returns to capital
b. The range over which there is decreasing returns to labor
c. The nature of returns to scale of production? Is it constant,
increasing, or decreasing returns to scale?
Solution:
1. c’(x)=3x2-18x, measures the slope of total cost curve
and to find the slope of marginal cost curve take the
derivative of MC, which is the second order derivative
of total cost function.
C”(x) =6X-18
C” (7) =6[7]-18=24
2. To answer question number two, you need to know
the concept of partial derivatives, which we will
present it below.

3.4 Partial derivatives and total derivatives

Though functions of singe variable have large simplicity, they are not
commonly used in economics. They are not common since the
functions that the real economic world consisted of are multivariate.
For example, our rational consumer will not consume one commodity
only; rather, s/he consumes different combinations to maximize
her/his utility. The representative production function of a firm is not
only a function of labor but it is also a function of capital. The overall
income of nations is not a function of physical capital stocks only. It is
also a function of human capital, factor productivity, institutions, and
government policies. Hence, we have to extend the concepts of
univariate calculus to multivariate ones. For the time being, let’s
begin with a function of two independent variables so that we can
simply extend it to n-variables case.
y = f ( x, z )
Let’s take the Cobb Douglas production function of a hypothetical firm.

y = ALα K β
Where:
Y-is the level of output
A-is the state of technology, which reflects productivity of factors
L & K represent the units of labor and capital
α & β are elasticity of y with respect to labor and capital respectively

57
Example
3 1
1. Suppose for a firm y = f ( L, K ) = 60 L 4 K 4 , then
a. How many units output will be produced if the firm uses 81 labor
and 16 capital
b. Show that the production function is constant returns to scale
Solution

a. Y=f (L, K)=60 (81)3/4(16)1/4=3240

b. To check for returns to scale, add the exponents of labor and

capital. If the sum is greater than one it is increasing returns to
scale, less than one it is decreasing returns to scale and if it is
equal to one it is constant returns to scale. Since 3/4 +1/4=1, it
is constant returns to scale. For the proof, you can see the
following table.

Optional♥

Proof: For a general production function, when we change all the

inputs by a given proportion if output changes by equal proportion,
then the production function illustrates constant returns to scale. If
it changes by greater/smaller proportion, the production function
illustrates increasing/decreasing returns to scale respectively.

For Y = ALα K β and, Let Yo= ALα K β

Assume labor and capital have changed by λ, the new labor and
capital will be λL, and λK respectively. So the new production
function will be,

YN = A(λL) α (λK ) β

=AλαLαλβKβ

=A(λα+β) Lα K β

=(λα+β)A Lα K β

=(λα+β)Y0

If α+β >1, it shows increasing returns to scale, if α+β <1, it shows

decreasing returns to scale, and if α+β =1, it is constant returns to scale.

♥
Optional points are points that are essential for knowledge but do not appear on
your exam.

58
3.4.1 Partial derivatives

Let y = f (x, z) . If we want to know the changes in y due to changes in x

or z, we have to find out the tow derivatives of y which are known as
partial derivatives with respect to each variable.

 ∂y 
The partial derivative of y with respect to x written as   is the
 ∂x 
derivative of y when z is considered as a constant and x is the only
independent variable.
 ∂y 
The partial derivative of y with respect to z is written as   is
 ∂z 
the derivative of y when x is considered as a constant and z is the
only independent variable.
∂f ∂f ∂f
= fx, = fz, = f y , etc
∂x ∂z ∂y

Example

1. If f ( x, y ) = 5 x 3 y 2 , find
∂f
a.
∂x
∂f
b.
∂y
∂ ∂f ∂2 f
c. f xx = ( ) = 2
∂x ∂x ∂x
∂ ∂f ∂2 f
d . f xy = ( ) =
∂x ∂y ∂xy
∂ ∂f ∂2 f
e. f yx = ( )=
∂y ∂x ∂yx
Solution
∂f
a. = 5 y 2 (3x 2 ) = 15 x 2 y 2
∂x
∂f 3 3 ∂2 f
b. = 10 x y and = 10 x
∂y ∂y 2
∂2 f ∂
c. = (15 X 2 y 2 ) = 30 xy 2
∂x 2
∂x

d. f xy = 30 x 2 y

59
 e. f yx = 30 x 2 y
Why don’t you try the following? Just try them!
2. If f ( x, y ) = (4 x + 3 y − 5) 8 , find f x , f y , f xx , f yy and f xy .
y
3. If f ( x, y ) = , find f x , f y , f xx , f yy , f xy
x + 3y
1 3
4.Given y = 60 L 4 K , find MPL /( L, K ) and MPK /( L, K ) where (L, K)=(81,16)
4

Moreover interpret your result.

3.4.2 Total derivatives (differentials)

We have seen that partial derivatives of f (x, y) indicate how much the
function changes with respect to small change in its variables. In
particular if σx and σy are small at the point (a, b), a change in x of
 ∂f 
σx units produces a change in f ( x, y ) of approximately  (a, b)σx units;
 ∂x 
and, a change in y of σy units produced a change in f ( x, y ) is
 ∂f 
approximated  (a, b)σy and the over all change is approximated
 ∂x 
using the sum of these terms. Thus,
 ∂f   ∂f 
f (a + σx, b + σy ) ≈ f (a, b) ≈  (a, b)σx +  (a, b)δy
 ∂x   ∂y 
The expression on the right hand side is known as total differential.

Note that,
 ∂f   ∂f 
For f ( x, y ) ⇒ total differential =  (a, b) σx +  (a, b) σy
 ∂x   ∂y 
 ∂f   ∂f   ∂f  
For f ( x, y, z ), total differential is  (a, b, c) σx +  (a, b, c) σy +  (a, b, c) σz 
 x   ∂y   ∂z  

Examples
3 1
1. Given a production function y = 60 L 4 K 4 . The total output when
L=81 and K=16 units is 3240 units, use the concept of total
differentials to approximate the effect of one unit decease in labour
and one unit increase in capital on output.

60
Solution
 ∂f   ∂f 
f (81 − 1,16 + 1) ≈ f (81,16) ≈  (81,16) (− 1) +  (81,16) (1)
 ∂L   ∂K 
1 3
 16  4  81  4
= 45  (− 1) + 15  (1)
 81   16 
405 5
= −30 + = 20
8 8
Thus a unit in L decreases y by 30 and a unit increase in K increases y
405 5
by and in total y increases by 20 units.
8 8
2. Suppose the profit for a firm selling two goods at prices p and q
respectively is by:
π ( p, q) = −100,000 + 5000 p + 10,000q − 50 p 2 − 100q 2 + 10 pq
when p = 11 & q = 50, π will be $200,000.
Use total differentials to approximate the effect of a $ 1 increase in p
and $0.20 increase in q. Compare your result with π (101,50.20)

Now it is time for you to answer the exercise number 2 on page 21.

3.5 Some Applications of Differential Calculus

3.5.1. L Hôpital’s Rule

In dealing with indeterminate limits, the best way to find limits is using
the “L Hôpital’s Rule”. To use this rule, let’s differentiate among the
different indeterminate limits.
f ( x)
We say Lim is in
x →a g ( x)

0
a. Indeterminate form if Lim f ( x) = Lim g ( x) = 0
0 x→a x→a

∞
b. Indeterminate form if Lim f ( x) = Lim g ( x) = ∞
∞ x→a x→a

The following theorem allows us to deal with 0 indeterminate form.

0

Λ
First form of L H o pital ' s Rule : Suppose that f (a) = g (a ) = 0 , that
f ( x) f ' (a)
f ' (a ) and g ' (a) exist, and g ' (a) ≠ 0, then Lim =
x →a g ( x) g ' (a)

61
 For example, see
x3 − 1 0
Lim =
x →1 x − 1 0
2
3X
Lim =3
x →1 1
Stronger form of L’: Suppose that f and g are differentiable on
an open interval I containing a, and that g ' ( x) ≠ 0 on I if x ≠ a .
f ( x) f ' (a)
Then Lim =
x →a g ( x) g ' (a)
Λ ∞
Note that the L H o pital ' s rule also applies to indeterminate forms.
∞

We have also other indeterminate forms and some these are

⇒ 0∞
⇒ 00
⇒ 1∞
⇒ ∞ 0 and
⇒ ∞ −∞
0 ∞
To deal with these forms change them into or forms by taking
0 ∞
logarithms and performing some algebra.

3.5.2. Homogeneous Functions and Euler’s Theorem

A function is said to be homogeneous of degree “r”, if multiplication of

each of its independent variables by a constant “j” will alter the value
of the function by the proportion “jr,” that is if:
f ( jx1 ,..............., jx n ) = j f ( x1 ,.........., x n )
r

In general j can take any value. However, in order for the above
equation to make sense, (jx1,……….,jxn) must not lie outside the
domain of the function f. For this reason, in economic applications the
constant j is usually taken to be positive, as most economic variables
do not admit negative values.

Example

x 2w
1. f ( x, y, w) = + , if we multiply each variable by j, we get
y 3x
jx 2 jw
f ( jx, jy, jw) = + = f ( x, y, w) = j 0 f ( x, b, w).
jy 3 jw

62
Thus f is a homogenous function degree of zero (r=0)

x 2 2w w j 2 x 2 2 j 2 w2
2. f ( x, y, w) = + ⇔ f ( jx, jy, jw) = +
y x jy jx
 x 2 2w 2 
= j  + 
 y x 
= j f ( x, y, w)
Thus f(x,y,w) is a homogeneous function degree of one (r=1).

3. h( x, y, w) = 2 x 2 + 3 yw − w 2
h(ax, ay, aw) = 2a 2 x 2 + 3a 2 yw − a 2 w 2
= a 2 (2 x 2 + 3ayw − w 2 )
= a 2 h( x, y, w)

Thus h is a homogeneous function degree of 2.

3.5.2.1 Linear Homogeneity

A function is linearly homogeneous if it is homogenous degree of one.

The Cobb Douglas production function is homogeneous function degree

of α & β , if α + β = 1 then it is a linear homogeneous function.

3.5.2.2. Properties of linearly homogeneous production function

For Y = ALα K β and α+β =1
1. Both APL & APK can be expressed as a function of (capital labor
ratio). If Y = ALα K 1−α , then
1−α
Y α −1 1−α K
APL = = AL K = A 
L L
−α
Y K
APK = = ALα K −α = A 
K L
2. Both MPL & MPK can be expressed as a function of capital labor
ratio.

3. If Y= f (K,L), then from the Euler’s theorem we can find

∂Y ∂Y
K +L =Y .
∂K ∂L
Let us formally and briefly state the Euler’s theorem.

63
3.5.2.3 Euler theorem
If Z = f = ( y, x ) be homogeneous function of x and y of degree n and
posses continuous partial derivatives, then
This is Euler theorem where n
∂z ∂z
x +y = nz represents the degree of
∂k ∂y homogeneity and z the original
function.
But this shouldn’t confuse you with the total differential which is given
∂z ∂z
by dx + dy = dz and different from the Euler’s theorem.
∂x ∂y

Example:
Let’s take the famous production function Y = ALα K β with α+β =1 so
that we can have Y = AL1− β K β .

∂Y
= (1 − β )( AL1− β −1 K β )
∂L
∂Y
and , L = (1 − β )( AL1− β K β )
∂L
∂Y
= β ( AL1− β K β −1 )
∂K
∂Y
K
∂K
( )
= β AL1− β K β , ~ FromEuler ' s``theorem, we ~ can ~ write ~ as

∂Y ∂Y
K
∂K
+L
∂L
( )
= (1 − β )( AL1− β K β ) + β AL1− β K β

( )
= (1)( AL1− β K β ) − ( β )( AL1− β K β ) + ( β ) AL1− β K β
= ( AL1− β K β )
=Y

64
 IV. APPLICATIONS OF DIFFERENTIAL CALCULUS
4.1. Introduction
In the previous sections we have dealt with the techniques of
differential calculus. Now it is essential to see its applications. The two
of the most important applications of derivatives are obtaining the
graphs of functions and finding the optimum values. Optimization is
the “quest for the best!” It involves either maximization or
minimization. Since the economic person is an optimizer, most part of
the chapter deals with the applications of derivates to static
optimization problems♥. However, this doesn’t mean the application of
derivatives to find out a graph of a function is irrelevant for
economists. I hope no one forgets the relevance of graphs in
economics for Paul. A. Samuelson has said it “a graph is worth more
than thousand words.” In fact, we will briefly see the applications of
derivatives to graphing functions first and divert our attention to
optimizations.

Curve sketching- involves identifying the x-intercept and y-intercepts,

whether a function is increasing or decreasing in a given range, finding
out its extreme points, and finally checking whether the function is
concave up or concave down at a point or a point is an inflection point.
Let’s see each of these concepts and tie them with the derivatives of
f(x). Note that we commonly read graphs from left to right.

4.1.1. Increasing, and Decreasing Functions

A function f(x) is increasing if the value of the dependent and the
independent variables are directly related. That is as the value of x
increases, the value of f(x) increases too and vice versa.
Mathematically speaking:

A function f(x) is increasing over an interval if whenever X1>X2,

then f(X1)> f(X2) for any X1 and X2 in that interval.

On the other hand, a function f(x) is decreasing if the value of the

independent and the dependent variables are inversely related. That
is, as the value of X increases, the value of f(x) decreases and vice
versa.
Thus f(x) is decreasing over an interval if for any X1 and X2, when
X1>X2 f(X1)< f(X2) over that interval.

♥
The more realistic and subtle form of optimization is dynamic optimization which is
left for advanced courses.

65
Graphically,
Y=f(x) Y=f(x)

F (x1) F (x2)

F (x2)
F (x1)

X1 X2 X1 X2
X1<X2 and f(X1)> f(X2), X2>X1 and f(X2)> f(X1),
therefore f(x) is decreasing therefore f(x) is increasing

The graphs of some functions, like that of marginal utility, have

increasing or decreasing property throughout. But, most of functions
are increasing in a given range and decreasing in another range. For
example, the graph of total product increases first, reaches maximum
and then decreases. Most of the economic functions are not increasing
or decreasing throughout. Graphs of marginal cost, average total cost,
average variable costs, average product, and marginal product are not
increasing or decreasing throughout. The point at which the graph
changes from decreasing to increasing or vice versa is known as an
extreme point or critical point or stationary point. An extrema point is
a point of relative maximum or relative minimum.

If a function is increasing, it has positive slope which implies that f ’(x)

is positive. If a function is decreasing, it has negative slope and
negative first derivative. If the point is an extreme pint, there is zero
slope or undefined slope at that point and f ' ( x) is zero or undefined.

In summary for f(x) =y

dy
• If < 0, f ( x) is decreasing
dx
dy
• If > 0, f ( x) is increasing
dx
dy
• If = 0, f ( x) has extrema value at this value of x
dx

66
4.1.2. Testing For Concavity
Knowing the range over which the function increasing, decreasing, and
achieving relative extrema point is important to draw graphs of
functions but they are not all about. A given function increases over a
range but it can increase at an increasing rate, a constant rate or at a
decreasing rate. Thus, we have to see the concavity of graphs to add
this dimension.

A given function is concave up (changes at increasing rate) if the

function lays above all its tangent lines.

As we move from left to right the slope of f(x) increases; i.e., f ’(x) is
an increasing function. Which means f " ( x) is positive. Thus, if a
function is concave up, then f " ( x) >0.

A given function is concave down (changes at a decreasing rate) if the

function lies below its tangent line.

As we move from left to right the slope of f(x) decreases to zero. This
shows that f ’(x) is a decreasing function implying that f " ( x) <0.
Hence, if a function is concave down, then its f " ( x) is negative.

In summary,
If
• f " ( x) < 0, the function is concave down
• f " ( x) > 0, the function is concave up
• f " ( x) = 0, it may or not be point of inflection. At point of
inflection the function transforms itself from changing at an
increasing rate to decreasing rate and vice versa.

67
Note that if a function changes at constant rate, f ' ( x) is a constant
function and f " ( x) =0. These curves are linear curves. Hence, getting
f “(x)=0 may not be a sufficient condition for a point to be an inflection
point.

Example
1. Given f ( x) = x 3 − 3 x − 2 find
a. The ranges of x over which f(x) is decreasing and increasing
b. The point at which f(x) attains its relative extreme value
c. The ranges of x over which f(x) is concave up and down
d. The point of inflection
e. Plot the graph

2. Given TVC ( x) = 18 x − 15 x 2 + 4 x 3 , find

a. MC function
b. The point of diminishing returns
c. The point at which stage II of production begins
d. Plot the graph

Solution

1. f ' ( x) = 3x 2 − 3 f " ( x) = 6 x
a. If f(X) is increasing if f ‘(x)>0. If f(X) is decreasing if
That is, 3x 2 − 3 >0 f ‘(x)<0. That is,
x2 −1 > 0 3x 2 − 3 <0
( x − 1)( x + 1) > 0 x2 −1 < 0
( x − 1)( x − (−1)) > 0 ( x − 1)( x − (−1)) < 0
x > 1 or x < −1 −1 < x < 1

Thus, the function f(x) is increasing for the value of x>1 or x<-1. It is
decreasing in the range − 1 < x < 1

b. At the point of extrema, f ' ( x) = 0 . That is

3x 2 − 3 = 0
x2 = 1
x = ±1
f (1) = −4 f (−1) = −1 + 3 − 2 = 0
Hence, points of extrema are (1,4) and (-1,0).
c. f " ( x) = 6 x

68
 If f(x) concave up, If f(x) concave down,
f " ( x) > 0 f " ( x) < 0
6x > 0 6x < 0
x>0 x<0

d. f " ( x) = 0 at the point of inflection. That is 6x=0 this occurs when

x=0

f ( x) = x 3 − 3 x − 2

-1 2

-2

-4
2. TVC = 18 x − 15 x + 4 x , find
2 3

a. Marginal cost function

b. Point of diminishing returns
c. The point at which stage II of production begins.
dTVC
a. MC = = 18 − 30 x + 12 x 2
dx
b. At the point of diminishing returns MP is maximum and MC is
minimum. Thus at this point
dMC
=0
dx
− 30 + 24 x = 0
30 10 5 × 25
x= = = = 1.25
24 8 4 × 25

c. At the point where stage II begins AP is maximum and AVC is

minimum. Thus,
AVC = 18 − 15 x + 4 x 2
dAVC
=0
dx
− 15 + 8 x = 0
15
x= = 1.875
8

69
4.2. Free (unconstrained) optimization
Optimization, as described earlier involves maximization or
minimization of a given objective function. Most of our problems in
economics are optimized subject to some constraints but to grasp the
basic idea, let’s begin how one can handle optimization problems
functions without any constraint. In the next section we will extend
the technique to the constrained one. We will begin by briefly
describing an optimum value and proceed to the next section.

4.2.1. The concept of optimum (extreme) value

An extreme value of a function is the maxima or minima value of that

function. We can have two types of extreme values.

Relative minima
a. Relative extreme values
(Local extrema)
Relative maxima
Absolute minima
b. Absoloute extrema values
(Global extrema)
Absolute maxima

4.2.1.1. Relative Extreme Value

Relative extreme value: is an extreme value when compared with the
values near to it. It can be relative maxima (when it is higher than
any other point near to it) or relative minima (when it is lower than
any value near to it)

1 3
For example, consider the function y = x − x 2 − 3x + 2 . It graph is:
3

Relative maxima
( -1, 1 )
3

-2 3 5

(3,-7)
Relative minima

70
In general

• If a function f has a relative maxima at x = x 0 if there exist an

interval around x0 on which f ( x0 ) ≥ f ( x) for all x in the interval
• If a function f has relative minima at x = x 0 if there exist an
interval around x0 on which f ( x0 ) ≤ f ( x) for all x in the interval

It is easy to find out relative extreme values using the derivative.

Previously we have seen that a function changes from increasing to
decreasing and vice versa. When its first derivative is zero or
undefined. Since a function changes from decreasing to an increasing
at a relative minima, we have f ' ( x) = 0 or f ' ( x) is undefined at relative
minima or maxima.

Hence, we can find relative maxima and minima for a function by

finding the values of x for which f ' ( x) = 0 or f ' ( x) is undefined. Once we
get the critical value, we can use the behavior of f ‘(x) near the
extreme point to know whether it is a relative minima or relative
maxima point.

To find out relative minima or maxima,

1st Find f ' ( x)

2nd Solve for the value of x which will make f ' ( x) = 0
3rd Use the above value to get the extreme value of the function (i.e,
insert the value you got in the 2nd step in f(x) to get the extreme
value).
th
4 Evaluate the first derivative function at some values of x to the teft
and the right of each critical point
a. If f ' ( x) > 0 to the left and f ' ( x) < 0 to the right of the critical value
of f(x), the critical point is relative maximum
b. If f ' ( x) < 0 to the left and f ' ( x) > 0 to the right of the critical value
of f(x), the critical point is relative minima.

Example

1. Find the relative extremas of the following functions

1
a. f ( x) = x 3 − x 2 − 3x + 2
3
b. f ( x) = 2 x 3 − 6 x 2 + 6 x + 1
c. f ( x) = x 3 − 3 x + 6

71
Solution

a. f ' ( x) = x 2 − 2 x − 3
f ' ( x) = x 2 − 2 x − 3 = 0
x 2 − 2x + 1 = 3 + 1
x − 1 = ±2
x = 3 or x = −1
1
f (3) = 33 − 3 2 − 3(3) + 2
3
=-7
1
f (−1) = − − 1 + 3 + 2
3
1 11
= 4− =
3 3

 11 
Thus, the critical values are (3,−7) and  − 1,  . To find out which
 3
one relative minima or maxima, use the number line.

f ' ( x) + 0 − f ' ( x) − 0 +
x −2 −1 2 X 0 3 4

 11  Thus, (3,−7 ) is relative minima.

Thus,  − 1,  is relative maxima.
 3

b. f ( x) = 2 x 3 − 6 x 2 + 6 x + 1
f ' ( x) = 6 x 2 − 12 x + 6
f ' ( x) = 6 x 2 − 12 x + 6 = 0
x 2 − 2x + 1 = 0
( x − 1) 2 = 0
x =1
f (1) = 2 − 6 + 6 + 1 = 3

The only critical value is (1,3) . To check whether it is relative minima

df ( x)
or maxima, let’s see the behavior of near 1.
dx

72
 f ' ( x) + 0 +
x 1
This shows that the function has no relative extreme value.

c. f ( x) = x 3 − 3 x + 6
f ' ( x) = 3 x 2 − 3
= 3( x − 1)( x + 1) = 0
x = 1 or x = −1
f (1) = 4 or f (−1) = 8

Critical values are (1,4) and (-1,8)

f ' ( x) + 0 − f ' ( x) − 0 +
f −1 f 1

Relative maxima at (-1,8). Relative minima at (1,4).

Note that instead of taking the number line to check for maxima or
minima, we can use the 2nd derivative. If λ is the critical value which
makes f ' ( x) = 0 , then the point (λ , f (λ )) is relative maxima if f " (λ ) < 0
and relative minima if f " (λ ) > 0 . But when f " (λ ) = 0 we can’t use the 2nd
derivative test and we have to use the behavior of the first derivative
to check for relative optima.

Example
Use the 2nd derivative test of the previous function to know whether
they are relative maxima or relative minima.

1 3
a. f ( x) = x − x 3 − 3x + 2
3
b. f ( x) = 2 x 3 − 6 x 2 + 6 x + 1
c. f ( x) = x 3 − 3 x + 6

 11 
a. f ' ( x) = x 2 − 2 x − 3 critical values are  − 1,  and (3,-7)
 3
 11 
f ' ' ( x) = 2 x − 2 f ' ' (−1) = −4 < 0 so  − 1,  is relative maxima
 3
f ' ' (3) = 6 − 2 = 4 > 0,∴ (3,−7) is relative minima

b. f ' ( x) = 6 x 2 − 12 x + 6

73
 f " ( x) = 12 x − 12
critical values x=1 (1,3)
f " (1) = 12 − 12 = 0
Use the number line in this case.

c. f ' ( x) = 3 x 2 − 3 f " ( x) = 6 x
critical values (1,4) and (-1,8)
 f " (1) = 0 
 f " (−1) = 0 Use the first derivative test in this case.
 

4.2.1.2. Absolute Extrema

Absolute extreme value: is an extreme value that the function attains

throughout its domain. The possible largest value that the function
attains in its domain is called absolute maxima of the function.
Similarly, the smallest possible value that the function attains
throughout its domain is known as the absolute minima of the
function. Note that a function may (or may not) have absolute
extreme value like there may or may not exist relative extreme value.

For example consider a function y=2x+1. As we increase the value of

x, the value of y increases continuously. Again, when we decrease the
value of x, the value of y decreases continuously. However, if the
domain is restricted to a closed interval [a, b] which means
a ≤ x ≤ b and the function is continuous everywhere within the interval,
the function will have an absolute minima and absolute maxima within
the interval. They may be located either:

a. At the end points; i.e., a and b or

b. At the interior points

To find absolute extrema within an interval

a. Find the critical values and critical points and check whether
they are within the interval
b. Calculate f(a) and f(b)
c. The largest value and smallest value obtained from the above
steps represent the absolute maxima and the absolute
minima of the function over the interval respectively.
Example
1. Find the absolute maxima and minima of f ( x) = x 2 − 6 x + 2 over the
interval 0 ≤ x ≤ 5.

74
Solution
f ' ( x) = 0
2x − 6 = 0
x=3 There is absolute maxima at x=0 and absolute
f (3) = −7 minima at x=3 within [0, 5].
f (0) = 2
f (5) = −3
2. Find the absolute extrema of f ( x) = x 3 + 3x 2 + 4 . Over the interval
[-4, 2]
Solution
f ' ( x) = 3x 2 + 6 x = 0
Absolute maxima at x=2 (2, 24) and absolute
3x( x + 2) = 0 minima at x=-4 (-4, -12)
x = 0 or x = −2
f (0) = 4 f (−2) = 8
f (2) = 24 f (−4) = −12
Try the following problems and check your answers with that of the
tutor!
3. Given TC = 5 x + 200 and TR = 10 x − 0.01x 2 , what is the value of x, which
will maximize profit

4. Given TC = 200 + 25 x + 2 x 2 , what is the level of x which minimizes

average cost?

5. A farmer uses his 1000 hectar land to raise two grain crops. Grain
A is sold at a market and grain B is fed to his herd of 1000 cattle.
Using the data below determine how many hectars of each crop
should be planted in order to realize greatest profit.
a. A hectar planted with grain A yields 150 bushels which is sold at
$5 per bushel.
b. Feeding the herd, the yield of x hectares of grain B will result in
each steer weighing 350+x-0.001x2 pound. Steer sells for $1
per pound.
6. A book publisher sells 100,000 copies of a certain book each year.
Setting up each run costs $2000, each book costs $5 in material
and labor, and storage fees are $1 per book per year. Assuming, a
uniform rate of sales throughout the year, find the number of runs
that minimize the publishers cost.
7. A travel agency will plan a group four for groups of size 25 or
larger. If the group contains exactly 25 people, the cost is $300
per person. However, each person’s cost is reduced by $10 for
each additional person above 25. What size group will produce the
largest revenue for the agency?

75
4.3. Constrained optimization
How strong the concepts of extrema are insightful, they worth nothing
for us unless they are tied to the theoretical and empirical problems
with which economists are concerned with. Consumers maximize
utility; but, under the constraints of income and prices. Producers
maximize profit; but, under the constraints of labor’s and capital’s
price. Since most of our optimization problems are some what
constrained it is necessary to extend the tools of calculus for
unconstrained to constrained optimization. This extension had first
emerged in the monumental work of Joseph Louis Lagrange. His
research has illustrated how one can obtain maxima or minima of a
function f(x, y) subject to a constraint g(x, y) =0. The method of
obtaining this optimum value is known as Lagrange Multiplier.

So as to find the critical values of f(x, y) subject to the constraint g(x,

y)=0, let’s introduce new variable " λ " to form the objective function.
That is:
f ( x, y , λ ) = f ( x, y ) ± λg ( x, y )

Lagrange has shown that the critical values of f ( x, y, λ ) will satisfy the
constraint g(x, y) and is the critical point of f(x, y).

To find out the critical values of f ( x, y, λ ) , find out the values of x, y,

and λ which will make ∂f with respect to each variable zero. That is

∂f 
= 0
∂x 
∂f 
= 0
∂y 
∂f 
= 0
∂z 

The values of x and y that simultaneously satisfy the three equations

are the critical values of λ and f.

76
Examples

1. Find the critical points and values of f ( x, y ) = x 2 + y 2 subject to

g ( x, y ) = 2 x + 3 y − 4

Solution
f ( x, y, λ ) = x 2 + y 2 + λ (2 x + 3 y − 4)
= x 2 + y 2 + 2λx + 3λy − 4λ

∂f
= 2 x + 2λ = 0 λ = −x
∂x
∂f 2
= 2 y + 3λ = 0 λ=− y
∂y 3
∂f
= 2x + 3 y − 4 = 0
∂y

x = −λ
3
y=− λ
2
2
−x=− y
3
2
3x = 2λ x= y
3
2x + 3y = 4
2 
2 y  + 3 y = 4
3 
4
y + 3y = 4
3
13 12
y=4 Y=
3 13
2  12  8
x=  =
3  13  13
2 2
 8 12   8   12  16
The optimum value is f  ,  =   +   =
 13 13   13   13  13
2. Minimize cost c( x, y ) = 21x + 14 y subject to g ( x, y ) ⇒ xy = 600, where x,
y>0

Solution
c( x, y ) = 21x + 14 y
g ( x, y ) = xy − 600 = 0

77
 f ( x, y, λ ) = 21x + 14 y + λxy − 600λ

21 λ =λ
21 + λy = 0 ⇔ λ = −
y
 21 − 14
14  − = ⇔ 3x = 2 y
14 + λx = 0 ⇔ λ = −  y x
x
2
xy = 600 x= y
2 3
2 x = × 30 = 20
y. y = 600 3
3
min . cos t (20,30) = 21(20) + 14(30)
3
y 2 = x600 = 420 + 420
2
y = 30 = 840
3 1
3. Given production function y = 60 L 4 K 4 . Assume price of labor is
$100 whereas price of capital is $200 and the firm has $30,000
available to spend on production. How many units of labor and
capital should be used to maximize output?

Solution

3 1
Maximum: y = 60 L 4 K 4
Subjected to: PL L + PK K = 30,000
100 L + 200 K = 30,000

3 1
f ( L, K , λ ) = 60 L 4 K 4
+ 100λL + 200λK − 30,000λ

∂f −1 1
= 45 L 4 K 4 + 100λ = 0
∂L
∂f 3 −3
= 15L 4 K 4 + 200λ = 0
∂K
∂f
= 100 L + 200 K − 30,000 = 0
∂K

1
K 4
45  = −100λ
L
1
45  K  4
λ=−   ........(1)
100  L 

78
 −3
K 4
15  = −200λ
L
−3
15  K  4
λ=−   ........(2)
200  L 
100 L + 200 K = 30,000.............(3)

45 14 −14 15 −3 4 3 4
− K L =− K L
100 200
6K = L
100(6 K ) + 200 K = 30,000
800 K = 30,000
300
K= = 37.5
8
Try the following problems and check your answers with that of the
tutor!
3 1
4. The production function of a firm is given by Q = 64 L 4 K 4 when labor
costs $96 per unit and capital costs $162 per unit the firm decides
to produce 3456 units of Q.
a. Determine the amount of labor and capital that should be utilized
so as to minimize cost.
b. Find the value of λ at the optimum value and give its economic
meaning
MPL P
c. Show that at the optimum level = L
MPK PK

5. On the basis of past experience a company has determined that its

sales revenue is related to its advertising according to the formula
s = 20 x + y 2 + 4 xy where x and y is the amount of money spent on
advertising on radio and TV respectively. If the company plans to
spend $30,000 on these two means of advertising, how much
should they spend on each method to maximize their sales
revenue?
6. Suppose the utility function for two goods x and y is given by
u = x 2 y, and a consumer purchases 6 units of x and 15 unit of y. If
the consumer purchases 60 units of y, how many units of x must be
purchased to retain the same level of utility.
7. If the utility function for two commodities is u = x 2 y, and the budget
constraint is 4 x + 5 y = 60 , find the values of x and y that maximize
utility.

79
4.4. Testing For the Nature of Optima
In case when we are dealing with objective functions with more than
one independent variable, the first order partial derivative shows us
the value that is optimum. In case when the objective function is not
clearly stating whether to maximize or minimize [both in the
constrained and unconstrained case], we can’t identify whether the
critical value is maxima or minima. The next course-Linear Algebra for
Economists has a lot to say about this point, but let us briefly state the
simplest case instead of leaving you without any information. To know
whether the optimum value is maximum value or minimum value we
undertake 2nd order partial derivative test.

Let a and b be the critical points calculated from first order Lagrange
multipliers function. Then,
Find f xx (a, b) = A
f xy (a, b) = B and,
f yy (a, b) = C
Then calculate D = AC − B 2
If:
1. D>0 and A<0, then f (a, b) is local maximum
2. D>0 and A>0, then f (a, b) is local minimum
3. D<0, the function has a saddle point at f (a, b).
4. D=0, the test fails. The function must be investigated
near the point.

80