Chapter Three: Univariate Calculus and Multivariate Calculus

3.1. Introduction
In the preceding chapter of derivatives, we have applied the key concepts of calculus, the
derivative to functions with one independent variable. And then we extended the key concepts
of calculus to functions with two independent variables. Practically, we might use functions with
two independent variables to study how production depends on both labor and capital.
Functions with two or more independent variable, in general, are considered as multivariable
functions.

For instance, a small manufacturing company produces good with quantity 𝑸. If fixed costs are
$500 per week and variable costs are $70 per output produced, the weekly cost function is given
by:
𝑪 = 𝒇(𝑸) = 𝟓𝟎𝟎 + 𝟕𝟎𝑸
where 𝑸 is the number of outputs produced per week. The cost function is a function of a single
independent variable 𝑸. For each value of 𝑸 from the domain of 𝑪, there exists exactly one value
of 𝑪(𝑸) in the range of 𝑪.
Now, suppose that the company decides to produce variety of products, 𝑸𝟏 and 𝑸𝟐 If the fixed
cost for the new product is $200 per week and the variable costs is $100 per output, then the
cost function above shall be modified to:
𝑪 = 𝒇(𝑸𝟏 , 𝑸𝟐 ) = 𝟕𝟎𝟎 + 𝟕𝟎𝑸𝟏 + 𝟏𝟎𝟎𝑸𝟐
The modified equation above is an example of a function with two independent variables 𝑸𝟏 and
𝑸𝟐 . Of course, as the company expands its product line even further, its weekly cost function
must be modified to include more and more independent variables, one for each new product
produced.

In general, an equation of the form 𝒛 = 𝒇(𝒙, 𝒚) describes a function of two independent variables
if, for each permissible ordered pair
(𝒙, 𝒚), there is one and only one value of 𝒛 determined by 𝒇(𝒙, 𝒚).The variables 𝒙 and 𝒚 are
independent variables, and the variable 𝒛 is a dependent variable.

Here the set of all ordered pairs of permissible values of 𝒙 and 𝒚 is the domain of the function,
and the set of all corresponding values 𝒇(𝒙, 𝒚) is the range of the function.

We can similarly define functions of three independent variables 𝒘 = 𝒇(𝒙, 𝒚, 𝒛); of four
independent variables, 𝒖 = 𝒇(𝒘, 𝒙, 𝒚, 𝒛)and so on.

It is impossible to provide any sort of graphical interpretation for functions of more than two
variables. For example, a function of, say, four variables would require five dimensions, one for
each of the independent variables and a further one for the dependent variable! In spite of this
setback we can still perform the task of differentiating functions of several variables and, as we
1
shall see in the remaining sections of this chapter, such derivatives play a vital role in analyzing
economic behavior.

3.2. Partial Derivatives

Partial derivative is a technique for deriving the rate of change of a function with respect to
increases in one independent variable when all other independent variables in the function are
held constant.

The basic rule for partial differentiation is that all independent variables, other than the one that
the function is being differentiated with respect to, are treated as constants. Apart from this,
partial differentiation follows the standard differentiation rules.

A symbol ∂ is used in a partial derivative to distinguish it from the derivative of a single variable
(total derivative) function where a normal letter d is used.

3.2.1. First-order Partial Derivatives

Given a function of two variables, 𝒛 = 𝒇(𝒙, 𝒚) we can determine two first-order derivatives.
 The first order partial derivative of 𝒇 with respect to 𝒙 is given as:
𝝏𝒛 𝝏𝒇
or or 𝒇𝒙
𝝏𝒙 𝝏𝒙
The above expression denotes differentiating 𝒇 with respect to 𝒙, with 𝒚 held constant
 The first order partial derivative of 𝒇 with respect to 𝒚 is given as:
𝝏𝒛 𝝏𝒇
or or 𝒇𝒚
𝝏𝒚 𝝏𝒚
Likewise, the expression above represents differentiating 𝒇 with respect to 𝒚 while holding
𝒙 constant.

Example: Find the first-order partial derivatives of the ff functions

a) 𝒛 = 𝒇(𝒙, 𝒚) = 𝒙𝟐 + 𝒚𝟑
b) 𝒛 = 𝒇(𝒙, 𝒚) = 𝒙𝟐 𝒚

Solution:
a) By the sum rule we know that we can differentiate each part separately and then add.
𝒇𝒙 = 𝟐𝒙 + 𝟎 = 𝟐𝒙
𝒇𝒚 = 𝟎 + 𝟑𝒚𝟐 = 𝟑𝒚𝟐

b) When we differentiate a multiple of two variables we differentiate each and then multiply by
the constant. Thus,
𝒇𝒙 = 𝟐𝒙𝒚 + 𝟎 = 𝟐𝒙𝒚
𝒇𝒚 = 𝟎 + 𝒙𝟐 = 𝒙𝟐

2
Self-Exercise: for the following given functions find the first order partial derivatives with respect
to both variables [i.e 𝒇𝒙 and 𝒇𝒚 ]
a) 𝒛 = 𝒇(𝒙, 𝒚) = 𝟓𝒙𝟒 − 𝒚𝟐
b) 𝒛 = 𝒇(𝒙, 𝒚) = 𝒙𝟐 𝒚𝟑 − 𝟏𝟎𝒙

3.2.2. Higher Order Partial Derivatives

There is a possibility of differentiating a multivariable function for the second and even more
times.

Consider the expressions below and their representation:

Notation Denotation

𝝏𝟐 𝒛 𝝏𝟐 𝒇 The function obtained by differentiating twice

or or 𝒇𝒙𝒙 with respect to 𝒙
𝝏𝒙𝟐 𝝏𝒙𝟐

𝝏𝟐 𝒛 𝝏𝟐 𝒇 The function obtained by differentiating twice

or or 𝒇𝒚𝒚
𝝏𝒚𝟐 𝝏𝒚𝟐 with respect to 𝒚

𝝏𝟐 𝒛 𝝏𝟐 𝒇 The function obtained by differentiating first

or or 𝒇𝒙𝒚
𝝏𝒙𝝏𝒚 𝝏𝒙𝝏𝒚 with respect to 𝒚 and then with respect to𝒙

𝝏𝟐 𝒛 𝝏𝟐 𝒇 The function obtained by differentiating first

or or 𝒇𝒚𝒙
𝝏𝒚𝝏𝒙 𝝏𝒚𝝏𝒙 with respect to 𝒙 and then with respect to 𝒚

Example: find expressions for the second-order partial derivatives 𝒇𝒙𝒙 , 𝒇𝒚𝒚 , 𝒇𝒙𝒚 and 𝒇𝒚𝒙 for the
following functions.
a) 𝒛 = 𝒇(𝒙, 𝒚) = 𝒙𝟐 + 𝒚𝟑
b) 𝒛 = 𝒇(𝒙, 𝒚) = 𝒙𝟐 𝒚
Answers:
a) 𝒇𝒙𝒙 = 𝟐 b) 𝒇𝒙𝒙 = 𝟐𝒚
𝒇𝒚𝒚 = 𝟔𝒚 𝒇𝒚𝒚 = 𝟎
𝒇𝒙𝒚 = 𝟎 𝒇𝒙𝒚 = 𝟐𝒙
𝒇𝒚𝒙 = 𝟎 𝒇𝒚𝒙 = 𝟐𝒙

Surprisingly, differentiating with respect to 𝒙 then 𝒚 gives the same expression as differentiating
with respect to 𝒚 then 𝒙 (i.e. 𝒇𝒙𝒚 = 𝒇𝒚𝒙 ). In fact, there are some exceptional mathematical
functions for which this result is not true (for some functions 𝒇𝒙𝒚 ≠ 𝒇𝒚𝒙 )

Self-Exercise: find expressions for the second-order partial derivatives 𝒇𝒙𝒙 , 𝒇𝒚𝒚 , 𝒇𝒙𝒚 and 𝒇𝒚𝒙 for
the following functions and check that it holds that 𝒇𝒙𝒚 = 𝒇𝒚𝒙 .

3
 a) 𝒛 = 𝒇(𝒙, 𝒚) = 𝟓𝒙𝟒 − 𝒚𝟐
b) 𝒛 = 𝒇(𝒙, 𝒚) = 𝒙𝟐 𝒚𝟑 − 𝟏𝟎𝒙
Although we have concentrated exclusively on functions of two variables, it should be obvious
how to work out partial derivatives of functions of more than two variables.

For the general function 𝒚 = 𝒇(𝒙𝟏 , 𝒙𝟐 , 𝒙𝟑 , … , 𝒙𝒏 ), there are 𝒏 first-order partial derivatives,
written as:

𝝏𝒚 𝝏𝒇
or or 𝒇𝒊 , for 𝒊 = 𝟏, 𝟐, 𝟑, … , 𝒏
𝝏𝒙𝒊 𝝏𝒙𝒊
which are found by differentiating with respect to one variable at a time, keeping the remaining
𝒏 − 𝟏 variables fixed. The second-order partial derivatives are determined in a similar way.

3.3. Total Differentials and Total Derivatives

3.3.1. Total Differentials
𝒅𝒚
When the concept of differentiation was introduced, you have learnt that the derivative
𝒅𝒙
measured the rate of change of 𝒚 with respect to 𝒙 for infinitesimally small (very small) changes
in 𝒙 and 𝒚. These infinitesimally small changes in 𝒙 and 𝒚 are referred to as differentials.
𝒅𝒚
For any non-linear function 𝒚 = 𝒇(𝒙) the value will change if 𝒙 and 𝒚 change. If the changes
𝒅𝒙
in variables 𝒙 and 𝒚 becomes infinitesimally small, then even for non-linear functions, it holds:
∆𝒚 𝒅𝒚 𝒅𝒚
=  ∆𝒚 = ∆𝒙
∆𝒙 𝒅𝒙 𝒅𝒙
It is, therefore, not possible to predict the effect of a given change in 𝒙 on 𝒚 with complete
accuracy. However, for a very small change in 𝒙 [∆𝒙], we can say that it will be approximately
true that the resulting change will be given by the differential
𝒅𝒚
∆𝒚 = ∆𝒙
𝒅𝒙
The closer the function 𝒚 = 𝒇(𝒙) to linearity, the more accurate will be the prediction. Consider
the cases below:
Example 4.3: for the three functions below assume that the value of 𝒙 increases from 𝟏𝟎 to 𝟏𝟏.
𝒅𝒚
Predict the effect on 𝒚 using the derivative and check the answers against the new value of
𝒅𝒙
the function.
a) 𝒚 = 𝟐𝒙 b) 𝒚 = 𝟐𝒙𝟐 c) 𝒚 = 𝟐𝒙𝟑
Solutions: in all the cases ∆𝒙 = 𝟏𝟏 − 𝟏𝟎 = 𝟏
𝒅𝒚
a) 𝒅𝒙 = 𝟐
Therefore, the predicted change in 𝒚 is computed by the differential given by:
4
 𝒅𝒚
∆𝒚 = ∆𝒙
𝒅𝒙
∆𝒚 = 𝟐(𝟏) = 𝟐
The actual values are 𝒚 = 𝟐𝟎 when 𝒙 = 𝟏𝟎
𝒚 = 𝟐𝟐 when 𝒙 = 𝟏𝟏
The actual change in 𝒚 = 𝟐𝟐 − 𝟐𝟎 = 𝟐 (with accuracy of prediction 100% - because predicted
and actual value are equal)
𝒅𝒚
b) 𝒅𝒙 = 𝟒𝒙
The predicted change in 𝒚 is given by:
𝒅𝒚
∆𝒚 = ∆𝒙
𝒅𝒙
∆𝒚 = 𝟒𝒙(𝟏) = 𝟒𝒙

The actual values are 𝒚 = 𝟐𝟎𝟎 when 𝒙 = 𝟏𝟎

𝒚 = 𝟐𝟒𝟐 when 𝒙 = 𝟏𝟏
The actual change in in 𝒚 = 𝟐𝟒𝟐 − 𝟐𝟎𝟎 = 𝟒𝟐
𝒅𝒚
c) 𝒅𝒙 = 𝟔𝒙𝟐
The predicted change in 𝒚 is given by:
𝒅𝒚
∆𝒚 = ∆𝒙
𝒅𝒙
∆𝒚 = 𝟔𝒙𝟐 (𝟏) = 𝟔𝒙𝟐

The actual values are 𝒚 = 𝟐𝟎𝟎𝟎 when 𝒙 = 𝟏𝟎

𝒚 = 𝟐𝟔𝟔𝟐 when 𝒙 = 𝟏𝟏
The actual change in in 𝒚 = 𝟐𝟔𝟔𝟐 − 𝟐𝟎𝟎𝟎 = 𝟔𝟔𝟐

When 𝒚 is a function of more than one independent variable such as 𝒚 = 𝒇(𝒙, 𝒛) and if there
are infinitesimally small changes in all variables, then the total effect will be:

𝒚 𝒚
∆𝒚 = ∆𝒙 + ∆𝒛
𝒙 𝒛
This is known as total differential, depicting the total effect of changes in all independent
variables on 𝒚.
It is usual to write 𝒅𝒚, 𝒅𝒙, 𝒅𝒛 to represent infinitesimally small changes instead of writing
∆𝒚, ∆𝒙, ∆𝒛. Thus, the total differential of the above form can be written as
𝒚 𝒚
𝒅𝒚 = 𝒅𝒙 + 𝒅𝒛
𝒙 𝒛
Example 4.4: find the total differential of the function
𝒚 = 𝒇(𝒙, 𝒛) = 𝟔𝒙𝟐 + 𝟖𝒛𝟐 − 𝟎. 𝟑𝒙𝒛
Solution: the total differential is
𝒚 𝒚
𝒅𝒚 = 𝒅𝒙 + 𝒅𝒛
𝒙 𝒛

5
 𝒅𝒚 = (𝟏𝟐𝒙 − 𝟎. 𝟑𝒛)𝒅𝒙 + (𝟏𝟔𝒛 − 𝟎. 𝟑𝒙)𝒅𝒛

The concept of total differential in economics

In production theory, the slope of an isoquant represents the marginal rate of technical
substitution (MRTS) between two inputs. The use of the total differential can help demonstrate
that the MRTS will equal the ratio of the marginal products of the two inputs.

The MRTSL,K is usually defined as the amount of Labor that would be added to compensate for
the reduction of one unit of K so that the production level remains unchanged. This is only an
approximate measure and more accuracy can be obtained when the MRTSL,K is defined at a point
on an isoquant.
For infinitesimally small changes in L and K the MRTSL,K measures the rate at which L needs to be
substituted for K to keep output unchanged.
𝐝𝐊
𝑴𝑹𝑻𝑺𝑳,𝑲 = −
𝐝𝐋
For the long-run production function given as 𝑸 = 𝒇(𝑳, 𝑲) total differential is given as:
𝑸 𝑸
𝒅𝑸 = 𝒅𝑳 + 𝒅𝑲
𝑳 𝑲
If we are looking at a movement along the same isoquant then output is unchanged and so 𝒅𝑸 =
𝟎 and thus:
𝑸 𝑸
𝒅𝑸 = 𝒅𝑳 + 𝒅𝑲 = 𝟎
𝑳 𝑲
𝑸 𝑸
⇒ 𝒅𝑳 = − 𝒅𝑲
𝑳 𝑲
𝐐
𝐝𝐊 
⇒ − = − 𝐋
𝐝𝐋 𝐐
𝐊
𝐐 𝐐
Provided that 𝑴𝑷𝑳 = ; and 𝑴𝑷𝑲 = 𝐊 it follows that
𝐋

𝐝𝐊 𝑴𝑷𝑳
− =
𝐝𝐋 𝑴𝑷𝑲
𝑴𝑷𝑳
𝑴𝑹𝑻𝑺𝑳,𝑲 =
𝑴𝑷𝑲
3.3.2. Total Derivatives
In partial differentiation it is assumed that one variable changes while all other independent
variables are held constant. However, in some instances there may be a connection between the
independent variables and so this ceteris paribus assumption will not apply. For example, in a
production function the amount of one input used may affect the amount of another input that

6
can be used with it. From the total differential of a function we can derive a total derivative which
can cope with this additional effect.

Assume that 𝒚 = 𝒇(𝒙, 𝒛) and also 𝒙 = 𝒈(𝒛) the change in 𝒛 [∆𝒛] will affect 𝒚 in the following two
ways:
One. Directly through the function 𝒚 = 𝒇(𝒙, 𝒛)
Two. Indirectly by changing 𝒙 via the function 𝒈(𝒛) which in turn will affect 𝒚 through the
function 𝒚 = 𝒇(𝒙, 𝒛)
The total differential of the function 𝒚 = 𝒇(𝒙, 𝒛) is given as
𝒚 𝒚
𝒅𝒚 = 𝒅𝒙 + 𝒅𝒛
𝒙 𝒛
Dividing both sides of the expression above by 𝒅𝒛 yields the total derivative as given below:
𝒅𝒚 𝒚 𝒅𝒙 𝒚
= +
𝒅𝒛 𝒙 𝒅𝒛 𝒛
𝒚 𝒅𝒙
In the expression above the first term [i.e. 𝒙 𝒅𝒛 ] shows the indirect effect of 𝒛, via its effect on
𝒚
𝒙, and the second term [i.e. ] shows the direct effect.
𝒛

Example 4.4: for a given Cobb-Douglas production function 𝑸 = 𝟐𝟓𝑳𝟎.𝟓 𝑲𝟎.𝟒 and 𝑲 = 𝟎. 𝟖𝑳𝟐
what is the total effect of a change in 𝑳 on 𝑸? Identify the direct and indirect effects.

Solution
The total differential is:
𝑸 𝑸
𝒅𝑸 = 𝒅𝑳 + 𝒅𝑲
𝑳 𝑲
The total derivative with respect to 𝑳 will be
𝒅𝑸  𝑸  𝑸 𝒅𝑲
= +
𝒅𝑳 𝑳  𝑲 𝒅𝑳
Accordingly, from the given function we can derive the following:
 𝑸  𝑸 𝒅𝑲
= 𝟏𝟐. 𝟓𝑳−𝟎.𝟓 𝑲𝟎.𝟒 ; = 𝟏𝟎𝑳𝟎.𝟓 𝑲−𝟎.𝟔 ; and = 𝟏. 𝟔𝑳
𝑳 𝑲 𝒅𝑳
And substituting these values in the total derivative we have
𝒅𝑸
= (𝟏𝟐. 𝟓𝑳−𝟎.𝟓 𝑲𝟎.𝟒 ) + (𝟏𝟎𝑳𝟎.𝟓 𝑲−𝟎.𝟔 )( 𝟏. 𝟔𝑳)
𝒅𝑳
By simplifying
𝒅𝑸
⇒ = (𝟏𝟐. 𝟓𝑳−𝟎.𝟓 𝑲𝟎.𝟒 ) + (𝟏𝟔𝑳𝟏.𝟓 𝑲−𝟎.𝟔 )
𝒅𝑳

The expression (𝟏𝟔𝑳𝟏.𝟓 𝑲−𝟎.𝟔 ) shows the indirect effect of changes in 𝑳 on 𝑸; and the term
(𝟏𝟐. 𝟓𝑳−𝟎.𝟓 𝑲𝟎.𝟒 ) shows the direct effect.

7
Exercise 1: derive the total differentials of the following production functions:
a) 𝑸 = 𝟐𝟎𝑳𝟎.𝟒 𝑲𝟎.𝟔
b) 𝑸 = 𝟔𝑲𝟎.𝟖 + 𝟓𝑳𝟎.𝟕 + 𝟎. 𝟖𝑳𝟐 𝑲𝟐

Exercise 2: if 𝑸 = 𝟒𝟎𝑳𝟎.𝟒 𝑲𝟎.𝟑 and 𝑳 = 𝟓𝑲𝟎.𝟐𝟓 , then find the total effect of change in 𝑲 on 𝑸.
Identify the direct and indirect effects.

3.4. Partial Derivative of Implicit and Inverse Functions (Rules)

3.4.1. Partial Derivatives of Implicit Functions
Let 𝒚 be related to 𝒙 by the equation

𝑭 = 𝒇(𝒙, 𝒚) = 𝟎
Here the function is an implicit function in which 𝒚 is NOT a function of 𝒙 since at a particular
value of 𝒙 there can be more than one value of 𝒚.

Then the partial derivative of 𝒚 with respect to 𝒙 is given as

𝝏𝑭
𝒚  𝒇𝒙
= − 𝝏𝒙 = −
𝒙 𝝏𝑭 𝒇𝒚
𝝏𝒚
Example: given an implicit function 𝑭 = 𝒇(𝒙, 𝒚) = 𝒙𝟐 𝒚𝟐 + 𝒙𝟑 𝒚 = 𝟎 then find the partial
𝒚
derivative of the implicit function given.
𝒙
Solution:
𝝏𝑭
𝒚 
= − 𝝏𝒙
𝒙 𝝏𝑭
𝝏𝒚
𝒚 𝟐𝒙𝒚𝟐 + 𝟑𝒙𝟐 𝒚
= −
𝒙 𝟐𝒙𝟐 𝒚 + 𝒙𝟑

3.4.2. Partial Derivatives of Inverse Functions

Recalling that a derivative is a ratio of differentials, we can then rearrange the terms to get the
implicit function rule:

𝒚 𝒇𝒙
= −
𝒙 𝒇𝒚

𝒚
Notice that the partial derivative 𝒙 is the negative of the reciprocal of the corresponding partials
𝒚 𝒇𝒙 𝟏
=− =−
𝒙 𝒇𝒚 𝒇𝒚
𝒇𝒙

8
Assuming the inverse function exists1, the inverse function rule states that the derivative of the
inverse function is the reciprocal of the derivative of the original function.
𝐏
For example, if 𝑸 = 𝒇(𝑷, 𝑰) is the demand function, the partial derivative of the function is  𝐐
the partial derivative of the inverse function:
𝐏 𝟏
=
𝐐 𝐐
𝐏
3.5. Optimization of Multivariable Functions
3.5.1. Unconstrained Optimizations
For a multivariable function such as 𝒛 = 𝒇(𝒙, 𝒚) to be to be at a maximum or at a minimum, the
first-order conditions which must be met are:

1st. The first-order partial derivatives must equal zero simultaneously.

𝒛 𝒛
i.e. = 𝟎; and =𝟎
𝒙 𝒚

2nd. The second-order direct partial derivatives, when evaluated at the critical point (a, b),
must both be negative for a relative maximum and positive for a relative minimum.
𝟐 𝒛 𝟐 𝒛
< 𝟎; and < 𝟎  for maximum
 𝒙𝟐  𝒚𝟐
𝟐 𝒛 𝟐 𝒛
> 𝟎; and > 𝟎  for minimum
 𝒙𝟐  𝒚𝟐

These are similar to the second-order conditions for the optimization of a single variable function.
The difference here is that these conditions must hold with respect to changes in both
independent variables.

The other second-order condition for multivariable function is that

𝟐
𝟐 𝒛 𝟐 𝒛 𝟐 𝒛
( )( )>( )
 𝒙𝟐  𝒚𝟐 𝒙𝒚
This must hold at both maximum and minimum stationary points.

Keeping the above conditions for optimizations of multivariable functions in mind consider the
example below:

Example 4.5: Assume a firm produces two products Q1 and Q2 which are sold in two separate
markets with the inverse demand functions:

1
An inverse function exists if each value of y yields one and only one value of x.

9
 𝑷𝟏 = 𝟔𝟎𝟎 − 𝟎. 𝟑𝑸𝟏 and 𝑷𝟐 = 𝟓𝟎𝟎 − 𝟎. 𝟐𝑸𝟐
And the firm’s total cost is given as:
𝑻𝑪 = 𝟏𝟔 + 𝟏. 𝟐𝑸𝟏 + 𝟏. 𝟓𝑸𝟐 + 𝟎. 𝟐𝑸𝟏 𝑸𝟐

Required: If the firm wishes to maximize total profits, how much of each product should it
produce? What will the maximum profit level be?

Solution: To compute the profit function, first derive the revenue functions and then find the
difference between TR and TC.

TR = TR1 + TR2 (where TR1=P1•Q1 and TR2 = P2•Q2)

𝑻𝑹𝟏 = (𝟔𝟎𝟎 − 𝟎. 𝟑𝑸𝟏 )(𝑸𝟏 ) = 𝟔𝟎𝟎𝑸𝟏 − 𝟎. 𝟑𝑸𝟏 𝟐
𝑻𝑹𝟐 = (𝟓𝟎𝟎 − 𝟎. 𝟐𝑸𝟐 )(𝑸𝟐 ) = 𝟓𝟎𝟎𝑸𝟐 − 𝟎. 𝟐𝑸𝟐 𝟐
𝑻𝑹 = 𝟔𝟎𝟎𝑸𝟏 − 𝟎. 𝟑𝑸𝟏 𝟐 + 𝟓𝟎𝟎𝑸𝟐 − 𝟎. 𝟐𝑸𝟐 𝟐
Profit is given as, 𝜫 = 𝑻𝑹 − 𝑻𝑪
𝜫 = (𝟔𝟎𝟎𝑸𝟏 − 𝟎. 𝟑𝑸𝟏 𝟐 + 𝟓𝟎𝟎𝑸𝟐 − 𝟎. 𝟐𝑸𝟐 𝟐 ) − (𝟏𝟔 + 𝟏. 𝟐𝑸𝟏 + 𝟏. 𝟓𝑸𝟐 + 𝟎. 𝟐𝑸𝟏 𝑸𝟐 )
𝜫 = −𝟏𝟔 + 𝟓𝟗𝟖. 𝟖𝑸𝟏 − 𝟎. 𝟑𝑸𝟏 𝟐 + 𝟒𝟗𝟖. 𝟓𝑸𝟐 − 𝟎. 𝟐𝑸𝟐 𝟐 − 𝟎. 𝟐𝑸𝟏 𝑸𝟐

The first –order conditions (F.O.C) for the maximization of this unconstrained profit function are:
𝜫
=𝟎
𝑸𝟏
⇒ 𝟓𝟗𝟖. 𝟖 − 𝟎. 𝟔𝑸𝟏 − 𝟎. 𝟐𝑸𝟐 = 𝟎 ••• (∗)
and
𝜫
=𝟎
𝑸𝟐
⇒ 𝟒𝟗𝟖. 𝟓 − 𝟎. 𝟒𝑸𝟐 − 𝟎. 𝟐𝑸𝟏 = 𝟎 ••• (∗∗)
Solving equations (*) and (**) above simultaneously help to compute the optimal values of 𝑸𝟏
and 𝑸𝟐 .

Rearranging the two equations yields

𝟎. 𝟔𝑸𝟏 + 𝟎. 𝟐𝑸𝟐 = 𝟓𝟗𝟖. 𝟖
𝟎. 𝟐𝑸𝟏 + 𝟎. 𝟒𝑸𝟐 = 𝟒𝟗𝟖. 𝟓
Calculation yields 𝑸𝟏 = 𝟔𝟗𝟗. 𝟏𝟎 units and 𝑸𝟐 = 𝟖𝟗𝟔. 𝟕 units

And the maximized profit at these output levels will be

𝜫 = −𝟏𝟔 + 𝟓𝟗𝟖. 𝟖(𝟔𝟗𝟗. 𝟏𝟎 ) − 𝟎. 𝟑(𝟔𝟗𝟗. 𝟏𝟎 )𝟐 + 𝟒𝟗𝟖. 𝟓(𝟖𝟗𝟔. 𝟕) − 𝟎. 𝟐(𝟖𝟗𝟔. 𝟕)𝟐
− 𝟎. 𝟐(𝟔𝟗𝟗. 𝟏𝟎 )(𝟖𝟗𝟔. 𝟕)
𝜫 = 𝟒𝟑𝟐, 𝟕𝟗𝟕. 𝟎𝟐
When checking for the second order condition (S.O.C) is met or not

10
𝟐 𝜫
= −𝟎. 𝟔 < 𝟎 S.O.C for maximization met
𝑸𝟏 𝟐

𝟐 𝜫
= −𝟎. 𝟒 < 𝟎 S.O.C for maximization met
𝑸𝟐 𝟐

Looking at the other S.O.C of multivariable functions’ optimization, that is

𝟐
𝟐 𝜫 𝟐 𝜫 𝟐 𝜫
( )( )>( )
𝑸𝟏 𝟐 𝑸𝟐 𝟐 𝑸𝟏 𝑸𝟐
𝟐
𝟐 𝜫
( ) = −𝟎. 𝟐
𝑸𝟏 𝑸𝟐
𝟐
𝟐 𝜫 𝟐 𝜫 𝟐 𝜫
( )( )>( ) ⇒ (−𝟎. 𝟔)(−𝟎. 𝟒) > (−𝟎. 𝟐)𝟐 , true
𝑸𝟏 𝟐 𝑸𝟐 𝟐 𝑸𝟏 𝑸𝟐

Exercise: a firm produces two products which are sold in separate markets with the respective
inverse demand functions given as:
𝑷𝟏 = 𝟐𝟏𝟎 − 𝟎. 𝟒𝑸𝟏 𝟐 and 𝑷𝟐 = 𝟒𝟗𝟏 − 𝟔𝑸𝟐
And the firm’s total cost is given as:
𝑻𝑪 = 𝟑𝟐 + 𝟎. 𝟖𝑸𝟏 𝟐 + 𝟎. 𝟕𝑸𝟐 𝟐 + 𝟎. 𝟏𝑸𝟏 𝑸𝟐
Required: how much should the firm sell in each market in order to maximize total profits? Check
all the conditions.

Exercise: a monopolist produces two products with the respective inverse demand functions
given as:
𝑷𝟏 = 𝟓𝟎 − 𝑸𝟏 + 𝑸𝟐 and 𝑷𝟐 = 𝟑𝟎 − 𝟐𝑸𝟏 − 𝑸𝟐
And the firm’s total cost is given as:
𝑻𝑪 = 𝟏𝟎𝑸𝟏 + 𝑸𝟏 𝑸𝟐 + 𝟏𝟎𝑸𝟐
Required: how much units of the goods maximize total profits? Check all the conditions.

3.5.2. Constrained Optimization-the Method of Lagrange Composite Function

Reading Assignment

3.6. Economic Application of Partial Derivatives

In a multivariable functions in economics partial derivatives have the following applications:

A) Marginal Productivity of Factors:

11
The marginal product of a factor is defined as the change in output brought about by a small
change in the factors under consideration when all the other factors of production are held
constant.

Example: for a given Cobb-Douglas production function of two factors – L and K, MPL and MPK
can be found by the help of partial derivatives.
If 𝑸 = 𝟑𝟔𝑳𝑲 − 𝟑𝑳𝟐 − 𝟐𝑲𝟐 then

𝑸 𝑸
𝑴𝑷𝑳 = = 𝟑𝟔𝑲 − 𝟔𝑳; and 𝑴𝑷𝑲 = = 𝟑𝟔𝑳 − 𝟒𝑲
𝑳 𝑲

B) Determination of Multipliers in Macroeconomic Models

The partial derivative can also be used to derive the various multipliers of an income
determination model.

For instance, given the Macroeconomic equation of National Income as 𝒀 = 𝑪 + 𝑰 + 𝑮 + (𝑿 −

𝑴) where
𝑪 = 𝑪𝟎 + 𝒃𝒀 𝑮 = 𝑮𝟎
𝑴 = 𝑴𝟎
𝑰 = 𝑰𝟎 + 𝒂𝒀 𝑿 = 𝑿𝟎

After simple substitution in the equation we have a simplified form as

𝟏
𝒀̅ = (𝑪 + 𝑰𝟎 + 𝑮𝟎 + (𝑿𝟎 − 𝑴𝟎 ))
𝟏−𝒃−𝒂 𝟎
Using partial derivative now it is easy to find the multiplier of any of the above parameters (a, b,
C, I, G, X and M). For example, finding the government multiplier using partial derivative can be
given as:
𝒀̅ 𝟏
=
𝑮𝟎 𝟏 − 𝒃 − 𝒂
The import multiplier will be
𝒀̅ 𝟏
=−
𝑴𝟎 𝟏−𝒃−𝒂
And the multiplier for a change in the marginal propensity to invest (a) is given by
𝒀̅ 𝑪𝟎 + 𝑰𝟎 + 𝑮𝟎 + (𝑿𝟎 − 𝑴𝟎 ) 𝒀̅
= =
𝒂 (𝟏 − 𝒃 − 𝒂)
𝟐
𝟏−𝒃−𝒂
Other multipliers can be found in similar ways
C) Computing Elasticity
For a multivariate demand function of good 𝒙 (related to its own price (𝑷𝒙 ); price of substitute
good (𝑷𝒚 ) and consumer income (𝒀) given as 𝑸𝒙 = 𝒂 − 𝒃𝑷𝒙 + 𝒄𝑷𝒚 + 𝒎𝒀

12
  The point price elasticity of demand (𝜺𝑷 ) can be calculated by the use of partial derivative
as:
𝑸𝒙 𝑷𝒙
𝜺𝑷 = •
𝑷𝒙 𝑸𝒙
 The point cross-price elasticity of demand (𝜺𝒙,𝒚) can be calculated by the use of partial
derivative as:
𝑸𝒙 𝑷𝒚
𝜺𝒙,𝒚 = •
𝑷𝒚 𝑸𝒙
 The point income elasticity of demand (𝜺𝑰 ) can be calculated by the use of partial
derivative as:
𝑸𝒙 𝑰
𝜺𝑰 = •
𝑰 𝑸𝒙
D) Optimization of Functions
Partial derivatives are also useful for optimization of constrained and unconstrained economic
functions as discussed in the preceding sections.
==============================================================================
==================================

13