Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

Chapter Four: Derivatives of Multivariate Functions

Partial Differentiation shows how to find the derivative of a function with respect to one of
the independent variables while the other variable(s) is/are held constant.
Let's consider a function with two variables Z = f (x, y)
The goal is to differentiate the function with respect to one of the variables in question while
keeping the other variable constant.
z
Therefore, the partial derivative of f with respect to x, denoted by or fx, or fx (x, y), is
x
z f ( x  x, h)  f ( x, y )
defined by  lim , provided the limit exists.
x x 0 x
z
Similarly, the partial derivative of f with respect to y, denoted by or fy or fy (x,y), is
y
z f ( x, y  y )  f ( x, y )
defined by  lim , which is the function obtained by differentiating
y y  0 y
f with respect to y, treating x as a constant. The process of taking partial derivative is called
partial differentiation.
The above definitions can easily be extended to functions of three or more variables.
Example1: Find the partial derivatives fx and fy if Z = f(x,y) = x2 + 2y2 + 3xy
Solution:
To find fx differentiate the function with respect to x, treating y as a constant.
z
fx   2x  3y
x
To find fy differentiate the function with respect to y, treating x as a constant.
z
fy   4 y  3x
y
Example 2 Find the partial derivatives fx and fy and fx (2,3) and fy (2,3) if
Z = f(x,y) = 3x2 – 2xy2 +1
Solution:
z z
fx   6 x  2 y 2 (Treating y as a constant), f y   4 xy (Treating x as a constant)
x y
fx (2,3) = 6(2) – 2(3)2 = 12 – 18 = -6
fy (2,3) = -4(2) (3) = -24
Note that the partial differentiation differs from the previous discussion of differentiation is
that we must hold (n – 1) independent variables constant while allowing one variable to vary.
Otherwise no new rules are needed for the computation of partial derivatives.
3x  2 x2
Example 3: Given Y  21 , find the partial derivatives
x1  3x2
Solution:
The partial derivatives can be found by applying the quotient rule.

1
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

By treating x2 as a constant
y 3( x1  3 x 2 )  2 x1 (3 x1  2 x 2 ) 3 x1  9 x 2  6 x1  4 x1 x 2
2 2 2

f1   
x1 ( x1  3 x 2 ) 2
2
( x1  3 x 2 ) 2
2

 3x1  4 x1 x 2  9 x 2
2


( x1  3 x 2 ) 2
2

By treating x1 as a constant
y  2( x1  3 x 2 )  3(3 x1  2 x 2 )  2 x1  6 x 2  9 x1  6 x 2
2 2

f2   
x 2 ( x1  3 x 2 ) 2
2
( x1  3 x 2 ) 2
2

 x1 (2 x1  9)

( x1  3 x 2 ) 2
2

Therefore, students can use the above techniques of partial differentiation in order to find the
partial derivatives of any form of functions.
Total Differentials
In the above discussion you learned how to use the differential of a function to approximate
the change in the function resulting from a small change in its independent variable. Now we
extend this concept of differential to a function of two or more independent variables.
Let Z be a function of two variables x and y and the function is smooth and differentiable
everywhere.
Z = f (x, y)
If ∆x and ∆y denote a small change in x and y respectively then the change in Z is given as
z z
Z  dx  dy
x y
z
The partial derivative measures the rate of change of Z with respect to an infinitesimal
x
z
change in x. Similarly measures the rate of change of Z with respect to an infinitesimal
y
change in y. Hence the change in Z due to a change in x may be represented by the expression
z z
dx and dy for that of a change in y.
x y
Therefore the total change in Z will be equal to the sum of these two partial changes.
z z
dz  dx  dy
x y
The process of finding such a total differential is called total differentiation.
The more general case of a function of n independent variables can be exemplified by
Z = f (x1, x2, x3, ---, xn).
The total differential of this function can be written as

2
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

f f f
dz  dx1  dx2       dxn
x1 x 2 x n
or dz  f1dx1  f 2 dx2      f n dxn
This shows dz or the total change in z is the sum of the n partial changes resulting from all
possible sources of change.
Examples 1: Find the total differential of the function Z = 3x2 + xy – 2y3
Solution: Applying the approximation formula
z z
dz  dx  dy we get
x y
dz  (6 x  y )dx  ( x  6 y 2 ) dy
Example 2: Find the total differential of the function
x1
y
x1  x2
Solution: Applying the approximation formula
y y
dy  dx1  dx2 , we obtain
x1 x 2
1( x  x 2 )  x1 (1)   0( x1  x 2 )  x1 (1) 
dy   1  dx1    dx2
 ( x1  x 2 )  ( x1  x 2 )
2 2
 
 x2    x1 
dy    dx   2 
dx2
 
2 1
 1
( x x 2 )   1
( x x 2 ) 
 x2   x1 
dy   2 
dx1   2 
dx2
 ( x1  x 2 )   ( x1  x 2 ) 
Let's consider the following function
y = f (x, r) and x and r are related, i.e. x = h( r).
The question is how we can find the rate of change of y with respect to r. The concept of total
derivative will enable us to answer this question.
In the above function the variable r can affect y in two ways:
1. Directly via the function f
 f 
You can use the partial derivative   f r  in order to find the direct effect of r on y.
 r 
2. Indirectly, via the function h and then f.
To obtain the total derivative, we first differentiate y totally, to get the total differential.
dy = fx dx + frdr
Dividing both sides by dr, we obtain

3
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

dy dx dr
 fx  fr
dr dr dr
dy dx dr
 fx  fr ( Since  1)
dr dr dr
dx
where, f x is the indirect effect of r on y and f r is the direct effect of r on y.
dr
dy
Therefore is the total derivative.
dr
The process of finding the total derivative dy/dr is called total differentiation of y with respect
t o r.
Example 1: Find the total derivative dy/dr given the function
y = f (x, r) = 5x2 – 3r + 1, where x = h(r) = 2r
dy dx
Solution: The total derivative  fx  fr
dr dr
dy/dr = 10x (2) + (-3) = 20x – 3. Substituting x = 2r, we get
dy/dr = 20 (2r) – 3 = 40r – 3.
Example 2: Find the total derivative, dz/dt, for the function
Z = f (x, t) = (x + t) (x - 2t), where x = h (t) = 2 – 7t
Solution:
dz dx f dx f
 fx  ft  . 
dt dt x dt t
f f
 1( x  2t )  ( x  t )(1) ,
dx
 7 ,  1( x  2t )  ( x  t )(2)
x dt t
Therefore,
 1( x  2t )  ( x  t )(1)(7)  1( x  2t )  ( x  t )(2)
dz
dt
dz
 ( x  2t  x  t )(7)  x  2t  2 x  2t
dt
dz
 (2 x  t )(7)  x  4t  14x  7t  x  4t
dt
dz
 15x  3t
dt
By Substituting x  2  7t , we get
dz
 15(2  7t )  3t  108t  30
dt
Note that in the above examples you can check the results obtained by substituting the
function x = h(r) and h (t) in the function y and z respectively.
 Second Order Partial Derivatives

Let f be a function of the two variables x and y.

4
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

f f
The first partial derivatives of f denoted by f x or and f y or . The second order partial
x y
deviates are fxx, fyy, fxy and fyx.
  Z   2 z   z   2 z
f xx    . f yy     2
x  X  x 2 y  y  y
  Z   2 Z   Z   2 Z
f xy    , f yx   
y  x  yx x  y  xy
Students can use any one of the notations which they like is convenient.
fxx measures the rate of change of fx with respect to x, while y remains fixed.
fyy denotes the rate of change of fy with respect to y, while x is heed constant.
The other two second order partial derivatives (fxy and fyx) are called the cross or mixed partial
derivatives and each measures the rate of change of one first-order partial derivative with
respect to the other variable.
Now let us see the following examples on how to find the four second order partial
derivatives
Example 1: Find the four second order partial derivatives of the function
Z = x3 – 2xy + 5y
Solution:
In order to find the second order partial derivatives, we have to find first the
first partial derivatives
Z
fx   3 x 2  2 y and
x
Z
fy   2 x  5
y
The partial derivatives of fx and fy with respect to x and y are
  Z  
f xx     (3x 2  2 y )  6 x
x  x  x
  z  
f yy     (2 x  5)  0
y  y  y
The other two mixed partial derivatives are
  Z  
f xy     (3x 2  2 y )  2
y  x  y
  Z  
f yx      2 x  5  2
x  y  x
From the above result, the mixed partial derivatives fxy and fyx are the same.
Young's theorem: As long as the two mixed partial derivatives are both continuous
fxy = fyx.
Example 2: Z = x – 2x2y2 – 4y – 2x + 15
3

5
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

Solution:
z
fx   3 x 2  4 xy2  2
x
z
fy   4 x 2 y  4
y
Therefore, the second order partial derivatives are:
  z  
f xx    
x  x  x
 
3 x 2  4 xy 2  2  6 x  4 y 3

  z  
f yy    
y  y  y
 
 4 x 2 y  4  4 x 2

  z  
f xy     (3x  4 xy  2)  8 xy
2 2

y  x  y
  z  
fyx     (4 x 2 y  4)  8 xy
x  y  x
As expected the mixed partial derivatives (fxy and fyx) are equal
 First Order Necessary Condition for an Extremum of a Function with Two
Variables
In this section we will see the necessary conditions for finding the relative extremum of a
function Z = f (x, y).
The necessary condition is to find the values of x and y which make the first partial
derivatives of the function Z = f (x, y) zero. That is, fx = fy = 0. These values play an
important role in finding the relative maxima, minima or saddle points of the function Z = f (x,
y) as we did for functions of one variable in chapter five. To find the critical points of the
function, set fx = fy = 0 and solve these equations simultaneously for x and y. After this we use
the second order test inorder to know that the solutions for the above equations corresponds to
the relative maximum, relative minimum or saddle points.
 Second Order Condition
The sufficient condition for finding relative extremum of the function f (x, y) is presented as
follows.
Let (a, b) is a critical point of the function f (x, y). Then
 f (a , b) is a relative maximum if and only if fxx < 0, fyy < 0 and fxx fyy > f2xy
 f (a, b) is a relative minimum if and only if fxx > 0, fyy > 0 and fxx fyy > f2xy.
▪ f (a , b) is a saddle point if fxx fyy < f2xy and fxx, fyy have different signs.
▪ f (a , b) is an inflection point if fxx fyy < f2xy and fxx, fyy have the same signs.
The test fails to give information about the behavior of the function if f2xy = fxx fyy.
Note that the second order partial derivatives (fxx, fyy and fxy) should be evaluated at the critical
points (a, b). The necessary and sufficient conditions for relative extremum are summarized
below.

6
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

Conditions Minimum Maximum

- First-Order fx = fy = 0 fx = fy = 0
necessary condition
- Second-Order fxx > 0 , fyy > 0 fxx < 0, fyy < 0
sufficient condition fxx fyy > f2xy and
fxx fyy > f2xy
Example 3: Find the extreme value(s) of the function Z = x2 + y2 +2
Solution:
Step 1: Find the critical points from the first partial derivatives
z
fx   2x
x
z
fy   2y
y
fx = fy = 0 is the necessary condition
2x = 2y = 0
x = 0 and y = 0
The only critical point is (0, 0)
Step 2: Check the Second-Order sufficient conditions
  z  
f xx     (2 x)  2
x  x  x
  z  
f yy    (2 y )  2
y  y  y
  z  
f xy     (2 x)  0
y  x  y
Evaluating the second order partial derivatives at the critical values of the function, we get:
fxx (0, 0) = 2 fyy (0, 0) = 2 fxy (0, 0) = 0
Therefore, since fxx > 0, fyy > 0 and fxx fyy > f2xy, the critical values (0 , 0) belongs to the
relative minimum value of Z.
Hence Z = 02 + 02 + 2 = 2 is the relative minimum.
Example 4: Find the extreme value of the function Z = 2x2 – 3y2.
Solution:
Step 1. The first order partial derivatives are
z z
fx   4x fy   6 y
x y
The necessary condition
fx = fy = 0
4x = 0 -6y = 0
x=0 y=0
The critical point is (0, 0)

7
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

Step 2. The second-order sufficient condition

  z  
f xx     (4 x)  4
x  x  x
  z  
f yy     (6 y )  6
y  y  y
  z  
f xy     (4 x)  0
y  x  y
f xx f yy  24 f 2
xy 0
Therefore, since fxxfyy < f2xy the point (0, 0) belongs to the saddle point.
Example 5: Z = -x2 – y2 + xy + 2x + y
Solution: Step 1. Find fx and fy
z
fx   2 x  y  2  0
x
z
fy   x  2y 1  0
y
 2 x  y  2      (1)
x  2 y  1        ( 2 )
Solving equations (1) and (2) simultaneously we get x =5/3 and y = 4/3.
That is (5/3, 4/3) is the critical point
Step 2: Check the second order partial derivatives
   z    z 
(2 x  y  2)  f xx     2 f yy     2
x x  x  y  y 
  z  
f xy     (2 x  y  2)  1
y  x  y
Evaluating the second order partial derivatives at the critical points we get
fxx < 0, fyy < 0, fxxfyy = 4 > 0
fxy = 1, f2xy = 1
Therefore, since fxx < 0, fyy < 0 and fxxfyy > f2xy the function Z has a relative maximum at
(5/3 , 4/3). At the maximum point Z = 7/3.
Example 6: Z = x4 – 8x2 + 6y2 – 12y + 3
Solution:
Step 1. Find fx and fy
z
fx   4 x 3  16 x  0
x
z
fy   12 y  12  0
y
Solving the above equations, we get x = 0 or 2 or -2 and y = 1. That is (0, 1), (2 , 1) , (-2 ,
1) are the critical points.

8
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

Step 2. The second order partial derivatives are

  z  
f xx     (4 x 3  16x)  12 x 2  16
x  x  x
  z  
f yy     (12 y  12)  12
y  y  y
  z  
f xy     (4 x  16 x)  0
3

y  x  y
At x = 0, fxx = -16 < 0, fyy = 12 > 0
At x = 2, fxx = 32 > 0, fyy = 12 > 0
At x = -2, fxx = 32 > 0, fyy = 12 > 0
fxy = 0 , f2xy = 0
Therefore, from the above result we have the following conclusions:
- At x = 0, fxx < 0, fyy > 0 and fxx fyy < f2xy the point (0 , 1) belongs to the saddle point
and Z = -3
- At x = 2 , fxx > 0 , fyy > 0 and fxxfyy > f2xy , the point (2 ,1) is the relative minimum
and Z = -19
- At x = -2 , fxx > 0 , fyy > 0 and fxxfyy > f2xy , the point (-2 , 1) is also another relative
minimum and Z = -19

 Economic Applications
Here let's see the application of the above discussions in solving economic problems with the
help of the following examples.
Example on problem of a multiproduct firm.
Let's consider the following problems of a firm.
A firm operating under pure competition produces two products. Since price of the products
are exogenous to the firm, the level of the two products should be determined to get the
maximum profit.
Let Q1 and Q2 represent the output level of the first and second products per unit of time.
Accordingly, the firm's revenue function is as follows:
R = P1Q1 + P2Q2
Let's assume that P1 = 12 and P2 = 18
R = 12Q1 + 18Q2
The firm's cost function is given by
C = 2Q2 + Q1Q2 + 2Q22
The profit function is therefore
Pr ofit    R  C
  12Q1  18Q2  (2Q1 2  Q1Q2  2Q2 2 )
  12Q1  18Q2  2Q1 2  Q1Q2  2Q2 2

9
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

The firm's objective is to choose the level of Q1 and Q2 that maximizes its profit. Hence as
usual we need to compute the first order partial derivatives first.

1   12  4Q1  Q2
Q1

2   18  Q1  4Q2
Q2
These are the necessary condition for finding the maximum profit.
Solving the equations 4Q1+Q2 =12 and Q1+4Q2 =18 simultaneously, we get Q1 = 2 and Q2 = 4
Substituting the values of Q1 and Q2 (i.e. 2 & 4) in to the profit functions, the optimal profit
becomes 48 per unit of time.
Use the second order partial derivatives to check the result obtained above.
       
 11     4  22     4
Q  Q1  Q2  Q2 
   
 12     1
Q2 
 1
Q
 11 22  4 ,  12 2  1
Therefore, since 11  0,  22  0 and 11 22  16  12  1, the maximum profit is 48.
2

Example 2:A monopolist sells two products x and y for which the demand functions are
x = 25-0.5Px
y = 30-Py
And the combined cost function is x2+2xy+y2+20, and then find
a) the profit maximizing level of output for each product
b) the profit maximizing price for each product
c) the maximum profit
Solution:
a) To find the profit maximizing level of output, first construct the profit function.
Pr ofit    TRx  TR y  C
But TRx = Pxx and TRy = Pyy and hence Pr ofit    Px x  Py y  C .
From the demand functions we have
Px = 50-2x and Py = 30-y
Substituting these values in to the profit function, we obtain
  (50  2 x) x  (30  y) y  ( x 2  2 xy  y 2  20)
  50x  3x 2  30 y  2 y 2  2 xy  20
The first order necessary conditions for maximizing profit are
 
  x  50  6 x  2 y  0 and   y  30  4 y  2 x  0
x y
Solving simultaneously we get x =7 and y = 4

10
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

Now check the second order conditions

 xx  6, yy  4and xy  2
Therefore since xx <0, yy 0 and  xxyxy > 2xy profit is maximized at x =7 and y = 4
b) To get the profit maximizing prices substitute x=7 and y=4 in to the demand functions
Px=50-2x and Py=30-y.
Px = 50-2(7) =36 and Py =30-4 =26
c) The maximum profit is therefore obtained by substituting the values of x and y in to the
profit function and   215
Example1: Suppose the total cost function of a firm is given by
c  3q12  7q1  1.5q1q 2  6q 2  2q 22 .
Then find the marginal costs of the firm’s different products.
Solution:
The marginal cost from the first product is calculated by taking the partial derivative of c with
respect to q1 .
c
Mc1 =  6q1  1.5q 2  7
q1
The marginal cost from the second product is calculated by taking the partial derivative of c
with respect to q 2 .
c
Mc2 =  1.5q1  4q 2  4
q 2
Example 2: The demand function for a good is given by Q=700-2p+0.02y, where P is the
price and Y is income. If price =25 and Y = 5000, compute
a) the price elasticity of demand and
b) the income elasticity of demand
Solution:
a) The price elasticity of demand is calculated by using the following formula
Q P
Ed p 
P Q
Q
 2 and substituting the value of P and Y in to the demand equation
P
we get Q = 700-2(25)+0.02(5000) = 750
Q P 25
Therefore, Ed p  = -2( ) = -0.067
P Q 750
b) Similarly, the income elasticity of demand can be found by using the formula
Q Y
Ed y  . This measures the percentage change in the demand for a good resulting from a
Y Q
certain percentage change in income.

11
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

Q
 0.02
Y
Q Y 5000
Thus, Ed y  = 0.02( ) = 0.133
Y Q 750
Example 3: The total cost function of a firm producing two commodities x and y is given by
C(x,y)= x 2  0.5 xy  y 2 . If initially x = 100, y = 60, what is the additional cost incurred by the
firm if the output of x is increased by 3 units.
Solution: Here we use the concept of differentials.
For a very small change in output the differential is given by
c
dc  dx  (2 x  0.5 y )dx
x
The change in the total cost for a large increment is given by
c
c  x
x
Therefore, c  (2 x  0.5 y )(3)
c  2(100)  0.5(60)(3)  510
 Hessian determinant
f xx f xy
H 
f yx f yy
H 1  f xx And H 2  f xx f yy  f yx f xy are the first and second principal minors respectively.
By now we can express the sign defiantness of d 2 Z using the sign of these principal minors
and thereby we are able to identify the second order sufficient conditions for the extremum of
the function Z  f ( x, y ) as
d 2 Z is positive definite if H 1  f xx  0 , and H 2  f xx f yy  f yx f xy  0 . In this case, the
function achieves its minimum.
d 2 Z is negative definite if H 1  f xx  0 and H 2  f xx f yy  f yx f xy  0 . This indicates
that the function is at its maximum.

Considering the function of n- choice variables

Z  f ( x1 , x 2 , x3 ,.... x n )
The first order condition for the extremum of the function is f1  f 2  f 3  .........  0 which
leads to the fact that dZ  f1 dx1  f 2 dx2  f 3 dx3  ......  f n dxn  0
The second order sufficient conditions are identified using the Hessian determinant
f11 f12. ........ f1n
H  . f 21 f 22 .......... f 2 n
f n1 f n 2 .......... f nn

12
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

The sufficient condition for the maximum of the function is satisfied when
H1  0; H 2  0; H 3  0,......( 1) n H n  0 .where as for the minimum of the function all of
the Hessian principal minors must be positive.
Example
1.Given the function Z= 160x – 3x2 - 2xy - 2 y2 + 120 y - 18, find the maximum value of the
function.
Solution
The first order conditions that should be satisfied for maximum are
Z Z
Zx   0 And Zy  0
x y
z
Zx = = 160 - 6x - 2y = 0
x
6x + 2y = 160

Z
Zy  = -2x - 4y + 120 = 0
y
2x + 4y = 120

6x +2y = 160
2 x  4 y  120
12x +4y = 320
2 x  4 y  120

10 x + 0 =200
10 x = 200
x= 20
Substituting 20 in place of x in either of the two equation we will get y = 20
Taking the second partial derivatives,
Z xx = - 6 < 0, Z yy = - 4 < 0
And
( Z xx ) ( Z yy ) > ( Z xy ) 2
(-6) (- 4) > (-2)2
24 > 4
6 2
Alternatively, H  , H 1  6  0 and H 2  20  0
2 4
Thus, the function Z is maximized when x = 20 and y = 20
The maximum value of Z is
Z= 160(20) - 3(20)2 - 2(20) (20) - 2(20)2 + 120 (20) - 18

13
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

Z = 2,782
Economic Applications
Example
Suppose the monopolist sells a certain product in three separate markets and the demand
functions facing the firm are
P1  63  Q1
P2  105  5Q2 And the cost function is C  20  15Q; Q  Q1  Q2  Q3
P3  75  6Q3
Determine the amount of output that should be sold to maximize profit .Identify the prices
charged in each market to maximize profit.
First we should construct the total revenues for each market. These are
TR1  63Q1  Q12
TR 2  105Q2  5Q22
TR3  75Q3  6Q32
Therefore, the total profit function is   TR 2  TR 2  TR3  C
  63Q1  Q12  15Q2  5Q22  75Q3  6Q32  (20  15Q1  15Q2  15Q3 )
  48Q1  4Q12  90Q2  5Q22  60Q3  6Q32  20
First order conditions

1   48  8Q1  0  Q1  6
Q1
 2  90  10Q2  0  Q2  9
 3  60  12Q3  0  Q3  5
P1  57, P2  60, P3  45
We can check the second order condition using the sign of the principal minors of the Hessian
determinant. The corresponding Hessian determinate is
 11  8,  12  0,  13  0
 21  0,  22  10,  23  0
 31  0,  32  0,  33  12

 11  12  13  8 0 0
H   21  22  23  0  10 0
 31  32  33 0 0  12
Evaluating the successive principal minors,
8 0
H 1  8  0, H 2   80  0, H 3  H  960  0
0  10

14
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

It indicates that d 2  is negative definite. Thus, the monopolist can maximize its profit when
it sells 20 units of output out of which 6 units at a price of 57 birr per unit in market 1, 9 units
at a price of 60 birr per unit in market 2 and 5 units at a price of 45 birr per unit in market 3.

Constrained Optimization
Here we will learn how to find the stationary values of a function with constraints. To
illustrate this let us consider the following utility function of a consumer.
Example 1: Find the stationary value of the utility function
U = x1x2 + 2x1
If we were to find the free extremum, with out constraint, then the consumer can purchase any
amount of x1 and x2. In reality the consumer has a fixed budget to be spend on the two goods
x1 and x2. And the consumer’s objective is to maximize his satisfaction subject to such a
limited income. The total amount of money to be spent on these two goods is birr 60. Assume
that the market price of x1 and x2 per unit are birr 4 and 2 respectively. Then the budget
constraint can be expressed by the following linear equation.
4x1 + 2x2 = 60
Therefore, the problem of the consumer is to maximize U subject to the above constraint.
i.e. Max U = x1x2 + 2x1
Subject to: 4x1 + 2 x2 = 60
The technique of substitution and elimination to solve such a problem involves reducing the
function to one variable. We solve the constraint equation for one of the variables and then
substitute the resulting expression in to the objective function (the function to be optimized).
Here is a procedure of the method.
From the constraint equation, x2 = 30-2x1.
Substituting this in to the function to be optimized, we get
U = x1 (30-2x1) + 2x1 = 32x1-2x12
Now we have got a utility function of one variable only. The next step is to find the critical
values by the method learned before.
Taking the first derivatives and setting the resulting expression equal to zero, we get
dU
 32  4 x1  0  x1  8
dx1
Since x2  30  2 x1 the value of x1 is 14
d 2U
The second order derivative is equal to –4< 0.
dx 2
Hence x1 = 8 corresponds to the maximum of U. The stationary value of the function is then
obtained by substituting the values of x1 and x2 in to the objective function.
U = 8(14) + 2(8)=128.
Example 2: Find the extreme value of the function
Z = xy subject to the constraint x + y =1
Solution: using the procedure learned in example 1 above solves the problem.

15
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

From the constraint equation x  y  1, y  1  x. Substituting this value of y in to the objective

function z  x1  x   x  x 2 . Next find the critical values.
dz 1 y  1  1  1 Since y  1  x 
 1  2 x Setting this to zero we get 1  2 x  0  x  2 2
dx 2
The extreme value of the function is then found by substituting the value of x and y in to the
objective function z  xy
_
1 1 1
i.e. z    (Note that this value is the maximum value of the function since
2 2 4
d 2z
 2  0 )
dx 2
The problem of using the substitution and elimination method in solving constrained
optimization arises when we have a complicated constrained function or when there are
several constraints to consider. In such cases solving the constraint equation for one of the
variables is difficult or even impossible. In the next section we will learn another method
called Lagrange multiplier method.

Lagrange - Multiplier Method

Lagrange Multiplier Method, we should involve the following steps.

Step 1 Rewrite the constraint function in its implicit form as

M  xPx  yPy  0
Step 2 Multiply the constraint function by the Lagrange multiplier 
 (M- x P x- y P y) = 0

Step 3 Add the above constraint to the objective function and thereby we get the Lagrange
function that is a modified form of the objective function which includes the constraints as
follows:
L( x, y,  )  Z ( x, y)   (M  xPx  yPy )
Necessary condition, i.e. the first orders condition for maximization is that the first order
partial derivatives of the Lagrange function should be equal to zero.

Differentiating L with respect to x, y, and  and equating it with zero gives us.
L z
  Px  0
x x
L z
  Py  0
y y

16
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

L
 M  xPx  yPy  0

Z Zy
 = x and  =
Px Py
Zx Zy Z x Px
This means,   or 
Px Py Z y Py
Sufficient condition -To get the second order condition, we should partially differentiate the
above equations. Representing the second direct partial derivatives by Z xx and Z yy and the
second cross partial derivatives by Z xy and Z yx, the border Hessian determinant bordered
with 0, g x and g y is
0 gx gy 0  Px  Py
H  gx Lxx Lxy   Px Z xx Z xy  o
gy L yx L yy  Py Z yx Z yy

d 2 Z is referred to as positive definite subject to dg = 0 if H <0,

d 2Z is referred to as negative definite subject to dg = 0 if H > 0 .

2
Negative definiteness of d Z implies that the function achieves its relative maximum point
where as a positive definite is a sufficient condition to satisfy the relative minimum of the
objective function.
 Maximization
Example
Given the utility function of the consumer who consumes two goods x and y as
U (x, y) = (x+ 2) (y+1)
If the price of good x is P x = 4 birr, that of good y is P y = 6 Birr and the consumers has a
fixed budget of 130 birr. Determine the optimum values of x and y using the Lagrange
multiplier method,
Solution
Maximize U (x, y) = x y + x+ 2y + 2
Subject to 4x + 6y = 130
Now we should formulate the Lagrange function to solve this problem. That is
L( x, y ,  ) = x y + x+ 2y + 2 +  (130 - 4x - 6y)
L L L
Necessary conditions for utility maximization are  0,  0, 0
x y 
L
 ( y  1)  4 = 0
x
y = -1 + 4 

17
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

L
 ( x  2)  6  0
y
x  2  6
L
 4 x  6 y  130  0

4x+6y= 130
4 (-2+ 6  ) + 6 (-1 +4  ) = 130
- 8 + 24  - 6 + 24  = 130
48  = 144
 =3
Therefore, x = -2+6(3)
x = -2 + 18 = 16
y = -1 + 4 (3)
y = 11
Second order sufficient condition for utility maximization is

0 gx gy
H  gx L xx L xy
gy L yx L yy
The second partial derivatives of the objective function and the first partial derivatives of the
constraint function are
2L
L xx = = 0, L yy = 0, L xy = L yx = 1
x 2
g g
ց x = = 4, and ց y = 6 =
x y
Therefore, the bordered Hessian determinant of this function is
0 4 6
H  4 0 1 = - 4(0-6) + 6 (4- 0) = 48 > 0
6 1 0
The second order condition, i.e., H > 0 is satisfied for maximization. Thus, the consumer
maximizes utility when he consumes 11 units of good y and 16 units of good x. The
maximum utility is U = (16+2) (11+1) = (18) (12) = 216 units which is similar to the value of
the Lagrange function at these values of x , y and  . The value of the Lagrange multiplier 
is 3. It indicates that a one unites increase (decrease) in the budget of the consumer increases
(decreases) his total utility by 3 units.

Suppose the monopolist sells two products x and y and their respective demand is
P x = 100 - 2 x and P y = 80 - y

18
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

The total cost function is given as TC = 20x + 20y, when the maximum joint product of the
two outputs 60 unit. Determine the profit maximizing level of each output and their respective
price.
Solution
We know that total profit (  ) = TR - TC, where TR represents total revenue and TC
represents total cost.
TR= x P x + y P y = (100x - 2x2) + (80y - y2)
Thus  = 100x - 2x2 + 80 y - y2 - 20x - 20 y
 = 80 x + 60 y – 2x2- y2
But this monopolist can maximize its profit subject to the production quota. Thus,
Maximize  = 80x + 60 y- 2x2- y 2
Subject to x+ y = 60
To solve this problem, we should formulate the Lagrange function,
L (x, y,  ) = 80x + 60y - 2x2 - y 2 +  (x+ y - 60)
First order conditions for maximum profit are
L x = 80 - 4x +  = 0
- 4x = - 80 - 
 = 20 + 1 4 
L y = 60 - 2y+  = 0
- 2y = - 60 - 
1
y = 30 + 
2
L  = x + y -60 = 0
X + y = 60
1 1
20 +  + 30+  = 60
4 2
3
50 +  = 60
4
3
 = 10
4
40

3
1 40 1 ( 40)
Thus, x = 20 + ( ) y = 30 +
4 3 2 3
= 20+ 3.33 = 30+6.67
x = 23.33 y = 36.67
Second order condition for maximum profit is
L xx = - 4, L y y = -2, L x y = L y x = 0

19
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

g x = 1, g y = 1
Therefore, the bordered Hessian determinant of the given function is
0 1 1
H = 1 -4 0 = -1 (-2 - 0) + 1 (0+4) = 6> 0
1 0 -2

The second order condition is satisfied for maximization of functions.

Px  100  2(23.33) Py  80  36.67
Px  53.34 Py  43.33
Therefore, the monopolist maximizes its profit when it sells 23.33 of good x at a price of
40
53.34 birr per unit and 36.67 units of good y at a price of 43.33 birr per unit.  = shows
3
that a one unit increase in total expenditure on inputs increases total profit of the monopolist
40
by units. In other words, if the constant of the constraint relaxes by one unit that is
3
x  y  61 , then the value of the objective function increases by the value the Lagrange
multiplier .
 Minimization
As we know, the firm can determine the least cost combination of inputs for the production of
a certain level of output Q. Given the production function Q= f (L, K) and the cost function of
the firm is C = LPL + KP k Where L = labor, K = capital, Q = output. Suppose the price of
both input to be exogenous, we can formulate the problem of minimizing the cost as
Minimizes C = PL L + P k k
Subject to Q = f (L, K)
To determine the amount of labor and capital that should be employed initially we should
formulate the Lagrange function. It is
L  LPL  KPK   (Q  f ( L, K )
First order conditions for a minimum cost are
LL  PL  QL  0
PL P
  L
QL MPL
LK  PK  Qk  0
PK P
  K
QK MPK
L  Q  f ( K , L)  0
Where QL and Qk represents marginal product of labor and capital respectively.

20
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

PL P
  K
MPL MPK

The above equation indicates that, at the point of optimal input combination the input - price
ratio and the marginal product ratio have to be the same for each input. This ratio shows the
amount of expenditure per unit of the marginal product of the input under consideration.
Thus, the interpretation the Lagrange multiplier is the marginal cost of product at the optimal
condition. In other words, it indicates the effect of change in output on the total costs of
production, i.e., it measures the comparative static - effect of the constraint constant on the
optimal value of the objective function.
The first order condition indicated in equation (21) can be analyzed in terms of iso-quants and
iso-costs as
P MPL
 = L =
Pk MPk
MPL
The represents the negative of the slope of the isoquant, which measures the marginal
MPK
rate of technical substitution of labor to capital (MRTS Lk ).
PL
The ratio shows the negative of the slope of the isocost. An isocost is a line which
PK
indicates the locus of input combinations which entail the same total cost. It is shown by the
equation
C PL
C= PL L + P k K or K = - L
PL Pk
PL MPL
= indicates the fact that the isocost and isoquant lines are tangent to each other at
Pk MPk
the point of optimal input combination.
Second order condition for minimization of cost.
As you know, a negative bordered Hessian determinant is sufficient to say the cost is at its
minimum value. That is
0 QL QK
H  QL LLL LLK
QK LKL LKK
Example
Suppose a firm produces an output Q using labor L and capital K with production
function Q  10K 0.5 L0.5 . If the output is restricted to 200 units, price of labor is 10 birr per
unit, the price of labor is 10 birr per unit and Price of capital is 40Birr per unit, and then

21
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

determines the amount of L and K that should be employed at minimum cost. Find the
minimum cost.
The problem is Minimize C = 10 L + 40K
Subject to 200  10K 0.5 L0.5
Formulating the Lagrange function
L( L, K ,  )  10L  40K   (200  10K 0.5 L0.5 )
First order conditions
LL  10  5K 0.5 L0.5  0
2 L0.5

K 0 .5
LK  40  5K 0.5 L0.5  0
8 K 0.5

L0.5
L  200  10K 0.5 L0.5  0
10K 0.5 L0.5  200
2 L0.5 8 k 0 .5
=
k 0 .5 L0.5
2L = 8K
L= 4K
K 0.5 (4 K ) 0.5  20
2K = 20
K = 10 and L = 4(10) = 40,  = 4
Second order condition
Now we should check the second order condition to verify that cost is least at K = 10 and L =
40.
For cost minimization the determinant of the bordered Hessian matrix must be less than zero.

0 QL QK
H  QL LLL LLK < 0
QK LKL LKK
At L = 40 and K = 10
Q K 10
QL = = (5)  (5)  2.5
L L 40
Q L 40
Q k= = (5)  (5)  10
k K 10
= 2.5  K
0 .5 1 .5
LLL L = 2.5(4)(10) 0.5 (40) 1.5
= 0.125

22
Quantitative Methods In Economics I: Chapter Four Derivatives of Multivariate Functions

L kk = 2.5 K 1.5 L0.5 = 2.5(4)(10) 1.5 (40) 0.5

=2
LKL  LLK  2.5K 0.5 L0.5  2.5(4)(10) 0.5 (40) 0.5
Therefore, the determinant of the bordered Hessian matrix is

0 2 .5 10
H  2.5 0.125  0.5
10  0.5 2
= - 2.5 (5+5) +10(-1.25 -1.25)
= - 2.5 (10) + 10 (-2.5)
H = -50 < 0
Thus, the firm can minimize its cost when it employs 10 units of capital and 40 units of labor
in the production process and the minimum cost is
C = 10 (40) + 40 (10)
Min. C = 400 + 400 = 800 birr
In this problem K, L and  are endogenous. The Lagrange multiplier  measures the
responsiveness of the objective function to a change in the constant of the constraint function.

23