Engineering Economic Analysis

2019 SPRING
Prof. D. J. LEE, SNU

Chap. 20
PROFIT MAXIMIZATION
 Introduction
 A model of how the firm chooses the amount to
produce and the method of production to employ
 Profit maximization problem of a firm that faces
competitive market for the factors of production it
uses and the output goods it produces
 Competitive market
• A collection of well-informed consumers
• Homogeneous product that is produced by a large
number of firms
• Price-taking behavior
 Exogenous variable: price
 Endogenous variable: levels of outputs and inputs
1
 Economic Profit

 A firm uses inputs j = 1…,m to make products i =

1,…n.
 Output levels are y1,…,yn.
 Input levels are x1,…,xm.
 Product prices are p1,…,pn.
 Input prices are w1,…,wm.
 Profit = Revenue – Cost
n m
=π ∑ pi yi − ∑ wi xi
=i 1 =i 1

• Economic definition of profit requires that all inputs

and outputs are valued at their opportunity cost
2
 Profit Maximization

 Profit maximization
n m
=
Max π ∑ pi yi − ∑ wi xi
y ,x
i i=i 1 =i 1

• Using production plan y ∈ Y , where y j ≥ (≤)0 if j is output (input)

Max π ( p )= p ⋅ y
y

such that y ∈ Y ,
where p is the vector of prices for all inputs and outputs

• 1-output case, we can use the production function y = f ( x )

m
p f ( x ) − ∑ wi xi
Max π =⋅
xi i =1

3
 Fixed and Variable factors

 Fixed factor: a factor of production that is in a

fixed amount for the firm
• Fixed factor must be expensed even at the state of zero
output
 Variable factor: a factor which can be used in
different amounts
 Short run: there are some fixed factors
 Long run: all factors are variable factors
• In the short run, the firm could make negative profits
• But in the long run, the least profit is zero since the firm
always free to decide to use zero inputs and produce
zero output
4
Short-run Profit Maximization(1-output & 2-inputs)

 Suppose the firm is in a short-run circumstance in

which x 2 : a fixed factor
 Its short-run production function is f ( x1 , x 2 )

 Profit-max. problem
p f ( x1 , x2 ) − w1 x1 − w2 x2
max π =⋅
x1

 F.O.C.
∂π ∂f ( x1* , x2 )
∂x1
=p ⋅
∂x1
− w1 =0
(
p ⋅ MP1 x1* , x 2 =
w1)
x1* ( p, w1 ) : factor demand function “The value of marginal product of
factor 1 should equal its price”

5
Short-run Profit Maximization(1-output & 2-inputs)

 Iso-profit curves
• Given profit function π =py − w1 x1 − w2 x 2
• The level set of profit function y = w1 x + 1 (π + w x )
1 2 2
p p

• all combinations of inputs and outputs that give a constant

level of profit
y

w1
Slopes = +
p

x1
6
 Short-Run Profit-Maximization

Short-run
production function

x1
7
 Short-Run Profit-Maximization

y Optimality condition
d w1
f ( x1 , x 2 ) =
*
Slopes = +
w1
dx1 p p

( )
p ⋅ MP1 x1* , x 2 =
w1 Short-run
production function
f ( x1 , x 2 )

d 2 f ( x1* , x 2 ) x1
Second order condition 2
≤ 0 : locally concave
dx1 8
 Short-Run Profit-Maximization

 Short-run Cobb-Douglas production function

9
Long-Run Profit-Maximization (1-output, 2-inputs)

 Now allow the firm to vary all input levels.

 Since no input level is fixed, there are no
fixed costs.
 Profit maximization
max π =⋅ p f ( x1 , x2 ) − w1 x1 − w2 x2
{x , x }
1 2

 F.O.C. • Optimality condition

∂π ( )
∂f x1* , x2* p ⋅ MP1 = w1 , p ⋅ MP2 = w2
=p ⋅ − w1 =0
∂x1 ∂x1 • Solution
∂π
=p ⋅
( )
∂f x1* , x2*
− w2 =0
x1* ( w1 , w2 , p )
:Factor demand function
∂x2 ∂x2 x2 ( w1 , w2 , p )
*

10
Long-Run Profit-Maximization (1-output, 2-inputs)

 Cobb-Douglas production function y = x1a x2b

• Profit-max. problem
max π ( x1 , x2 )= px1a x2b − w1 x1 − w2 x2

• F.O.C. • Multiplying xi
 ∂π
= pax1a −1 x2b =
− w1 0
 1∂x  pax1a x2b − w1 x1 =
0 pay = w1 x1
  pby = w2 x2
= ∂π  pbx1 x2 − w2 x2 =
a b
0
pbx1a x2b −1 −=w2 0
 ∂x2

• Factor demand function

apy
x1* ( w1 , w2 , p ) =
w1
bpy
x2* ( w1 , w2 , p ) =
w2
11
 Long-Run Profit-Maximization (1-output, 2-inputs)

• Inserting factor demand functions into the

production function gives
a b a b
 apy   bpy   ap   bp  a + b
=y =        y
 w1   w2   w1   w2 
a b
 ap   bp 
∴ y1− a −b =
   
 w1   w2 

• Supply function
a b
 ap  1− a − b  bp  1− a − b
y ( p, w1 , w2 ) =    
 w1   w2 

12
Long-Run Profit-Maximization (1-output, n-inputs)

• Output y, Input bundle x

• Profit maximization
max π (=
x ) pf ( x ) − w ⋅ x
x

• F.O.C.
∂f ( x )
p⋅ = wi i =1,..., n
∂xi

x * ( p, w ) : factor demand function

f ( x * ( p, w )) : supply function

13
Long-Run Profit-Maximization (1-output, n-inputs)

 Exceptional case
1) When production function is not differentiable
 Leontief technology

2) Corner (boundary) solution case ( xi* = 0 for some i )

 F.O.C.

∂f ( x )
p⋅ − wi = 0 if xi* > 0
∂xi
∂f ( x )
p⋅ − wi ≤ 0 if xi* = 0
∂xi

14
Long-Run Profit-Maximization (1-output, n-inputs)

 Optimization with constraints

• When equality constraints
max f ( x )
s.t. h ( x ) = c

• Lagrangian function
L ( x , λ ) ≡ f ( x ) − λ  h ( x ) − c 

• Kuhn-Tucker condition
( )
Suppose that x * = x1* ,..., xn* is a solution.
Suppose further that ∂h ∂xi xi = xi*
≠ 0 (critical point)
Then there exits a real number λ * such that
∂L ∂L
= = 0 at x * , λ *
∂xi ∂λ
( )
15
Long-Run Profit-Maximization (1-output, n-inputs)
• When inequality constraints
max f ( x ) Lagrangian function
s.t. g ( x ) ≤ b L ( x , µ ) ≡ f ( x ) − µ  g ( x ) − b 

• Kuhn-Tucker condition
( )
Suppose that x * = x1* ,..., xn* is a solution.

(
If g x* , y* ) b (binding), then further suppose that ∂g ∂xi xi = xi*
≠ 0.

Then there is a multiplier µ * ≥ 0 such that

∂L
∂xi
(
= 0 at x * , µ * )
µ *  g ( x * ) − b  =
0 (complementary slackness)

( )
g x* , y * ≤ b

16
Long-Run Profit-Maximization (1-output, n-inputs)
• Example
max f ( x, y ) = xy
s.t. x 2 + y 2 ≤ 1

17
 Long-Run Profit-Maximization

 Generalized optimality condition for 2-input

max p ⋅ f ( x1 , x2 ) − ( w1 x1 + w2 x2 )
s.t. xi ≥ 0 ⇒ − xi ≤ 0

• Lagrangian
L=p ⋅ f ( x1 , x2 ) − ( w1 x1 + w2 x2 ) + µi xi

• K-T condition
∂L ∂f
= p − wi + µ= 0,
∂xi ∂xi
i

µi xi 0 (complementary slackness), xi ≥ 0, µi ≥ 0

• Optimality condition
If xi* 0, then µi* ≥ 0.
= If xi* > 0, then µi* =
0.
∂f ∂f
Thus, if=
xi* 0, then p ⋅ − wi ≤ 0 Thus, if xi* > 0, then p ⋅ − wi =
0
∂xi ∂xi

18
Long-Run Profit-Maximization (1-output, n-inputs)

 Exceptional case
3) No optimal solution case
 When f(x)=x. If p > w, then x* =
∞

 When CRS technology

Let x be the optimal and assume that
*
( )
p ⋅ f x * − w ⋅ x * = π * > 0
Scale up production by t > 1
( ) ( )
Since CRS, f tx * = tf x *

Then p ⋅ f ( tx ) − w ⋅ ( tx )=

* *
( )
t  p ⋅ f x * − w ⋅ x * = tπ * > π *
Contradiction!
→Thus the only nontrivial profit-max position for a
CRS firm is zero-profits

19
Long-Run Profit-Maximization (1-output, n-inputs)

 Exceptional case
4) Multiple (Infinite) number of optimal solutions
Let x * be the optimal for a CRS technology which
gives zero profit, i.e., p ⋅ f ( x* ) − w ⋅ x* = π * = 0

Then scale up production by t > 0

( ) ( )
p ⋅ f tx * − w ⋅ tx * = tπ * = 0

Thus tx * is also an optimal for any t > 0 !!

20
 Comparative Statics

 One-input & One-output

Max pf ( x ) − wx
x

F.O.C.: pf ′ ( x* ( p, w ) ) − w =
0
x* ( p, w ) : factor demand function
S.O.C.: pf ′′ ( x* ( p, w ) ) ≤ 0

• Differentiating F.O.C. with respect to w

dx* ( p, w )
pf ′′ ( x ( p, w ) )
*
−1 ≡ 0
dw
• Assuming that f ′′ ≠ 0
1) Sign ∂x* ( p, w )
dx *
( p, w ) = 1  Since f " < 0 → <0
dw pf ′′ ( x* ( p, w ) ) ∂w
2) Magnitude
∂x*
 As f ′′ increases, decreases.
∂w

21
 Comparative Statics

 Two-input & One-output

Max pf ( x1 , x2 ) − ( w1 x1 + w2 x2 )
{ x1 , x2 }

• F.O.C.
∂f  x1 ( w1 , w2 ) , x2 ( w1 , w2 ) 
p ≡ w1
∂x1
xi* ( w1 , w2 ) :factor demand function
∂f  x1 ( w1 , w2 ) , x2 ( w1 , w2 ) 
p ≡ w2
∂x2
• Differentiating F.O.C. with respect to w1 and w2 (let p=1)

∂x1 ∂x ∂x1 ∂x
f11 + f12 2 =1 f11 + f12 2 =0
∂w1 ∂w1 and ∂w2 ∂w2
∂x1 ∂x ∂x1 ∂x
f 21 + f 22 2 =0 f 21 + f 22 2 =1
∂w1 ∂w1 ∂w2 ∂w2

22
 Comparative Statics

• Rearranging by matrix form

 ∂x1 ∂x1 
 f11 f12   ∂w1 ∂w2   1 0 
   = 
 f 21 f 22  ∂x2 ∂x2   0 1 
 ∂w ∂w2 
 1
 f11 f12 
• Note that   is Hessian matrix.
 f 21 f 22 
 By Young’s theorem, Hessian matrix is symmetric f12 = f 21

• Then the substitution matrix can be obtained

 ∂x1 ∂x1 
− f12 
 ∂w  f 22
∂w2   f11
−1
f12   
 1 = −f
=  21
 f11 
 ∂x1 ∂x2   f 21 f 22 
 ∂w ∂w2 
H
 2

23
 Comparative Statics

• If we assume that S.O.C. is satisfied, it is

equivalent to the fact that the Hessian matrix is ND
 Thus, f11 < 0, H = f11 f 22 − f12 f 21 > 0

• Comparative results
 The changes of factor demand with respect to the
change of its own price
∂xi f jj
= <0
∂wi H
 The changes of factor demand with respect to the
change of other price
∂xi fij f ji ∂x j
=
− =
− = :indeterminate and symmetric
∂w j H H ∂wi

24