Lecture Notes for Chapter 11

Kevin Wainwright
April 26, 2014

1 Optimization with More than One Vari-

able
Suppose we want to maximize the following function

z = f (x; y) = 10x + 10y + xy x2 y2

Note that there are two unknowns that must be solved for: x and
y. This function is an example of a three-dimensional dome. (i.e. the
roof of BC Place)
To solve this maximization problem we use partial derivatives.
We take a partial derivative for each of the unknown choice variables
and set them equal to zero
@z
@x = fx = 10 + y 2x = 0 The slope in the ”x” direction = 0
@z
@y = fy = 10 + x 2y = 0 The slope in the ”y” direction = 0

This gives us a set of equations, one equation for each of the un-
known variables. When you have the same number of independent
equations as unknowns, you can solve for each of the unknowns.
rewrite each equation as

1
 y = 2x 10
x = 2y 10
substitute one into the other

x = 2(2x 10) 10
x = 4x 30
3x = 30
x = 10
similarly,
y = 10

REMEMBER: To maximize (minimize) a function of many

variables you use the technique of partial di¤erentiation. This produces
a set of equations, one equation for each of the unknowns. You then
solve the set of equations simulaneously to derive solutions for each of
the unknowns.

1.0.1 Second order Conditions (second derivative Test)

To test for a maximum or minimum we need to check the second partial
derivatives. Since we have two …rst partial derivative equations (fx ,fy )
and two variable in each equation, we will get four second partials
( fxx ; fyy ; fxy ; fyx )
Using our original …rst order equations and taking the partial deriv-
atives for each of them (a second time) yields:

2
 fx = 10 + y 2x = 0 fy = 10 + x 2y = 0

fxx = 2 fyy = 2
fxy = 1 fyx = 1

The two partials,fxx , and fyy are the direct e¤ects of of a small
change in x and y on the respective slopes in in the x and y direction.
The partials, fxy and fyx are the indirect e¤ects, or the cross e¤ects
of one variable on the slope in the other variable’s direction. For both
Maximums and Minimums, the direct e¤ects must outweigh the cross
e¤ects

1.1 Rules for two variable Maximums and Mini-

mums
1. Maximum
fxx < 0
fyy < 0
fyy fxx fxy fyx > 0

2. Minimum
fxx > 0
fyy > 0
fyy fxx fxy fyx > 0

3. Otherwise, we have a Saddle Point

From our second order conditions, above,
fxx = 2 < 0 fyy = 2 < 0
fxy = 1 fyx = 1
3
 and
fyy fxx fxy fyx = ( 2)( 2) (1)(1) = 3 > 0
therefore we have a maximum.

1.2 Using Di¤erentials Approach

Given
z = f (x; y)
Then
dz = fx dx + fy dy
if
dx 6= 0; dy 6= 0
and
dz = 0 (critical point)
Then it must be true that
@z @z
fx = fy = 0 or @x = @y =0

Fx = 0: Means z is not changing in the x-direction

Fy = 0: Means z is not changing in the y-direction
This is the First Order Necessary Condition for a max or min

1.3 Second Order Conditions

Given
z = f (x; y)
The …rst derivative (di¤erential) is

dz = fx dx + fy dy

4
 Take the total di¤erential a second time, treating dx and dy as constants

d2 z = fxx dxdx + fyy dydy + fxy dxdy + fyx dydx

= fxx dx2 + fyy dy 2 + fxy dxdy + fyx dydx

where
fxx = 2nd partial derivative with respect to x
fyy = 2nd partial derivative with respect to y
@z
fxy = Change in @x from a in y
@z
fyx = Change in @y from a in x
| {z }
fxy; fyx are cross partial derivatives

1.4 Example: Two Market Monopoly with Joint

Costs
A monopolist o¤ers two di¤erent products, each having the following
market demand functions
1
q1 = 14 4 p1
1
q2 = 24 2 p2

The monopolist’s joint cost function is

C(q1 ; q2 ) = q12 + 5q1 q2 + q22

The monopolist’s pro…t function can be written as

= p1 q 1 + p2 q2 C(q1 ; q2 ) = p1 q1 + p2 q2 q12 5q1 q2 q22

which is the function of four variables: p1 ; p2 ; q1 ;and q2 . Using

the market demand functions, we can eliminate p1 and p2 leaving us

5
with a two variable maximization problem. First, rewrite the demand
functions to get the inverse functions

p1 = 56 4q1
p2 = 48 2q2

Substitute the inverse functions into the pro…t function

= (56 4q1 )q1 + (48 2q2 )q2 q12 5q1 q2 q22

The …rst order conditions for pro…t maximization are

@
@q1 = 56 10q1 5q2 = 0
@
@q2 = 48 6q2 5q1 = 0

Solve the …rst order conditions using Cramer’s rule. First, rewrite
in matrix form
10 5 q1 56
=
5 6 q2 48
where jAj = 35

56 5
48 6
q1 = = 2:75
35
10 56
5 48
q2 = = 5:7
35
Using the inverse demand functions to …nd the respective prices,
we get
p1 = 56 4(2:75) = 45
p2 = 48 2(5:7) = 36:6

6
 From the pro…t function, the maximum pro…t is

= 213:94

Next, check the second order conditions to verify that the pro…t
is at a maximum. The various second derivatives can be set up in a
matrix called a Hessian The Hessian for this problem is

11 12 10 5
H= =
21 22 5 6

The su¢ cient conditions are

jH1 j = 11 = 10 < 0 (First Principle Minor of Hessian)
jH2 j = 11 22 12 21 = ( 10)( 6) ( 5)2 = 35 > 0 (determinant)

Therefore the function is at a maximum. Further, since the signs

of jH1 j and jH2 j are invariant to the values of q1 and q2 , we know that
the pro…t function is strictly concave.

1.5 Example: Pro…t Max Capital and Labour

Suppose we have the following production function

q = Output
1 1
q = f (K; L) = L + K
2 2 L = Labour
K = Capital

Then the pro…t function for a competitive …rm is

= P q wL rK P = Market Price
or w = Wage Rate
1 1
= P L 2 + P K 2 wL rK r = Rental Rate
7
 First order conditions

General Form
@ P 1
1: @L = 2L 1
2 w=0 P fL w = 0
@ P
2: @k = 2K
2 r=0 P fK r = 0
Solving (1) and (2), we get

L = ( 2w
P )
2
K = ( 2r
P)
2

Second order conditions (Hessian)

3
= P fLL = 4P L 2 < 0
LL
P 3
KK = P fKK = 4 K 2 < 0
LK = KL = P fLK = P fKL = 0
or, in matrix form
P 3
LL LK L 2 0
H = = 4 P 3
KL KK 0 4 K 2

P 3 P 3
P fLL fKK (fLK )2 = L 2 K 2 0>0
4 4
Di¤erentiate …rst order of conditions with respect to capital (K)
and labour (L)
=)Therefore pro…t maximization

Example: If P = 1000, w = 20, and r = 10

1. Find the optimal K, L, and

2. Check second order conditions

8
1.6 Example: Cobb-Douglas production function
and a competitive …rm
Consider a competitive …rm with the following pro…t function

= TR TC = PQ wL rK

where P is price, Q is output, L is labour and K is capital, and

w and r are the input prices for L and K respectively. Since the …rm
operates in a competitive market, the exogenous variables are P,w and
r. There are three endogenous variables, K, L and Q. However output,
Q, is in turn a function of K and L via the production function

Q = f (K; L)

which in this case, is the Cobb-Douglas function

Q = La K b

where a and b are positive parameters. If we further assume de-

creasing returns to scale, then a + b < 1. For simplicity, let’s consider
the symmetric case where a = b = 41
1 1
Q = L4 K 4

Substituting Equation 3 into Equation 1 gives us

1 1
(K; L) = P L 4 K 4 wL rK

The …rst order conditions are

@ 1 3 1

@L =P 4 L 4K 4 w=0
@ 1 1 3

@K =P 4 L4 K 4 r=0

9
 This system of equations de…ne the optimal L and K for pro…t
maximization. But …rst, we need to check the second order conditions
to verify that we have a maximum.
The Hessian for this problem is
" #
3 7 1 1 2 3 3
LL LK P ( 16 )L K P 4 L K
4 4 4 4
H= = 2 3 3 1 7
KL KK P 14 L 4 K 4 P 3
16 L K
4 4

The su¢ cient conditions for a maximum are that jH1 j < 0 and
jHj > 0: Therefore, the second order conditions are satis…ed.
We can now return to the …rst order conditions to solve for the
optimal K and L. Rewriting the …rst equation in Equation 5 to isolate
K 3 1
P 14 L 4 K 4 = w
3
K = ( 4w
p L 4 )4

Substituting into the second equation of Equation 5

3
1 3 1 3
4 4
P P 4w
4L K = L p L =r
4 4 4 4
4
1 4
= P4 4 w 3L 2 =r

Re-arranging to get L by itself gives us

P 3 1
L =( w 4 r 4 )2
4
Taking advantage of the symmetry of the model, we can quickly
…nd the optimal K
P 3 1
K = ( r 4 w 4 )2
4
L and K are the …rm’s factor demand equations.

10
1.7 Young’s Theorem
For cross partial "e¤ects" the order of di¤erentiation is immaterial.
Therefore:
fxy = fyx
As long as the cross partials are continuous.

In the case of GENERAL FUNCTIONS , this will always be as-

sumed to be true!

1.7.1 Example:
z = x3 + 5xy = y 2

fx = 3x2 + 5y fy = 5x 2y
fxx = 6x fyy = 2
fxy = 5 fyx = 5
| {z }
fxy =fyx

1.8 Quadratic Form

The function
q = ax2 + 2bxy + cy 2
is a quadratic form. A quadratic can be written in matrix form as:

a b x
q = x y
1x1 1x2
c d y
2x2 2x1

Det = (ac b2 )

11
1.9 Positive and Negative De…niteness
q is said to be

1. (a) i. positive de…nite

ii. positive semi-de…nite
iii. negative semi-de…nite
iv. negative de…nite

IF q is always > 0; 0; 0; < 0 (for all x, y)

positive de…nite a>0

q is if and (ac b2 ) > 0
negative de…nite a<0
The second order total di¤erential
d2 z = (fxx )dx2 +(2fxy )dxdy+(fyy )dy 2 fY oung0sT heorem fxy = fyx g
Therefore
f f dx
d2 z = dx dy xx xy
fxy fyy dy

The matrix of 2nd partial derivatives is called the Hessian

fxx fxy 2
H= where jHj = fxx fyy fxy
fxy fyy
Second Order Conditions
positive de…nite
q is
negative de…nite
2
if ffxx > 0; fxx < 0g, and fxz fyy fxy >0
2
Note: fxz fyy fxy > 0 implies that fyy must have the same sign as
fxx
Therefore if:

12
 1. fx = fy = 0 (FOC), and if:

2. SOC
negative de…nite a maximum
d2 z is then z is
positive de…nite a minimum

1.10 n-Variable Case

Given
z = f (x1 ; x2 ; :::xn )
For an Extremum (max or min):

f1 = f2 = f3 = ::: = fn = 0 (First Order Conditions)

Then d2 z in Matrix Form is

2 32 3
f11 f12 f13 ::: f1n dx1
6 f21 f22 ::: ::: f2n 7 6 7
6 76 dx2 7
dx1 dx2 ::: dxn 6
6 f31 f33 ::: ::: 76 7
76 dx3 7
(1xn) 4 ::: ::: ::: ::: ::: 54 ::: 5
fn1 ::: ::: ::: fnn dxn
(nxn) (nx1)

For a Max: (Principal minors alternate signs)

2
jH1 j = f11 < 0; jH2 j = f11 f22 f12 > 0; jH3 j < 0; :::( 1)n jHn j > 0
For a Min: (Principal minors have the same sign)
jH1 j > 0; jH2 j > 0; ::: jHn j > 0

13
1.10.1 Example: Output Maximization
Let
Q = 10L + 10K + LK L2 K2
F.O.C.’s
@Q
@L = 10 + K 2L = 0 2L K = 10
@Q OR
@K = 10 + L 2K = 0 L + 2K = 10

2 Equations with 2 unknowns from FOC. Matrix Form:

2 1 L 10
= Det = 3
1 2 K 10

Cramer’s Rule

10 1
10 2 20 + 10
L = = = 10
3 3
2 10
1 10 20 + 10
K = = = 10
3 3
Now check 2nd order conditions from F.O.C
10 + K 2L = 0 when K,L=10 FOC’s
10 + L 2K = 0 are identities

S.O.C.

dK 2dL = 0
dL 2dK = 0

14
 Find the Hessian
2 1 dL 0
=
1 2 dK 0

QLL QLK 2 1
H= =
QKL QKK 1 2

QLL = 2<0 QLL QKK Q2KL = ( 2)( 2) (1) > 0

| {z }
Therefore Q is Max at K=10, L=10

1.11 Economic Interpretation of the 2nd Order

Conditions
Given the production function

Q = Q(K; L)

QL > 0; QLL < 0 implies the "Law of Diminishing Returns"

The condition
QLL QKK Q2KL > 0

1. (a) says that for a Maximum the direct e¤ects (QLL ;QKK ) must
outweigh the indirect e¤ects (QKL QLK )
(b) a production function can have the properties of "the law
of diminishing returns" and "increasing returns to scale"
at the same time
(c) Therefore Q(K,L) has no unconstrained maximum

15
1.12 Pro…t Maximization and Comparative Stat-
ics
Let q = f (x1 ; x2 )be the production function.
The pro…t function is

= pf (x1 ; x2 ) w 1 x1 w 2 x2

where p = output price, w1 ,w2 = factor prices of x1 and x2 respectively

The FOC‘s for pro…t max

)
@
@x1 = pf1 w1 = 0 This gives us 2 equations and
@ 2
@x2 = pf2 w2 = 0 2 unknowns, x1 ,x2

16
 Solving the F.O.C’s we get:
x1 = x1 (w1 ; w2 ; P ) x2 = x2 (w1 ; w2 ; P )
fx1 ,x2 as functions of exogenous variablesg

2nd order conditions

From the F.O.C.’s

pf1 w1 = 0
at x1 =x1 ; x2 =x2
pf2 w2 = 0
Di¤erentiate again for the Hessian

pf11 pf12 dx1 0

=
pf21 pf22 dx2 0
pf11 pf12 f f
H= = p 11 12
pf21 pf22 f21 f22
For Maximum
2
jH1 j = f11 < 0 jH2 j = f11 f22 f12 >0

1.13 Comparative Statics

By substituting
x1 = x1 (w1 ; w2 ; P ) x2 = x2 (w1 ; w2 ; P )
back into the F.O.C.’s
P f1 = (x1 (w1 ; w2 ; P ); x2 (w1 ; w2 ; P )) w1 = 0
P f2 = (x1 (w1 ; w2 ; P ); x2 (w1 ; w2 ; P )) w2 = 0
The F.O.C.‘s become identities that implicitly de…ne x1 ,x2 as functions
@x1 @x2
of w1 ,w2 ,and P. Therefore to …nd @w ;
1 @w1
etc. we can use the implicit
function theorem by …nding the Jacobian of the F.O.C.’s

17
 @x1 @x2
Find: @w ;
1 @w1
Totally di¤erentiate with respect to w1

@x1 @x dw1 dw1

P f11 + P f12 2 = 0 =1
@w1 @w1 dw1 dw1
@x @x
P f21 1 + P f22 2 = 0
@w1 @w1
Matrix Form:
!
@x1
P f11 P f12 @w1 1
@x2 =
P f21 P f22 @w1
0

The Jacobian determinant

2
jJj = P (f11 f22 f12 )>0

The Jacobian of the F.O.C.’s is also the Hessian of the S.O.C.’s

1.13.1 Solving by Cramer’s Rule

1 P f12
@x1 0 P f22 P f22 f22
= = 2 = 2 <0
@w1 jHj P (f11 f22 f12 ) f11 f22 f12
P f11 1
@x2 P f21 0 f22
@w1
=
jHj
== 2 ? 0?
f11 f22 f12
@x1 @x2
@w1 < 0 implies downward sloping factor demand curve. For @w 1
this
sign depends on the relationship in production between x1 and x2

18
1.13.2 Example: Pro…t Maximization
Suppose we have the following production
8 9
1 1
< q = output =
q = f (K; L) = L 2 K 2 L = labour
: ;
K = capital
Then the pro…t function for a competitive …rm is
8 9
< P = market price =
= P q wL rK w = wage rate
: ;
r = rental rate
1 1
or = P L2 + P K 2 wL rK

First Order Conditions

General Form
@ P 1 z }| {
(1) = L2 w=0 fP fL w = 0g
@L 2
@ P 1
(2) = K2 r=0 fP fK r = 0g
@K 2
Solving (1) and (2) we get
2w 2 2n 2
L = P K = P

Second Order Conditions (Hessian)

Di¤erentiate First Order Conditions with respect to K, L
General

P fLL dL + P fLK dK = 0
P fKL dL + P fKK dK = 0

19
 Hessian
P fLL P fLK dL
P fKL P fKK dK
jH1 j = P fLL < 0
jH2 j = P fLL fKK (fKK )2 > 0
Speci…c

P 3
L 2 dL + (0)dK = 0
4
P 3
K 2 dL + (0)dL = 0
4
Hessian
P 3

4L 0 dL
2

P 3
0 4K
2 dK
P 3
H1 = L2
4
P 3 P 3
jH2 j = L2 K 2 0>0
4 4
jH2 j for both general and speci…c >0, therefore Pro…t Max
From the FOC’s we know:
2 2
L = 2w P K = 2r P
by subbing K and L into the pro…t function, we get:
1 1
= P L 2 + P K 2 wL rK
" #1 " # 21
2 2 2 2 2
2w 2r 2w 2r
= P +P w r
P P P P
P2 P2 P2 P2
= +
2w 2r 4w 4r
20
 Finally:
P2 P2
= (w; r; P ) = +
4w 4r
where (w; r; P ) is "Maximum pro…ts as a function of w,r, and P"

1.14 Hotelling’s Lemma

Hotelling’s Lemma states the following conditions about the pro…t
function:
@ (w;r;P )
1: @P =q
@ (w;r;P ) @ (w;r;P )
2a: @w =L 2b: @r =K

Using the pro…t function:

P2 P2
(w; r; P ) = +
4w 4r
Condition 1:
@ 2P 2P P P
= + = +
@P 4w 4r 2w 2r
Check:
" # 12 " # 21
2 2
1 1 2w 2r
q = L2 K 2 = +
P P
1 1
2w 2r P P
= + = +
P P 2w 2r

Condition 2a
2
@ (w; r; P ) @ P2 P2 P2 2w
= + = =
@w @w 4w 4r 4w2 P

21
Therefore @@w = L
Condition 2b
2
@ (w; r; P ) P2 2r
= = =K
@r 4r2 P

1.14.1 Factor Demand Curves

L and K are the …rms demand curves for labour and capital

P2 @h P2
L = =) = <0
4w2 @w 4w3
P2 @K P2
K = =) <0
4r2 @r 4r3
Therefore: Downward sloping factor demand curves

1.15 Iso-Pro…t Curves (Level Curves)

Take the total di¤erential of (w; r; P ); let d =0

P2 P2
d = dw + dr = 0
4w2 4r2
P2
dr 4w2 r2
= P2
= 2
< 0 (slope of Iso-Pro…t Curve)
dw 4r2
w

Concave or Convex?

d dr r2 r2
= 2 =2 >0
dw dw w3 w3

22
 dr
Therefore the slope of the Iso-Pro…t curve is negative dw but the
2
d r
slope is becoming less negative: dw2 > 0 Therefore: Convex

23
1.16 Pro…t Maximization
Developing the pro…t function

= TR TC
where
= PQ C(Q)
Therefore pro…t max is:

@ @T R @C
= = MR MC = 0 Q is the choice variable
@Q @Q @Q
Now suppose
Q = f (K; L)
Then

= P f (K; L) (wL + rK) where TC=wL+rK

Now pro…t max is:

@
(1) @L = P fL w=0
@
(2) @K = P fK r=0

Now K,L are the choice variables.

K = K (w; r; P )
The solution are demand curves
L = L (w; r; P )
Now suppose the …rm is a monopolist, then he faces a downward
sloping demand curve
P = D(Q)
Pro…t function is
= D(Q)Q wL rK

24
 where
Q = f (K; L)
Di¤erentiate using the Chain Rule
F.O.C.

L : [D(Q) + QD0 (Q)] fL w=0

K : [D(Q) + QD0 (Q)] fK r=0
OR
M R M PL w = 0
M R M PL r = 0
OR
[D(f (K; L)) + f (K; L)D0 (f (K; L))] fL w = 0
[D(f (K; L)) + f (K; L)D0 (f (K; L))] fK r = 0
Giving
K = K (w; r) L = L (w; r)

2 Chapter 11: Part 2 - Price Disc.

2.0.1 Example
Let
P1 = 100 q1 P2 = 150 2q2 Mkt. AR Functions
Let
T C = 100 + (q1 + q2 )2
= P1 q1 + P2 q2 100 (q1 + q2 )2
= 100q1 q12 + 150q2 2q22 100 (q1 + q2 )2

25
 FOC‘s

1 = 100 2q1 2(q1 + q2 ) = 100 4q1 2q2 = 0

2 = 150 4q2 2((q1 + q2 ) = 150 2q1 6q2 = 0

100 2
150 6 600 300
q1 = = = 15
4 2 20
2 6
4 100
2 150 600 200
q2 = = = 20
20 20
P1 = 85 P2 = 110

SOC’s

4 2
jHj = H1 = 4 <0 H2 = 20 > 0
2 6
Therefore a Max
**At home, verify that the Inverse Elasticity Rule holds here!

2.1 Concavity and Convexity

Let
y = f (x)
where
x = [x1 ; :::xn ] and let x^ = [^
x1 ; :::^
xn ]
such that x̄ 6= x̂

26
 De…nition 1:
y=f(x̄) is a concave function if

f (k x + (1 k) x^) kf (x) + (1 k)f (^

x)
| {z } | {z }
Point on Dome Line Segment

De…nition 2:
y=f(x̄) is convex if

f (kx + (1 k)^
x) kf (x) + (1 k)f (^
x)

for strict concavity/convexity replace the weak inequalities with strict

inequalities.
If the function y=f(x̄) is twice di¤erentiable, then the following
holds:

Theorem 1: y=f(x̄) is concave/convex if and only the Hessian, jHj

is negative/positive semide…nite

Theorem 2: If the Hessian is negative de…nite/positive de…niate for

all x, then y=f(x) is concave/convex
NOTE: Theorem 2 is a su¢ cient condition for strict concavity/convexity
but it is not a necessary condition

2.2 Limit Output Model

Suppose a monopolist faces the following demand curve

p=a q a is a constant > 0

His cost function is

T C = k + cq where K = set up costs, cq = variable costs

27
 Therefore
k
AT C = +c f= AF C + AV Cg
q
The pro…t function is

= pq (K + cq)

Maximize
@ a c
=a 2q c=0 ! q=
@q 2

a c a+c
p=a 1=a =
2 2
Set MR=MC

a 2q = c
a c
q =
2
Monopolists pro…t max graphically
1

Now consider a potential entrant to the monopolist‘s market

Assumption: Entrant takes monopolist‘s output as given
Let

qe = Entrant0 s Output
qm = M onopolist0 s Output
1
Graph - page 5 Cha. 11 part 2

28
 If entrant does enter, market price will be:

p=a (qm qe )

Entrant’s pro…ts

= pqe k cqe
e = (a qe qm )qe k cqe
@ e
= a qm 2qe c=0
@qe
a c qm
qe = Entrant’s output is a function of the monopolist’s output.
2
Entrant’s output
a c qm
qe =
2
Sub into pro…t function

e = (a qe qm )qe k cqe
2
a c qm a c qm a c qm
e = (a qm ) k c
2 2 2
Entrant’s pro…t function is a function of a, c, k, and qm
He will enter if: e > 0 OR if: (a qm qe )qe cqe > k
Which says: If an entrant’s pro…ts (gross) can cover …xed costs (k)
then he will enter the market of the monopolist.
Graphically:
Entrant takes monopolist’s qm as given
Entrant maximizes pro…ts o¤ the residual demand curve
MONOPOLIST’S DEMAND CURVE
2

2
insert …rst graph on page 8 chap 11 part 2

29
 B=Entrants pro…t above variable costs
if B>k then the entrant will enter
if B<k then there will be no entry
RESIDUAL DEMAND CURVE
3

The monopolist knows that

a c qm
qe =
2
or generallyqe = f (qm )Therefore the monopolist can e¤ect the en-
trant’s choice qe
The monopolist can choose qm such that when the entrant chooses
the optimal qe he will not earn any pro…ts
Therefore the monopolists maximization problem is:
MAX:
m = (a qm ) qm k cqm
Subject to:
e = (a qm qe )qe cqe k
Substitute
c qm a
qe =
2
into the monopolist’s max problem, Max
2
aqm qm cqm k

subject to
2
a c qm a c qm a c qm
(a qm ) c =K
2 2 2
Notice that there is now only one choice variable, qm :
3
insert second graph page 8 ch. 11 part 2

30
 There qm is determined by the constant
Without di¤erentiating solve the constraint for qm

Answer: p
2
qm = a c k
4

2.3 Cournot Duopoly

Suppose the monopolist decides to allow entry. The result: Duopoly
Assumption: Each …rm takes the other …rms output as exongenous
and chooses the output to maximize its own pro…ts
Market Demand:

P = a bq
or P = a b(q1 + q2 ) fq1 + q2 = qg

where qi is …rm i’s output fi = 1; 2g

Each …rm faces the same cost function

T C = K + cqi fi = 1; 2g

Each …rm’s pro…t function is:

i = pqi cqi K

Firm 1:

1 = pq1 cq1 K
1 = (a bq1 bq2 )q1 cq1 K
4
GRAPH page 11

31
 Max 1, treating q2 as a constant
@ 1
= a bq2 2bq1 c=0
@q1
2bq1 = a c bq2
a c q2
q1 = ! "Best Response Function"
2b 2
Best Response Function: Tells …rm 1 the pro…t maximizing q1 for
any level of q2
For Firm 2:

2 = (a bq1 bq2 )q2 cq2 K

Max 2 (treating q1 as a constant) gives

a c q1
q2 = Firm 2’s Best Response Function
2b 2
The two "Best Response" Functions
a c q2
(1) q1 = 2b 2
a c q1
(2) q2 = 2b 2

gives us two equations and two unknowns.

The solution to this system of equations is the equilibrium to the
"Cournot Duopoly" game
Using Cramer’s Rule:

(1) q1 = a3bc
(2) q2 = a3bc
2(a c)
Market Output : q1 + q2 =
3b
Best Response Functions Graphically

32
2.4 Stackelberg Duopoly
In the Cournot Duopoly , 2 …rms picked output simultaneously. Sup-
pose …rm 1 was able to choose output …rst, knowing how …rm 2‘s output
would vary with …rm 1‘s output.

2.4.1 Firm 1‘s Max Problem

M ax q1 : (a bq1 bq2 )q1 cq1 K
Subject to:
a c q1
q2 = f2’s Response Functiong
2b 2

33
Sub in for q2

a c q1
M ax q1 : aq1 bq12 bq1 cq1 K
2b 2
@ 1 a c
= a 2bq1 + bq1 c=0
@q1 2b
a c
q1 =
2b
Firm 2:
a c q1
q2 =
2b 2
Sub in
a c
q1 =
2b
a c 1 a c a c
q2 = =
2b 2 2b 4b

34
Graphically: Stackelberg and Cournot Equilibrium

35