The   U nofficial   S olution   M anual to

 A Primer in Game Theory

by RA Gibbons
Unfinished Draft

N avin   K umar
Delhi School of Economics
 

This version is an unreleased and unfinished manuscript.

The author can be reached at navin.ksrk@gmail.com

Last Updated: January  20 ,  2013

  A
Typeset using L AT  X and the Tufte book class.
E

This work is not subject to copyright. Feel free to reproduce, distribute

or falsely claim authorship of it in part or whole.
 

This is strictly a beta version. Two thirds of it are missing and there are er-
rors aplenty. You have been warned.

On a more positive note, if you do find an error, please email me at

navin.ksrk@gmail.com, or tell me in person.

- Navin Kumar
 

Static Games of Complete Information

A nswer 1 . 1  See text.

A nswer 1 . 2  B is strictly dominated by T. C is now strictly dom-

inated by R. The strategies (T,M) and (L,R) survive the iterated
elimination of strictly dominated strategies. The Nash Equilibrium
are (T,R) and (M,L).

A nswer 1 . 3  For whatever value Individual  1  chooses (denoted

 by S
 by  S 1 ), Individual  2 ’s best response is S is  S 2   =   B2 (S1 ) =   1  S 2 . −
Conversely, S
Conversely,  S 1   =   B1 ( S2 ) =   1 S1 . We know this because if  S  S 2   <
1 S1 , then there is money left on the table and Individual  2  could
− −
increase his or her payoff by asking for more. If, however, S
however,  S 2   >
−
1 S1 , Individual  2  earns nothing and can increase his payoff by
reducing his demands sufficiently. Thus the Nash Equilibrium is
S1 + S2  =  1.

A nswer 1 . 4  The market price of the commodity is determined by

formula  P  =   a Q  in which Q
the formula P − determined  Q  =   q1 +  ... + q n
which  Q  is determined Q
by  C i   =   c qi . The
The cost for an individual company is given by C ·
profit made by a single firm is

π i  = ( p − c) · qi  = (a − Q − c) · qi  = (a − q1∗ − ... − q∗n − c) · qi

 q j∗ is the profit-maximizing quantity produced by firm j
Where q
Where firm  j  in
equilibrium. This profit is maximized at
dπ i
= ( a q∗1 ... qi∗ ... q∗n
−  − −  − −  − c) − qi∗  = 0
dqi
⇒  a − q∗1 − ... − 2 · qi∗ − ... − q∗n − c =  0
⇒  a − c =  q 1∗ + ... + 2 · q∗i   + ... + q∗n
For all i=1, ... ,n. We could
We  could solve
 solve this question using matrices and
Cramer’s rule, but a simpler method would be to observe that since
all firms are symmetric, their equilibrium quantity will be the same
i.e.

q1∗   =  q 2∗   =  ...  =  q ∗i   =  ...  =  q ∗n

Which means the preceding equation becomes.

a − c = (n + 1)qi∗  ⇒ q ∗i   =   na +− c1

 

6   Static Games of Complete Information

A similar argument applies to all other firms.

A nswer 1 . 5  Let  q m  be the amount produced by a monopolist.

Thus, if the two were colluding, they’d each produce

q1m  =  q 2m  =
  qm   a c
 =
−
2 4
In such a scenario, the profits earned by Firm  1  (and, symmetrically,
Firm  2 ) is

1
π mm  = (P − c) · q1m   =
( a − Q − c) · a q1m  =
−  a

−  c − c · q1
m
2
  a − c   a − c   ( a − c )2
=  ·  =   =  0.13 · (a − c)2
2 4 8
If both are playing the Cournot equilibrium quantity, than the prof-
its earned by Firm  1  (and Firm  2 ) are:   1   1
The price is determined by
P  =  a −Q
1
π cc  = (P − c) · q1c   = (a −  2 (a 3− c ) − c) · q1c   = ( a +3 2 c − c) · q1c =  a − q1cc − q2cc
 a − c
  ( a − c )2 =  a − 2 ·
= 3
9   a + c
2 =
=  0.11 (a c) 2

· −
What if one of the firms (say Firm  1 ) plays the Cournot quantity
and the other plays the Monopoly quantity?   2 Firm  1 ’s profits are:   2
The price is given by
P  =  a−Q
1
π cm  = (P
− · c) q1c   = (
5a   7c
 +
12 12
 − · −3 c  =  365  · (a − c)2
c)
 a
=  a − q1c − q2m
 a − c   a − c
=  0.14 · ( a − c)2 =  a −  − 4
3
 7
And Firm  2 ’s profits are: =  a −  · ( a − c )
12
  5 a   7c
=  +
2
π cm  = ( P
5 a   7c
c) q2m  = (  +
− · 12 12
 − c) · −4 c  =  485  · (a − c)2
 a 12 12

2
· −
=  0.10 ( a c)

For notational simplicity, let

α ≡ ( a − c)2
Their profits are reversed when their production is. Thus, the pay-
offs are:

Player  2
qm   qc
qm   0.13α, 0.13α   0.10α, 0.14α
Player 1
qc   0.14α, 0.10α   0.11α, 0.11α

As you can see, we have a classic Prisoner’s Dilemma: regardless of 

the other firm’s choice, both firm’s would maximize their payoff by
choosing to produce the Cournot quantity. Each firm has a strictly
qm
dominated strategy ( 2  ) and they’re both worse off in equilibrium
(where they make 0.11 ( a c)2 in profits) than they would have
· −
 

 been had they cooperated by producing q

producing  q m  together (which would
have earned them 0.13 ( a c)2 ). · −
by  P  =  a Q  and
A nswer 1 . 6  Price is determined by P  Q  is determined
 and Q −
 Q  =  q 1 + q2 . Thus the profit generated for Firm  1  is given by:
 by Q
 by

π 1  = (P − c1) · q1  = (a − Q − c1 ) · q1  = (a − q1 − q2 − c1) · q1

At the maximum level of profit,
dπ 1
dq 1
− − −
= ( a c1 q1 q2 ) + q1 ( 1) =  0 ·−  ⇒ q 1 =   a − c21 − q2
And by a similar deduction,

q2  =
 a − c2 − q1
2
Plugging the above equation into it’s predecessor,

q1  =
 a − c1 −   a−c2 −q 2 1
=
 a − 2c1 + c2
2 3
And by a similar deduction,

q2  =
 a − 2c2 + c1
3
Now,

2c2   > a + c1  ⇒ 0 > a − 2c2 + c1 ⇒ 0 >

a − 2c2 +  c1 ⇒ 0 > q2
3
⇒ q 2  =  0
Since quantities cannot be be negative. Thus, under certain condi-
tions, a sufficient difference in costs can drive one of the firms to
shut down.

A nswer 1 . 7  We know that,

− a  p i   if   pi   <   p j ,
qi  = −
a  pi
  if   pi   =   p j ,
2

0 if   pi   >   p j .

We must now prove  prove   pi   =   p j   =   c  is the Nash Equilibrium of this
game. To this end, let’s consider the alternatives exhaustively.
If   pi   >   p j   =   c,  q i   =  0 and   π  j   =  0. In this scenario, Firm j can
increase profits by charging  charging   p j  +  ε  where   ε   >  0 and   ε   <   pi  p j ,  −
raising profits. Thus this scenario is not a Nash Equilibrium.
If   pi   >   p j   >   c,  q i   =  0 and   π i   =  0 and Firm i can make pos-
itive profits by charging   −
charging   p j ε   >   c. Thus, this cannot be a Nash
Equilibrium.
If   pi   =   p j   >   c,  π i   = ( pi c) −2  i . Firm i can increase
  a  p
 − · increase profits
profits by by
 −
charging   pi ε  such that 
charging  −
that   p j   >   pi ε > 0, grab the entire market and
a larger profit, provided

( pi  − ε − c) · (a − pi − ε) > ( pi  − c) ·  ( a −2 pi )

 

8   Static Games of Complete Information

Therefore, this is not a Nash Equilibrium

If   pi   =   p j   =   c, then   π i   =   π  j   =  0. Neither firm has any reason
to deviate; if FFirm i were to reduce   pi ,   π i  would become negative.
If Firm i were to raise   pi ,  q i   =   π i   =  0 and he would be no better
off. Thus Firm i (and, symmetrically, Firm j) have no incentive to
deviate, making this a Nash Equilibrium.

A nswer 1 . 8  The share of votes received by a candidate is given by

 xi  +   21 ( x j
 ·  − xi )   if   xx i   < x j ,
Si  = 1   if   xx i  =  x j ,
 2
(1 − xi ) +   21 · (xi − x j )   if   xx i  > x j .

We aim to prove the Nash Equilibrium is  ( 12 ,  12 ). Let us exhaustively

consider the alternatives.
Suppose x
Suppose  x i   =   21   and x
and  x j   >   21  i.e. One candidate is a centrist
while the other (Candidate j) isn’t. In such a case, Candidate j can
increase his share of the vote by moving to the left i.e. reducing x reducing  x j .
  1
If   xx j   < 2 , Candidate j can increase his share of the vote by moving
to the right.
Suppose x
Suppose  x i   >   x j   >   21 ; Candidate i can gain a larger share by
moving to a point x point  x j   >  x i   >   21 . Thus this is not a Nash Equilibrium.
Suppose x
Suppose  x i   =   x j   =   21 ; the share of i and j are   21 . If Candidate i were
to deviate to a point (say) x (say)  x i   >   21 , his share of the vote would de-
cline. Thus  ( 12 ,  12 )  is a unique Nash Equilibrium. This is the famous
Median Voter Theorm, used extensively in the study of politics. It
explains why, for example, presidential candidates in the US veer
sharply to the center as election day approaches.

A nswer 1 . 9  See text.

A nswer 1 . 10  (a) Prisoner’s Dilemma

Player  2
− p) Fink 
( p) Mum   (1
(q) Mum   −1, −1   −9, 0
Player  1
(1 − q) Fink    0, −9   −6, −6
Prisoner’s Dilemma

In a mixed strategy equilibrium, Player  1  would choose q

choose  q  such that
Player  2  would be indifferent between Mum and Fink. The payoff 
from playing
playing Mum and Fink must be equal.
equal. i.e.

−1 · q + −9 · (1 − q) = 0 · q + −6 · (1 − q) ⇒ q  = −3.5

This is impossible.
 

(b)

Player 2
Le f t (q0 )   Middle(q1 )   Right(1 − q0 − q1 )
U p( p)   1, 0 1, 2 0, 1
Player  1
Down(1 − p )   0, 3 0, 1 2, 0
Figure  1 .1.1.

Here, Player  1  must set p so that Player  2  is indifferent between

Left, Middle and Right. The payoffs from Left and Middle, for
example, have to be equal. i.e.

 p 0 + ( 1
· − p) · 3 =   p · 2 + (1 − p ) · 1
⇒   p =  0.5
Similarly, the payoffs from Middle and Right have to be equal

·
2  p  + 1 (1· − p) = 1 · p + 0 · (1 − p )
⇒   p = −0.5
Which, besides contradicting the previous result, is quite impossi-
 ble.
(c)

Player  2
L( q0 )   C(q1 )   R(1 − q0 − q1 )
T ( p0 )   0, 5 4, 0 5, 3
Player  1  M( p1 )   4, 0 0, 4 5, 3
B (1 − p0 − p1)   3, 5 3, 5 6, 6
Figure  1 .1.4.

In a mixed equilibrium, Player  1  sets and   p1  so that Player  2

 sets   p0   and 
would be indifferent between L, C and R. The payoffs to L and C
must, for example, be equal i.e.

4  p 0 + 0  p 1 + 5 (1
· · · − p0 − p1) = 0 · p0 + 4 · p1 + 5 · (1 − p0 − p1)
⇒   p0  =   p1
Similarly,

0  p 0 + 4  p 1 + 5 (1
· · · − p0 − p1) = 3 · p0 + 3 · p1 + 6 · (1 − p0 − p1)
⇒   p1  =  2.5 − p0
violates   p0  =   p1 .
Which violates 
 

10   Static Games of Complete Information

A nswer 1 . 11  This game can be written as

Player  2
L( q0 )   C(q1 )   R(1 − q0 − q1 )
T ( p0 )   2, 0 1, 1 4, 2
Player  1  M( p1 )   3, 4 1, 2 2, 3

B (1 − p0 − p1)   1, 3 0, 2 3, 0

 sets   p0  and   p1  so that the

In a mixed Nash Equilibrium, Player  1  sets
expected
expected payoffs
payoffs from L and C are the same. i.e.

E2 ( L) =  E 2 (C)
⇒ 0 · p0 + 4 · p1 + 3 · (1 − p0 − p1 ) = 1 · p0 + 2 · p1 + 2 · (1 − p0 − p1 )
⇒   p1  =  2 · p0 − 1
Similarly,

E2 ( C) =  E 2 ( R)
 1  p 0 + 2  p 1 + 2 (1  p 0  p 1 ) =  2  p 0 + 3  p 1 + 0 (1  p 0  p 1 )
⇒ · · · − −⇒   p1  =   2 −
·  p0 · · − −
3
Combining these,

· − 1 =   32 − p0 ⇒   p0 =   95

2  p 0
 5  1
∴ p1  =  2 ·  p 0 − 1  =  2 ·  − 1  =
9 9
 5  1   3
∴ 1 −  p 0 −  p 1  =  1 −  −   =
9 9 9
calculate  q 0  and  q 1 . Player  2  will set them such that
Now we must calculate q

E1 ( T ) =  E 1 ( M)
 2 q0 + 1 q1 + 4 (1 q0 q1 ) =  3 q0 + 1 q1 + 2 (1 q0 q1 )
⇒ · · · − −⇒ q 1  =  1 −· 1.5 · q·0 · − −
And

E1 ( M) =  E 1 ( B)
⇒ 3 · q0 + 1 · q1 + 2 · (1 − q0 − q1) = 1 · q0 + 0 · q1 + 3 · (1 − q0 − q1)
Qed

A nswer 1 . 12

( q ) L2   ( 1 − q) R 2
( p) T 1   2,  1 0,  2

(1 − p)B1   1,  2 3,  0
 

11

Player  1  will set p such that

E2 ( L) =  E 2 ( R)

⇒ 1 · p + 2 · (1 − p ) = 2 · p + 0 · (1 − p)

 2
⇒   p  = 3
Player  2  will set q such that

E1 ( T ) =  E 1 ( B)

⇒ 2 · q + 0 · (1 − q) = 1 · q + 3 · (1 − q)

⇒  q  =   43
A nswer 1 . 13

(q)Apply1 to Firm  1   (1 − q)Apply1 to Firm 2

( p)Apply2 to Firm  1   1 1
2 w1 , 2 w1   w1 , w 2
(1 − p )Apply2 to Firm 2   w2 , w1 1 1
2 w2 , 2 w2

There are two pure strategy Nash Equilibrium (Apply to Firm  1 ,

Apply to Firm  2 ) and (Apply to Firm  2 , Apply to Firm  1 ). In a
mixed-strategy equilibrium, Player  1  sets p such that Player  2  is
indifferent between Applying to Firm  1  and Applying to Firm  2 .

E2 (Firm 1) =  E 2 (Firm 2)

⇒   p ·  12 w1 + (1 − p) · w1  =   p · w2 + (1 − p) ·  12 w2

⇒   p =   2ww11 +− ww22
2w1   >   w2 , 2w1 w2  is positive and 
Since 2w  − and   p   >  0. For 
For   p   <  1 to be
true, it must be the case that
−
2 w1 w2
< 1  ⇒  21 w1  < w2
w1 +  w2
Which is true. And since the payoffs are symmetric, a similar analy-
sis would reveal that

q  =
−
  2 w1 w2
w1 + w2

A nswer 1 . 14
 

Dynamic Games of Complete Information

A nswer 2 . 1  The total family income is given by

I C ( A) + I P ( A)

This is maximized at
d( I C ( A) + I P ( A))
dA
  =  0  ⇒  d I Cd A( A)  = − dI dP (A A)
The utility function of the parents is given by

V ( I P
 − B) + kU ( I C + B)
This is maximized at

k    +
 d V ( I P B)
dU ( I C  +  B )  dV   −
  =  0
dB db

( I C  +  B )  d ( I C +  B )   dV ( I P − B)   I P − B

⇒ k dU 
d( I C +  B )
  · dB   + d( I P − B)   · dB   = 0

⇒ kU  ( I C + B) − V  ( I P − B) = 0

⇒ V  ( I P − B∗ ) = kU  ( I C + B∗ )
Where  B ∗  is the maximizing level of the bequest. We know it exists
Where B
 because a) there are no restrictions on B and b) V() and U() are
concave and increasing
by  U ( I C ( A) B∗ ( A)) . This is
The child’s utility function is given by U  −
maximized at
dU ( I C ( A) + B ∗ ( A))
  =  0
dA

d I C ( A)  d B∗ ( A)

⇒ U  ( I C ( A) + B∗ ( A)) ·
   +
 =  0
dA dA

∗
⇒   d I CdA( A)  +  d BdA( A)  = 0 ⇒  I C ( A) = −B∗ ( A)
prove  B ∗ ( A) =   I P ( A)
We now have only to prove B

dV ( I P ( A) B∗ ( A

−  )) =  0
dA
 

14   Dynamic Games of Complete Information

 d B∗ ( A)
⇒ V  ( I P ( A)∗B ( A)) ·
 dI P ( A)
 −
 =  0
dA dA

⇒   I P ( A) =  B ∗ ( A)

by  V ( I P
A nswer 2 . 2  The utility function of the parent is given by V   −
B) + k [U 1 ( I C S) + U 2 ( B + S )]. This is maximized at
 −
d · {V ( I P − B) + k [U 1 ( I C − S) + U 2 ( B + S )] }
  =  0
dB

⇒ −V  + [U 1 (−SB ) + U 2 (SB + 1)] = 0 ⇒ V   =  −kU 1 S + U 2 S + U 2
by  U 1 ( I C − S) + U 2 ( S +  B ). This is
The utility of the child is given by U 
maximized at:
d[U 1 ( I C  − S) + U 2 (S +
 
 B )]
=  0  ⇒  U   =  U  (1 + B  )
1 2
dS
Total utility is given by

V ( I  p  − B) + k (U 1( I C − S) + U 2(B + S)) + U 1( I C − S) + U 2(B + S)

 − )(U 1 ( I C
=  V ( I  p B) + (1 + k )(  − S) + U 2(B + S))
This is maximized (w.r.t S) at:

V  ( BS ) + (1 + K ) [U 1 ( 1) + U 2 (1 + BS )] =  0

 · − · −
 
⇒ U 1  =  U 2 (1 + BS ) −  V 1 +B k S
to  U 1  =  U 2 (1 + BS ), which is the equilibrium condition.
as opposed to U 
  V  BS
Since 1+k    >  0, equilibrium  U 1  is ‘too high’ which means that S
the equilibrium U 

- the level of savings - must be too low (since   dU 
dS   <  0). It should be
higher.

A nswer 2 . 3  To be done

suppose  c 2   =   R c1 , partner  2 ’s payoff is V 

A nswer 2 . 4  Let’s suppose c − is  V   −
− 2
( R c1 ) . If   cc 1   R, partner  2 ’s best response is to put in  0  and
 ≥
pocket V. If   cc 1   <   R  and partner  2  responds with some c
some  c 2  such
that c
that − 2  −
 c 2   <   R c1 , his payoff is c2 . If he puts in nothing, he will,
receive a payoff of zero. There is, therefore, no reason to put in such
a low amount. There is - obviously - also no reason to put in any
 −
c2   >   R  c 1 . He will put in R  −
in  R  c 1  if it’s better than putting in
nothing i.e.
√ 
− (R − c1)2 ≥ 0 ⇒ c 1 ≥ R − V 
V 

For player  1 , any c

√    is dominated by c
any  c 1  R − V  is >
√ 
by  c 1  =  R − V . Now,
player  1  will do this if the benefit exceeds the cost:

δV   ≥ ( R − √ V )2

 ≥
 

15

  2
 R
⇒  ≥ √  −
 δ
V 
1

R2
If   R ≥   4V ,   δ would have to be greater than one, which is impossi-
 2 
 R
 ble. Therefore, if   δ ≥ √   − 1 and  R 2 ≥   4V  (i.e.
and R   (i.e. the cost is not
√  V  √ 
high’),  c   =   R − V  and
‘too high’), c 1 Otherwise,  c   =  0 and
 and  c   = V . Otherwise, c
2 2
c1  =  0.

A nswer 2 . 5  Let the ‘wage premium’ be be   p   =   wD  w E , where  −

 p   ( ∞, ∞ ). In order to get the worker to acquire the skill, the
 ∈ −
firm has to credibly promise to promote him if he acquires the skill
- and not promote him if he doesn’t.
Let’s say that he hasn’t acquired the skill. The firm will not
will  not pro-
 pro-
him  iff  the
mote him iff   the returns to the firm are such that:

 yD0 − wD ≤ y E0 − wE ⇒ y D0 − yE0 ≤ w D − wE   =   p

he  does acquire
If he does  acquire the skill, the firm will promote if the returns to
the firm are such that:

 y DS  − wD ≥ y ES − wE ⇒ y DS − yES ≥ w D − wE  =   p

Thus the condition which the firm behaves as it ought to in the
desired equilibrium is:

 yD0 − yE0 ≤  p ≤ y DS − yES

Given this condition, the worker will acquire the promotion  iff    iff   he
acquires
acqu ires the skill. He will acquire skill  iff  the
acquire the skill iff   the benefit outweighs
the cost i.e.

wD  − C ≥ w E ⇒ w E +  p − C ≥ w E ⇒   p ≥ c

That is, the premium paid by the company must cover the cost of 
training. The company wishes (obviously) to minimize the pre-
mium, which occurs at:
 C   if   C
C ≥ y D0 − yE0,
wD  − wE  =   p =  C  y D0 − y E0
 y D0 − y E0   if   C <

A final condition is that the wages must be greater than the alterna-
tive i.e.   ≥
i.e.   wE  0 and w  ≥
and  w D  0. The firm seeks to maximize y
maximize  y ij wi ,  −
which happens at wat  w E  =  0 and w
and  w D  =   p.

A nswer 2 . 6  The price of the good is determined by

P( Q) =  a − q1 − q2 − q3
The profit earned by a firm is given by   π i   = ( p − c) · qi . For Firm 2 ,
for example

π 2  = (a − q1∗ − q2 − q3 − c) · q2
 

16   Dynamic Games of Complete Information

which is maximized
maximized at
dπ 2
dq 2
= (a − q1∗ − q2 − q3 − c) + q2 (−1) = 0

⇒  q 2 =  a − q1∗ − q3 − c
2
Symmetrically,

q3  =
 a − q1∗ − q2 − c
2
Putting these two together,

q2  =
 a − q1∗ −  a−q∗ −2 q −  c − c ⇒ q  =   a − c − q1∗
1 2

2
2 3
to  q 3 .
Which, symmetrically, is equal to q

∴  π 1  = ( a − q1 − q2 − q3 − c ) · q1  = ( a − q1 − 2 ·
− c −  q1 − c) · q1  =  a
 − − ·
a q1 c
q1
3 3
This is maximized at
dπ 1  a q1 c q1  a c
dq 1 = − 3 −  + −3   = 0 ⇒ q 1∗  = −2
Plugging this into the previous equations, we get

q2  =  q 3  =
 a −c
6

A nswer 2 . 7  The profit earned by firm i is

π i  = ( p − w) · Li
Which is maximized at
dπ i   d( p w) Li   d( a
− · − Q − w) Li = 0
= =
dL i dL i dL i

⇒   d(a − L1 − . . . − LdLi −i . . . − Ln − w) · Li = 0

⇒  ( a − L1 − . . . − Li − . . . − Ln − w) + Li (−1) = 0

⇒  L 1 + . . . + 2Li + . . . + Ln  =  a − w ∀ i  =  1 1,, . . . , n
This generates a system of equation similar to the system in Ques-
tion (1.4)
 2 1 ... 1
 L1
 − a w

 1 2 ... 1
 L2
 − a w

.. ..   . .   .. · .. = ..
. . . . . .
 1 1 ... 2
  Ln
  −a w

 

17

Which resolves to

Li  =
−
 a w
n + 1
Thus, total labor demand is given by

L  =  L 1 +  L 2 + . . . + L n  =

 a w −   + . . . +
 a w
 =
−
  n
(a  · − w)
n + 1 n + 1 n + 1

The labor union aims to maximize

U  =
 = ( w − wa ) · L = (w − wa ) · n   +n 1 · (a − w) = n   +n 1 · (aw − awa − w2 + wwa )
The union maximize U by setting w. This is maximized at

dU 
dw
  = (a − 2w + wa ) · n   +n 1   = 0 ⇒ w  =   a +2 wa
The sub-game perfect equilibrium is  and  w   =   a+2w a .
is   Li   =   an−+w1   and w
Although the wage doesn’t change with n
with  n,, the union’s utility (U  ( U )
  n
is an increasing function of  n+1 , which is an increasing function
of  n.
 n . This is so because the more firms there are in the market, the
greater the quantity produced: more workers are hired to produce
this larger quantity, increasing employment and the utility of the
labor union.

A nswer 2 . 8

A nswer 2 . 9  From section  2 .2.C, we know that exports from coun-

try i can be driven to  0  iff 

e∗i   =
a − c − 2 t j = 0 ⇒ t  =   a − c
 j
3 2
which, symmetrically, is equal to t model  c   =
to  t i . Note that in this model c
wi  and we will only be using  w i  from now, for simplicity. What
happens to domestic sales?

 a w  +  t  a wi  +   a−wi   a w
 h i  = i   i =
3  2 = 2 i
∴
− 3 − −
Which is the monopoly amount. Now, in the monopoly-union bar-
gaining model, the quantity produced equals the labor demanded.
Thus,

Lei = 0 =
 a − wi
2
The profit earned by firm i in this case is given by
 −  −  −  −   −  2
e =0 a wi a wi a wi
π i   = ( pi  − wi )hi  = (a − wi − hi )hi  = a wi =
2 2 2

The union function’s payoff is defined by

2
U e=0 = ( wi − wa −2wi   + wi wa
 − wa )Li  = (wi − wa ) a −2 wi   =   wi a
 

18   Dynamic Games of Complete Information

The union sets wages to maximize utility with the following condi-
tion:
dU e=0   a − 2wi + w  a = 0 ⇒ w i   =   a + wa
=
dwi 2 2
situation,  t j  =  0 and therefore
Now, tariffs decline to zero. In this situation, t

h j  =  h i  =
 a − wi
3
and

e j  =  e i  =
 a − wi
3
Due to this, prices fall:

Pi  =  P j   =  a − Q =  a − hi − e j  =  a −   a −3 w i −   a −3 w i =   a +32wi

So what happens to profits?

a + 2wi a + 2wi

t =0
π i   = ( pi  − wi )hi + ( p j − wi )e j  =
   − wi
 −   a wi
+   − wi
 − 
a wi
3 3 3 3

2
t= 0 a wi e =0
  π i   =  2 =  2
⇒ −  3 · π i
Thus, profits are higher at zero tariffs. What happens to employ-
ment?

Lit= 0 =  q i  =  h i  + ei  =
 2
(a  · − wi ) =
 4
 ·
 − 
a wi
=
  4 e =0
 ·
L
3 3 2 3 i
Employment rises. And what happens to the wage? That depends
on the payoff the union now faces:

U t=0 = (wi  − wa ) · Li   =   32 (wi − wa )( a − wi ) =   32 · (wi a − wa a − w2i   + wa wi )

This is maximized at
dU t=0
 2   a + w a
= ( a 2w  +  w ) =  0  w  =
dwi 3 i a −  ⇒ i 2
Which is the same as before.

A nswer 2 . 10  Note that  ( P1 , P2 ),  ( R1 , R2 )  and  ( S1 , S2 )  are Nash

Equilibrium.

P2   Q2   R2   S2
P1   2,  2   x,  0   -1,  0 0,0

Q1   0, x   4,  4   -1,  0 0,  0

R1   0,  0 0,  0 0,  2 0,  0

S2   0, -1 0, -1   -1, -1 2,  0

So what are player  1 ’s payoff from playing the strategy? Let PO

Let  PO ji ( X 1 , X 2 )
denote player i’s payoff in round j when player  1  plays  X 1   and
 plays X 
 

19

player  2  plays  X 2 . If player  2  doesn’t deviate, he earns the sum of 
 plays X 
payoffs from two rounds of gaming:

PO11 ( Q1 , Q2 ) + PO12 ( P1 , P2 ) =  4 + 2  =  6

And if player  2  deviates from the strategy, player  1  earns:

PO11 ( Q1 , P2 ) + PO12 (S1 , S2 ) =  0 + 2  =  2

Now, let’s look at his payoff from deviating, when  2  doesn’t:

PO11 ( P1 , Q2 ) + PO12 ( R1 , R2 ) =  x  + 0  =  x

And when they both deviate:

PO11 ( P1 , P2 ) + PO12 ( P1 , P2 ) =  2 + 2  =  4

Thus, if player  2  deviates, player  1 ’s best response is to deviate for

a payoff of   44  (as opposed to the  2  he’d get when not deviating). If,
however,, player 2  doesn’t deviate, player  1  gets a payoff of   66  when
however
playing the strategy and x when he doesn’t. Thus, he (player  1 )
will play the strategy i
strategy  i f f x   <  6. A symmetric argument applies to
player  2 .
Thus the condition for which the strategy is a sub-game perfect
Nash Equilibrium is

4<x < 6

A nswer 2 . 11  The only pure-strategy subgame-perfect Nash Equi-

librium in this game are  ( T , L)  and  ( M
 M,, C).

L C R
T   3 ,1 0 ,0 5,0

M   2 ,1 1 ,2 3,1

B   1 ,2 0 ,1 4,4

Unfortunately, the payoff (4,4) - which comes from the actions (B,R)
- cannot be maintained. Player  2  would play R, if player  1  plays B -
player  1  however would deviate to T to earn a payoff of  5 . Consider,
however, the following strategy for player  2 :
• In Round  1 , play R.
• In Round  2 , if Round  1  was (B,R), play L. Else, play M.
Player 1 ’s best response in Round  2  is obviously T or M, depend-
ing on what player  2  does. But what should he do in Round  1 ? If 
he plays T, his playoff is:   3   3
If you do not understand the nota-
tion, see Answer  2 .10
PO11 ( T , R) + PO12 ( M
 M,, C) =  5 + 1 =  6

If he plays B:

PO11 ( B, R) + PO12 ( T , L) =  4 + 3 =  7

Thus, as long as player  2  is following the strategy given above, we

can induce (B,R).
 

20   Dynamic Games of Complete Information

To get a intuitive idea of how we constructed the strategy, note

that there are two Nash equilibrium that player  2  can play in the
second round: a "reward" equilibrium in which player  1  gets  3  and
"punishment" equilibrium in which player  1  gets  1 . By (credibly)
threatening to punish player  1  in Round  2 , player  2  induces "good"
 behavior in Round  1 .

A nswer 2 . 12  See text.

A nswer 2 . 13  The monopoly quantity is   a− c

2  . The monopoly price
is, therefore:

P  =  a − Q =  q −
 −  a c
=
  a + c
2 2

The players can construct the following strategy.

• play   pi  =   a+
 In Round  1 , play  2
c

 In Round t =1, if the previous Round had   p j =   a+

•  c
play   pi   =  c
2   , play   
  a+ c
(the Bertrand equilibrium). Else, play   pi  = 2
Now, if player i deviates by charging   a− 2
c
ε  where   ε   >  0, he  −
secures the entire market and earns monopoly profits for one round
and Bertrand profits (which are  0 ) for all future rounds which totals
to

π deviate  =
−
a
 a
 
− c  − c · a − c +  δ 0 +  δ 2 0 + . . .  =
· ·   (a − c)2
2 2 4

The payoff from sticking to the strategy (in which both firms pro-
duce   a− c
4  ) the payoff is
−  a
   · −
−  c − c · a − c  a
 
−  c − c · a − c
lo w  =
π  f ol low a +  δ a + . . .
2 4 2 4

=
 · − · −   ·
 1 a c a c
=
 1  ( a − c )2
1 −δ 2 4 1 −δ 8

The strategy is stable if 

π deviate  ≤   π  f ol low

2  · − c )2
⇒ −4 c)  ≤
  (a
1
 1
−δ
 ( a
2

⇒   δ ≥  21
Q.E.D.

A nswer 2 . 14  The monopoly quantity when demand is high is

a H −c
2   which makes
makes the monopoly price

=   a H  +  c
 p H   =  a H   −  a H  c
2
 −  2
 

21

This is the price that the firms have to maintain when demand is
high. Conversely, when demand is low
  a L + c
 p L  =
2
Let   p M  be the monopoly price, defined as
Let

 p  a   =  a
  if  a
 p M  =
  H 
 p L  a i   =  a L
  if  a
i  H 

Consider the following strategy for firm i:

• set   pi  =   p M .
 In Round  1 , set 
• Round  t =  1, if   p j  =   p M  in the previous round, play   p M , else
 In Round t  
play   pi  =  c
play
The payoff received from deviating is the monopoly profit for
one round  4 and then zero profits in all future rounds:   4
See the previous question for the
derivation of one round monopoly

π deviate  =
  ( ai  − c)  2 + δ · 0 + δ2 · 0 + . . . =   (ai − c)2 profits
4 4
If he follows the strategy, he earns:
2 2 2
π  f f ol low  =   ( ai  −8 c)  +  δ ·  ( a H  −8 c ) + (1 − π ) ·  ( aL −8 c) + . . .
·  π 

⇒   π  f ol low  =   ( ai  − c)  2 +   δ  ·  π  ·  ( a H   − c )2 + (1 − π ) ·  ( aL − c)2 
8 1−δ 8 8

The strategy is stable if 

π deviate  ≤   π  f ol low

2 2  − c )2 + (1 − π ) ·  ( aL − c)2
⇒  −4 c)  ≤   (ai −8 c)  + 1  −δ δ ·
  ( ai  ( a H 
 π  ·

8 8

A nswer 2 . 15  If the quantity produced by a monopolist is   a− c

2   , the
quantity produced by a single company in a successful cartel is

qm
 a −c
n =
2n
Therefore, the profit earned by one of these companies is
−   −   ·  −  2
π m  = ( p c )q m c )q m
 a − c  − c a c  1 a c
− n = (a −Q− n = a =
2 2n n 2
−c  which means
The Cournot oligopoly equilibrium quantity5 is   1a+   5
See Answer  1 .4
n
that the profit earned at this equilibrium is
 −  − − · −   −  2
a c a c a c
π c  = (a −Q− c)qcn  = a n c =
1 + n 1 + n 1 + n

A grim trigger strategy for a single company here is

• produce  q m
 In Round t=1, produce q n
 

22   Dynamic Games of Complete Information

• In Round t
Round  t 1, if the total quantity produced in  t 1 is
 >  n qm
is n n , − ·
produce q
produce m
 q n , else produce q c
produce  q n .
Now, the best response to everyone else producing  q m n   is deter-
mined by finding q 
finding  q  which renders profit

 
π   = ( a − Q − c) q  =
− a
 a − c  · (n − 1) − q − c  q  =
  n + 1
(a   · − c)q − q2
2n 2n

Which is maximized at
dπ    n + 1
dq 
 =
2n
(a   · − c) − 2q  =  0 ⇒ q    =   a +
4n
1
  ( a − c)

at  q   (cheating gain) is
The profit at q
−−     2

π   = a c
 a − c  · (n − 1) −   n + 1  · (a − c) · n + 1
(a   · − c) =
n + 1
(a   · − c)
2n 4n 4n 4n
If the firm deviates, it earns the cheating gain for one round and
Cournot profits for all future rounds i.e. the gain from deviating
from the strategy is
      2 2 2
π deviate  =
n + 1
(a  · − − c)
− +  δ
a c
+ δ 2
a c
+ . . .
4n 1 + n 1 + n

    
n + 1 2
  δ  1 2
⇒   π deviate = − + (a c)2
4n 1 +  δ 1 + n

If the firm follows the strategy, it’s payoff is   π m  for all rounds:
− 2 − 2 − 2 − 2
 1 a c  1 a c 2  1 a c   1  1 a c
lo w  =
π  f ol low
n 2
+  δ · n 2
+  δ · n 2
+ . . .  =
1 − δ · n 2
The strategy is stable if 

π  f ol low  ≥   π deviate

2 2 2
  1  1 a −c n + 1   δ   1
+ (a c )2
⇒ 1 − δ · n 2
  ≥   4n 1 −δ  n + 1 −
∗   n2 + 2n + 1
⇒   δ  ≤ n2 + 6n + 1
Thus as n  rises,   δ∗  falls. If you want to know more, see Rotember
as  n rises,
and Saloner (1986).

A nswer 2 . 16

A nswer 2 . 17

A nswer 2 . 18  See text.

A nswer 2 . 19  In a one period game, player  1  would get a payoff 

of   11  and player  2  would get a payoff of  0
 0  i.e. (1,0). In a two period
 

23

game, if the two players can’t agree, the game goes to the second
stage, at which point, player  2  gets  1  and player  1  gets  0 . This pay-
off of   11  in the second round is worth   δ  to player  2  in the first round.
If player  1  offers   δ  to player  2  in the first round, player  2  will accept,
− −
getting a payoff of 1 δ  i.e. (1 δ,   δ).
In a three period game, if player  2  rejects the offer in the first
round, they go on to the second round, at which point it becomes
a two period game and player  2  gets a payoff of 1 δ  and player  1
−
gets   δ. This payoff ( δ,1 δ) is worth ( δ2 ,δ[1 δ]) to the players in
− −
the first round. Thus, if player  1  makes an offer of (1 δ[1 δ],δ[1
− − −
δ ])=(1− δ +  δ 2 ,δ
− δ2 ), player  2  would accept and player  1  would
secure a higher payoff.

A nswer 2 . 20  In round  1 , player A offers  ( 1  −

1 ,   δ )  to player B,
δ 1+ δ
  δ ∗
which player B accepts since 1+δ   δs  =   δ 1−δ .
 ≥  1

What if A deviates from the equilibrium, offers less and B re-

fuses? The game then goes into the next round and B offers A  1+δ δ ,
which will be accepted, leaving  1+1 δ  for B. This is worth   δ  1+ · 1
δ
  to
B in the first round (which is why B will refuse anything less than
2
this amount) and  1δ+δ  to A in the first round. This is less than the
1
δ
  A  would  have made had he not deviated.
 would have
1+

A nswer 2 . 21

A nswer 2 . 22  In the first round, investors can either withdraw (w)
as  x 1 x2  where
or not (d). A strategy can represented as x  x 1  is what the
 where x
and  x 2  is what the investor does in
investor does in the first round and x
the second round. The game can be represented by the following
table:

ww wd dd dw
ww r,r r,r D,2r-D D,2r-D
wd r,r r,r D,2r-D D,2r-D
dd   2r-D,D   2r-D,D R ,R D,2R-D
dw   2r-D,D   2r-D,D   2R-D,D R,R

There are  5  Nash Equilibria: (dw,dw) , (ww,ww) , (ww,wd) , (wd,ww)

and (wd,wd). Of these, (ww,wd), (wd,ww) and (wd,wd) are not
Subgame Perfect Equilibria since there is no subgame in which both
or either player doesn’t withdraw his or her funds in the second
round.

A nswer 2 . 23  The optimal investment is given by

d(v +  I   p  I 2 )
− −  1
  =  0   I ∗  =
 ⇒
dI  2
The boost in the value added is If the buyer had played ‘Invest’,
 buyer will buy if 

− − I 2 ≥ −I 2 ⇒   p ≤ v + I 

v +  I   p
 

24   Dynamic Games of Complete Information

Thus, the highest possible price that the buyer will pay at this point
is   p   =   v +  I .
is   I . If, however, the buyer doesn’t invest, the buyer will
 buy if 

v − p ≥ 0 ⇒ v ≥  p
Thus the buyer would be willing to pay   p   =   v. Thus investment is
I  0,  12 . The price is drawn from 
from   p v, v +   21 . There is no gain
 ⊆ { }
from charging anything other than these prices.  ⊆ { }
v +   21   v
1,
2  A   − 14 , v +   21 14 , v
1
2, R   − 14 , 0   − 14 , 0
0,  A   − 12 , v +   21   0, v
0, A
0, R   0, 0
0, 0, 0

if   I   =   21 , A weakly

As you can see from this (complicated) table, if  
dominates R. And if   I   =  0 R weakly dominates A. Thus we can
collapse the above table into a simpler one:

(q)v +   21   (1 q)v
−
=   12 Accept
( p) I   =   − 1 ,  +   1
4 v 2
1,
4 v

(1 − p ) I   ==  0 Reject   0, 0 0, 0

The only pure Nash Equilibrium is for the buyer to not invest and
the
 

Static Games of Incomplete Information

A nswer 3 . 1  See text

A nswer 3 . 2  Firm  1  aims to maximize

π 1  = ( p − c)q1 = (ai − c − q1 − q2 )q1

Which is done by

dπ 1
dq1
=  a i  − c − q1 − q2 + q1(−1) = 0 ⇒  q 1  =   ai − c2 − q2
Thus, the strategy for firm  1  is
 −−
a H  c q2
   a i   =  a H 
if  a
2
q1  = −−
a L c q2
 2    a i   =  a L
if  a

Now, the firm  2  aims to maximize

π 2  = ( p − c)q2  = (a − c − q1 − q2)q2

This is maximized at
dπ 2
dq 2
=  a − c − q1 − q2 + q2 (−1) = 0 ⇒ q 2  =   a − c2− q  1 =   θ a H  + ( 1 − θ2)aL − c − q1
in  q 1 , we get
Plugging in q

θ a H  + ( 1 − θ)aL − c −
 θ a H +( 1
− θ ) a L − c − q2
2

q2  =
2

⇒ q 2  =   θ a H  + ( 1 −3 θ)aL − c

 a i   =
Now, we need to find out what firm  1 ’s output would be. If  a
a H ,

  a H   − c −   θ a H +( 1 −θ ) a L −c   (3 − θ)a H  − (1 − θ)aL − 2c

q H 
1 =
3
  =
2 6
But what if   aa i  =  a L ?
 θ

q1L  =   a
L  − c − −θ  −
a H +(13 )a L c
=   (2 + θ )a L  − θ a H  − 2c
2 6
 

26   Static Games of Incomplete Information

Now, based on these results, the constraints for non-negativity are:

q2  ≥ 0 ⇒  θ a H  + (1 − θ)aL − c ≥ 0 ⇒  θ ≥  a H c − −aaLL

Which also requires

θ ≤ 1 ⇒ 1 ≥ c − aL ⇒ a L ≥ c − 1

Furthermore,

q1L ≥  0 ⇒  (2 + θ )aL −6θ a H  − 2c ≥ ≥ 0 ⇒  θ ≤ 2 ·  a H c − −aaLL

Which subsumes the last-but-one result. And finally,

a H  +  a L
1  ≥ 0 ⇒  θ ≤   2c −a H 3 −
q H 
aL

A nswer 3 . 3  The profits earned by firm  1  is given by

π 1  = ( p1 − c)q1 = ( p1 − c)( a − p1 − b1 p2)

This is maximized at

dπ 1 =  a  + ( 1 − θ ) p L ]
− p1 − b1 p2 +  p1 (−1) = 0 ⇒  p1  =   a − 2b1 p 2 =   a − b1[θ p H  +
d p1 2

Now, what if   bb 1  =  b H ? To start with, 

with,   p1  =   p H   and

 p H   =
 a − b H [θ p H  + ( 1 − θ) pL ] ⇒   p H   =   a − (1 − θ)b H  pL
2 2 + θ b H 

And if   bb 1  =  b L :

 p L  =
 a − bL [θ p H  + ( 1 −  θ) pL ] =   a − θbL p H 
2 2 + b L (1 − θ )

Which means that,

a (1 θ)
−
  a θ bL p H 
2+(1 θ )b L
 p H   = − −2 + θ b H  −
 
⇒   p H   = 4  +a2(1(1−−[1θ)−bLθ +]b H 2)θb H 
Similarly,

 p L  =
  a (1 θ b L )−
−
4 + 2(1 θ )bL + 2θ b H 

A nswer 3 . 4  Game  1  is played with  0 .5  probability:

L R
(q)T   1,1 0 ,0

(1-q)B   0,0 0 ,0
 

27

If nature picks game  2 , which  0 .5  probability:

L R
(q)T   0,0 0 ,0

(1-q)B   0,0 2 ,2

If nature picks game  2 , player  1  will always play B, since it weakly

dominates T and player  2  will play R, since it weakly dominates L.
Now, if nature chooses game  1 , player  1  will play T. If nature
chooses game  2 , player  1  will play B. Furthermore, if player  2  plays
L with probability p:
1  1
π 2  =   p [
2
0+
 · 2
1] + ( 1  · − p)[ 12 · 0 +  12 · 2] = 1 −  12 · p
at   p  =  0 i.e. player  2  will always play R. Thus the
This is maximized at 
Pure-strategy Bayesian Nash equilibrium is

 {
PSNE  = (1, T , R), (2, B, R) }
A nswer 3 . 5

nswer 3 6
A .  The payoff is given by
  −
vi bi   if   bb i   > b j  j  =  1  ∀
 1,, 22,, . . . , i − 1, i + 1 , . . . , n
ui  =
−
vi b j
  if   bb i   =  b j
 m
0 any   j  =  1
if   bb i   < b j  for any   1,, 22,, . . . , i − 1, i + 1 , . . . , n
are:   v j  is uniformly distributed on [0,1]. Actions are
The beliefs are: 
by  b i  [ 0, 1]  and types are given by  v i  [ 0, 1]. The strategy is
given by b  ⊆  ⊆
by  b i   =  a i +  ci vi . Thus, the aim is to maximize
given by b

π i  = ( vi  − bi ) · P ( bi   > b j ∀ j =  1 1,, . . . , i − 1, i + 1 , . . . , n) = (vi − bi ) · [ P ( bi   > b j )]

−
n 1

n 1 −
 bi  − a j
 −
n 1   −  −
bi a j
n 1
⇒   π i   = (vi − bi ) · [ P (bi   > a j  +  c j v j )] = ( vi  − bi ) · P v j   <
c j
= ( vi  − bi ) c j

This is maximized at

dπ i
  −  −bi a j
n 1  −    −  −
n 1 bi a j
n 2

dbi
= ( 1) − · c j
+ ( vi  − bi ) c j c j
=  0

  −  − 
bi a j
n 2
a j + ( n − 1)vi − nbi

· =  0
c j c j

This requires that either

bi  − a j = 0 ⇒ b  =  a
i  j
c j
Or that

a j  + ( n −c1)vi − nbi = 0 ⇒ bi  = an j   +  n −n 1  · vi

 j
 

28   Static Games of Incomplete Information

that  c i  =   n−
that  b i  =  a i  + ci vi . Here, we know that c
Now, we know that b 1
n   and

a j
ai  =
n
  ⇒  a 1  =   an2  =   an3  = . . . =   ann
 a 1  =  a 2  =  . . .  =  a n  =  0. Thus,
Which is only possible if  a

bi  =
 n − 1  · vi
n

A nswer 3 . 7

A nswer 3 . 8
 

Dynamic Games of Incomplete Information

A nswer 4 . 1  (a)

(q)L’ (1-q)R’
L   4 ,1 0 ,0

M   3 ,0 0 ,1

R   2 ,2 2 ,2

The Nash Equilibria are  ( L, L ), ( R, R )  and they are both sub-game

playing  L    is
perfect equilibria. Now, the payoff to player  2  from playing L

π 2 ( L ) =  1  p  + 0 (1  p ) =   p
· · −
playing  R 
The payoff from playing R

π 2 ( R
 ) = 0 ·  p + 1 · (1 − p ) =  1 − p
play  L    if 
Player  2  will always play L

π 2 ( L
 ) > π 2 ( R ) ⇒  p > 1 − p ⇒  p > 1
2
The playoff to player  1  from playing L is

π 1 ( L) =  4 q + 0 (1
· · − q) = 4 4qq
And the payoff from playing M is
π 1 ( M ) =  3 q + 0 (1
· · − q) = 3 3qq
Player  1  will always play L if 

π 1 ( L) > π 1 ( M ) ⇒ 4 4qq > 3q

Which is true. Thus   p  =  1. In which case, player  2  will always play
L . Thus the outcome  ( R, R )  violates Requirements  1  and  2 .
(b)

L    M   R
L   1,3 1 ,2 4 ,0

M   4,0 0 ,2 3 ,3

R   2,4 2 ,4 2 ,4
 

30   Dynamic Games of Incomplete Information

The expected values of the payoffs to player  2  are:

π 2 ( L
 ) =  3 · p + 0 · (1 − p ) =  3
 3 p
 p

π 2 ( M
 ) =  2 · p + 2 · (1 − p ) = 2

π 2 ( R
 ) = 0  p  + 3 (1  p ) =  3 3 p

And the payoffs to player  1  are:

· · − −
π 1 ( R) =  2

π 1 ( L) =  1 q1 + 1 q2 + 4 (1

· · · − q1 − q2) = 4 − 3q1 − 3q2
π 1 ( M) =  4 q1 + 0 q2 + 3 (1
· · · − q1 − q2 ) = 3 + q1 − 3q2
The only Nash Equilibrium is  ( R, M ); it is also sub-game perfect.
To be a Perfect Bayesian Equilibrium, player  2  must believe that

π 2 ( M
 ) > π 2 ( L ) ⇒ 2 > 3 p ⇒  2   >   p
3

and
π 2 ( M
 ) > π 2 ( R ) ⇒ 2 > 3 − 3 p ⇒  p > 1
3
Furthermore, player
player  1  must believe
2
π 1 ( R) > π 1 ( L)  ⇒ 2 > 4 − 3q1 − 3q2 ⇒ q 1  >  − q2
3
Since q
Since  q 1   > 0, this implies that
2
3
 − q2  > 0  ⇒  32  > q2

and

π 1 ( R ) > π 1 ( M )  ⇒ 2 > 3 + q1 − 3q2 ⇒ 3 3qq2 − 1 > q1

Which, in turn, requires

2 5
3q2 −1 >
3
 − q2 ⇒ q 2  >
12
The pure Bayesian Nash equilibrium is
  1 2
( R, M ),   >   p > , 3q2 −1 > q1   >
2
 −  2
q2 ,   > q2   >
5

3 3 3 3 12

A nswer 4 . 2

( q ) L    (1 − q ) R
( p) L   3 ,0 0 ,1

(1  p ) M   0,1 3 ,0
−R   2 ,2 2 ,2
 

31

As you can see, there is no Pure Nash Equilibrium. But, we need

rigorous proof: A pure strategy Nash Equilibrium exists if 
either   L  or   M. For example, player  1
(a) Player  1  always picks either 
will always play L
play  L  if 
1
π 1 ( L) > π 1 ( M )  ⇒ 3 · q + 0 · (1 − q) > 0 q + 3 (1
· · − q) ⇒ q >
2
0.5,   p  =  1.
Thus if   qq > 0.5,
either   L   or
(b) Player  2  always picks either   R  ; player  2  will always play
or R
L   if 

π 2 ( L
 ) > π 2 (R ) ⇒ 0 · p + 1 · (1 − p ) > 1 ·  p + 0 · (1 − p ) ⇒   1   >   p
2
Thus, if   p   <  0.5,
 0.5, q q  =  1. This violates the condition we uncovered in
part (a), proving that there is no PSNE.
In a mixed strategy BE, player  1  plays  plays L L  with probability 
probability   p  and
player  2  plays
 plays L  L   with probability q
probability  q.. In equilibrium, player  2  is
between  L  and  R  :
indifferent between L 
   0  p + 1 (1  p ) = 1  p + 0 (1  p )   p =   1
π 2 ( L ) =   π 2 ( R )  ⇒ · · − · · −  ⇒ 2
And similarly, for player  1 :

π 1 ( L)  ⇒ 3 · q + 0 · (1 − q) = 0 · q + 3 · (1 − q) ⇒ q  =   12

=   π 1 ( M)

Thus, in a mixed strategy equilibrium, player  1  plays

 plays R  with   p  =  0.5
 R  with
and player  2  plays
 plays L  probability  q  =  0.5.
 L  with probability q

A nswer 4 . 3  (a) Let’s start with the pooling equilibrium  ( R, R). In

situation,   p  =  0.5. Now, the payoff to the receiver is
this situation, 

π R ( R, u) · ·
=  0.5 (1) + 0.5 (0) =  0.5

π R ( R, d) ·
=  0.5 (0) + 0.5 (2) =  1·
plays  R,, the receiver will play d
Thus, if the sender plays R play  d.. We have to test
two strategies for the receiver:   ( u, d)  and (d, d). Under the strategy
( d, d)
π 1 ( L, d) =  2 and   π 1 (R, d) =  3

π 2 ( L, d) =  3 and   π 2 (R, d) =  2

There is no incentive for type  1  to deviate and play L

play  L,, but there is
an incentive for type  2  to do so. Under the strategy  ( u, d),

π 1 ( L, u) =  1 and   π 1 ( R, d) =  3

π 2 ( L, u) =  0 and   π 2 ( R, d) =  2

to  L  instead of  R.
Neither type  1  nor type  2  have any reason to L  R . Thus,
we have the following pooling equilibrium:

[( R, R), (u, d), p  =  0.5,1  ≥ q ≥ 0 ]

 

32   Dynamic Games of Incomplete Information

(b) We must find a pooling equilibrium in which the sender

plays  ( L, L, L). For the receiver, the payoffs are

 1  1  1
π R ( L, u) = 1+
 · 1+ · 1  =  1
 ·
3 3 3

 1  1  1
π R ( L, d) = 0+
 · 0+ · 0 =  0 ·
3 3 3
There are two strategies:   (u, u)  and  ( u, d). Under  ( u, u):

π 1 ( L, u) =  1 and   π 1 (R, u) =  0

π 2 ( L, u) =  2 and   π 2 (R, u) =  1

π 3 ( L, u) =  1 and   π 3 (R, u) =  0

send  R  instead of  L.

None of the three types have an incentive to send R  L .
Thus, we have the following equilibrium:

 1
[( L, L, L), (u, u), p  = ,1  ≥ q ≥ 0 ]
3

A nswer 4 . 4  (a) Let’s examine pooling equilibrium  ( L, L).   p   =

0.5.   π R ( L, u) =   π R ( L, d). Thus, it doesn’t matter for the receiver
whether he/she plays u plays  u or or d
 d..
• Under  ( u, u),   π 1 ( L, u) =  1 < π 1 ( R, u) =  2, making it unsustain-
able.
• Under  ( u, d),   π 2 ( L, u) =  0 < π 2 ( R, d) =  1, making it unsustain-
able.
• Under  ( d, d),   π 2 ( L, d) =  0 < π 2 ( R, d) =  1, making it unsustain-
able.
• Under  ( d, u),   π 2 ( L, u) =  0 < π 2 ( R, d) =  1, making it unsustain-
able.
Thus, ( L, L)  is not a sustainable equilibrium.
Let’s examine separating equilibrium  ( L, R). The best response
to this is  ( u, d)6 . Let’s see if either of the types have an incentive to   6
Since  π R (1, L, u)   >   π R (1, L, d)  and
π R ( 2, R, u) > π R (2, R, d)
deviate:
• For type  1 ,   π 1 ( L, u) =  1 > π 1 ( R, d) =  0 i.e. no reason to play R
play  R
instead of  L.  L .
• For type  2 ,   π 2 ( L, u) =  0 <   π 2 ( R, d) =  1 i.e. no reason to play L
play  L
instead of   R. R.
Let’s examine pooling equilibrium  ( R, R).   π R ( R, u) =   1   >
π R ( R, d ) =  0.5. Therefore, the two strategies that can be followed by
the receiver are  ( u, u)  and  ( d, u).
• Under  ( u, u),   π ( L, u) < π ( R, u)  for both types.
• Under  ( d, u),   π ( L, d)   π ( R, u)  for both types.
 ≤
Let’s examine pooling equilibrium  ( R, L). The best response to
this is  ( d, u)7 .   7
Since  π R (1, R, u) =  2 >  π R (1, R, d) =
• For type  1 ,   π 1( L, d) = 2 ≥  π 1(R, u) = 2, i.e. will play 
play   L. 0 and  π R (2, L, u) =  0 < π R (2, L, d ) =  1
• For type  2 ,   π 2( L, d) = 0 ≤  π 2(R, u) = 1, i.e. will play R
play  R..
 

33

Thus the perfect Bayesian equilibrium are:

[( L, R), (u, d), p

 p,, q]

[( R, R), (u, u), p

 p,, q  =  0.5 ]

[( R, R), (d, u), p,

 p, q  =  0.5 ]

[( R, L), (d, u), p

 p,, q]

(b)  Let’s examine pooling equilibrium is  ( L, L).   π R ( L, u) =  1.5   >

(b) Let’s
π R ( L, d) =  1, therefore player  2  will respond to L to  L  with
 with u
 u.. The two
strategies are  ( u, u)  and  ( u, d).
•  Under  ( u, u),   π ( L, u) > π ( R, u)  for both types.
•  Under  ( u, d),   π 1 ( L, u) =  3 < π 1 ( R, d) =  4, making it unsustain-
able.
Let’s examine separating equilibrium  ( L, R). The best response to
this is  ( d, u).
•  For type  1 ,   π 1 ( L, d) =  1 > π 1 ( R, u) =  0 i.e. type  1  will play 
play   L.
  π  < π 
 For examine
•Let’s 2 ( L, d) =  0
type  2 , pooling 2 ( R, u) =  1 i.e. type  2  will play R
equilibrium  ( R, R).   π R (R, u) =   1play  R..
 >

π R ( L, d) =  0.5, therefore player  2  will respond to R to  R  with  u.. The two
 with u
strategies are  ( u, u)  and  ( d, u).
• Under  ( u, u),   π 1 ( L, u) =   3   >   π 1 ( R, u) =  0, making  ( R, R)
unsustainable.
• Under  ( d, u),   π 1 ( L, d) =   1   >   π 1 ( R, u) =  0, making  ( R, R)
unsustainable.
Let’s examine separating equilibrium  ( R, L). The best response to
this is  ( d, u).
• For type  1 ,   π 1 ( L, d) =   1   >   π 1 ( R, u) =   0, making  ( R, L)
unsustainable.
• For type  2 ,   π 2 ( L, d) =  0   <   π 2 ( R, u) =  1, which doesn’t conflict
with the equilibrium.
The perfect Bayesian Equilibria are

[( L, L), (u, u), p  =  0.5, q]

[( L, R), (d, u), p  =  1, q  =  0 ]

A nswer 4 . 5  Let’s examine  4 .3(a). We’ve already tested equilib-

rium  ( R, R). Let’s try another pooling equilibrium  ( L, L).   q   =  0.5.
π R ( L, u) =   1   >   π R ( L, d ) =   0.5. Thus, the receiver’s
receiver’s response
response to
L  will always be ube  u.. We have to test two strategies for the receiver:
(d, u)  and (u, u).
•  Under the strategy  ( d, u),   π ( L, d) =  2 > π ( R, u) =  0 i.e. there is
no incentive for either type to play R play  R  instead of   L.
L.
•  Under the strategy  ( u, u),   π 2 ( L, u) =  0 <   π 2 ( R, u) =  1, making
( L, L)  unsustainable.
 

34   Dynamic Games of Incomplete Information

Let’s examine  ( L, R). The best response to this is  ( u, d). In re-

sponse to this,
•
 For type  1 ,   π 1 ( L, u) =   1   <   π 1 ( R, d) =  3 i.e. type  1  will play R
play  R
which violates the equilibrium
 For type  2 ,   π 2 ( L, u) =   0   <   π 2 ( R, d) =  2 i.e. type  2  will play R
• play  R
which doesn’t violate the equilibrium.
Let’s examine  ( R, L). The best response to this is  ( u, d). In re-
sponse to this,
•
 For type  1 ,   π 1 ( L, u) =  1 < π 1 ( R, d) =  3, i.e. type  1  will play R
play  R..
 For type  2 ,   π 2 ( L, u) =  0 <   π 2 ( R, d) =  2, i.e. type  2  will play R
• play  R,,
violating the equilibrium.
The perfect Bayesian Equilibrium is

[( L, L), (d, u), p,

 p, q  =  0.5 ]

Now, let’s examine  4 .3(b). There is one pooling equilibrium

other than the  ( L, L, L):   ( R, R, R). There are six pooling equilib-
ria:   1.( L, L, R),  2 .( L, R, L),  3 .( R, L, L),  4 .( L, R, R),  5 .( R, L, R)  and
6. ( R, R, L ) .
Let’s start with pooling equilibrium  ( R, R, R).   π R ( R, u) =   32   >
 1
π R ( R, d ) = 3 . Thus, receiver will play the strategy  ( u, u)  or  ( d, u).

 For strategy  ( u, u),   π ( L, u)   <   π ( R, u)  for all types, making it
•
unsustainable.
 For strategy  ( d, u),   π 1 ( L, d) < π 1 ( R, u), making the equilibrium
•
unsustainable. Let’s examine the various separating the equilib-
rium.
1.  ( L, L, R). The best response to this is  ( u, d),   8   8
Since  π R (3, R, u) =  0 <  π R (3, R, d) =
 For type  1 ,   π S (1, L, u) =  1 > π S (1, R, d) =  0 i.e. type  1  will play
•  ·
1 and 0.5  π R (1, L, u) +  0.5  ·
π R ( 2, L, u) =  1 > 0.5 π R (1, L, d ) + 0.5
· ·
L. π R ( 1, L, d) =  0
 For type  2 ,   π S (2, L, u) =  2 > π S (2, R, d) =  1 i.e. type  2  will play
•
L.
•
 For type  3 ,   π S (3, L, u) =  1 < π S (3, R, d) =  2 i.e. type  3  will play
R.
Thus, this is a viable equilibrium.
2  ( L R L)  ( u u)  9   9
since  π  (2, R, u) =  1   >   π  (2, R, d) =
. For, type
• ,  1. The
,   π S (best
1, L,response
u) =  1 > to this is , ,
π S (1, R, u) =  0 i.e. type  1  will play
0 and 0.5R  π R (1, L, u) +  0.5R
 ·  ·
·
π R ( 3, L, u) =  1 > 0.5 π R (1, L, d ) + 0.5 ·
L. π R ( 3, L, d) =  0
•
 For type  2 ,   π S (2, L, u) =  2 > π S (2, R, u) =  1 i.e. type  2  will play
L, instead of   R.
R.
 For type  3 ,   π S (3, L, u) =  1 > π S (3, R, u) =  0 i.e. type  3  will play
•
L.
Thus, this is not a viable equilibrium.
3.  ( R, L, L). The best response to this is  ( u, u),  10   10
since  π R (1, R, u) =  1 < π R (1, R, d) =
•
 For type  1 ,   π S (1, L, u) =  1 > π S (1, R, u) =  0 i.e. type  1  will play 0 and 0.5  π R (2, L, u) +  0.5
 ·  ·
π R ( 3, L, u) =  1 > 0.5 π R (2, L, d ) + 0.5
· ·
L, instead of   R.
R. π R ( 3, L, d) =  0
 For type  2 ,   π S (2, L, u) =  2 > π S (2, R, u) =  1 i.e. type  2  will play
•
L.
 For type  3 ,   π S (3, L, u) =  1 > π S (3, R, u) =  0 i.e. type  3  will play
L. •
Thus, this is a not viable equilibrium.
 

35

4.  ( L, R, R). The best response to this is  ( u, u)  and  ( u, d)   11   11

since  π R (1, L, u) =  1 >  π R (1, L, d) =
Let’s test  ( u, u):  ·
0 and 0.5  π R (2, R, u ) +  0.5  ·
π R ( 3, R, u) =  0.5  =  0.5 π R (2, R, d ) +
·
• For type  1 ,   π S (1, L, u) =  1 > π S (1, R, u) =  0 i.e. type  1  will play ·
0.5 π R (3, R, d) =  0.5
L.
• For type  2 ,   π S (2, L, u) = 2 > π S (2, R, u) =  1 i.e. type  2  will play
L, instead of   R.
R.
 For type  3 ,   π S (3, L, u) =  1
• > π S (3, R, u) =  0 i.e. type  3  will play
L, instead of   R.
R.
Thus, this is not a viable equilibrium.
Let’s test  ( u, d)
 For type  1 ,   π S (1, L, u) =  1 > π S (1, R, d) =  0 i.e. type  1  will play
•
L.
 For type  2 ,   π S (2, L, u) =  2 > π S (2, R, d) =  1 i.e. type  2  will play
•
L, instead of   R.
R.
•
 For type  3 ,   π S (3, L, u) =  1 < π S (3, R, d) =  2 i.e. type  3  will play
R.
Thus, this is not a viable equilibrium.
5.  ( R, L, R). The best response to this is either  ( u, u)  or  ( u, d)   12   12
since  π R (2, L, u) =  1 >  π R (2, L, d) =
Let’s test  ( u, u). 0 and 0.5  π R (2, L, u) +  0.5
 ·  ·
·
π R ( 3, L, u) =  0.5  =  0.5 π R (2, L, d ) +
 For type  1 ,   π S (1, L, u) =  1 > π S (1, R, u) =  0 i.e. type  1  will play
• 0.5 π R (3, L, d) =  0.5
·
L, instead of  R .
 For type  R.
•  2 ,   π S (2, L, u) =  2 > π S (2, R, u) =  1 i.e. type  2  will play
L.
• For type  3 ,   π S (3, L, u) = 1 > π S (3, R, u) =  0 i.e. type  3  will play
L, instead of   R.
R.
Thus, this is not a viable equilibrium.
Let’s test  ( u, d)
•
 For type  1 ,   π S (1, L, u) =  1 > π S (1, R, d) =  0 i.e. type  1  will play
L, instead of   R.
R.
•
 For type  2 ,   π S (2, L, u) =  2 > π S (2, R, d) =  0 i.e. type  2  will play
L.
 For type  3 ,   π S (3, L, u) =  1  =   π S (3, R, d) =  1 i.e. type  3  can play
•
R.
Thus, this is not a viable equilibrium.
6.  ( R, R, L). The best response to this is  ( u, u),   13   13
since  π R (3, L, u) =  1 >  π R (3, L, d) =
 For type  1 ,   π S (1, L, u) =  1 > π S (1, R, u) =  0 i.e. type  1  will play
•  ·
0 and 0.5  π R (1, R, u) +  0.5  ·
π R ( 2, R, u) =  0.5   >  0.5 π R (1, R, d ) +
·
L, instead of   R
R ·
0.5 π R (2, R, d) =  0
 For type  2 ,   π S (2, L, u) =  2 > π S (2, R, u) =  1 i.e. type  2  will play
•
L, instead of   R
R
•
 For type  3 ,   π S (3, L, u) =  1 > π S (3, R, u) =  0 i.e. type  3  will play
L.
Thus, this is not a viable equilibrium.
The only other perfect Bayesian Equilibrium is
  1
 p , q0  = , q1  =
( L, L, R), (u, d), p,
 1

3 3

A nswer 4 . 6  Type  2  will always play R since   π S (2, R, a) > π S (2, R, u)  and   π S (2, R, a) >
π S (2, R, d ). Thus if the Receiver gets the message L
message  L,, he knows that
plays  u 14 , creating   14 in fact,  π R ( x, L, u)   >   π R ( x, L, d)  for
it can only be type  1 . In such a case, the Receiver plays u
 both types, so the Receiver will always
 u
play u
play
 

36   Dynamic Games of Incomplete Information

a payoff of  (
 ( 2, 1). This gives type  1  a higher payoff than if he played
R, which would have given him a payoff of  1 . Thus, the perfect
Bayesian Equilibrium is

[( L, R), (u, a), p  =  1, q  =  0 ]