Download as pdf or txt
Download as pdf or txt
You are on page 1of 22

University of California, Davis - Department of Economics

SPRING 2005 ECN 200C: MICRO THEORY Professor Giacomo Bonanno


First Midterm Exam Version 1
======================================================================
ANSWER ALL QUESTIONS (total 100 points)
1. [50 points] The pivotal mechanism is used to decide whether a new park should be built.
There are 5 individuals. According to the proposed project, the cost of the park would be
allocated as follows:
individual 1 2 3 4 5
share of cost c1 = $30 c2= $25 c3 = $25 c4 = $15 c5 = $5
For every individual i = 1, .., 5, let vi be the perceived gross benefit (if positive; perceived
gross loss, if negative) from having the park built. The vi’s are as follows:
individual 1 2 3 4 5
gross benefit v1 = $60 v2= $15 v3 = $55 v4 = $− 25 v5 = $− 20

(Thus the net benefit (loss) to individual i is vi − ci. )


Individual i = 1,…,5 has the following utility of money function (where mi denotes the

amount of money of individual i): Ui(mi) =


RS mi if the project is not carried out
T m +v i i if the project is carried out
Let ei be the initial endowment of individual i and assume that ei is large enough that it
exceeds ci plus any tax that the individual might have to pay.
(a) [5 points] What is the Pareto-efficient decision: to build the park or not?
Assume that the pivotal mechanism is used, so that each individual i is asked to state a number
wi which is going to be interpreted as the gross benefit to individual i from carrying out the
project. The are no restrictions on the number wi: it can be positive, negative or zero. Suppose
that the individuals make the following announcements:
individual 1 2 3 4 5
stated benefit w1 = $70 w2= $10 w3 = $65 w4 = $− 30 w5 = $5

(b) [5 points] Would the park be built based on the above announcements?
(c) [10 points] Using the above announcements and the rules of the pivotal mechanism, fill in the
following table:
individual 1 2 3 4 5
Pivotal? (Yes/No)
Tax he/she has to pay
(d) [10 points] If all the individuals played their dominant strategies, would the park be built? [It is
not enough to say “Yes” or “No”: you have to explain how you got your answer].
(e) [10 points] If all the individuals played their dominant strategies, who would be pivotal?
(f) [10 points] Show that if every other individual reports his/her true benefit, then it is best for
individual 1 to also report his/her true benefit.

Page 1 of 2
2. [25 points] Consider Hotelling’s model. On a street of length L miles, two gas stations sell an
identical brand of gasoline. One gas station is located at point x1 and the other located at
point x2. There are N consumers uniformly distributed along this street (thus, for every point
t on the street, the fraction of consumers who live between 0 (the beginning of the street)
t
and t is equal to ). Each consumer buys exactly one unit, from the gas station that offers
L
the least delivered price, which is given by the sum of the actual price charged by the firm
and the transportation cost faced by the consumer. All the consumers have the same
roundtrip transportation cost equal to α d where d is the one-way distance (thus if the
consumer lives at point A and goes to point B which is d miles away from A, then the
transportation cost for the entire roundtrip from A to B and back to A is αd). The two gas
stations have the same cost function given by C(q) = c q.
Assume that: L = 8, x1 = 1, x2 = 5, N = 480, α = 3, c = 1.
(a) [10 points] Find the demand functions for the two gas stations (denote the price of
the gas station located at x1 by p1 and the price of the gas station located at x2 by
p2).

(b) [10 points] Find the Nash equilibrium in prices. Calculate the equilibrium profit
of each firm.

(c) [5 points] If both firms had been located in the same spot, namely in the center of
the street, what would the equilibrium prices have been?

3. [25 points] There are three firms in a homogeneous-product industry, where inverse
demand is given by P = a − bQ (a > 0, b > 0, Q = total industry output). All firms have the
same cost function given by C(x) = F + cx (where F ≥ 0 is fixed cost, c > 0 is marginal cost,
with c < a, and x is output). Firms compete in output levels.

(a) [10 points] Calculate the Cournot-Nash equilibrium.


The owners of firms 1 and 2 meet to discuss the possibility of merging the two
firms into a single firm that would have the same cost function, namely C(x) = F + cx. If
they decide to merge, they will each have 50% of the shares of the new firm. They reason
as follows: "If we don't merge the two firms, we maintain the status quo, which is given
by the three-firm Cournot-Nash equilibrium. If we do merge, then we will be in a two-
firm industry (assume that new entry is impossible) and the outcome will be a Cournot-
Nash equilibrium between the two firms".

(b) [5 points] Assume that F = 0 (no fixed costs). Should they merge?

(c) [10 points] Assume now that F > 0. Does your answer to question (b) change?

Page 2 of 2
University of California, Davis -- Department of Economics
SPRING 2005 ECON. 200C: GAME THEORY Giacomo Bonanno

First Midterm Exam ANSWERS : VERSION 1

5
1. Let neti = vi − ci. Then, since ∑ neti = −15 < 0,
i= 1

(a) The Pareto efficient decision is: Pareto = "No to project"

5
Let ni = wi − ci. Then, since ∑ ni = 20 > 0,
i= 1

(b) Based on the announcements: decision = "Yes to project"

⎛ "individual" "Pivotal?" "Tax" ⎞



⎜ 1 "pivotal" 20 ⎟
⎜ 2 "not pivotal" 0 ⎟
(c) Based on the announcements: Table = ⎜ ⎟
⎜ 3 "pivotal" 20 ⎟
⎜ 4 "not pivotal" 0 ⎟

⎝ 5 "not pivotal" 0 ⎠

(d) The dominant strategy is to report the true benefit. Thus if all individuals do
that the decision will be the Pareto efficient one, namely: Pareto = "No to project"

⎛ "individual" "Pivotal?" "Tax"⎞



⎜ 1 "not pivotal" 0 ⎟
⎜ 2 "not pivotal" 0 ⎟
(e) Based on the true benefits: Table = ⎜ ⎟
⎜ 3 "not pivotal" 0 ⎟
⎜ 4 "pivotal" 25 ⎟

⎝ 5 "pivotal" 10 ⎠

Page 1 of 3
(f) Assume that everybody else reports the true benefit. Then the project is not carried
out and individual x = 1 is not pivotal. His utility is equal to his initial endowment of en . Any
other announcement w that leaves the decision at “no project” gives him the same utility. On the
other hand, if he states a large enough w that turns the decision into a “Yes to project” will make
him pivotal and he will have to pay a tax of t = 45 and his utility would be U ( en ) → en − 15
and thus he would be worse off.

2.
(a) Given prices p1 and p2, the location of the consumer who is indifferent between
the two gas stations is given by the solution to the equation: p1 + 3 (t − 1) = p2 +

p2 − p1
3(5 − t). The solution is: t = 3 + . The fraction of consumers who go to
6
p2 − p1
3+
6 3 p − p1
firm 1 is therefore: = + 2 and the fraction of consumers
8 8 48
5 p2 − p1
who go to firm 2 is − . Thus the demand function for firm 1 is:
8 48
FG 3 + p − p IJ = 180 +10p − 10p and that of firm 2 is :
H 8 48 K
2 1
D1(p1,p2) = 480 2 1

F 5 p − p IJ = 300 + 10p − 10p .


D (p ,p ) = 480G −
H 8 48 K
2 1
2 1 2 1 2

(b) The profit function of firm 1 is: Π1(p1,p2) = p1D1(p1,p2) − D1(p1,p2) and the profit

function of firm 2 is Π2(p1,p2) = p2D2(p1,p2) − D2(p1,p2). The Nash equilibrium is

∂Π1 ∂Π 2
given by the solution to = 0 and = 0 which is p1 = 23 and p2 = 27.
∂p1 ∂p2

Profits are Π1 = 4,840 and Π2 = 6,760.

(c) When the two firms are located on the same spot the two products are
homogeneous. Thus by Bertrand's theorem there is a unique Nash equilibrium
where both firms charge a price equal to marginal cost. Thus p1 = p2 = c = 1 .

Page 2 of 3
3. (a) The profit function of firm i (i = 1,2,3) is given by:
πi(q1,q2,q3) = qi [a − b(q1+q2+q3)] − c qi − F.

∂π
i
The Cournot equilibrium, given by the solution to = 0 (i=1,2,3), is:
∂qi

a−c a + 3c ( a − c )2
q1 = q2 = q3 = , P= , π1 = π2 = π3 = − F.
4b 4 16b

(b) and (c) In a two-firm industry, the profit function of firm i (i=1,2) is given by:

πi(q1,q2) = qi [a − b(q1+q2)] − c qi − F.

∂π
i
The Cournot equilibrium, given by the solution to = 0 (i=1,2), is:
∂qi

a−c a + 2c ( a − c )2
q1 = q2 = , P= , π1 = π2 = − F.
3b 3 9b

Thus merger is profitable if and only if:

2& − F' <


( a − c )2 ( a − c )2
−F,
16b 9b

that is, if and only if

( a − c )2
F > [This is clearly not satisfied when F = 0].
72b

Page 3 of 3
University of California, Davis - Department of Economics
SPRING 2005 ECN 200C: MICRO THEORY Professor Giacomo Bonanno
Second Midterm Exam Version 1
======================================================================
ANSWER ALL QUESTIONS (total 100 points)

1. [35 points] There are two firms producing a homogeneous product whose inverse demand
function is P = 4 − Q (where Q is industry output). Firm 1 acts as Stackelberg leader: it
chooses its output first; firm 2 chooses its output after observing the output of firm 1. Firm 2
decides to enter (i.e. chooses a positive level of output) if and only if it expects to make
positive profits. Consider two scenarios.

(a) [10 points] Both firms have zero costs. Find the backward-induction solution of this
perfect-information game.

From now on assume that both firms have zero marginal cost, but a positive fixed cost F
(which is incurred by a firm if and only if the firm produces a positive level of output).

(b) [5 points] Calculate and draw the reaction curve of firm 2 when F = 0.16.

(c) [10 points] What is the backward-induction solution when F = 0.16?

(d) [10 points] What is the backward-induction solution when F = 0.04?

2. [15 points]. Two players play the following perfect information game. Player 1 starts by

choosing a number from the set {1,2,3,4,5,6,7}, then player 2 chooses a number from this set,

then player 1 again, followed by player 2, etc. The first player who brings the cumulative

sum of all the numbers chosen (up to and including the last one) to 50 or more wins. In this

game player 1 has a winning strategy. Find this strategy and explain it fully.

Page 1 of 2
3. [50 points] You have to go to a part of town where many people have been mugged recently.
You consider whether you should leave your wallet at home or carry it with you. Of the four
possible outcomes, your most preferred one is having your wallet with you and not being
mugged. Being mugged is a very unpleasant experience, so your second favorite alternative
is not carrying your wallet and not being mugged (although not having any money with you
can be very inconvenient). If, sadly enough, your destiny is to be mugged, then you prefer to
have your wallet with you (possibly with not too much money in it!) because you don't want
to have to deal with a frustrated mugger. A typical potential mugger, on the other hand, does
not care whether or not you are carrying a wallet, in case he decides not to mug you (that is,
he is indifferent between the corresponding two outcomes). Of course his favorite outcome is
the one where you have your wallet with you and he mugs you. His least preferred outcome
is the one where he attempts to mug you and you don't have your wallet with you (he risks
being caught for nothing). Denote the possible outcomes as follows:
Potential mugger
Not mug Mug
You Leave wallet at home z 1
z2
Take wallet with you z3 z4

(a) [5 points] What is the ordinal ranking of the outcomes for each player?

(b) [10 points] Suppose now that both players have von Neumann-Morgenstern utility
functions. You are indifferent between the following lotteries:
⎛ z1 z2 z3 z4 ⎞ ⎛ z1 z2 z3 z4 ⎞
L1 = ⎜ 3 14 3 ⎟ and L2 = ⎜ 1 1 ⎟
; furthermore, you are
⎝ 20 20 20 0 ⎠ ⎝0 2 0 2 ⎠
⎛ z1 z2 z3 z4 ⎞ ⎛ z1 z2 z3 z4 ⎞
indifferent between L3 = ⎜ 2 1 ⎟ and L 4 = ⎜1 1 ⎟ . The
⎝0 3 3 0⎠ ⎝2 2 0 0⎠
⎛ z1 z2 z3 z4 ⎞
potential mugger is indifferent between the two lotteries L5 = ⎜ 1 1 1 1 ⎟
⎝4 4 4 4⎠
⎛ z1 z2 z3 z4 ⎞
and L6 = ⎜ 8 67 16 37 ⎟ . For each player find the normalized von Neumann-
⎝ 128 128 128 128 ⎠
Morgenstern utility function that assigns utility 1 to the best outcome and utility 0
to the worst outcome.

You have to decide whether or not to leave your wallet at home. Suppose that, if you
leave your wallet at home, with probability p (with 0 < p < 1) the potential mugger will
notice that your pockets are empty and with probability (1 − p) he will not notice (i.e. he
will not know whether your pockets are empty or contain a wallet). In either case the
potential mugger has to decide whether to mug you or not.

(c) [10 points] Represent this situation as an extensive game with imperfect information.

(d) [10 points] Write the corresponding normal form.

(e) [15 points]Find all the subgame-perfect equilibria (including the mixed-strategy ones,
if any). [Hint: your answer should distinguish between different values of p].

Page 2 of 2
University of California, Davis -- Department of Economics
SPRING 2005 ECON. 200C: GAME THEORY Giacomo Bonanno

Second Midterm Exam ANSWERS : VERSION 1

1.
(a) The profit function of firm 2 is given by Π 2 ( q1 , q2 ) = q2 (4 − q1 − q2 ) . The reaction
∂Π 2
function of firm 2 is given by the solution, w.r.t. q2, of = 0 . Thus
∂q2

q1
R2 (q1 ) = 2 − . Hence the profit function of firm 1 (using backward induction)
2
∂Π1
is Π1 ( q1 ) = q1 (4 − q1 − R2 ( q )) = 12 (4q1 − q12 ) . Taking = 0 gives q1 = 2 . The
∂q1

output of firm 2 is thus R2(2) = 1.

q1
(b) When there is a fixed cost F > 0, the reaction function of firm 2 is still R2 (q1 ) = 2 −
2
provided that Π 2 (q1 , R2 ( q1 )) > F where the function Π 2 (q1 , q2 ) is as given in part

(a). Solving Π 2 (q1 , R2 ( q1 )) = F when F = 0.16 gives q1 = 3.2. Thus the reaction

⎧ q1
⎪2 − if q1 < 3.2
function of firm 2 is R2 (q1 ) = ⎨ 2 . It is shown in the following
⎪⎩0 if q1 ≥ 3.2

diagram:
q2

0.4

0 q1
3.2

(c) When F = 0.16, the profit function of firm 1 becomes:

Page 1 of 5
⎪⎧ 1 (4q1 − q1 ) − 0.16 if q1 < 3.2
2
Π1 (q1 ) = q1 (4 − q1 − R2 (q )) − 0.16 = ⎨ 2
⎪⎩q1 (4 − q1 ) − 0.16 if q1 ≥ 3.2

it is sketched below:

4
2.4

2
Π 1( y , 0.16)
0

− 0.16 2
0 2 4
0 y 4

As is clear from the picture, the backward induction outcome is q1 = 3.2 (and q2 = 0).
Alternatively, this can be established by comparing the maximum value of the function
1
2 (4q1 − q12 ) − 0.16 which is 1.84 (achieved at q1 = 2) with the maximum value of the

function q1 (4 − q1 ) − 0.16 in the range q1 ≥ 3.2, which is 2.4 and is achieved at q1 = 3.2.

(d) When F = 0.04, the profit function of firm 1 becomes:

⎪⎧ 1 (4q1 − q1 ) − 0.04 if q1 < 3.6


2
Π1 (q1 ) = q1 (4 − q1 − R2 (q )) − 0.04 = ⎨ 2
⎪⎩q1 (4 − q1 ) − 0.04 if q1 ≥ 3.6

it is sketched below:

2
1.96

1
Π 1( y , 0.04)
0

− 0.04 1
0 2 4
0 y 4

As is clear from the picture, the backward induction outcome is q1 = 2 (and q2 = 1).
Alternatively, this can be established by comparing the maximum value of the function

Page 2 of 5
1
2 (4q1 − q12 ) − 0.04 which is 1.96 (achieved at q1 = 2) with the maximum value of the

function q1 (4 − q1 ) − 0.04 in the range q1 ≥ 3.6, which is 1.4 and is achieved at q1 = 3.6.

2. First find a losing position. If player i with his choice can bring the sum to 42 then he can win
(the other player with her next choice will take the sum to a number between 43 and 49 and then
player i can win with his next choice). Working backwards, the previous losing position is 34
(from here the player who has to move will take the sum to a number between 35 and 41 and
after this the opponent can take it to 42). Reasoning backwards, the earlier losing positions are
26, 18, 10 and 2.

Thus player 1’s winning strategy is: start by choosing 2 and then at every turn if the
opponent’s last choice was n then player 1 should choose (8 − n).

⎛ z3 ⎞
⎜ ⎟
z
3. (a) Your ranking is (best at the top, worst at the bottom) ⎜ 1 ⎟ while the potential mugger’s
⎜ z4 ⎟
⎜ ⎟
⎝ z2 ⎠
⎛ z4 ⎞
⎜ ⎟
ranking is ⎜ z1 , z3 ⎟
⎜ z ⎟
⎝ 2 ⎠

(b) Let U be your utility function. Let U(z3) = 1, U(z1) = a, U(z4) = b and U(z2) = 0, with

0 < b < a < 1. The expected utilities are as follows: EU ( L1 ) = 3


20 a + 203 ,

EU ( L2 ) = 12 b , EU ( L3 ) = 13 and EU ( L4 ) = 12 a . From EU ( L3 ) = E ( L4 ) we get that

a = 23 . Substituting this into the equation EU ( L1 ) = EU ( L2 ) gives b = 12 . Thus

2 1
U(z3) = 1, U(z1) = , U(z4) = and U(z2) = 0.
3 2
Let V be the mugger’s utility function. Let V(z4) = 1, V(z1) = V(z3) = c and V(z2) =

0. The expected utilities are as follows: EV ( L5 ) = 14 (2c + 1) and

EV ( L6 ) = 128
1
(24c + 37) . Solving EV ( L5 ) = EV ( L6 ) gives c = 18 . Thus,

1
V(z4) = 1, V(z1) = V(z3) = and V(z2) = 0.
8

(c) The extensive game is as follows:

Page 3 of 5
You
leave
take
NATURE
not notice
notice
p 1-p

Mugger Mugger
not
mug
not mug not mug

z1 z2 z1 z2 z 3 z4
2/3 0 2/3 0 1 1/2
1/8 0 1/8 0 1/8 1

(d) The normal form is as follows (for the mugger’s strategy the first refers to the left
node, the second to the information set)

Potential Mugger
NN NM MN MM
L 2/3 , 1/8 (2/3)p , (1/8)p 2/3(1-p) , 1/8 (1-p) 0,0
You
T 1 , 1/8 1/2 , 1 1 , 1/8 1/2 , 1

(e) At a subgame-perfect equilibrium the mugger will choose not to mug when he notices
your empty pockets. Thus the normal form can be simplified as follows:

Potential Mugger
NN NM
L 2/3 , 1/8 (2/3)p , (1/8)p
You
T 1 , 1/8 1/2 , 1

Page 4 of 5
• If p < ¾ then Take is a strictly dominant strategy for you and therefore there is a
unique subgame-perfect equilibrium given by (Take, Not mug / Mug).

• If p = ¾ then there is a continuum of equilibria where the Mugger chooses “Not


mug / Mug” with probability 1 and you choose L with probability q and T with
7
probability (1− q) for any q with 0 ≤ q ≤
8− p

• If p > ¾ then there is no pure-strategy subgame-perfect equilibrium. Let q be the


probability that you choose L and r the probability that the mugger chooses NN.
Then the unique mixed strategy equilibrium is given by the solution to:

2 2 1
r + p (1 − r ) = r + (1 − r )
3 3 2

1 1
= pq + (1 − q)
8 8

7 4p −3
which is q = and r = . Thus the unique subgame-perfect equilibrium
8− p 4 p −1
is:

⎛ L T NN NM MN MM ⎞
⎜ ⎟
⎜ 7 1− p 4p −3 2
⎜8− p 0 0 ⎟⎟
⎝ 8− p 4 p −1 4 p −1 ⎠

Page 5 of 5
University of California, Davis - Department of Economics
SPRING 2005 ECN 200C: MICRO THEORY Professor Giacomo Bonanno
Final Exam Version 1
======================================================================
ANSWER ALL QUESTIONS (total 100 points)
1. [30 points] A monopolist faces nA = 50 type A consumers, nB = 60 type B consumers and nC =
80 type C consumers. Each type A consumer has the demand function DA ( P ) = 60 − 2 P , each
type B consumer has the demand function DB ( P ) = 60 − 4 P and each type C consumer has
the demand function DC ( P) = 60 − 8 P . The monopolist is considering selling the good in
packages. The pair (Q,V) represents a package containing Q units at a total price of V (thus V
is the price of the entire package, not the price per unit). The monopolist is considering the
following options.
OPTION 1. Sell only one type of package (Q1 = 20, V1 = 170)
OPTION 2. Sell two types of packages: (Q21 = 20, V21 = 170) and (Q22 = 30, V22 = 211).
OPTION 3. Sell two types of packages: (Q31 = 30, V31 = 168) and (Q32 = 56, V32 = 388).
(a) [6 points] Calculate the monopolist’s profits for Option 1.
(b) [8 points] Calculate the monopolist’s profits for Option 2.
(c) [8 points] Calculate the monopolist’s profits for Option 3.
(d) [8 points] What would the monopolist’s profits be if it were able to use first-
degree price discrimination?
2. [40 points] Three scientists, A, B and C agree that the space of “possible worlds” is Ω = {1, 2,
3, 4, 5}. Initially they have a common prior that attaches equal probability to every state. Now
they agree that A will perform the experiment represented by the following partition: {1,2,4},

{3,5} , B will perform the experiment represented by the following partition: {1,3,5}, {2,4}

and C will perform the experiment represented by the following partition: {2,3,4}, {1,5} .
They are interested in establishing the probability of event F = {1,2,3}. After the experiment,
each scientist will communicate her posterior probability of F to a common friend. The friend
will compute the average of the posteriors and make a public announcement of this average (so
that it becomes common knowledge). Each scientist will then use this information to update
her posterior of F and communicate the new posterior to the common friend, who will then re-
compute the average and announce it publicly, and so on.
(i) [20 points] Replace the question marks in the following tables (e.g. in the second table, in the
cell corresponding to state 2 and column A you have to write the posterior probability that
scientist A attaches to event F given the information that she obtains if the true state is 2).
(ii) [10 points] If the true state is 1, which of the scientists (if any) will discover in the end that the
true state is 1?

Page 1 of 4
(iii) [10 points] If the true state is 2, which of the scientists (if any) will discover in the end that the
true state is 2?
Time: AFTER EXPERIMENT
A’s partition 1 2 4 3 5

B’s partition 1 3 5 2 4

C’s partition 2 3 4 1 5

Common knowledge partition ?

Scientist → A B C Average
State ↓
1 ? ? ? ?
2 ? ? ? ?
3 ? ? ? ?
4 ? ? ? ?
5 ? ? ? ?

Time: AFTER FIRST ANNOUNCEMENT


A’s partition ?
B’s partition ?
C’s partition ?
Common knowledge partition ?

Scientist → A B C Average
State ↓
1 ? ? ? ?
2 ? ? ? ?
3 ? ? ? ?
4 ? ? ? ?
5 ? ? ? ?

Page 2 of 4
Time: AFTER SECOND ANNOUNCEMENTT
A’s partition ?
B’s partition ?
C’s partition ?
Common knowledge partition ?

Scientist → A B C Average
State ↓
1 ? ? ? ?
2 ? ? ? ?
3 ? ? ? ?
4 ? ? ? ?
5 ? ? ? ?

Page 3 of 4
3. [30 points]: Two complementary objects (e.g. a left shoe and a right shoe) are auctioned to 3
bidders: Ann, Bob and Carla. Each bidder assigns zero value to each individual object but
values the pair at $100. The bidding takes place as follows. There are two rounds. In each
round each player can bid either $0 or $10 or $20. At the end of the first round the following
happens: (1) if one player bid more than the others, he/she gets the first object and everybody
(including those who lost the first round) has to pay his/her bid; (2) if the highest bid was
submitted by at least two players then Ann gets the object (even if she was not one of the
highest bidders) and everybody has to pay his/her bid (even those who lost). Then we proceed
to the next stage where the same rules apply. Thus, at the end of the second round the
following happens: (1) if one player bid more than the others, he/she gets the second object
and everybody (including those who lost the second round) has to pay his/her bid; (2) if the
highest bid was submitted by at least two players then Ann gets the second object (even if she
was not one of the highest bidders) and everybody has to pay his/her bid (even those who
lost). [Note that in the second round the bids that are used to determine the outcome of the
second round, i.e. who gets the second object, are not the cumulative bids (first-round bid +
second-round bid) but just the second-round bids.]

(a) [8 points] Write the normal form of the second-round game in the case where in
the first round Ann bid $0, Bob bid $10 and Carla bid $10. Note: write the payoffs
of the entire game, not only of the second stage.
Do any of the players have a dominant strategy?

(b) [4 points] Does any of the players have a dominant strategy in the second round ?
If Yes, explain why, if No explain why not. [You have to cover all cases, not only
the one given above under (a).]

(c) [8 points] What is the iterated dominant-strategy equilibrium in the second round?
[Once again, you have to cover all possible second-round games.]

(d) [8 points] Write the normal form for the first-round game assuming that
everybody predicts the outcome of the second round to be the one given in part
(c).

(e) [2 points] What are the pure-strategy Nash equilibria of the game under (d)?

Page 4 of 4
University of California, Davis -- Department of Economics
SPRING 2005 ECON. 200C: GAME THEORY Giacomo Bonanno

Final Exam ANSWERS : VERSION 1

1.
Let Wi(Q) be the willingness to pay of customer of type i ∈ {A,B,C} for Q units. Then
Q2 Q2 15 Q2
WA (Q) = 30Q − , WB (Q) = 15Q − and WC (Q) = Q − .
4 8 2 16

(a) Since WC(Q1) − V1 = − 45, WB(Q1) − V1 = 80 and WA(Q1) − V1 = 330, only


customers of type A and B buy. Thus

The firm’s profits will be Π1 = (n A + nB )V1 = 18, 700

(b) Since WC(Q21) − V21 = − 45, WC(Q22) − V22 = − 42.25, WB(Q21) − V21 = 80,
WB(Q22) − V22 = 126.5, WA(Q21) − V21 = 330, WA(Q22) − V22 = 464, C-customers do
not purchase anything, while the others purchase the second package. Thus the
firm’s profits would be

Π 2 = (nA + nB )V22 = 23, 210

(c) Since WC(Q31) − V31 = 0.75, WC(Q32) − V32 = − 164, WB(Q31) − V31 = 169.5,
WB(Q32) − V32 = 60, WA(Q31) − V31 = 507, WA(Q32) − V32 = 508, B and C customers
purchase the first package, while A customers purchase the second package. Thus
the firm’s profits would be

Π 3 = (nC + nB )V31 + n AV32 = 42,920


Since the third option yields the highest profits, of the three the firm would choose the
third option.
(d) With first degree price discrimination the firm would set price equal to marginal
cost and charge different fixed fees to the different types of individuals. Thus it
would set P = 0 and sell 60 units to every type. The corresponding tariffs would
be TA = 900, TB = 450 and TC = 225. The firm’s profits would be Π = 90,000.

1
2. (i)
Time: AFTER EXPERIMENT

A’s partition 1 2 4 3 5

B’s partition 1 3 5 2 4

C’s partition 2 3 4 1 5

Common knowledge partition 1 2 3 4 5

The following table is obtained as follows. Take scientist A if the true state is 1, his
information is {1,2,4} and, using Bayes’ rule, P(F | {1,2,4}) = P({1,2,3} | {1,2,4}) = P(1)
+ P(2) = 13 + 13 = 23 . If the true state is 3, his information is {3,5} and P(F | {3,5}) =
P({1,2,3} | {3,5}) = P(3) = 12 . Similarly for the other cases.
Scientist →
A B C Average
State ↓

1 2/3 2/3 1/2 11/18

2 2/3 1/2 2/3 11/18

3 1/2 2/3 2/3 11/18

4 2/3 1/2 2/3 11/18

5 1/2 2/3 1/2 5/9

1
Based on the above table, the announcement will be if the true state is 1, 2, 3 or 4 and
18
5 1
if the true state is 5. Thus if the announcement is everybody learns (it becomes
9 18
5
common knowledge) that the true state is not 5. Similarly, if the announcement is then
9
it becomes common knowledge that the state is 5. Thus the information partitions are
refined as follows

2
Time: AFTER FIRST ANNOUNCEMENT

A’s partition 1 2 4 3 5

B’s partition 1 3 2 4 5

C’s partition 1 2 3 4 5

Common knowledge partition 1 2 3 4 5

The probabilities in the following table are computed as before, but using the new information
partitions. For example, if the true state is 3 now scientist A knows (because he was able
to eliminate state 5 on the basis of the first announcement) and therefore he attaches
probability 1 to event F: P({1,2,3} | {3}) = 1.
Scientist →
A B C Average
State ↓

1 2/3 1 1 8/9

2 2/3 1/2 2/3 11/18

3 1 1 2/3 8/9

4 2/3 1/2 2/3 11/18

5 0 0 0 0

8 11
Now the announcement is if and only if the true state is either 1 or 3 and it is if and
9 18
8
only if the true state is either 2 or 4. So, if the announcement is , scientist A, who was
9
uncertain as to whether the state was 1, 2 or 4, can now eliminate state 1. Similarly for the
other cases.

3
Time: AFTER SECOND ANNOUNCEMENTT

A’s partition 1 2 4 3 5

B’s partition 1 3 2 4 5

C’s partition 1 2 4 3 5

Common knowledge partition 1 3 2 4 5

Now I am tired of always repeating the same things! It should be clear how the following table
is obtained.
Scientist →
A B C Average
State ↓

1 1 1 1 1

2 1/2 1/2 1/2 1/2

3 1 1 1 1

4 1/2 1/2 1/2 1/2

5 0 0 0 0

(ii) Only A and C will, B won’t (according to B’s information, state 3 is a possibility)

(iii) No one. However, it will be common knowledge that the state is either 2 or 4.

4
3.
(a) Given the rules, since there were two highest bids, the first object went to Ann (even though she bid less than the others).
Thus the normal form of this second-round game is:
Bob Bob Bob
$0 $10 $20 $0 $10 $20 $0 $10 $20

100,− 10,− 10 0,− 20,− 10 0,− 30,− 10 0,− 10,− 20 100,− 20,− 20 0,− 30,− 20 0,− 10,− 30 0,− 20,− 30 100,− 30,− 30
A $0 A $0 A $0

$10 90,− 10,− 10 90,− 20,− 10 − 10,− 30,− 10 N $10 90,− 10,− 20 90,− 20,− 20 − 10,− 30,− 20 N $10 − 10,− 10,− 30 − 10,− 20,− 30 90,− 30,− 30

N
$20 80,− 10,− 10 80,− 20,− 10 80,− 30,− 10 N $20 80,− 10,− 20 80,− 20,− 20 80,− 30,− 20 N $20 80,− 10,− 20 80,− 20,− 30 80,− 30,− 30

Carla: $0 Carla: S10 Carla: $20

For both Bob and Carla, bidding $0 is a strictly dominant strategy. Ann does not have a dominant strategy.

(b) For every player it is a dominant choice to bid 0 in round 2 if he/she did not win object 1 in round 1. There is no dominant

choice for any player if he/she won the first round (e.g. for Ann bidding 0 is worse than bidding 20 if only one other person bids 20).
(c) The iterated dominant-strategy equilibrium in the second round is as follows: (1) if Amy won the first round, everybody
bids 0; (2) if Bob won the first round, Bob bids 10 and the others 0; (3) if Carla won the first round, Carla bids 10 and the others 0.
(d) The normal form for the reduced first-round game is as follows. First figure out who wins the first object:

Page 5 of 6
Bob Bob Bob
$0 $10 $0 $10 $20 $0 $10 $20
$20

Ann Bob Bob Carla Ann Bob Carla Carla Ann


A $0 A $0 A $0

$10 Ann Ann Bob N $10 Ann Ann Bob N $10 Carla Carla Ann

N
$20 Ann Ann Ann N $20 Ann Ann Ann N $20 Ann Ann
Ann
N

Carla: $0 Carla: S10 Carla: $20

Recalling the conclusions from part (b) we get the following normal-form:

Bob Bob Bob


$0 $10 $0 $10 $20 $0 $10 $20
$20

100,0,0 0, 80,0 0,70,0 0,0,90 100,− 10,− 10 0,70,− 10 0,0,70 0,− 10,70 100,− 20,− 20
A $0 A $0 A $0

$10 90,0,0 90,− 10,0 − 10, 70,0 N $10 90,0,− 10 90,− 10,− 10 − 10,70,− 10 N $10 − 10,0,70 − 10,− 10,70 90,− 20,− 20

N
$20 80,0,0 80,− 10,0 80,− 20,0 N $20 80,0,− 10 80,− 10,− 10 80,− 20,− 10 N $20 80,0,− 20 80,− 10,− 20 80,− 20,− 20

Carla: $0 Carla: S10 Carla: $20

(e) There are no pure-strategy Nash equilibria (e.g. everybody bidding 0 is not a Nash equilibrium: Bob can do better by
increasing his bid to 10, etc.).

Page 6 of 6

You might also like