Game Theory Problem Set 3 Answers

Download as pdf or txt
Download as pdf or txt
You are on page 1of 2

ECON 7732 Problem Set 3

1. Consider a simultaneous move game where an agent chooses the probability of complying
with a regulation, and 1- of violation. The cost of compliance is T. The regulator invests
effort e in inspection, where e 0. The probability of detecting a violator is m = min{e, 1}.
The regulators reward R is stochastic, and takes the value 1 with probability , and
otherwise. The cost of expending effort e to inspect is e2/2. Inspection of a violator leads to
her penalization, where the penalty is denoted as F (> T). The agent and the regulator play a
simultaneous move game in which the latter knows the realized value of R, but not the
former. Find all the possible Bayes-Nash equilibrium of this game, and the conditions under
which they exist.

Answer: Agents BR: = 0 if Em < T/F, [0, 1] if Em = T/F, = 1 if Em > T/F. Regulator
maximizes e(1-)R e2/2, when m = min{(1-)R, 1}= (1-) with probability , and (1-)
otherwise. Then, Em = 7(1-)/8. As shown in class, plotting the agents BR and Em = 7(1-)/8
reveals that, if 8T<7F, * = 1-8T/7F. if 8T7F, * = 0. The corresponding Em* and each m* can
then be derived by plugging in * in the appropriate expressions for Em* and m*s as per above.

2. Consider a Cournot duopoly where each firm has a constant unit cost of production
represented by c. The market demand of the product is represented by p = A Q, where p is
the price, and Q the total industry output. Firm 2 suffers from the following additional
complication: with probability it may suffer a setback and fail to produce any output (in
which case it does not incur any production cost). Firm 1 cannot observe the output of firm 2
in this simultaneous move game. Find the equilibrium outputs.

Answer: With probability 1-, firm 2 produces the profit maximizing output q2=(A-q1-c). To
firm 1, Eq2 = (1-)( A-q1-c). Firm 1 maximizes E(A-q1-q2-c)q1, when q1 = ( A-Eq2-c). From
this and the expression for Eq2, we derive q1 = (1+)(A-c)/(3+), when q2 = (A-c)/(3+).

3. Consider a Cournot duopoly where, in addition to production costs, each firm incurs a fixed
cost F. Firm 1s constant unit cost of production is 1. Firm 2s unit cost is stochastic: it is 1
with probability , and 2 otherwise. (1) Find the Bayes Nash equilibrium of the game for
each of the following values of the fixed cost: (i) F = 2, and (ii) F = 2.7, and the market
demand function as p = 10 Q, where p is the market price, and Q represents the total output
from both firms. (2) With p = A Q, the fixed cost as F for the firms, and the unit costs as
above, find a condition involving F such that the informed firm produces no output in one
state of nature. What is the output of firm 1 if this condition is satisfies, and how does it
compare to its output when firm 2 produces positive outputs in all states of nature?

Answer: I shall work out the answer to the general case (2). Starting with the assumption that
firm 2 produces a positive output in both states of nature, q2i = (A-q1 ci), where cH=2, and
cL=1. Eq2 = (A-q1-3/2) to the uninformed firm 1, and q1 = (AEq21). We can now derive the
equilibrium outputs as q1 = (A-0.5)/3, q2H = (A-11/4)/3, q2L = (A-5/4)/3. As demonstrated in
class, the profits are the respective squares of the outputs (for firm 1 this give the expected profit)
minus any fixed cost. His implies that, if A > 11/4, and (A-11/4)2/9 F, firm 2 produces in both
states of nature, and the equilibrium outputs are as per above.
Suppose that, either A 11/4, or (A-11/4)2/9 < F, or both. Then, start again and find the
equilibrium outputs when firm 2 only produces when the cost is low. Following the exact method
of the previous question, but with c = 1 and = , we derive q1 = 3(A-1)/7, q2 = 2(A-1)/7.
Clearly, if A>1, and 4(A-1)2/49 F, firm 2 produces when its cost is low. If 4(A-1)2/49 < F, firm
1 is a monopoly, and produces (A-1). If A<1, no firm produces a positive output.

4. Find the answer to exercise 3 when F = 9.

Answer: Assume, as in question 2, that A=10. Then, using the expressions derived in question 3,
we find that firm 1 will be a monopoly.

5. Consider an extension of exercise 1 above when (i) R=2 with probability , and R=1
otherwise, and (ii) R=4 with probability , and R=2 otherwise. In each case, find all the
possible Bayes-Nash equilibria, and the conditions under which they exist.

Answer: (i) Here, the agent has the same BR, and m = min{(1-)R, 1}. This means that, for R =
2, m=2(1-) if > , and m=1 if . For R=1, m = 1-. Then, Em = 7(1-)/4 if > , and 1-
/4 if . Em plotted against has a kink at = and Em = 7/8. Then, as the possible plots
of the BRs of the two players show (demonstrated in class) it T/F 7/8, * = 1- 4T/7F. If T/F >
7/8, * = 4(1 T/F).

(ii) Here, when R = 4, m = 4(1-) if > , and 1 when . When R=2, m=2(1-) if > ,
and m=1 if . Em plotted on has two kinks, one at = , and the other at = , and can
be represented as: Em = 1 if , Em = 5/4 /2 if (, ), and 7(1-)/2 if > . Then, if
T/F < 7/8, from T/F = 7(1-)/2 we derive * = 1 2T/7F. If T/F 7/8, * = 5/2 2T/F.

6. Consider a simultaneous move game where an agent chooses the probability of complying
with a regulation, and 1- of violation. The cost of compliance is C. The regulator of type i
chooses either to inspect or not to inspect. His cost of an inspection is ri, which is distributed
uniformly over the interval [0, 1]. Inspection of a violator leads to her penalization, where the
penalty is denoted as F (> C). A regulator obtains the reward R (= 5) upon apprehending a
violator. The agent and the regulator play a simultaneous move game in which the latter
knows the realized value of ri, but not the former. Find the Bayes-Nash equilibrium of this
game.

Answer: As demonstrated in class, m = 5(1-) if > 4/5, and m = 1 if 4/5. Then, * = 1-C/5F
and m* = C/F, and is also the cutoff ri.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy