Problem Set 9
Problem Set 9
Submission instructions:
1
1. (Poverty Traps) Alex is deciding whether or not to make a loan to Brian who is very
poor and who has a bad credit history. Simultaneous to Alex making this decision, Brian
must decide whether or not to buy gifts for his grandkids. If he buys gifts, he will be
unable to repay the loan. If he does not buy gifts, he will repay the loan. If Alex refuses
to give Brian a loan, then Brian will have to go to a loan shark.
The payoffs in this game are as follows: if Alex refuses to make a loan to Brian and Brian
buys gifts then both Alex and Brian get 0. If Alex refuses to make a loan to Brian and
Brian does not buy gifts then Alex gets 0 and Brian gets -1. If Alex makes a loan to Brian
and Brian buys gifts then Alex gets -2 and Brian gets 7. If Alex makes a loan to Brian
and Brian does not buy gifts, then Alex gets a payoff of 3 and Brian gets a payoff of 5
(a) (15 points) Suppose this game is played only once. Write down the payoff matrix
and find the Nash equilibria of this game.
Now suppose that the game is repeated. Suppose that (for all players) a euro tomorrow
is worth 2/3 of a euro today. In addition, suppose that, after each period (and regardless
of what happened in the period), Brian has a 1/2 chance of escaping poverty. Assume
that, if Brian escapes poverty then he will not need a loan from either Alex or a loan
shark: in effect, Brian will exit the game. Assume that, if Brian escapes poverty, he will
never return. Thus, after each period, there is only 1/2 chance of the game continuing.
Given this, the effective discount factor for the game between Alex and Brian is δ =
(1/2) × (2/3) = (1/3).
Consider the following strategy profle. In period one, Alex makes Brian a loan. There-
after, Alex continues to make Brian loans (if he is still poor) as long as Brian has always
got a loan and repaid it in the past. But if Brian ever does not repay (or does not get a
loan) then Alex never makes a loan to Brian again. In period one, Brian does not buy
gifts (and hence repays the loan if he gets one). Thereafter (as long as he is still poor),
Brian does not buy gifts (and hence repays the loan if he gets one) as long as he has
always got a loan and repaid it in the past. But if Brian ever does not repay (or does not
get a loan) then he will return to buying gifts and hence never repay a loan again.
(b) (15 points) Is this strategy profile an SPE of the repeated game?
(c) (10 points) Suppose that the government introduces regulation of loan sharks. As
a consequence, Brian’s payoff in each period in which he still needs a loan but does
not get it from Alex is 1 if he does not buys gifts (instead of -1 before) and 2 if he
buys gifts (instead of 0 before). Is the strategy profile outlined above still an SPE?
2
(d) (10 points) Suppose that the government abandons its loan-shark policy and re-
places it with a job scheme that increases the probability after each period of Brian
escaping poverty to 2/3 (i.e., 1/3 chance of returning to the loan game). Explain
the likely consequences of this policy for the business relationship between Alex
and Brian.
Assume that if both leaders choose the same strategy then each wins with pi = 12 ; while
if one leader chooses Moderate and the other chooses Radical, the Moderate (Radical)
leader wins with probability pi = 1 (pi = 0).
(a) (5 points) Write down the payoff matrix of this game and find the Nash equilibria.
Now assume that, in addition to utility from winning the election, both leaders are Ide-
ological, and receive some intrinsic payoff from announcing Radical platforms; they re-
ceive this payoff if they choose Radical regardless of the electoral outcome. Denote
this intrinsic payoff as I, and again assume that if leaders choose the same strategy then
1
each wins with pi = 2; while if one leader chooses Moderate and the other chooses
Radical, the Moderate (Radical) leader wins with probability pi = 1 (pi = 0).
(b) (10 points) Write down the payoff matrix of this game and find the Nash equilibria
when β = 5 and I = 3.
Now, assume that the electorate is biased against leader 2. Formally speaking, assume
that if both leaders choose Moderate then each wins with pi = 21 , but if they both choose
Radical leader 1 wins the election with probability p1 = 1. As before, if one leader
chooses Moderate and the other chooses Radical, the Moderate (Radical) leader wins
with probability pi = 1 (pi = 0).
(c) (10 points) Write down the payoff matrix of this game and find the Nash equilibria
of this game when β = 5 and I = 3.
3
Continue to assume that the electorate is biased against leader 2, but now change this
1
bias such that if both leaders choose Moderate then leader 1 wins with probability p̂1 > 2
and leader 2 wins with probability p̂2 = 1 − p̂1 ; and if they both choose Radical leader 1
wins the election with probability p̃1 > p̂1 and leader 2 wins with probability p̃2 = 1− p̃1 .
As before, if one leader chooses Moderate and the other chooses Radical, the Moderate
(Radical) leader wins with probability pi = 1 (pi = 0).
(d) (5 points) Assume leader 2 moves first, and that β = 10 and I = 3. Draw the
extensive form of this game, and mark the expected payoffs for both players at the
end nodes.
7
(e) (10 points) Suppose that p̃1 > 10 . What is 1’s optimal move when 2 chooses
Moderate? What about when 2 chooses Radical?
(f) (10 points) What is the critical difference p̃1 − p̂1 that leaves leader 2 indifferent
between choosing Moderate and Radical in the first stage? What is the SPE if p̃1 − p̂1
is greater than this critical difference? What about when p̃1 − p̂1 is lower than this
critical difference?