Problem Set 1

Economics of Information and Uncertainty

ECON 447, Winter 2023

The solution is due on January 31, 11.59pm. Please upload your solution
as a single pdf-file to the corresponding assignment on mycourses before the
deadline. Late submissions will not be accepted. You can work in groups,
but everyone has to hand in her/his own solution. Copied solutions and
”failed to submit” will receive 0% for this particular problem set. I will post
a sample solution shortly after the submission due date. The TA will use the
TA office hours on Feb 3 for a conference to discuss the solution.

1. (based on Zandt, Exercise 2.10) Suppose that the set of outcomes X

has exactly three outcomes, z1 , z2 and z3 . Then each lottery can be
specified by three numbers: the probabilities p1 , p2 and p3 of the three
outcomes.

(a) Write down the mathematical definition of the set of lotteries, as

a subset of R3 .
(b) Draw the set of lotteries in R3 . This set should be a 2-dimensional
triangle, even though it is sitting in R3 . It is called a simplex.
(c)

1
 three numbers: the probabilities p  , p  and p  of the three outcomes.

a. Write down the mathematical definition of the set of lotteries, as a subset of R .

Given an example of an expected utility representation, and use it to explain that
the utility function over lotteries is linear in probabilities.

b. Draw the set of lotteries in R the best that your 3D-drawing skills allow. This
set should be a 2-dimensional triangle, even though it is sitting in R . It is called a
simplex.
We can redraw the 2-dimensional triangle of lotteries flat on the page:
p = 

p =  p = 
For example, at the p  =  vertex, the probability of z  is 1 and the probability of
the other outcomes is zero. Along the side opposite this vertex, the probability of
z  is zero. For all points in a given line parallel to this size, the probability of z  is
the same.
30 Lotteries and objective expected utility Chapter 2

Specify two lotteries P and Q, and plot them on the simplex. Indicate the posi-
tion of the lottery (/)P + (/)Q, as defined in class.
For the utility representation you gave in Problem 2.10, draw two indifference
curves on the simplex.

2. Preference relations over lotteries:

(a) Similarly to our definition of a preference relation over lotteries

that seeks to ”minimize the number of possible outcomes”, as de-
scribed in the lecture slides, formalize a preference relation which
seeks to ”maximize the number of possible outcome”. That is, an
individual wants the number of prizes that might be realized (the
number of prizes in the support of the lottery) to be as large as
possible.
(b) (Osborne & Rubinstein, ch.3, problem 3) ”comparing the most
likely prize”:

2
 A decision maker has in mind a basic preference relation %∗ over
the set of outcomes X. Assume X contains at least three prizes.
Whenever each of two lotteries has a single most likely prize the
decision maker compares the two lotteries by comparing the most
likely prizes using %∗ . That is, he prefers a lottery P over a lottery
Q if he prefers the most likely outcome in P over the most likely
outcome in Q.

Check, for each preference relation, if Continuity and/or Independence

are satisfied.

3. Fairness Preferences. Suppose you can award exactly one person with
a prize. Person A and person B both equally deserve the prize and you
are indifferent between giving the prize to A or B. However, you strictly
prefer to toss a fair coin to decide who gets the prize over giving the
prize to either of the two individuals directly.

(a) What are three different lotteries for which a preference ranking
is specified in the question? Write down the preference ranking
between these three lotteries.
(b) Show that such preferences violate one of the vNM axioms and
hence are not compatible with expected utility representation.
(c) Now assume by contradiction that the preferences are consistent
with expected utility representation, state what the preference
ranking implies for the utility values and derive a contradiction.

4. Consider the following variant of the Ellsberg (1961) paradox. There

are two urns marked I and II, each of which contains 100 red and black
balls. The individual knows that urn I contains 45 black and 55 red
balls exactly. However, he does not know the proportion of black and

3
 red balls in urn II. One ball is drawn randomly from each urn. Call
these BI and BII respectively. The color of these balls is not disclosed.
Now consider the following 2 choice situations. In both situations he
has to choose either BI or BII. After he chooses, the color will be
disclosed. No money changes hands if he loses the bet.
(a) A prize of $100 is won if the chosen ball is black.
(b) A prize of $100 is won if the chosen ball is red.
Suppose he chooses BI in both situations, i.e. strictly prefers choosing
BI to BII in both situations. Show that there cannot exist any (sub-
jective) probabilities α and 1 − α of drawing a red versus a black ball
from urn II that generate these strict preferences under an expected
utility evaluation.

5. (based on Bikhchandani et al., Exercises and Excursions 1.5, Q2).

Three individuals are expected utility maximizers with respective vNM
utility functions v1 (x) = 5x − 2, v2 (x) = ln(x + 1), and v3 (x) = x3 .
They each can choose any one of the three following money lotteries:
P1 with P1 (480) = 1.
P2 with P2 (850) = P2 (200) = .5.
P3 with P3 (1000) = P3 (0) = .5.

(a) Characterize the risk attitude of each individual.

(b) Describe each lottery in words.
(c) Determine for each individual, the lottery which the individual
will choose.
(d) Give a short interpretation of the choices of the individuals, re-
ferring to their risk attitudes and the expected outcomes of the
lotteries.

4
6. Insurance. Suppose an individual has initial wealth w. With prob.
p an accident occurs such that she will lose an amount of money L.
The individual can buy insurance which pays her q dollars in case the
accident happens. The price for insurance is πq where π is the premium
per dollar of coverage. The expected profit of the insurance is given by
πq − pq which we assume to be zero due to competition; thus π = p.
The individual is a risk averse expected utility maximizer with vNM
utility v(x) = x.5 where x is her final wealth level. How much coverage
q does the individual choose to buy? Depict the individual’s situation
in a 2-states-of-the-world diagram.