Lecture Notes Handout

MicroeconomicsReview Material
ECO202 Monsoon 2018
Ashoka University
Lecture : Decision Under Uncertainty
Ref: Varian Ch 12, Nicholson and Snyder Ch 7.
The broad outline of this chapter parallels a standard presentation of microeconomic theory for determin-
istic situations. It …rst considers the problem of an individual consumer facing an uncertain environment.
It shows how preference structures can be extended to uncertain situations and describes the nature of the
consumer choice problem. We then processed to derive the expected utility theorem, a result of central
importance. In the remaining sections, we discuss the concept of risk aversion. We will also discuss some
paradoxes that violate expected utility theory, and then extend the basic theory by allowing utility to depend
on states of nature underlying the uncertainty as well as on the monetary payo¤s.

Choice Under Uncertainty

So far, we have discussed consumer’s decision problem with sure outcomes. But life is full of uncertainty! In
real life, most of the decisions that a consumer makes are uncertain in nature. For example, in the morning
before you leave the house, you do not know whether it will rain or not, so you have to decide whether
to carry an umbrella in an environment of uncertainty. When one saves, usually she does not know what
the interest rate will be in future. So, we will have to consider how a consumer takes this uncertainty into
account and makes her choice.
Example: Lottery
Suppose you have $100 now, and you are contemplating whether to buy a lottery ticket number 13: If
number 13 is drawn in the lottery, the holder will be paid $200: The ticket costs $5:
Clearly, the two outcomes of interest are the event that the ticket number 13 is drawn and the event that
it isn’t. We structure our analysis using the following de…nitions:
De…nition 1 (States of Nature) Let us think di¤ erent outcomes of some random event as being di¤ erent
states of nature. So, state of nature is essentially a "random variable".
(Check the de…nitions and concepts of random variables from your statistics class!)
De…nition 2 (Contingent Plan) A Consumer’s consumption decision that depends on (or, contingent
on) the …nal outcome of the state of nature.
In this example, the states of nature are:
= fnumber
| drawn
{z = 13}; number drawn 6= 13g
| {z }
1 2

Contingent on the realization of the outcome, your wealth level will be:
w( 1 ) = 100 + 200 5 = 295
w( 2 ) = 100 5 = 95
We also need the probability of the realization of each of the states of nature, i.e. the probability
distribution of the random variable . Suppose the lottery involves drawing an integer from the interval
(0; 20] with equal probability. So,
1
Pr obability( = 1) =
20
19
Pr obability( = 2) =
20
The expected payo¤ from buying this lottery is thus:
1 19
Expected P ayof f = (295) + (95) = 105
20 20
Now, just like people have preference over di¤erent consumption bundles, they also have preference over
di¤erent consumption plans. To discuss the preference under uncertainty, let us discuss the axioms underlying
the preference.

1
Expected Utility Theory
Suppose there are n states of the nature/ outcomes:

= f 1; 2 ; ::; n g; i 2R

The probability distribution associated with the states of the nature is:

Pr ob( = i) = pi 8i = 1; :::; n

This set of probabilities along with the outcome space is termed as a "lottery":

L = ( 1; 2 ; ::::; n ; p1 ; p2 ; :::; pn )

Such that
n
X
pi = 1; pi 2 [0; 1]
i=1

Then, according to the expected utility theory, consumer’s preferences over lotteries L( 1 ; 2 ; ::::; n ; p1 ; p2 ; :::; pn )
must satisfy the following axioms:
0 0 0 0
Completeness: For any lotteries L and L ; either L L ; or L L or L L
0 0 00 00
Transitivity: If L L and L L ; then L L :
To state the next two axioms, we need the term "compound lotteries." Suppose there are two lotteries:
0 0 0 0
L( 1 ; 2 ; ::::; n ; p1 ; p2 ; :::; pn ) and L ( 1 ; 2 ; ::::; n ; p1 ; p2 ; :::; pn ) suppose the decision maker faces L with
probability and the other lottery with probability 1 :
Now we e¤ectively have a lottery of lotteries, or, a "compound lottery". The probabilities of the
outcomes ( 1 ; 2 ; ::::; n ) from this compound lottery are:
0 0
( p1 + (1 )p1 ; ::::::; pn + (1 )pn )

We now state the last two axioms:

0 00
Continuity: If L L L ; then there must exist a 2 [0; 1] such that
0 00
L L + (1 )L

Axiom of Independence
0 00
If any lotteries L; L ; L ; for any 2 [0; 1];
0
L L
00 0 00
if f L + (1 )L L + (1 )L

your preference over two lotteries isn’t a¤ected by mixing in a third.

Theorem of Expected Utility (Due to John von Neumann and Oskar Morgenstern)
If preferences satisfy Axioms 1-4, then it is possible to assign a real number (utility) ui to each outcome
i ; i = 1; :::n; such that 0 0
L L , U (L) U (L )

where

2
 Figure 1: Risk Aversion

U (L) = p1 u1 + p2 u2 + ::: + pn un
Let us apply the expected utility framework to a simple choice problem. Suppose that a decision maker
currently has $10 of wealth and is contemplating a gamble that gives him a 50% chance of winning $5 ; and
50% chance of losing $5:
His expected value of wealth=15 :5 + 5 :5 = 10
His expected utility is=
1 1
u(15) + u(5)
2 2
This is depicted in Figure 1. The expected utility of wealth is the average of the two numbers u($15)
and u($5), labeled :5u(5) + :5u(15) in the graph. We have also depicted the utility of the expected value of
wealth, which is labeled u($10). Note that in this diagram the expected utility of wealth is less than the
utility of the expected wealth. That is,
1 1
u(10) > u(15) + u(5)
2 2
In this case we say that the consumer is risk averse since he prefers to have the expected value of his
wealth rather than face the gamble. Of course, it could happen that the preferences of the consumer were
such that he prefers a random distribution of wealth to its expected value, in which case we say that the
consumer is a risk lover. An example is given in Figure 2.
Note the di¤erence between Figures 1 and 2. The risk-averse consumer has a concave utility function— its
slope gets ‡atter as wealth is increased. The risk-loving consumer has a convex utility function— its slope
gets steeper as wealth increases. Thus the curvature of the utility function measures the consumer’s attitude
toward risk. In general, the more concave the utility function, the more risk averse the consumer will be,
and the more convex the utility function, the more risk loving the consumer will be.
The intermediate case is that of a linear utility function. Here the consumer is risk neutral: the expected
utility of wealth is the utility of its expected value. In this case the consumer doesn’t care about the riskiness
of his wealth at all— only about its expected value.
Example of a risk-averse consumer:
u(x) = xa ; a 2 (0; 1)
Risk-loving:
u(x) = xa ; a > 1
Risk-neutral:
u(x) = x

3
 Figure 2: Risk-Loving Consumer

Risk-Aversion, Certainty Equivalent and Risk Premium

As we have seen from the previous example, Expected utility for a risk-averse consumer from a lottery is
strictly less than the expected payo¤ from the lottery. So, we need the following concept:

De…nition: Certainty Equivalent of a lottery L is any outcome x such that

L x

where x is a degenerate lottery that pays out x irrespective of the state of the nature.
i.e. at x;
U (x) = EU (L)
p
From the previous example, if u( ) = ;
then certainty equivalent from the lottery is given by:
1p 1p
U (x) = 15 + 5
2 p2 p
p 15 + 5
, x=
2
p p !2
15 + 5
, x=
2

This x is typically denoted by CE(L)

Due to continuity assumption, every lottery has at least one certainty equivalent.
De…nition: Risk-Premium for a lottery L is the di¤erence between the expected payo¤ of the lottery
and the certainty equivalent:

Risk premium(L) = R(L) = E(L) CE(L)

Conceptually, this is the cost of risk aversion. A risk averse consumer wold be willing to accept a
certain payment of lesser value than the expected payment from a lottery because of her aversion to
the risk involved in the lottery. Consequently, for a risk-averse consumer, R(L) > 0

4
 In the above example,
p p !2
15 + 5
R(L) = 10 >0
2

Absolute and Relative Risk Aversion

To measure risk-aversion, there are several ways.

1. A more concave utility function leads to higher degree of risk-aversion.

2. For two risk averse consumers with utility functions U and V respectively, if CE(L)U < CE(L)V ; then
U is more risk-averse than V:
3. Similarly, if R(L)U > R(L)V ; then U is more risk-averse.

However, an exact unit-free measure of risk aversion for a twice-continuously di¤erentiable utility function
U is given by the Arrow-Pratt Coe¢ cient of Risk Aversion at point x :
00
U (x)
(x) =
U 0 (x)

For a risk-averse consumer, > 0: As becomes higher, the consumer is more risk-averse.
Also, may depend on the wealth level. If increases in x; then the consumer gets more risk-averse as
wealth level increases. If is constant for all x; then the risk-aversion is independent of wealth. To measure
risk-aversion relative to the wealth level, we use the measure called Coe¢ cient of Relative Risk-Aversion
at point x given by:
00
R U (x)
(x) = x 0
U (x)

Example: If the utility for money is given by:

u(x) = ln x

then
00
U (x)
(x) =
U 0 (x)
( 1=x2 )
=
1=x
1
=
x

Thus, the decision maker becomes less risk averse if wealth level goes up. From the relative risk aversion
coe¢ cient:
00
R U (x)
(x) = x
U 0 (x)
1
= x:
x
= 1

We see the risk aversion relative to wealth stays constant.

5
Example: Demand for Insurance
Let’s apply the expected utility structure to the demand for insurance that we considered earlier. Recall
that in that example the person had a wealth of $35,000 and that he might incur a loss of $10; 000. The
probability of the loss was 1%, and it cost him K to purchase K dollars of insurance.
Let be the probability that the loss will occur, and 1 be the probability that it won’t occur.
Let state 1 be the situation involving no loss, so that the person’s wealth in that state is:

c1 = 35000 K

and let state 2 be the loss situation with wealth

c2 = 35000 10000 + K K

The expected utility from buying this insurance is:

EU (bu y) = u(c2 ) + (1 )u(c2 )

= u(35000 10000 + K K) + (1 )u(35000 K)

Now let us look at the insurance contract from the viewpoint of the insurance company. With probability
they must pay out K, and with probability 1 they pay out nothing. No matter what happens, they
collect the premium K. Then the expected pro…t, P , of the insurance company is

P = K K (1 ):0
= ( )K

Let us suppose that on the average the insurance company just breaks even on the contract. That is, they
o¤er insurance at a “fair” rate, where “fair” means that the expected value of the insurance is just equal to
its cost. Then we have

P = ( )K = 0
) =

Then, the expected utility of the consumer from buying this insurance is:

EU (bu y) = u(35000 10000 + K K) + (1 )u(35000 K)

As the consumer is assumed to be risk-averse, this is a concave objective function, ensuring that SOCs are
satis…ed. If consumer decides to buy this insurance, she solves the UMP for the amount of coverage K:

max EU
K2(0;10000]

The FOC gives: 0 0

u (c2 )[1 ] + (1 )u (c1 )[ ]=0
i.e.
0 0
u (c2 ) = u (c1 )
, c1 = c2
, 35000 10000 + K K = 35000 K
, K = 10000

This means that when given a chance to buy insurance at a “fair” premium, a risk-averse consumer will
always choose to fully insure.
This happens because the utility of wealth in each state depends only on the total amount of wealth the
consumer has in that state— and not what he might have in some other state— so that if the total amounts
of wealth the consumer has in each state are equal, the marginal utilities of wealth must be equal as well.
To sum up: if the consumer is a risk-averse, expected utility maximizer and if he is o¤ered fair insurance
against a loss, then he will optimally choose to fully insure.

6
Failure of Expected Utility: Some Paradoxes
Allais Paradox
Allais in 1953 found an experimental evidence of a possible violation of the expected utility paradigm.
Let us consider the outcome space:

= fW in Nothing, W in $1; 000; 000; W in $5; 000; 000g

So, we de…ne two sets of lotteries over this same outcome space as follows:

L1 = (0:9; 0; 0:1); L2 = (0:89; 0:11; 0)

L3 = (0:01; 0:89; 0:1); L4 = (0; 1; 0)

When asked to choose between L1 and L2 ; most people prefer L1 ; But when choosing between L3 and
L4 ; the same people choose L4 : This observation violated independence axiom (How?)

Ellsberg Urn Paradox

This paradox was popularized by Ellsberg, but previously Keynes did mention a version of this.
Suppose you have an urn containing 90 balls: 30 of them are red balls and 60 other balls are either black
or yellow. You don’t know how many black or how many yellow balls there are, but that the total number
of black balls plus the total number of yellow equals 60. The balls are well mixed so that each individual
ball is as likely to be drawn as any other. You have to draw one ball randomly out of the urn. You are now
given a choice between two gambles:

L1 pays $100 if the ball is Red

L2 pays $100 if the ball is Black

Next, you are given a choice between two other lotteries:

L3 pays $100 if the ball is Red or Yellow

L4 pays $100 if the ball is Black or Yellow

Most people strictly prefer L1 to L2 ; but L4 to L3 ; thus violating Expected utility theory (How?). This
case is referred to as Knightian Uncertainty, where the probability of drawing a black or yellow ball is
anything between 31 and 23 :

There are many other behavioral …ndings that can not be explained by the expected utility theory.
However, as a standard way to deal with risky situation, it remains the most popularly used tool, used
widely in applications ranging from stock markets, to the design of insurance contracts.