Session 12
Framing and Decision Making
Asian Disease Problem
Imagine that the US is preparing for the outbreak of an
unusual Asian disease, which is expected to kill 600
people. Two alternative programs to combat the disease have
been proposed. Assume that the exact scientific estimates of
the consequences of the programs are as follows.
Program A: If Program A is adopted,
200 people will be saved.
Program A: If Program A is adopted,
400 people will die.
Program B: If Program B is adopted,
there is 1/3 probability that 600 people
will be saved, and 2/3 probability that no
people will be saved.
Program B: If Program B is adopted,
there is 1/3 probability that 1/3
probability that nobody will die, and 2/3
probability that 600 people will die.
Which of the two programs would you favor?
Which of the two programs would you favor?
Ticket versus Lottery

Ticket


$100 for traffic ticket and be done with it, or challenge it,
and take a 50-50 chance of getting it dismissed or paying
double the ticket value (for court fee).
Lottery

Two stage game in which, at end of Stage 1, you have
won $100. Now you choice is to take your cash and leave
or go to Stage 2, taking a 50-50 chance of making an
additional $100 or losing $100.
Framing and Risk Preference

Tendency to choose riskier options under loss frames rather than gain
frames.
 A pure framing effect (Asian Disease Problem)
 A genuine reflection effect (Ticket/Lottery)

Has to do with shape and curvature of the utility functions in the domain
of losses and gains.

This is described in Prospect Theory (Kahneman and Tversky, 1979).

Violates the general assumption of economic behavior that people should
respond to absolute outcomes not relative outcomes.
Prospect Theory

Over weighting of certain outcomes


Value-function



Certain losses are more aversive, certain gains are more
attractive.
Concave in gains
Convex and steeper in losses
Decreasing marginal utility and disutility

The psychological intensity of outcomes of increasing
magnitude is progressively smaller.
Prospect Theory and Reflection
Effect
Value of Outcome
V(Risk-free)
V(Risky)
-90
-50
-10
10
50
V(Risky)
V(Risk-free)
90
Magnitude of
Outcome
Why then…?

Do people buy lotteries?

Do people buy insurance?

Limitations of Prospect Theory


Holds better in low involvement decisions
Holds better when outcomes are amenable to mathematical
computation
S-P/A Theory

Security-Potential




Aspiration Level


Security orientation leads to risk aversion
Potential orientation leads to risk seeking
A personality trait
The minimum payoff/maximum loss rules the risky option in or
out
No clear predictions under losses and gains
Where’s the Fat?

Package A states that product contains 75% fat.

Package B states that product is 25% fat free.

Which would be preferred more?
Attribute Framing

Varying the description of the option by framing the attribute of the
option.

Specifically, focusing on presence of undesirable versus desirable
attributes, or absence of desirable versus undesirable attributes.

Positive attribute framing results in higher persuasion compared to
negative attribute framing.

Works by trigger differential elaboration.
 Positive frames cause contemplation of positive states of affairs relative to
negative frames.
Skin Exam

By taking the skin exam now, you will give yourself
an opportunity to detect and treat any problems.


Positive goal frame or outcome frame
By not taking the skin exam now, you give up an
opportunity to detect and treat any problems.

Negative goal or outcome frame.
Loss Aversion Explanation for Goal
Framing Effects


Everything else being
equal, people are more
averse to losing than they
are attracted toward gains
of equal magnitude.
Hence, greater motivation
to avoid losses.
Value of Outcome
Magnitude of
Outcome
Choosing versus Rejecting




What if we asked people to add options to a base computer
versus remove options from a fully loaded computer.
Principle of task invariance suggests, nothing should
change.
People should wind up with very identical final
configurations.
Study:



Simulated computer purchase scenario.
Online study
Subjects randomly assigned to choose/reject task
Effect on Size of Final Option Set

Rejecting (paring down)
results in a significantly
larger option set compared
to choosing (building up). 5
4

Y-axis = Number of
options.
3
2
1
0
Reject
Choose
Effect on Value of Choice Set

Rejecting results in significantly
higher value of final option set
than choosing.
250

Y-axis = dollar value of the
options in final set.
200
150
100
50
0
Reject
Choose
Biases in Decision Making
Enhance Production Capacity?






9 months back your team decided to go ahead with major modification to
your production facilities.
The budget was $10m.
The project is on time, and scheduled to be completed next month.
The project is on budget, and you have invested $9m to date.
You are now informed that a major shift in the market place will put your
product on a downward path to extinction, and there is nothing you can
do about it.
The question before you is, should you invest the remaining $1m and
complete the planned plant modification, and continue or should you
abandon?
Results



A significant majority said,
invest.
You are not alone.
Many examples





Vietnam
Concorde
Hubble
And this is called, sunk
cost fallacy.
Why does this happen?
70
60
50
40
30
20
10
0
Abandon
Continue
The Source of Sunk Cost Fallacy
Investment at T1
Investment at T2
Investment at T3
Investment at T4
• This is a prospect theory account
• Incremental losses look smaller in comparison with the potential
gain.
• As you invest more, it looks like a bigger waste if you abandon,
and incremental investment looks miniscule in comparison with
the increasingly remote gain.
• The killer is that the decision-maker starts abandoning the
probability that modifies the gain.

A cost is sunk if it is irretrievable, and has no effects on future payoffs.
Alternate Explanation for Sunk Cost
Fallacy

Cognitive Dissonance

What is it?

Tendency to engage in revisionist history while looking
back at decision.

Engage in behavior, especially information search that:


Strengthens the reasons in favor of decision
Weakens reasons against.
More on Sunk Cost Fallacy



Reverse of ‘Hindsight Bias’?
Also known as ‘irrational escalation of
commitment’.
Many anecdotes, somewhat weak evidence.


Bought two tickets for Shows A and B worth $50 and $100, and
prefers Show A, but finds out that they are for the same day/time,
and people choose to go to Show B.
Shuttle mission justified on investments and lives lost.
Source: Friedman et al (2003), preliminary version of manuscript
Trading Scenario






Let us say, a slightly intoxicated fellow with plenty of money shows up in
class and says “I am feel like giving stuff away”.
I tell him, do the following:
 Give half of this class $2.00
 Give the other half, a lottery ticket worth some unspecified amount
 Randomly decide who belongs in which half.
He does just that.
But then, the Donor thinks a bit and says, “Hey this is good only for those
whose preferences for the lottery/money matches what they got…what if
they don’t like what they got?”
To which, I say, “Okay, give them an opportunity to switch to the other
gift”.
What do you think happened?
Endowment Effect




Mere possession of an object causes it to be over valued.
Result, very few people traded.
Replicated in trades involving coffee mugs for $6.00.
Replicated in trades involving coffee mugs for chocolate,
89% and 90% respectively chose to retain what was given to
them.



People are willing to pay less than what they would expect others’ to
pay.
Cannot be attributed to a simple economic market rationality.
The dropping of the atomic bomb was justified by the $2b spent on
the Manhattan Project.
Indifference Curves
Desirable Attribute A
Trade down
Attribute A
Trade down
Attribute B
Desirable Attribute B


Indifferences curves represent the points in satisfaction space that are equivalent for two combinations
of items/attributes of value.
For example, a person might have in indifference curve between engine power and upholstery.


Equally satisfied with a car with a powerful engine with cloth interior as with a care with a less powerful engine but
a leather interior.
Indifference curves are not supposed to intersect, regardless of whether you start with high interiors
and trade down to get more power, or start with high power and trade down to get a better interior.
More on Endowment Effect

Endowment effect occurs because indifference
curves intersect.

It affects inexperienced traders more than it does
experienced traders.


As experience accumulates and trading intensity picks up,
endowment effects gives way.
Can be thought of as behavioral inertia.