BEHAVIORAL ECONOMICS

ECON F345

Dushyant Kumar
BITS Pilani, Hyderabad Campus
Decision Making Under Certainty

Decision Making Under Certainty

Introduction

I Decision theory or choice theory vs. game theory.

I Abstract modeling- robustness of a model..
I Two basic assumptions of choice theory- rationality and,
common knowledge.
I Rationality: An individual is rational if she has well-defined
objectives or preferences over the set of possible outcomes and
she implements the best available strategy to pursue them.
I Common knowledge: An information I is common
knowledge if everybody knows it, everybody knows that
everybody knows it, if everybody knows that everybody knows
that everybody knows it, and so on!
Theory of Rational Choice

I The decision maker chooses the best possible choice given the
information set and preferences.
I Rationality just imposes some ‘consistency’ restrictions on the
agent’s behavior, it doesn’t put any directional and qualitative
restrictions.
I Actions: set of all possible actions- A.
I In any situation, a subset of A will be available to agent.
I Like a consumer will be restricted to the budget set- feasibility
constraints..
I Example: To watch a late night soccer or cricket game or
have a routine sleep.
Consume now or save-invest and consume later..
Preferences

I Basic assumption: preferences to be complete and transitive.

Definition
A preference relation  is complete if for all x, y ∈ X , either
x  y or y  x or both.

A preference relation  is transitive if for all x, y , and z ∈ X

x  y . and y  z, x  z.

A preference relation  is rational if it is complete and transitive.

I Fundamental to all analysis, however some of the behavioral

models even relax these ones!
Payoff functions

I We can describe a preference simply as an ordering-

a  b  e ∼ c  d, and work with it.
or, we can work with payoff/value function-
u(A, ) : A → R.
I A payoff function associate a number to an action in such a
way that-
u(a) > u(b) if and only if the decision-maker prefers a to b.
I Examples: utility function in consumption case, profit function
in production case..
I It is important to stress here that the payoff function is just a
way to represent agent’s preference. Hence its ordinal.
Preferences & Payoff Functions
I In standard approach, it tells nothing about the ‘intensity’ of
preferences..
I So, if u() represents an agent’s preferences, then any
increasing function of u also represents the preference..
I Sometimes, like in the case of a firm, its natural to first think
in terms of payoffs (rather than ‘preferences’). Although here
as well, we are essentially working with underlying preferences-
more money is better..
I Theory of rational choice-
Definition (Choice Set)
The action chosen by a decision-maker is at least as good as every
other available action, as per her preferences.

C (A; ) = {x ∈ A|x  y for all y ∈ A}

Theory of Rational Choice

I The optimal choice need not be singleton.

I If A is finite, then C (A; ) is non-empty.
I Suppose A is infinite, for example A = {x|x ∈ [0, 1)}.
I Here consider the preference ‘more is better’ or, x  y if
x > y , then C (A; ) =?
I Consistency- Suppose the agent is indifferent between a and
b. So, when faced with a choice between a and b, its fine that
the agent sometime chooses a and sometime choose b.
I But suppose that the agent always chooses a from {a, b} and
chooses b from {a, b, c}. Is it consistent?
Theory of Rational Choice

I Formally,
Proposition
Suppose  is complete and transitive. Then (1) for every finite
non-empty set A, C (A; ) 6= ∅ and, (2) if x, y ∈ A ∩ B, and
x ∈ C (A; ) and y ∈ C (B; ); then x ∈ C (B; ) and
y ∈ C (A; ).
I We can construct the payoff function using this result.
I Most of the time additional assumptions of continuity,
monotonicity, and convexity is assumed on the preferences
to get a ‘nice structured’ payoff function.
Theory of Rational Choice & Empirical Support

I Empirical support: theory of rational choice is is not ‘always

practiced’ by decision makers-
experiments, observations, surveys do suggest deviations..
I Bounded rationality, behavioral economics..
I Still rational choice remains quite (or most!) widely used-
best bet among available options?.
I Also, the rational choice theory remains the starting as well as
reference point for alternative theories..
I Nudge to what?
I Decision or choice under uncertainty requires further
structure, we are going to analyze it at latter stage..
Utility Maximisation, , Profit Maximisation,...

I max U(X , Y ) such that Px X + Py Y ≤ M.

I Marginal cost equals marginal benefit..
I max π(Q, PQ , Wi ) subject to Q(PQ ).
I Marginal cost equals marginal benefit..

I Optimisation implies that one should focus on opportunity

cost rather than just explicit cost..
Opportunity Costs
Opportunity Costs

I Out-of-pocket or explicit cost vs. implicit cost..

I Opportunity cost- max value of alternative paths..
I Example: You won a free ticket to see an Eric Clapton
concert (which has no resale value). Bob Dylan is performing
on the same night and is your next-best alternative activity.
Tickets to see Dylan cost $40. On any given day, you would
be willing to pay up to $50 to see Dylan. Assume there are no
other costs of seeing either performer. Based on this
information, what is the opportunity cost of seeing Eric
Clapton?
1. $0
2. $10
3. $40
4. $50
Opportunity Costs

I Suppose there are n available actions- a1 , a2 , · · · , an .

I Respective utilities/values (adjusting for ‘explicit’ costs)-
u(a1 ), u(a2 ), · · · , u(an )..
I Opportunity cost of an action ai (OC (ai ))
= max{u(a1 ), u(a2 ), · · · , u(ai−1 ), u(ai+1 ), · · · , u(an )}
I Here we are assuming away any ‘explicit costs’ attached to an
action.
I ai is going to be a rational choice if u(ai ) ≥ OC (ai )
I To maximise the one’s welfare, the opportunity cost must be
taken into account rather than just accounting cost..
I Many a times people ignore it- maybe attidude, maybe its not
that easy to use opportunity cost- complexity...
Opportunity Costs

I Many a times its just that getting at opportunity cost require

‘too much time and effort’- trade-off between accuracy and
cost!
? search theory-optimal stopping model..
? observe this is a rational approach once we incorporate the
search cost!
? Not a bias..
I Windfall gain- best way to spend a rupee should be
independent of how it is earned! often violated- mental
accounting..
I Spending through contingency- structural issues.. opportunity
cost? business class air tickets pricing..
I Room service in high-end hotels.
Opportunity Costs

I Many a times we think in terms of percentage -

Jacket/calculator problem - in an study-
... 68 percent of respondents were willing to make the drive
to save $5 on the $15 calculator. Yet only 29 percent were
willing to make the drive to save $5 on the $125 calculator.
I BIG APPLICATION- international trade..
I Everyday example- driving to office in a personal car vs
using public transport..
personal cost vs. social cost
I Yet another example: Time is money..
I Opportunity cost- not just money, time as well..
Opportunity Costs

Opportunity cost of going for a higher studies

Sunk Costs

I Fixed Costs should not matter in the ‘level’ of operation..

I Sunk Costs- irrecoverable costs
I Relationship between sunk costs and fixed costs.
I Once a (sunk) cost is incurred, it should not matter!
I Consider a project requiring current investment of Rs. 5 lakhs.
Return- Rs. 17 lakhs with probability 0.10... will you take this
project?
Sunk Costs

I Now suppose you have already invested Rs. 8 lakhs in this

project..

I Pressure to make it right?

I Season tickets
Sunk Costs

I Now suppose you have already invested Rs. 8 lakhs in this

project..

I Pressure to make it right?

I Season tickets
Sunk Costs

I Now suppose you have already invested Rs. 8 lakhs in this

project..

I Pressure to make it right?

I Season tickets
Sunk Costs & Public Policy
Sunk Costs

I Some other examples:

I A nonrefundable, nontransferable ticket.
I The time already spent in a particular career.
I The time, effort and money spent in a relationship.
I Nonrefundable money invested into research and development
projects.
I The loss of soldier’s lives in an ongoing war or conflict.
The decision to invest additional resources in a losing ac-
count, when better investments are available, is known
as the sunk-cost fallacy, a costly mistake that is observed
in decisions large and small...The sunk-cost fallacy keeps
people for too long in poor jobs, unhappy marriages, and
unpromising research projects.
Thinking Fast and Slow by Daniel Kahneman.
Sunk Costs

I However many-a-times, ‘seemingly’ sunk cost fallacy serves

very crucial role in better decision making..
I Sunk cost fallacy can be modeled as an optimal response to
memory and computational constraints.. commitment device-
season tickets
I Sunk cost can be just a response to limited memory- since
initially we agreed to take this project at this cost, there must
be some ‘reason’ to it..
I Example- New CEO agreeing to continue with the project
initiated by former CEO, even though not able to see the
benefits clearly..
I Revenge- can it be rational?
I Baliga & Ely papers..
Menu Dependence and Decoy Effect
I If x is always chosen from {x, y }, then y should not be chosen
from {x, y , z}.
I This consistency requirement is violated in practice
many-a-times..
I Consider following subscription prices of The Economist-

notice option 2 and option 3- no one is going to pick option 2,

right?
Menu Dependence and Decoy Effect

I Why is option 2 there in the first place?

I Option 2 is there to make option 3 look more attractive!
I In experiments-
between 1 and 3- individuals typically choose 1.
from 1,2, and 3- people mostly choose 3! Inconsistent?
I Standard consumer theory take preference to be given,
particularly independent of frame or reference.. not true
here, not true in many cases..
I Market players and others can exploit it to their beneifit!
I We observe similar kind of strategy often in politics,
marketing and other areas..
I Decoys..
Menu Dependence and Decoy Effect

I Why is option 2 there in the first place?

I Option 2 is there to make option 3 look more attractive!
I In experiments-
between 1 and 3- individuals typically choose 1.
from 1,2, and 3- people mostly choose 3! Inconsistent?
I Standard consumer theory take preference to be given,
particularly independent of frame or reference.. not true
here, not true in many cases..
I Market players and others can exploit it to their beneifit!
I We observe similar kind of strategy often in politics,
marketing and other areas..
I Decoys..
Menu Dependence and Decoy Effect

I Why is option 2 there in the first place?

I Option 2 is there to make option 3 look more attractive!
I In experiments-
between 1 and 3- individuals typically choose 1.
from 1,2, and 3- people mostly choose 3! Inconsistent?
I Standard consumer theory take preference to be given,
particularly independent of frame or reference.. not true
here, not true in many cases..
I Market players and others can exploit it to their beneifit!
I We observe similar kind of strategy often in politics,
marketing and other areas..
I Decoys..
Menu Dependence and Decoy Effect
Menu Dependence and Decoy Effect
Menu Dependence and Decoy Effect

I Let x-axis be speed, and y-axis be safety..

I So in order to sell the target, one need to introduce a decoy
such that it is
? inferior than competitor in safety but better in speed..

I Essentially you want customers to focus more on speed and

then look for (more) safety (from the decoy point)!
I Notice decoy should be inferior to the target from both
dimensions- speed and safety..
Menu Dependence and Decoy Effect
Menu Dependence and Decoy Effect
Menu Dependence and Decoy Effect

where should the decoy be located?

Menu Dependence and Decoy Effect
I Decoy effect is one form of menu dependence; another type of
menu dependence- extremeness aversion..
I An expensive brand company might introduce a inasanely
expensive product to make other products ‘look good
(cheap)’..
I Customers kind of look for reason to pick some products or
reject some products- decoy effect as well as extremeness
aversion effect sort of provide it!
I Simple enough idea- explains a significant amount of observed
behaviour..
I These are all related to reference dependence and framing
effects.
I Under rational choice theory, the choice should be
independent of reference and frame.
I Many a times, this is violated in practice.
Endowment Effect

I Another common bias is endowment effect- once you

have/own something, you hate to loose it.
A wine-loving economist we know purchased some nice
Bordeaux wines years ago at low prices. The wines have
greatly appreciated in value, so that a bottle that cost only
$10 when purchased would now fetch $200 at auction.
This economist now drinks some of this wine occasionally,
but would neither be willing to sell the wine at the auction
price nor buy an additional bottle at that price. Thaler
(1980) called this pattern- the fact that people often de-
mand much more to give up an object than they would be
willing to pay to acquire it- the endowment effect.
Kahneman, Knetsch, and Thaler (1991)
Endowment Effect
I Many a times the endowment effect is confused with
diminishing marginal utility. They are not same or, even
related.
I Example-Suppose you own a coffee mug, whats the min price
that you ready to sell it for? (A)
I Whats the price that you are ready to pay for another exactly
same mug? (B)
I Typically A R B? rationale?
I A > B is not endowment effect. Since you already have a
mug, you may not have much of a usage/value for the second
one..
I So you need to compare it with the max price that you were
ready to pay for the first mug, controlling for quality and all
other issue.. (C)
I A > C is endowment effect.
Endowment Effect
I Many a times the endowment effect is confused with
diminishing marginal utility. They are not same or, even
related.
I Example-Suppose you own a coffee mug, whats the min price
that you ready to sell it for? (A)
I Whats the price that you are ready to pay for another exactly
same mug? (B)
I Typically A R B? rationale?
I A > B is not endowment effect. Since you already have a
mug, you may not have much of a usage/value for the second
one..
I So you need to compare it with the max price that you were
ready to pay for the first mug, controlling for quality and all
other issue.. (C)
I A > C is endowment effect.
Endowment Effect
I Many a times the endowment effect is confused with
diminishing marginal utility. They are not same or, even
related.
I Example-Suppose you own a coffee mug, whats the min price
that you ready to sell it for? (A)
I Whats the price that you are ready to pay for another exactly
same mug? (B)
I Typically A R B? rationale?
I A > B is not endowment effect. Since you already have a
mug, you may not have much of a usage/value for the second
one..
I So you need to compare it with the max price that you were
ready to pay for the first mug, controlling for quality and all
other issue.. (C)
I A > C is endowment effect.
Endowment Effect

I WTA (wilingness-to-accept) vs. WTP (willingness-to-pay)..

I In the previous wine collection example- the economist is not
buying additional bottle at $ 200, so WTP(C ) < $200.
I Also not willing to sell at $ 200, so WTA(A) > $200.
hence, A > C .
I Social psychology explanation- when one owns something, it
becomes a part of “extended self”..
I The term endowment effect was coined by Thaler in his 1980
paper.
I Before proceeding further, to understand and appreciate the
relevance of the endowment effect, we need to cover the
Coase theorem!
Endowment Effect

I Ronald H. Coase-The Problem of Social Cost (1960)

I Coase Theorem: Under ‘certain conditions’, economic actors
can arrive at an efficient solution to an externality without
direct government involvement.
I To understand the Coase theorem, consider the baker-doctor
example originally given by Coase.
I Suppose a baker and a doctor operate their business from
same building.
I Issue- the loud noise created by baker’s operations create
disturbance for doctor’s practice..
I Externality- quite common in many situations..
Coase Theorem and Endowment Effect

I How to resolve the issue- should baker be treated strictly and

be asked to install noise cancellation? who is victim here?
I Typically a tendency to blame the baker- the source of noise!
I Why shouldn’t the doctor be asked to shift to some other
place or install noise proof cabins?
I Its difficult to say that which one is ’justified’ here..
I Coase response- from efficiency point of view, from
social welfare point of view, it does not matter!
I Policy or rule-wise two setups possible here-
1. doctor has the right to operate in peaceful environment.
2. baker has the right to create noise as its a input for his
production.
Coase Theorem and Endowment Effect

I How to resolve the issue- should baker be treated strictly and

be asked to install noise cancellation? who is victim here?
I Typically a tendency to blame the baker- the source of noise!
I Why shouldn’t the doctor be asked to shift to some other
place or install noise proof cabins?
I Its difficult to say that which one is ’justified’ here..
I Coase response- from efficiency point of view, from
social welfare point of view, it does not matter!
I Policy or rule-wise two setups possible here-
1. doctor has the right to operate in peaceful environment.
2. baker has the right to create noise as its a input for his
production.
Coase Theorem and Endowment Effect

I What we are going to argue here is that the government

should not waste too much effort on finding which setup is
optimal..
I To access the optimality, suppose that the noise cancellation
equipment is going to cost Rs. 35,000/- to the baker and the
noise proof cabins are going to cost 45,000/- to the doctor.
I Lets consider these setup one-by-one.. First, say the doctor
has the right.
I What is going to be the outcome, who is going to install the
new setup, who is going to pay for it?
I As the doctor has the right of peaceful operation, the baker
need to install the equipment.. its going to cost him Rs. 35k.
I Can the baker somehow convince the doctor to install the
noise proof cabin? NO. (why?)
Coase Theorem and Endowment Effect

I What we are going to argue here is that the government

should not waste too much effort on finding which setup is
optimal..
I To access the optimality, suppose that the noise cancellation
equipment is going to cost Rs. 35,000/- to the baker and the
noise proof cabins are going to cost 45,000/- to the doctor.
I Lets consider these setup one-by-one.. First, say the doctor
has the right.
I What is going to be the outcome, who is going to install the
new setup, who is going to pay for it?
I As the doctor has the right of peaceful operation, the baker
need to install the equipment.. its going to cost him Rs. 35k.
I Can the baker somehow convince the doctor to install the
noise proof cabin? NO. (why?)
Coase Theorem and Endowment Effect

I Now consider the second setup- the baker has the right to
create noise!
I Now what is going to be the outcome, who is going to install
the new setup, who is going to pay for it? Will the doctor
install the noise proof cabin? NO!
I The doctor can ’negotiate’ with the baker- they can have the
following ’deal’-
the doctor pays Rs. x, x ∈ (35k, 45k) to the baker, the baker
in turns install the noise cancellation equipment..
I Both gains from such deal!
I So does not matter what the rule is, the outcome is same- the
baker install the noise cancellation equipment, and this
outcoem is efficient/optimal from social welfare point of
view...
Coase Theorem and Endowment Effect

I Now consider the second setup- the baker has the right to
create noise!
I Now what is going to be the outcome, who is going to install
the new setup, who is going to pay for it? Will the doctor
install the noise proof cabin? NO!
I The doctor can ’negotiate’ with the baker- they can have the
following ’deal’-
the doctor pays Rs. x, x ∈ (35k, 45k) to the baker, the baker
in turns install the noise cancellation equipment..
I Both gains from such deal!
I So does not matter what the rule is, the outcome is same- the
baker install the noise cancellation equipment, and this
outcoem is efficient/optimal from social welfare point of
view...
Coase Theorem and Endowment Effect

I The Coase theorem implies that the focus of the government

should be on (1.) well defining property rights, and (2.)
making negotiations easy and costless..
I ‘certain conditions’- well-defined property rights, absence of
transaction cost..
I Obviously in many cases, these conditions can’t be met, there
the government needs to intervene directly.. environment,
traffic, etc..
I Now, why is this seemingly fun example such important for
econ theory? and how are Coase theorem and endowment
effect related?
I Coase theorem is crucial for policy making, it tells us how
government should approach such issues, gives a lot of
independence and comfort to the policy makers..
Coase Theorem and Endowment Effect

I However in the presence of endowment effect, as we have seen

in earlier examples, the negotiations fails..
I Individuals value objects and rights differently when they own
vs. they don’t own..
I Hence the outcome is not going to be independent of a
particular property rights- in the presence of endowment
effect..
trouble for policy makers!
I We are going to take some examples- first one, consider the
issue of urbanisation and the required land acquisition..
I Typically, one can say that we should go for urbanisation if it
improves social welfare..
Endowment Effect

I The environmental and resource economists studying issues

involving environmental policy often need to find a value for
goods that cannot be traded in the market.
I A new shopping mall in the neighborhood- is it worth it?
I Potential consumers could be asked their willingness-to-pay
(WTP) for the new mall..
I Those who reside nearby might lose their scenic view, might
face discomfort due to increased traffic, noise, bright lights
and all.. willingness-to-accept (WTA) for this discomfort..
I If the WTP > WTA, the new constrution is going to improves
the social welfare, we should go for it..
I Most of the times, it has been observed that in such surveys,
willingness-to-accept are on higher side compared to
willingness-to-pay..
Endowment Effect

I It seems to imply that such ‘new developments’ are not worth

it!!
I Many such protests and all..
I Jack Knetsch, a environmental and resource economist, begin
investigating the endowment effect on an experimental basis
in a series of papers culminating in his work with Daniel
Kahneman and Richard Thaler.
I They found a crucial role of endowment effect.
For example, if you are putting a certain valuation of say, Rs.
X on the discomfort in the sense that you are willing to go
through the discomfort if you are paid any amount ≥ X , are
you also willing to pay X for those comfort in similar
situation?
NO
Endowment Effect

I It seems to imply that such ‘new developments’ are not worth

it!!
I Many such protests and all..
I Jack Knetsch, a environmental and resource economist, begin
investigating the endowment effect on an experimental basis
in a series of papers culminating in his work with Daniel
Kahneman and Richard Thaler.
I They found a crucial role of endowment effect.
For example, if you are putting a certain valuation of say, Rs.
X on the discomfort in the sense that you are willing to go
through the discomfort if you are paid any amount ≥ X , are
you also willing to pay X for those comfort in similar
situation?
NO
Endowment Effect
Endowment Effect
Endowment Effect

I So its difficult to assess the optimality..

I Kink in the indifference curves at the endowment point ⇒
indifference curves intersecting!
I Observe that in the above graph- if you are at point x, you
strictly prefer x to y .
if you are at point y , you strictly prefer y to x.
Status quo bias
I The ‘endowment’ can as well be thought of expectations and
aspirations.
need not be just what ’actually’ you have..
I Another example- leg space trading in airlines..
Endowment Effect
Endowment Effect

Source: The Economist, May 15th 2017

Next- Anchoring and Adjustment Bias

Anchoring and Adjustment
I Many a times its not feasible to do proper calculation and
analysis- time and other resource constraints..
I People try to rely of anchors- reference points..
I Not that bad an idea if anchors are chosen properly and
optimal (sufficient) adjustment is done..
I Sampling, Newton-Raphson method..
I Problem arises when (1) the anchor itself depends on contexts
and frames, and (2) not sufficient adjustment is done..
I Example: Suppose you are told that the average price of a
car on Hyderabad road is Rs. 50 lakhs..
Response-

Compare this with the statement that the average price of a

car on Hyderabad road is Rs. 5 lakhs..
Response-
Anchoring and Adjustment
I Many a times its not feasible to do proper calculation and
analysis- time and other resource constraints..
I People try to rely of anchors- reference points..
I Not that bad an idea if anchors are chosen properly and
optimal (sufficient) adjustment is done..
I Sampling, Newton-Raphson method..
I Problem arises when (1) the anchor itself depends on contexts
and frames, and (2) not sufficient adjustment is done..
I Example: Suppose you are told that the average price of a
car on Hyderabad road is Rs. 50 lakhs..
Response-

Compare this with the statement that the average price of a

car on Hyderabad road is Rs. 5 lakhs..
Response-
Anchoring and Adjustment

I (1 × 2 × · · · × 8) vs. (8 × 7 × · · · × 1)- often we violate

procedure invariance or path independence..
I Often starting point serves as an anchor..
I So starting statement, first impression, etc. matter!
I MRP and discounts..
I The Winner’s Curse and Anchoring and Adjusting
I Winner’s curse: uncertainty around the valuation, highest
bidder i.e. the winner also has the highest probability err in
bidding on the upper side.. might result in loss going ahead..
I More importantly, the winning bid is often taken as ‘market
value’, a proxy for ‘fair valuation’, etc..
I So going ahead others might also form upward biased beliefs
about the valuation!
Heuristics-and-Biases Program

I All these biases that we discussed so far, are a part of the

heuristics-and-biases program
I Heuristics-and-biases program: Kahneman and Tversky.
I Kahneman and Tversky ran a series of experiments, mostly in
1970s- explore the process of individual’s decision making..
I Individuals often use heuristics to save time and effort
required to solve everyday problems- good many-a-times, but
leads to biases many-a-times as well..

Next- Strategic Decision Making