LES 1: STANDARD ECONOMIC MODEL
STANDARD ECONOMIC MODEL
Neo-classical economic model is the way most economists think about consumer welfare & choice
a) People act with full information -> full external knowledge
b) People have known preferences -> full internal knowledge
c) People choose the best option available -> rational choices
Agents are assumed to be fully rational, and be driven purely by their self-interest
not true: satisfice rather than maximise + information is not always available = BE
Theories are usually normative + descriptive; this may lead to tensions if they fail descriptively
Normative theories: tell us how we should behave to obtain a certain goal (utility max.)
Descriptive theories: how people really behave, may (not) be the same as the normative theory
o Probability
= a number between 0 and 1 that indicates a likelihood that a particular outcome will occur,
0 means the event is impossible, 1 means it is certain
Probability is known = RISK (bv: flipping a coin is 0,5)
Probability is unknown = UNCERTAINTY
the probability of all possible events sum to 1
binary prospects with two prospects (x , y) and probability (p):
∼ = indifference
≻ = strict preference
o Expected value
= the value of each possible outcome times the probability of that outcome
= 𝐸𝑉(𝑥, 𝑝; 𝑦) = 𝑝𝑥 + (1 – 𝑝)𝑦
bv: Suppose you are planning to play at an outdoor concert. The probability of rain tomorrow is
0.30, and thus the probability of no rain is 0.70. Suppose you will make €500 if it doesn’t rain , but
only €100 if it rains: EV = (0.70) (500) + (0.30) (100) = €380
bv: You have the option to participate in a game where:
With a 50% chance you win €100. With a 30% chance you win €50. With a 20% chance you win
nothing (€0). You want to calculate the expected value of your winnings:
EV(game) =(0.5×100)+(0.3×50)+(0.2×0) =50+15+0 = € 65
o St Petersburg paradox
= A coin is tossed. If it comes up heads, you are paid €2. Then the coin is tossed again. If it comes up
heads again, you are paid €4= 22; and so on. When the coin comes up tails the game is over.
overall people want max to pay €25 to play this gamble
expected value is infinite:
but still some people don’t want to play… -> expected
utility
1
, o Expected utility (a solution to the St. Petersburg paradox by Daniel Bernoulli)
= the satisfaction or pleasure a person derives from consuming a good, service, level of wealth
(ex: the first euro’s winning means more than the last ones// poor people get more satisfaction
from winning money)
utility = If you prefer eating apples (A) to eating chocolate (C ), than U(A)=2 and U(C )=1
choice = revealed preference : if you choose X then this “reveals” that you prefer X to Y ( X ≻ Y)
Decreasing marginal utility: utility increases as consumption increases but at a diminishing rate =
expected utility = 𝐸𝑈(𝑥𝑖, 𝑝𝑖)= ∑ p i U (x i )
expected value = 𝐸𝑉(𝑥𝑖, 𝑝𝑖) = ∑ p i x i
o Certainty equivalent
= to find the sure amount of money that makes a decision maker indifferent between playing the
prospect and obtaining that amount (that’s the amount that makes you indifferent from playing or
taking the money)
=
Economic agents in the standard economic model:
motivated by expected utility maximization
= p(st)
The utility is governed by selfish concerns, it does not take into consideration the utility of
other
t
= U (x i l st)
They are Bayesian probability operators = I update my probabilities every time I receive new
info
= p(st)
They have consistent time preferences according to the discounted utility model = I always
stick to the decision I made
t
=
According to the standard model, individual i at time t = 0 maximises expected utility subject to a
probability distribution p(s) of the states of the world s ∈ S
2
, LES 2: HEURISTICS AND EXPECTED UTILITY PARADOXES
HEURISTICS AND BIAS
o Definitions and introductory concepts
heuristics = ‘Rule of thumb’ or a simple rule of behaviour by which a person solves a problem.
(bv: buying what you usually do) = mental shortcut to solve problems and make quick judgments.
bias = Systematic suboptimal judgments that can result as a consequence of the heuristic process
search heuristics (= complementary heuristics):
try until aspiration level is met (satisficing)
+ eliminate by aspects that don’t meet your aspiration level
+ directed cognition: try each product and treat it as if it’s the last one
utility and search
preferences tell you how much utility you would get from a combination of goods and money
to maximize the utility, you should do a search heuristic
Bayes rule: describes the probability of an event, based on prior knowledge of conditions that
might be related to the event
= P(H|E) = [P(E|H)*P(H)] / P(E)
P(H|E) The posterior probability: the probability of the hypothesis H being true given the evidence E.
P(E|H) The likelihood: the probability of observing the evidence E given that the hypothesis H is true.
P(H) The prior probability: the initial probability of the hypothesis H being true before considering the evidence.
P(E) The marginal likelihood or evidence: the total probability of observing the evidence E, regardless of the
hypothesis
bv: As part of a clinical study, you are being tested for a rare disease, which affects 1 in 10,000
people. The test correctly detects the disease when it is present 99% of the time; it also correctly
detects the absence of the disease 99% of the time (it has 1% false positives). Imagine now the test
comes back positive; what is the probability that you indeed have the disease?
-> 𝑃(𝑡𝑒𝑠𝑡 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒) = 𝑃(𝑡𝑒𝑠𝑡 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑑𝑖𝑠𝑒𝑎𝑠𝑒) × 𝑃(𝑑𝑖𝑠𝑒𝑎𝑠𝑒) + 𝑃(𝑡𝑒𝑠𝑡 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑛𝑜 𝑑𝑖𝑠𝑒𝑎𝑠𝑒) × 𝑃(𝑛𝑜 𝑑𝑖𝑠𝑒𝑎𝑠𝑒)
-> P(test positive) =(0.99 x 0.0001) + (0.01 x 0.9999) = 0.010098
The posterior probability of having the disease: [(0.99x 0.0001) / 0.010098] = 0.0098 = 0.98%
Despite the test being highly accurate, the low prevalence of the disease leads to a high
probability that the positive result is a false positive
o Heuristic 1: Representativeness
= evaluate the likelihood that object A belongs to category B, by the extent to which A
resembles characteristics of B
The issue arising from this, is that similarity may be determined by many elements not affecting
probabilities, leading to bias: Systematic biases may therefore result from this heuristic
Bv: “John is very shy and withdrawn, invariably helpful, but with little interest in people, or in the
world of reality. He is very tidy, he has a need for order and structure, and a passion for detail”.
What is the probability that John is a) a farmer; b) a salesman; c) a librarian; d) a physician ?
most people guess that he is a librarian
The issue is that there are many more salesmen than librarians, this ignores the prior probability
(base rate) = the failure of Bayesian updating
3
STANDARD ECONOMIC MODEL
Neo-classical economic model is the way most economists think about consumer welfare & choice
a) People act with full information -> full external knowledge
b) People have known preferences -> full internal knowledge
c) People choose the best option available -> rational choices
Agents are assumed to be fully rational, and be driven purely by their self-interest
not true: satisfice rather than maximise + information is not always available = BE
Theories are usually normative + descriptive; this may lead to tensions if they fail descriptively
Normative theories: tell us how we should behave to obtain a certain goal (utility max.)
Descriptive theories: how people really behave, may (not) be the same as the normative theory
o Probability
= a number between 0 and 1 that indicates a likelihood that a particular outcome will occur,
0 means the event is impossible, 1 means it is certain
Probability is known = RISK (bv: flipping a coin is 0,5)
Probability is unknown = UNCERTAINTY
the probability of all possible events sum to 1
binary prospects with two prospects (x , y) and probability (p):
∼ = indifference
≻ = strict preference
o Expected value
= the value of each possible outcome times the probability of that outcome
= 𝐸𝑉(𝑥, 𝑝; 𝑦) = 𝑝𝑥 + (1 – 𝑝)𝑦
bv: Suppose you are planning to play at an outdoor concert. The probability of rain tomorrow is
0.30, and thus the probability of no rain is 0.70. Suppose you will make €500 if it doesn’t rain , but
only €100 if it rains: EV = (0.70) (500) + (0.30) (100) = €380
bv: You have the option to participate in a game where:
With a 50% chance you win €100. With a 30% chance you win €50. With a 20% chance you win
nothing (€0). You want to calculate the expected value of your winnings:
EV(game) =(0.5×100)+(0.3×50)+(0.2×0) =50+15+0 = € 65
o St Petersburg paradox
= A coin is tossed. If it comes up heads, you are paid €2. Then the coin is tossed again. If it comes up
heads again, you are paid €4= 22; and so on. When the coin comes up tails the game is over.
overall people want max to pay €25 to play this gamble
expected value is infinite:
but still some people don’t want to play… -> expected
utility
1
, o Expected utility (a solution to the St. Petersburg paradox by Daniel Bernoulli)
= the satisfaction or pleasure a person derives from consuming a good, service, level of wealth
(ex: the first euro’s winning means more than the last ones// poor people get more satisfaction
from winning money)
utility = If you prefer eating apples (A) to eating chocolate (C ), than U(A)=2 and U(C )=1
choice = revealed preference : if you choose X then this “reveals” that you prefer X to Y ( X ≻ Y)
Decreasing marginal utility: utility increases as consumption increases but at a diminishing rate =
expected utility = 𝐸𝑈(𝑥𝑖, 𝑝𝑖)= ∑ p i U (x i )
expected value = 𝐸𝑉(𝑥𝑖, 𝑝𝑖) = ∑ p i x i
o Certainty equivalent
= to find the sure amount of money that makes a decision maker indifferent between playing the
prospect and obtaining that amount (that’s the amount that makes you indifferent from playing or
taking the money)
=
Economic agents in the standard economic model:
motivated by expected utility maximization
= p(st)
The utility is governed by selfish concerns, it does not take into consideration the utility of
other
t
= U (x i l st)
They are Bayesian probability operators = I update my probabilities every time I receive new
info
= p(st)
They have consistent time preferences according to the discounted utility model = I always
stick to the decision I made
t
=
According to the standard model, individual i at time t = 0 maximises expected utility subject to a
probability distribution p(s) of the states of the world s ∈ S
2
, LES 2: HEURISTICS AND EXPECTED UTILITY PARADOXES
HEURISTICS AND BIAS
o Definitions and introductory concepts
heuristics = ‘Rule of thumb’ or a simple rule of behaviour by which a person solves a problem.
(bv: buying what you usually do) = mental shortcut to solve problems and make quick judgments.
bias = Systematic suboptimal judgments that can result as a consequence of the heuristic process
search heuristics (= complementary heuristics):
try until aspiration level is met (satisficing)
+ eliminate by aspects that don’t meet your aspiration level
+ directed cognition: try each product and treat it as if it’s the last one
utility and search
preferences tell you how much utility you would get from a combination of goods and money
to maximize the utility, you should do a search heuristic
Bayes rule: describes the probability of an event, based on prior knowledge of conditions that
might be related to the event
= P(H|E) = [P(E|H)*P(H)] / P(E)
P(H|E) The posterior probability: the probability of the hypothesis H being true given the evidence E.
P(E|H) The likelihood: the probability of observing the evidence E given that the hypothesis H is true.
P(H) The prior probability: the initial probability of the hypothesis H being true before considering the evidence.
P(E) The marginal likelihood or evidence: the total probability of observing the evidence E, regardless of the
hypothesis
bv: As part of a clinical study, you are being tested for a rare disease, which affects 1 in 10,000
people. The test correctly detects the disease when it is present 99% of the time; it also correctly
detects the absence of the disease 99% of the time (it has 1% false positives). Imagine now the test
comes back positive; what is the probability that you indeed have the disease?
-> 𝑃(𝑡𝑒𝑠𝑡 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒) = 𝑃(𝑡𝑒𝑠𝑡 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑑𝑖𝑠𝑒𝑎𝑠𝑒) × 𝑃(𝑑𝑖𝑠𝑒𝑎𝑠𝑒) + 𝑃(𝑡𝑒𝑠𝑡 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑛𝑜 𝑑𝑖𝑠𝑒𝑎𝑠𝑒) × 𝑃(𝑛𝑜 𝑑𝑖𝑠𝑒𝑎𝑠𝑒)
-> P(test positive) =(0.99 x 0.0001) + (0.01 x 0.9999) = 0.010098
The posterior probability of having the disease: [(0.99x 0.0001) / 0.010098] = 0.0098 = 0.98%
Despite the test being highly accurate, the low prevalence of the disease leads to a high
probability that the positive result is a false positive
o Heuristic 1: Representativeness
= evaluate the likelihood that object A belongs to category B, by the extent to which A
resembles characteristics of B
The issue arising from this, is that similarity may be determined by many elements not affecting
probabilities, leading to bias: Systematic biases may therefore result from this heuristic
Bv: “John is very shy and withdrawn, invariably helpful, but with little interest in people, or in the
world of reality. He is very tidy, he has a need for order and structure, and a passion for detail”.
What is the probability that John is a) a farmer; b) a salesman; c) a librarian; d) a physician ?
most people guess that he is a librarian
The issue is that there are many more salesmen than librarians, this ignores the prior probability
(base rate) = the failure of Bayesian updating
3
