IA 350, Intermediate MicroeconomicsPart IV – Behavioral Economics

A Methodological Approach …

IndividualChoices

ObservedAggregateOutcomes

Are the individuals “humans” or are they “econs”?*

* This distinction between the way psychologists and economists model the agents in their models was originally suggested by Richard Thaler (see Kahneman, 2011, p269).

Simple Ultimatum Games

Mean offer: ~.37x

Modal offer: .50x

Offers < .20x often rejected

Strategic Interaction and Social Norms

Consider a two-stage ultimatum game in which player 1 allocates $100 in round 1. If player 2 accepts player 1’s offer, the game ends with the players keeping the respective amounts. If player 2 refuses player 1’s offer in round 1, the game moves to round 2 with the player’s switching roles, and player 2 allocates $25. The same rules apply as in round 1. What is player 1’s optimal offer in round 1?

($75, $25)

($1, $24)

($76, $24)

($1, $24)

($77, $23)

($1, $24)

x = $25

x = $24

x = $23

Ultimatum Games and Rationality

Uncertainty and Loss Aversion

+ (Gain)+ (Loss)

+ v

- v

+ $30

- $10

Value = V($)

“Most respondents in a sample of undergraduates refused to stake $10 on the toss of a coin if they stood to win less than $30.”

Uncertainty and Loss Aversion

+ (Gain)+ (Loss)

+ v

- v

+ $30

- $10

Value = V($)

“Most respondents in a sample of undergraduates refused to stake $10 on the toss of a coin if they stood to win less than $30.”

You are offered the following bet: A fair coin will be tossed. If it is heads, you will win x, while if it is tails you will lose y. 

If y is $1, what value of x would you require to accept the bet? ______________ 

If y is $100, what value of x would you require to accept the bet? ______________ 

If y is $1,000, what value of x would you require to accept the bet? ______________

If y is $10,000, what value of x would you require to accept the bet? ______________ 

If y is $100,000, what value of x would you require to accept the bet? ______________

$1.63

$931

$3,212

$100,647

$463,824

A Mean Reverting Walk Down Wall StreetWhat implications do the concepts of “endowment effects” and “loss aversion” have for our understanding of investor behavior and stock market performance?

Loss aversion also explains one of the most common investing mistakes: investors evaluating their stock portfolio are most likely to sell stocks that have increased in value or, during a week like this (September 2008), have gone down the least amount. Why? Because it hurts to take a loss. Unfortunately, this means that people end up holding on to their depreciating stocks. Over the long term, this strategy is exceedingly foolish, since it ultimately leads to a portfolio composed entirely of shares that are losing money. (A study by Terrance Odean, an economist at UC-Berkeley, found that the stocks investors sold outperformed the stocks they didn't sell by 3.4 percent). Even professional money managers are vulnerable to this bias, and tend to hold losing stocks twice as long as winning stocks. Because selling shares that have decreased in value makes the loss tangible - and losing sucks - we try to postpone the pain for as long as possible.

Jonah Lehrer, author of How We Decidehttp://scienceblogs.com/cortex/2008/09/loss_aversion_and_the_stock_ma.php

0 1 2 3 4Delay in Years

0.50

0.45

0.40

0.35

0.30

.025

.020

0.15

0.10

0.05

0

Impl

ied

Dis

coun

t Rat

e

Adapted from Thaler, The Winner’s Curse (1992), Figure 8 – 1. Original source given as Benzion et al. (1989), “Discount Rates Inferred from Decisions: An Experimental Study,”Management Science 35: 270 – 284.

Discounting as a Function of Time Delay and Money Amount

Three strong patterns:1. Discount rates decline with the length of time to be waited.2. Discount rates decline with the size of reward.3. Discount rates for gains are higher than for losses.

Time Inconsistency

$40

$200

$1,000$5,000

Time Inconsistency6. You are working at a job and are currently earning $10,000 a month. You receive a memo that says

a new payment plan is under consideration for next year. Instead of receiving monthly paychecks, you can start the new year by receiving your annual salary in a lump sum. (If you quit during the year, you will have to pay back an amount equal to the percentage of the year that you will not be working there, plus interest.) There is a catch, however: if you choose this new plan, your salary will be deduced. What annual salary would you be willing to accept to receive the your entire year's salary in a lump sum at the beginning of the year?

 

7. You are working at a job and are currently earning $10,000 a month. You receive a memo that says a new payment plan is under consideration for next year. Instead of receiving monthly paychecks, you can receive your annual salary in a lump sum at the end of the year. (If you quit during the year, you will be paid a lump sum equivalent to an amount equal to the percentage of the year that you worked there.) To induce you to sign up for this plan, your salary will be increased. What annual salary would you require to sign up for this end-of-year lump sum payment plan?

AverageMinimumMaximum

$120,000$68,136

$200,000

AverageMinimumMaximum

$143,000$100,000$200,000

Implied d~ 0%120%- 85%

Implied d32%

- 34%92%

The psychology of intertemporal choice complicates the already complicated question of selecting the proper social rate of discount (the rate at which the government should discount future costs and benefits). The standard view is that the market rate of interest, corrected for tax distortions, represents an aggregation of individual time preferences, and is the appropriate social rate of time discounting … But, if individuals do not discount everything at a single rate, then which rate is the one that is appropriate for social discounting?

Rhichard Thaler, The Winner’s Curse,1992, pp. 105 – 106.

Time Inconsistency – One Implication

Some observations about the “correctness” of risk assessments:Risk analysis and the assumption of normality

-0.2

50-0

.242

-0.2

34-0

.226

-0.2

18-0

.210

-0.2

02-0

.194

-0.1

86-0

.178

-0.1

70-0

.162

-0.1

54-0

.146

-0.1

38-0

.130

-0.1

22-0

.114

-0.1

06-0

.098

-0.0

90-0

.082

-0.0

74-0

.066

-0.0

58-0

.050

-0.0

42-0

.034

-0.0

26-0

.018

-0.0

10-0

.002

0.00

60.

014

0.02

20.

030

0.03

80.

046

0.05

40.

062

0.07

00.

078

0.08

60.

094

0.10

20.

110

0.11

80.

126

0.13

40.

142

0.15

00.

158

0.16

60.

174

0.18

20.

190

0.19

80.

206

0.21

40.

222

0.23

00.

238

0.24

60%

1%

2%

3%

4%

5%

6%

7% S&P 500 – Frequency Distribution of Daily Returns

MedianMeanStandard DeviationSkewnessKurtosis

0.05%0.03%0.96%

-0.677823.0759

Properties of Distribution

October 19, 1987Daily Change: -20.47%Probability: 2.038 X 10101

October 21, 1987Daily Change: 9.10%Probability: 1.716 X 1021

October 13, 2008Daily Change: 11.58%Probability: 1.195 X 1033

Mean: 0.0323%Std. Dev.: 0.9653%

Range +/- s Low High Probailbity 1-Probability1 -0.9329% 0.9976% 68.2689492% 31.7310508%2 -1.8982% 1.9629% 95.4499736% 4.5500264%3 -2.8634% 2.9281% 99.9487524% 0.0512476%4 -3.8287% 3.8934% 99.9936658% 0.0063342%5 -4.7939% 4.8586% 99.9999427% 0.0000573%6 -5.7592% 5.8239% 99.9999998% 0.0000002%7 -6.7244% 6.7891%8 -7.6897% 7.7544%9 -8.6550% 8.7196%10 -9.6202% 9.6849%11 -10.5855% 10.6501%12 -11.5507% 11.6154%13 -12.5160% 12.5807%14 -13.4812% 13.5459%15 -14.4465% 14.5112%16 -15.4117% 15.4764%17 -16.3770% 16.4417%18 -17.3423% 17.4069%19 -18.3075% 18.3722%20 -19.2728% 19.3374%21 -20.2380% 20.3027%22 -21.2033% 21.2680%23 -22.1685% 22.2332%24 -23.1338% 23.1985%25 -24.0990% 24.1637%    

October 19, 1987

October 13, 2008

October 21, 1987

Risk Analysis and the Assumption of Normality

IA 350, Intermediate MicroeconomicsPart IV – Externalities

D

P*

P

Q

S

Q*Optimal Output

-- Private Market’s View

[= Private Value]

[= Private Cost]

MSC [Social Cost]

S’ [= Private + Social Cost]

Optimal Output-- Society’s View

P’

Q’

Market prices reflect only private valuationsand private costs of production.

The “Dead Weight Loss” associatedWith a negative externality.

Externalities

The Problem of Social Cost (1960)“It is strange that a doctrine as faulty as that developed by Pigou should have been so influential …”

Ronald Coase1910 -

“The traditional approach [to the externality problem, the ‘Pigovian’ approach] has tended to obscure the nature of the choice that has to be made. The question is commonly thought of as one in which A inflicts harm on B. But this is wrong. We are dealing with a problem of a reciprocal nature. To avoid the harm to B would be to inflict harm on A … The problem is to avoid the more serious harm.”

The Coase Theorem: In the absence of transaction costs, all government allocations of property rights are equally efficient, because interested parties will bargain privately to correct any externality.

Implication: Any market plagued by externalities can be made efficient if property rights are clearly defined and assigned (the “missing market” interpretation), and if transaction (bargaining) costs are negligible.

Externalities

A

B

Good 1 (Income)

Goo

d 2

(Ext

erna

lity)

·

Pure Exchange and Externalities

·

·

·

X

X’

A

B

Good 1

Goo

d 2

·

Pure Exchange and Externalities : The Coase Theorem

· ·

·

X X’

IA 350, Intermediate MicroeconomicsPart IV – Asymmetric Information

Markets with asymmetric information

Varian’s “market for plums and lemons”

spplums QPS 44 :

plumsP2200

50

| |

| |

lemons

plums

Q

QQ

P

Adapted from Varian, Intermediate Microeconomics, 8th ed., 719 – 720. His example is itself an adaptation from a famous paper by George Akerlof, “The Market for Lemons: Quality Uncertainty and the Market Mechanism.” Quarterly Journal of Economics 84 (1970): 485 – 500.

dpplums QPD 444400 :

sllemons QPS 441100 :

dllemons QPD 443300 :

lemonsP1100

lemonsplums

qualityuncertain

DlemonsprDplumspr

D

**

lemonsP̂1375

plumsP̂1925

56.25

| |

ˆlemonsQ

43.75

| |

ˆplumsQ

lemonsplums

qualityuncertain

DlemonsprDplumspr

D

**

Markets with asymmetric information

Signaling in a Labor Market

e

w

Adapted from Varian, Intermediate Microeconomics, 8th ed., 726 – 730. His example is itself an adaptation from a famous paper by Michael Spense, “Job Market Signaling.” Quarterly Journal of Economics 87 (1973): 355 – 374.

b(e)

ch(e)

cl(e)

e*

wh

wl

Markets with asymmetric information General categories of problems of asymmetric information:

Adverse Selection: A situation in which individuals possess hidden information, leading to a market selection process that results in a pool of individuals with economically undesirable characteristics.

Moral Hazard: A situation in which one party to a contact takes a hidden action that creates benefits at the expense of another party to the contract.

Social institutions that help solve these market inefficiencies:

Private agent responses to problems of asymmetric information:

Signaling – the party possessing private information takes action to generate a credible signal, such as warranties or guarantees.

Screening – the party with imperfect information takes action to induce a advantageous sorting of the market, such as an incentive-driven menu of options.

Public / Private sector responses to problems of asymmetric information:

Compulsory purchase plans (insurance markets)

Licensing and certification (note that this may also be done by private agents).

Market regulation, including “truth” laws and “insider trading” laws.

Agency costs

What is the “agency problem” and what are agency costs?

Efficiency costs which arise when there is a divergence of interests between parties to a transaction.

When one party to a transaction cannot observe the effort of another party, upon whose effort the value of the transaction depends, we say there is a principal-agent problem. Examples:

In such cases, the principal cannot determine whether a bad outcome was the result of the agent’s low effort or due to bad luck.

• Shareholders (principal) cannot perfectly monitor the efforts of a CEO (agent)

• CEO (principal) cannot perfectly monitor the efforts of managers (agent)

Basic Principal-Agent Model

Uncertainty

Agent’s Actions

Observable Outcome Payoffs

Principal’s Share

Agent’s Share

Contract

The Agency Challenge. To structure a contract that specifies division of payoffs in such a way that the agent’s actions will maximize those payoffs, given two complicating factors:

Uncertainty Unobservability of effort

There is an implicit tradeoff between efficient risk bearing and creation of incentive effects.

The difference in utility outcomes may be interpreted as the cost of inducing a risk-averse agent to accept some risk

Inability to observe effort entails efficiency costs (in terms of lower utility or surplus generated from the transaction.

Implications of Agency Theory Summary