Chapter 11

The Economics of Information

Copyright © 2013 by The McGraw-Hill Companies, Inc. All rights reserved.McGraw-Hill/Irwin

11-2

Learning Objectives1. Explain how middlemen add value to market

transactions2. Use the concept of rational search to find the

optimal amount of information market participants should obtain

3. Define asymmetric information and describe how it leads to the lemons problem

4. Discuss how advertising, conspicuous consumption, statistical discrimination, and other devices are responses to asymmetric information

11-3

Information and the Invisible Hand

• All parties have all relevant information– Without free information, market results are not

efficient• Bargaining for a bowl in Kashmir

• Parties must decide how much information to gather– Information gathering strategies differ

11-4

How The Middleman Adds Value

• Buyers sometimes choose among several version of a product– Each has complex feature sets

• Research options– Company web site– Ask friends and family– Consumer Reports, online product reviews– Visit stores, ecommerce sites

11-5

Consumer Choice: Buying Skis• Skis R Us recommend $600 Salomon X-Scream

9 skis– Sales rep seems knowledgeable

• Your next move is– Thank them and do more research– Trust the sales rep and buy them– Go home and buy at the best price online ($400)

• Evaluate the importance of– Immediate possession– Best price– Post-sales service and support

11-6

The Value of the Middleman• Sales representatives supply information to

buyers– Manufacturers can offer direct sales to bypass

middlemen• Information makes markets more efficient

– Purchasing the bowl in Kashmir

11-7

Selling Babe Ruth• Ellis wants to sell a Babe Ruth baseball card.

– His reservation price is $300– An ad in the local newspaper cost $5– eBay cost is 5% of the Internet auction price– The maximum price in the local market is $400– Two eBay shoppers have secret reservation prices

of $800 and $900, respectively

11-8

Selling Babe Ruth• Benefits of eBay

– Card sells for $800 on eBay less $40 commission• Ellis nets $760, $460 above his reservation price• Buyer surplus is $100

• Local option is inferior– Card sells for $400 less $5 cost of ad– Ellis nets $395, $95 more than his reservation price– Buyer surplus is $0

• Economic surplus is increased when a product goes to the person who values it the most

11-9

$/un

itUnits of information

MB

The Optimal Amount of Information

• More information is better than less– Gathering information has a cost

• Marginal benefit starts high, then falls rapidly– Low-Hanging Fruit Principle

• Marginal cost starts low, then increases

• Optimal amount of information is I* where MC = MB

MC

I *

Optimal

11-10

Free Rider Problem• A free-rider problem exists when non-payers

cannot be excluded from consuming a good– Interferes with incentives– Market quantity is below social optimum

• Stores bear the cost of training sales reps on merchandise– Shoppers use sales reps as information source

• Then some shoppers buy elsewhere– Store is unable to capture some of the value it

delivered to the shopper: a free-rider problem

11-11

Example: The Last Bookstore• Independent bookstores differentiate themselves

with personalized service– Offer more information and recommendations than

Barnes & Nobles or Borders• Chain bookstores carry large inventory and shopping

center location can erode local store base– Ecommerce sites such as Amazon.com and

Overstock.com offer reviews and recommendations• Large inventory; quick delivery• Online sales further reduce sales in independent

stores

11-12

Rational Search Guidelines• Additional search time is more likely to be

worthwhile for expensive items than cheap ones– Apartment search in Paris, Texas involves less time

than Paris, France• Texas has lower rents and narrower price range

• Prices paid will be higher when the cost of a search is higher– Two buyers, only one with a car

• Buyer with the car will look at more pianos before buying

11-13

Gamble Inherent in Search• Additional search has costs that are certain

– Benefits are uncertain benefits– Additional search has elements of a gamble

• A gamble has a number of possible outcomes– Each outcome has a probability that it will occur

11-14

Gamble Inherent in Search• The expected value of a gamble is the sum of

(the possible outcomes times their respective probability)– A fair gamble has an expected value of zero– A better-than-fair gamble has a positive expected

value

11-15

Risk Preferences• A risk-neutral person would accept any gamble

that is fair or better-than-fair– A risk-averse person would refuse any fair gamble

11-16

The Gamble in the Search• You need a one-month sublet in San Francisco

– One type of apartment rents for $400 and it is 80% of the available market

– The other type rents for $360 and makes up 20% of the market

– You must visit the apartment to get the rental rate• Cost per visit is $6

– You are risk-neutral

11-17

San Francisco Apartment Search

• The first apartment you visit is the $400 version• Look at the next apartment if the gamble is at

least fair– Two outcomes to the gamble

• You find a lower-priced apartment and your net benefit is $34 with 20% probability

• You find another $400 apartment and your net benefit is – $6 with 80% probability

– Expected value of the gamble is (34) (0.20) + (– 6) (0.80) = $2

– Keep searching

11-18

Commitment Problems and Search

• Some searches are for circumstances requiring commitment over some period of time– Leasing an apartment– Taking a job– Getting married

• Search is costly and therefore limited– People end their searches when the marginal cost

of searching exceeds the marginal benefit• BUT… what if you fall into a better option?

11-19

Commitment Problems and Search

• If information were freely available, there would be no commitment problem– Contracts are used to bind parties together AND– Contracts carry penalties for breaking the

arrangement• People terminate their search because

information gathering is costly• Under some circumstances, one party may

rationally choose to terminate the agreement and pay the penalties

11-20

Asymmetric Information• Asymmetric information occurs when either the

buyer or seller Is better informed about the goods in the market– Mutually beneficial trades

may not occur– A seller might know that

a murder was committed in a house offered for sale

• Buyer does not know

Buyer Seller

11-21

Private Sale of a Used Car• Jane's Miata is in excellent condition

– Jane's reservation price is $10,000• Blue Book value is $8,000

• Tom wants to buy a Miata– His reservation price is $13,000 for one in excellent

condition and $9,000 for one in average condition– Determining the condition of Jane's car has a cost

and the results are uncertain– Tom cannot verify that Jane's Miata is superior

• Tom buys another Miata for $8,000; Jane's is unsold

11-22

Surplus Loss and Asymmetric Information

• Tom's loss is $1,000– Pays $8,000 and has a gain of $1,000– Tom’s loss from buying an average car instead of

Jane's• $13,000 – $11,000 = $2,000

– Tom's net loss is $1,000

• Jane’s loss from losing Tom as a customer is $1,000

• Total loss is $2,000

11-23

The Lemons Model• People who have below average cars (lemons),

are more likely to want to sell them– Buyers know that below average cars are likely to

be on the market and lower their reservation prices• Good quality cars are withdrawn from the market

– Average quality decreases further and reservation prices decrease again

• The lemons model says that asymmetric information tends to reduce the average quality of goods for sale

11-24

The Lemons Model in Action• Your aunt offers you her 4-year old Accord

• The asking price of $10,000 is the blue book value• You believe the car is in good condition

• Blue book value is the equilibrium price for below average cars

• You should buy the car for $10,000– It is in better condition than the average Accord of

the same vintage and mileage

11-25

Naïve Buyer• Two kinds of cars: good cars and lemons

– Owners know what kind they have– Buyers can't determine a car's quality– Buyers are risk neutral

• What would the buyer offer for a used car?– Expected value of a car is

(0.90) ($10,000) + (0.10) ($6,000) = $9,600

• The buyer gets a lemon

Good Cars LemonsProbability 90% 10%Value $10,000 $6,000

11-26

Credibility Problem• Parties gain if they find a way to communicate

information truthfully• If Jane can convince Tom her Miata is in

excellent condition, Tom will buy– Statements are not credible– Jane offers Tom a six-month warranty on all car

defects at the time of purchase• A warranty for a lemon would cost more than the

economic surplus gained• Only sellers of good quality cars would offer the

warranty

11-27

The Costly-to-Fake Principle• To communicate information credibly, a signal

must be costly or difficult to fake– Sellers have an incentive to exaggerate the quality

of their product– Buyers value objective information about quality

11-28

Costly Signals• Television advertising is expensive

– In print advertising, "As seen on TV" signals a company's commitment to its product

• Potential signal of quality

• Educational institutions' brands and students' grades signal quality– An A+ student from MIT is more likely to be offered

a job than a C student from an average academic institution

11-29

Conspicuous Consumption• Choose a lawyer

– Lawyer A wears inexpensive suits and drives a 10-year old Dodge Neon

– Lawyer B wears custom-tailored suits and drives a new BMW 745i

– No other information is available• Conspicuous consumption signals success

– Choose Lawyer B ….… and pass on Ben Matlock!

11-30

Statistical Discrimination• Statistical discrimination uses group

characteristics to infer individual characteristics– Can be applied to people as well as to goods and

services– Results from observed differences between groups

• Example– This candidate for employment is in her late

twenties– Women have babies in their late twenties – This candidate will have a baby in the next few

years• High cost compared to other candidates

11-31

Dangerous Drivers• Men under 25 years of age pay more than other

drivers for auto insurance– Expected cost of insuring a driver depends the

probability and size of claims• Individual assessments are not possible

– Rates are based on demographic groups and the claim history of those groups

• Individual rates are adjusted upward as more information becomes available

11-32

Adverse Selection• Adverse selection occurs because insurance

tends to be purchased more by those who are most costly for companies to insure– Insurance is most valuable to those with many

claims• Adverse selection increases insurance

premiums– Reduces attractiveness of insurance to low-risk

customers• "Best" insurance risk customers opt out

– Rates increase– Repeat

11-33

Moral Hazard• Moral hazard is the tendency of people to

expend less effort protecting insured goods– People take more risk with insured goods or

activities– Deductibles give policy holders an incentive to be

more cautious• Suppose a car owner has a $1,000 deductible

policy– The owner pays the first $1,000 of each claim

• Strong incentive to avoid accidents– Claims less than $1,000 are not reported– Insurance premiums go down

11-34

Disappearing Political Discourse

• Disappearing political discourse theory holds that politicians who support a policy will remain silent to avoid being misunderstood– Opposing the death penalty could be interpreted vy

voters as being soft on crime• No necessary relationship between the two

• Assumes voters implicitly assign a position to a politician who has not made public statements

11-35

Politicians and the Death Penalty

• Arguments against the death penalty– Expensive relative to life in prison without parole– Irreversible for people later found innocent– Does not deter capital crimes

• Politicians avoid taking a public position on capital punishment

11-36

Politicians and the Death Penalty

• Voters want politicians are tough on crime– Broader issue than the death penalty

• Two groups of politicians: tough on crime and soft on crime

• Voters use information about a politician's views on the death penalty to infer the politician's stand on crime

11-37

Politicians and the Death Penalty

• Politicians tough on crime and opposed to capital punishment lose votes– They have an incentive to remain silent

• Move to blue box– Probability that a politician in the red box is tough on crime

decreases• More defections

• Public statements do not accurately reflect aggregate views of politicians on capital punishment

Favor Death Penalty PLUS Remain Silent95% tough on

crime

Oppose only the Death Penalty

Oppose Punishing Criminals

80% tough on crime

11-38

Legalized Drugs• Laws against buying and selling certain

substances are intended to reduce the harm to society from drug use

• Laws have a cost– Price of illegal drugs increases– Addicts commit crimes to pay for drugs– Diverts people from productive employment– Externalities of turf battles– High cost to law enforcement, legal, and prison

systems

11-39

Legalized Drugs• Legalization solves most problems

– With lower drug prices, quantity demanded may increase

• Not supported by evidence in UK and Netherlands

• An outspoken supporter is seen by voters as more likely to be crazy than an outspoken opponent– Rational supporters do not speak out

Rational Opponent

Rational Supporter

Crazy Opponent Crazy Supporter

Opponents Supporters

11-40

The Economics of Information

Information

Mid

dlem

en

Free

-Rid

er P

robl

em

Rational Search• Optimal amount of

information• Search as a gamble

• Risk preferences• Commitment

problems

Asymmetric Information• Credibility problem• Costly to fake

principle• Statistical

discrimination• Adverse selection• Moral hazard• Political discourse