Top Banner
1 Ka-fu Wong University of Hong Kong The Economics of Information
62

1 Ka-fu Wong University of Hong Kong The Economics of Information.

Dec 20, 2015

Download

Documents

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

1

Ka-fu WongUniversity of Hong Kong

The Economics of Information

Page 2: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

2

Invisible hand theory and information

The invisible hand theory assumes that buyers are fully informed.

Given that consumers are not fully informed, they must employ strategies for gathering information.

Page 3: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

3

Example 13.1. Searching for a racquet

Suppose you want a new tennis racquet, but aren't sure which brand & model to buy.

Page 4: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

4

Example 13.1. Searching for a racquet

Dick’s has a large selection, so you go there and ask a salesperson for advice.

Page 5: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

5

Example 13.1. Searching for a racquet

After some discussion about your experience and style of play, the salesperson recommends the "Prince Thunder O3 Silver."

You buy the O3 Silver for $300.

Then a friend tells you could have bought the same racquet online for only $260.

Page 6: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

6

Example 13.1. Searching for a racquet

How do you feel about having spent $300 (instead of $260)?

How do the costs of the two suppliers differ?

How do the services they provide differ?

Were the extra services you got at Dick’s worth it?

Page 7: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

7

The free-rider problem and "fair trade" laws:

Manufacturers may want to maintain minimum resale prices so that their retailers can earn enough to provide good service.

Without resale price maintenance, consumers can shop at a full service store for information, then order online.

Laws, which prohibited resale price maintenance (such as Fair Trade laws in the US), made it impossible for manufacturers to guarantee availability of high-service retailers.

Page 8: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

8

Optimal Investment in Acquiring Information

Keep acquiring more information until the expected marginal benefit equals the marginal cost.

$/unitMarginal costof Information

Marginal Benefitof Information

Units ofInformationI*

I* = the optimal quantity of information.

Page 9: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

9

Example 13.2.

You know that different stores charge different prices for the particular HP calculator you have decided to buy.

Should you buy from the first store you visit, or should you search all the stores in your area?

Page 10: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

10

Example 13.2.

Suppose you visit the first store and find that they charge $500.

Should you buy or continue searching?

To search further is costly.

What are the expected benefits?

Page 11: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

11

Example 13.2.

Suppose you have a good idea of the price distribution.

500

If you think $500 is already near the low end of the price distribution, then you don't stand to gain much.

Page 12: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

12

Example 13.2.

But additional search will be more attractive if you think $500 is near the high end of the distribution.

500

Page 13: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

13

Example 13.2.

Based on your estimate of the distribution of available prices and the cost of search, you should choose an “acceptance price.”

Then keep searching until you find a calculator priced at least as low as the “acceptance price.”

Continue to searchBuy it

Acceptance price

Page 14: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

14

Example 13.3.

There is a spinning wheel numbered from 0 to 100.

You have won the right to buy a new HP calculator for a price in dollars equal to 10 multiplied by the number you get on the spinning wheel when you spin the pointer.

You spin and get 50, i.e., 500 dollars for the calculator

50

0100

10

20

30

4060

70

80

90

A spin is like “randomly visiting a store”.

Page 15: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

15

Example 13.3.

Now suppose someone offers you a chance to spin again for a fee of $100.

If you end up with a higher number, you still pay only $500.

But if you get a lower number, you pay the lower amount.

50

0100

10

20

30

4060

70

80

90

The chance to spin again is like “the chance of randomly visiting another store”.

Page 16: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

16

Example 13.3.

If you accept this offer, will you end up better off on the average?

C = cost of spinning again = $100

B = expected benefits of spinning again

Page 17: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

17

Example 13.3.

50% of the time, the number you get will be greater than 50. In those cases your benefit is zero.

50% of the time, the number you draw is less than 50. In these cases your benefit will lie between 0

and $500, for an average savings of $250. On the average, your benefit from spinning again

will therefore beB = .5(0) + .5($250) = $125 > C = $100.

So it is worth your while to spin again.

50

0100

10

20

30

4060

70

80

90

Page 18: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

18

Search Is a Gamble

The fact that you will do better on the average by searching further does not guarantee that your next search will make you better off.

For instance, in Example 13.3, suppose you spin again and get 60, i.e., $600. Then you will have incurred a $100 cost and gotten nothing in return.

Page 19: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

19

Example 13.4:

In Example 13.3, suppose you spin again and get 60, i.e. $600. Then you will have incurred a $100 cost and gotten nothing in return. Should you spin for a cost of $100 again?

You should view the cost of your previous search as a sunk cost. So, yes, you should search again.

You made a good decision, but just happened to get a bad outcome.

Page 20: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

20

Example 13.5.

Same as Example 13.3 except now the cost of another spin is $150. Should you spin again?

This time B = $125 < C = $150, so you should not spin again.

Moral: Be prepared to pay a higher price when the cost of searching is high.

Page 21: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

21

Moral: Be prepared to pay a higher price when the cost of searching is high.

Who is more likely to search for additional information?

Page 22: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

22

Example 13.6.

Same as Example 13.3 except now your first spin lands on 40. Should you spin again?

60% of the time, the number you get will be greater than 40. In those cases your benefit is zero.

40% of the time, the number you draw is less than 40. In these cases your benefit will lie between 0 and $400, for an average savings of $200.

On the average, your benefit from spinning again will therefore be B = .6(0) + .4($200) = $80 < C = $100. So it is not worth your while to spin again.

Page 23: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

23

Example 13.6.

Moral: If you find a sufficiently low price, it will not be worth searching further.

Exercise: What is the highest price for which further search would not pay?

Page 24: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

24

Example 13.7

Same as Example 13.3 except now the price you pay for a flash disk is exactly the number you get on the spinning wheel.

If you start with a 50 and extra chances cost $100, does it pay to spin again?

50

0100

10

20

30

4060

70

80

90

Page 25: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

25

Example 13.7

50% of the time, the number you get will be greater than 50. In those cases your benefit is zero.

50% of the time, the number you draw is less than 50. In these cases your benefit will lie between 0 and $50, for an average savings of $25.

On the average, your benefit from spinning again will therefore be B = .5(0) + .5($25) = $12.5 < C = $100.So it is not worth your while to spin again.

Moral: Other things equal, additional search is more likely to be worthwhile for expensive items than for cheap items.

Page 26: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

26

Moral: Other things equal, additional search is more likely to be worthwhile for expensive items than for cheap items.

… than for a bicycle.

For example, you should spend more time searching for a car…

Page 27: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

27

Do "middlemen" create additional value the same way production

workers do?

Page 28: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

28

Example 13.8.

Jones has a house for sale. It is an updated 1895 Queen Anne Victorian with polished walnut woodwork, inlaid oak floors, and a distinctive corner turret.

Page 29: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

29

Example 13.8.

Although he is hoping to get much more, Jones's reservation price for his house is $200,000.

There are two possible local buyers:Adams, who generally likes modern houses and

would be willing to pay up to $400,000 for a house that suited his tastes. He would be willing to pay only $205,000 for Jones's house...

…and Baker, who generally likes Greek Revival Houses, and would be willing to pay up to $350,000 for a house that suited her tastes. She would be willing to pay $250,000 for Jones's house.

Page 30: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

30

Example 13.8.

Modern houses Greek Revival houses

Page 31: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

31

Example 13.8.

Both Adams and Baker are Madisonites, and Jones can hope to attract their attention with a classified ad.

Q. If Adams were the only buyer Jones found, would he sell to him?

A. Yes, because Adams is willing to pay $205,000 -- $5,000 more than Jones's reservation price ($200,000).

Page 32: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

32

Example 13.8.

Q. Would Jones sell to Baker if Baker and Adams were the only two potential buyers.

A. Yes, because Baker is willing to pay $250,000 -- $45,000 more than Adams ($205,000).

Page 33: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

33

Example 13.8.

Q. Obviously Jones would prefer to sell to Baker than to Adams. But is it more efficient for this to happen?

A. Yes. Since Baker values Jones's house more than Adams does, it is more efficient for Baker to have it.

If Adams happened to have bought it, both he and Baker would be made better off if Adams then sold it to Baker. The gain that results is no less valuable than if an additional $45,000 worth of goods and services had been produced.

Page 34: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

34

Example 13.9.

Now suppose there is a real estate agent who has a national computer network that puts her in touch with potential buyers in other cities.

Page 35: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

35

Example 13.9.

One such buyer is Cummings, who is planning to move to Madison soon.

Cummings is a passionate fan of Queen Ann Victorians and would be willing to pay $400,000 for Jones's, which is the only such house in the Madison area.

Page 36: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

36

Example 13.9.

Q. Does the total value of output increase if Jones hires the services of this realtor?

A. Yes, because an additional $150,000 of value is created when Jones sells his house to Cummings rather than to Baker.

Here, too, the gain is equivalent to the production of an additional $150,000 worth of goods and services.

Page 37: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

37

The Commitment Problems When Search is Costly

What happens when, by chance, a more attractive option comes along after the search has ceased?

When information is costly and the search must be limited, a relationship may dissolve.

Commitment agreementsLease agreementsEmployment contractsMarriage contracts

Reason for higher divorce rate? Globalization has suddenly decreased the search cost.

Page 38: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

38

Asymmetric information

Situations in which buyers and sellers are not equally well informed about the characteristics of goods and services for sale in the marketplace.

Page 39: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

39

The Market for Lemons (George Akerlof, 1970)

The Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel in 2001 (shared with Michael Spence and Joseph E. Stiglitz).

Page 40: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

40

Example 13.10

Suppose there are two kinds of cars: Good cars and “Lemons”.

The owner of a used car knows whether it is a lemon but prospective buyers cannot tell.

Page 41: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

41

Example 13.10.

Suppose 10% of all new cars produced are lemons. Good used cars are worth $10,000 to their owners.Lemons are worth only $6000.

Suppose cars do not depreciate with use. If the used cars for sale had the same quality distribution as new cars (i.e., 90% good, 10% lemons), how much would risk-neutral people be willing to pay for a used car?

P = .90($10,000) + .10($6000) = $9600

Page 42: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

42

Example 13.11

If you were the owner of a good used car in the previous example, what would your car be worth to you? $10,000, by assumption.

Under the assumptions of that example, how much would you get if you sold it? $9600. [P = .90($10,000) + .10($6000) =

$9600]

Would you sell it? No, because it is worth $10,000.

If you had a lemon, would you sell it? Yes.

Page 43: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

43

Example 13.11

Moral: The only used cars for sale will be lemons. Asymmetric information tends to reduce the average quality of goods offered for sale.

They will sell for $6000.

Page 44: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

44

Example 13.12

Suppose there are two kinds of computers:Reliable ones, and ones that crash frequently

Page 45: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

45

Example 13.12

Suppose that some fraction z of all personal computers are defective.

The defective ones, however, cannot be identified except by those who own them.

Consumers are risk-neutral and value nondefective computers at $2000 each.

Computers do not depreciate physically with use.

New computers sell for $1000, used ones sell for $500.

What is z?

Page 46: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

46

Example 13.12

Because of the lemons principle, we know that all the used computers that are for sale must be defective.

The price of a used computer must thus be the same as the value of a new defective one.

Because consumers are risk neutral, the price of a new computer--$1000--is simply a weighted average of the values of nondefective and defective computers, where the weights are the respective probabilities:

$1000 = $500 z + $2000 (1-z) z = 2/3.

Page 47: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

47

Communication and signaling

When information is asymmetric as in the used car market, can we communicate and signal to the potential buyers that the car we are selling is actually a good car? How?

Page 48: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

48

Example 13.13.: Should a toad fight for a mate?

When a toad and his rival vie for the same mate, each faces an important strategic decision.

Should he fight for her or set off in search of another?

To fight is to risk injury. But to continue searching has costs as well. At the very least, fighting will consume time. And there is no guarantee that the next potential

mate will not herself be the object of some other toad's affections.

Page 49: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

49

Example 13.13.: Should a toad fight for a mate?

In deciding between these alternatives, each toad's assessment of the other's fighting capabilities plays an important role.

If one's rival is considerably larger, the likelihood of prevailing will be low, and the likelihood of injury high, so it will be prudent to continue searching.

Otherwise, it may pay to fight.

Page 50: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

50

Example 13.13.: Should a toad fight for a mate?

Many of these decisions must be made at night, when it is hard to see.

Toads have therefore found it expedient to rely on various nonvisual clues.

The most reliable of these turns out to be the pitch of the rival's croak.

In general, the larger a toad is, the longer and thicker are its vocal cords, and hence the deeper its croak.

In a hot summer night …….

Page 51: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

51

Example 13.13.: Should a toad fight for a mate?

Hearing a deep croak in the night, a toad may reasonably infer that a big toad made it.

Experiments have shown that the common toad is much more likely to be intimidated by a deep croak than a high-pitched one.

Page 52: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

52

Signaling issues

Similar signaling issues arise in ordinary market exchange. The interests of buyers and sellers are not

identical, and indeed are often in direct conflict. The seller, for example, sometimes has an

incentive to overstate the quality of his product.The buyer, likewise, often has an incentive to

understate the amount she is willing to pay for it.

And the potential employee may be tempted to misrepresent his qualifications for the job.

Page 53: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

53

Example 13.14.

A street vendor in Manhattan is offering what he claims are genuine Rolex wristwatches for sale for $199.95, a normal retail value, he says, of $9,000.

Is his claim credible?

Page 54: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

54

Example 13.15.

The Springer family jewelry shop has been a successful local business in Portland, Maine, since 1870.

In Springer’s annual sale they are offering Rolex wristwatches for only $7,000, a normal retail value, they advertise, of $9,000.

Is their claim credible?

Page 55: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

55

The Costly-to-Fake Principle

For a signal to be credible, it must be costly (or, more generally, difficult) to fake.

If small toads could costlessly imitate the deep croak that is characteristic of big toads, a deep croak would no longer be characteristic of big toads. But they cannot. Big toads have a natural advantage, and it is that fact alone that enables deepness of croak to emerge as a reliable signal.

Page 56: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

56

Example 13.16. Product Quality Assurance

Company A sponsors an expensive national advertising campaign on behalf of its television set, claiming that it has the clearest picture and the best repair record.

Company B makes similar claims but does not advertise its product.

With no additional information, which company's claim is more credible?

Page 57: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

57

Example 13.17. Choosing a Hardworking, Smart Employee

Microsoft is looking for a hardworking, smart employee to manage a new division. There are two applicants, Smith and Jones, who seem alike in every respect but one:

established 1972

Jones graduated with a C+ average from Podunk University.

Smith is an honors graduate from CalTech.

Whom should Microsoft hire?

Page 58: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

58

Statistical Discrimination

The practice of making judgments about the quality of people, goods, or services based on the characteristics of the groups to which they belong.

Why do males under 25 years of age pay more than other drivers for auto insurance?

Page 59: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

59

Adverse Selection

The pattern in which insurance tends to be purchased disproportionately by those who are most costly for companies to insure

Raises premiumsReduces the number of low-risk policy holders Increases the risk level of the insured

Page 60: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

60

Moral Hazard

The tendency of people to expend less effort protecting those goods that are insured against theft or damage

Deductibles are used to reduce moral hazard and adverse selection.Lower rates Increase the incentive to drive safelyReduce the number of claims, which lowers

cost and premiums

Page 61: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

61

Product warranties

Producers of products with high quality can disclose that fact by offering liberal guarantees (because of the costly-to-fake principle).

Page 62: 1 Ka-fu Wong University of Hong Kong The Economics of Information.

62

End