2.0K views

Uploaded by Tom Holden

save

- Mock End of Term Test 2013
- Problem Set 4
- Lecture 6
- Lecture 1
- Lecture 7
- Lecture 2: Basic oligopoly
- Problem Set 5
- Lecture 7
- Power System Economics - Unilag Lecture Notes
- CCP602
- Lecture 9: More on advertising
- Lecture 1
- Cns Simulation.docx
- Mock End of Term Test 2012
- Lecture 8: Vertical differentiation and advertising
- tpmpresentation-090602104911-phpapp02.ppt
- Monetary liquidities to shrimping economy
- ETAR.Lecture 2-Demand, Supply, and Equilibrium.E.pdf
- 4.3
- H2 Econs Essay 1
- 237406_633882287082580595
- Managerial Economics1
- 155040101111062_DIAN DWI UTAMI
- Managerial Economics
- {955a4502-ea04-4b1d-9e86-fd60b129cb3c}_WP7_-_Do_You_See_A_Price_War
- ECON 1000 Midterm Practice
- 10622513
- Pricing
- Idea Generation Sample Assignment
- Adv Planning
- Lecture 7: More on product differentiation and Introduction to the economics of advertising
- Answers to 2012 Mock mid-term test
- Lecture 5: Price discrimination and nonlinear pricing
- Lecture 6: Profit reducing price discrimination, and product differentiation
- Lecture 3: Further static oligopoly
- Lecture 2: Basic oligopoly
- Mock mid-term multiple choice test
- Lecture 8 Auctions
- Problem Set 5
- Lecture 4: Will profits in reality be higher or lower than under Cournot?
- Lecture 9: More on advertising
- Lecture 1
- Lecture 8: Vertical differentiation and advertising
- Lecture 7 Asymmetric Information 2
- Problem Set 3
- Problem Set 2
- Lecture 3.2 Externalities
- Problem Set 1
- Lecture 8
- Lecture 3.1 Welfare
- Lecture 2 Uncertainty 2
- Lecture 5 Intertemporal Choice
- Lecture 1 Uncertainty 1
- Lecture 6 Asymmetric Information 1
- Mock End of Term Test 2012
- Lecture 4 Public Goods

You are on page 1of 16

org

More on advertising:

◦ Models. ◦ Welfare. ◦ Empirics.

Additional reference for the advertising material: Bagwell (2005)

If advertising genuinely acts by changing people’s preferences, surely it is wrong to use a fixed standard. Standard alternative is to use valuations before and after.

◦ Even this is only valid if the advert has not changed the value they put on other goods.

Fits in naturally with the complementary and informative views.

P

Lost

Q

Non-fixed standards will moderate results about excess advertising, since there is an additional positive effect to counteract the negative DixitNorman effect.

◦ Hence when prices don’t change there will always be insufficient advertising. ◦ It may be shown (see Bagwell) that there may be insufficient advertising even when increasing advertising would push up prices.

Sufficient conditions are that 1) when quantities are higher, the effect of advertising on prices is smaller and 2) increasing advertising increases quantities. These conditions mean that the marginal consumer gets the least benefit from increased advertising, so the firm provides too little.

An example with informative advertising:

◦ 𝑄 𝑃, 𝐴 = 𝐹 𝐴 𝐺 𝑃 as we had before. ◦ Then consumer surplus at the optimal price 𝑃∗ is ∞ ∞ ∞ 𝑄 𝑃, 𝐴 ⅆ𝑃 = 𝑃∗ 𝐹 𝐴 𝐺 𝑃 ⅆ𝑃 = 𝐹 𝐴 𝑃∗ 𝐺 𝑃 ⅆ𝑃 𝑃∗

**◦ So total surplus is given by: 𝑊 𝐴 = 𝐹 𝐴 Π 𝑃∗ , 𝐴 .
**

◦ Hence, 𝑊 ′ 𝐴 = 𝐹 ′ 𝐴

∞ 𝐺 𝑃∗

∞ 𝐺 𝑃∗ 𝑃

ⅆ𝑃 + 𝑃

ⅆ𝑃 +

◦ But when 𝐴 is chosen optimally, (i.e. 𝐴 = 𝐴 ), =0 (from the firm’s FOC). ∞ ◦ So 𝑊 ′ 𝐴∗ = 𝐹 ′ 𝐴 𝑃∗ 𝐺 𝑃 ⅆ𝑃 > 0 (as advertising increases demand, and demand is always non-negative). ◦ So there is too little advertising.

∗
𝜕

Π 𝑃∗ ,𝐴 𝜕𝐴 𝜕

Π 𝑃∗ ,𝐴 𝜕𝐴

There are a large number of firms, each of which can produce at most one instance of the same good, for a cost of 𝑐. There is no entry cost, but no one will buy from a firm unless they receive an advert from them. Sending an advert to one random consumer costs 𝑎. Each advert lists the firm’s price. Consumers will buy from any firm that sends them an advert with a price below their valuation 𝑣. Consumers who receive adverts from multiple firms buy from the cheapest.

If a firm sends an advert listing a price 𝑃, with some probability 𝑋 𝑃 it will be the cheapest advert that consumer receives, and they will make profits of 𝑃 − 𝑐. Because there are a large number of firms (equivalently, no entry costs), each firm must make zero profits.

◦ Thus total expected profits from sending an advert are 𝑃 − 𝑐 𝑋 𝑃 − 𝑎.

◦ If there was a firm making positive profits, then I would want to send out adverts offering a price just below the one it had chosen. ◦ But then my rival faces a lower probability of selling at his posted price, so must be making lower profits. 𝑎

Hence: 𝑎 = 𝑃 − 𝑐 𝑋 𝑃 for all 𝑃 firms set, so 𝑋 𝑃 = 𝑃−𝑐.

◦ Since 𝑋 𝑃 is a probability the price can never be below the level at which 𝑎 1 = 𝑋 𝑃 = , i.e. 𝑃 ≥ 𝑎 + 𝑐. 𝑃−𝑐 ◦ Since no one will buy if 𝑃 > 𝑣, no firm will advertise a price above 𝑣. But 𝑎 𝑎 since 𝑋 𝑣 = > 0 there must be a probability that a consumer will 𝑣−𝑐 𝑣−𝑐 only receive one advert, meaning firms can still sell at 𝑣. ◦ Indeed, in equilibrium, there are firms setting a price at every point between 𝑎 + 𝑐 and 𝑣.

𝑋

𝑃 looks a lot like the demand curve faced by each firm. Intuitively then, we might expect monopolisticcompetition style distortions. In fact, this is efficient (welfare optimal).

◦ Price is a transfer, so it’s irrelevant. ◦ The social benefit to reaching a new consumer (for sure) is 𝑣 − 𝑐. ◦ Thus the social benefit from sending another advert is 𝑣 − 𝑐 times the probability that the consumer had not received 𝑎 any other adverts. But this probability is , in equilibrium. 𝑣−𝑐 ◦ So social benefit to another ad equals the cost! ◦ However, when consumers have heterogeneous valuations it may be shown that advertising is inadequate.

Two firms, Hotelling set-up, fixed locations (0 and 1), linear transport cost 𝑡, zero MC. Firm 𝐴 (𝐵) sends adverts to a proportion 𝑧𝐴 (𝑧𝐵 ). 𝑟 2 𝑟 2 This costs them 𝑧𝐴 ( 𝑧𝐵 ). 2 2 Adverts are randomly distributed over consumers so, e.g. a proportion 1 − 𝑧𝐴 1 − 𝑧𝐵 receive no ads so do not buy. As in the standard Hotelling model, of those consumers who received two ads, the indifferent 1 𝑝 −𝑝 one is located at 𝑥 ∗ = + 𝐵 𝐴. 2 2𝑡 Demand faced by firm 𝐴 is then: 𝑧𝐴 1 − 𝑧𝐵 + 𝑧𝐴 𝑧𝐵 𝑥 ∗ .

**So firm 𝐴’s profits are: 𝑧𝐴 1 − 𝑧𝐵 + 𝑧𝐵 FOC 𝑧𝐴 : 0 =
**

◦ I.e. 𝑧𝐴 =
𝑝𝐴

𝑟

1 𝑝 −𝑝 + 𝐵 𝐴 2 2𝑡

2 𝑝𝐴 − 𝑧𝐴 . 𝑟

2

1 − 𝑧𝐵 + 𝑧𝐵

1 − 𝑧𝐵 + 𝑧𝐵

1 2

+ 𝑝𝐵

−𝑝𝐴 2𝑡

1 𝑝 −𝑝 + 𝐵2𝑡 𝐴 2 𝑝𝐴

− 𝑟𝑧𝐴 .

.

FOC 𝑝𝐴 : 0 = 𝑧𝐴 1 − 𝑧𝐵 + 𝑧𝐵

◦ I.e. 𝑝𝐴 =

2𝑡 𝑧𝐵

1 − 𝑧𝐵 + 𝑧𝐵

1 2

+ 𝑝𝐵

−𝑝𝐴 2𝑡

1 𝑝 −𝑝 + 𝐵2𝑡 𝐴 2

−

. 𝑧𝐴

𝑧𝐵 𝑝𝐴 . 2𝑡

**Solution must be symmetric, with 𝑝 ≔ 𝑝𝐴 = 𝑝𝐵 and 𝑧 ≔ 𝑧𝐴 = 𝑧𝐵 . Hence:
**

◦ 𝑧 =

◦ I.e.
𝑝

𝑟 𝑝𝑧

2𝑡

1−

=
𝑧

2 𝑟𝑧 𝑝

and 𝑝 =

2𝑡 𝑧

1− 𝑧

2

.

. So 𝑝 = 2𝑡𝑟 and 𝑧 = 𝑝

𝑟 1𝑝 1+ 2 𝑟

=

2𝑝 2𝑟+𝑝 𝑡 2

=

2 2𝑡𝑟 2𝑟+ 2𝑡𝑟

=

2 1+

**◦ For this to be valid we need 𝑧 < 1. 𝑟 > ◦ Profits then are:
**

2 1+

2𝑟 𝑡

**is necessary and sufficient for this.
𝑟**

2 2 1+

2𝑟 𝑡

2𝑟 𝑡

.

1−

1 2

2 1+

2𝑟 𝑡

2𝑡𝑟 −

2

=

2 2𝑡𝑟 1+

2𝑟 𝑡

−2 2𝑡𝑟−2𝑟

2𝑟 𝑡 2

=

2𝑟 1+

2𝑟 𝑡 2

1+

So…

◦ Price is higher than without the need for advertising. (𝑟 > implies 2 𝑃 = 2𝑡𝑟 > 𝑡.) ◦ When products are more differentiated (𝑡 is high), there is more advertising.

◦ Expensive advertising actually increases profits.
𝑡

So even if we observe higher differentiation in industries with a lot of advertising, it does not mean that advertising caused the differentiation. High costs reduce the amount of advertising performed, reducing the proportion of consumers who see two adverts, pushing up prices.

◦ Advertising cost and differentiation have the same (positive) effect on profits, but opposite effects on the amount of advertising performed. ◦ There may be too much or too little advertising.

** Thus we should not be surprised by finding either a positive or a negative correlation between advertising and profits.
**

If extra advertising reaches a new consumer, then the social benefit exceeds the private benefit to the firm (non-appropriability). But firm 𝐴 has an incentive to advertise more in order to expand its market share (business stealing).

A model of complementary advertising will begin with specifications for agent’s utility functions under which viewing adverts (or others viewing adverts) is a complement for the good. A very simple model is the following.

◦ If I have not seen an advert, then I value the good at zero. ◦ If I have seen an advert, then I value the good at 𝑣.

Thus every model of informative advertising may be reinterpreted as a model of complementary advertising.

Read Bagwell (2005)!

Consistent with the informative/search view.

**◦ Conclusion is that different views are valid in different industries.
**

◦ Benham (1972) found eyeglass prices were higher where advertising was banned. ◦ Kwoka (1984) found a similar result for optometry. ◦ Milyo and Waldfogel (1999) look at the end of a ban on liquor price advertising and find firms cut the prices of only those goods that either they advertise or their rival does.

Other important papers:

◦ Comanor and Wilson (1967) find profits, advertising and differentiation move together. (Possible in Grossman and Shapiro (1984) model.) ◦ Nelson (1974), Porter (1974), Esposito et al. (1990) – product characteristics are important. Experience goods different to search goods etc. Some evidence for an inverse-U relationship between concentration and advertising (but e.g. Willis and Rogers (1998) find the opposite result.)

**Advertising is not unambiguously bad.
**

All three views (persuasive, informative, complementary) have something going for them.

◦ But the persuasive view is unpopular these days for methodological reasons.

Empirical evidence is hard to interpret, since differentiation, entry, advertising and profits are all endogenous.

OZ Ex. 11.7

◦ Question 1,2

**OZ Extra exercises:
**

◦ http://ozshy.50webs.com/io-exercises.pdf ◦ Set #16

- Mock End of Term Test 2013Uploaded byTom Holden
- Problem Set 4Uploaded byTom Holden
- Lecture 6Uploaded byTom Holden
- Lecture 1Uploaded byTom Holden
- Lecture 7Uploaded byTom Holden
- Lecture 2: Basic oligopolyUploaded byTom Holden
- Problem Set 5Uploaded byTom Holden
- Lecture 7Uploaded byTom Holden
- Power System Economics - Unilag Lecture NotesUploaded byeresuyi
- CCP602Uploaded byapi-3849444
- Lecture 9: More on advertisingUploaded byTom Holden
- Lecture 1Uploaded byTom Holden
- Cns Simulation.docxUploaded byItee Ashok
- Mock End of Term Test 2012Uploaded byTom Holden
- Lecture 8: Vertical differentiation and advertisingUploaded byTom Holden
- tpmpresentation-090602104911-phpapp02.pptUploaded byLorenzo Owens
- Monetary liquidities to shrimping economyUploaded by7609dd61
- ETAR.Lecture 2-Demand, Supply, and Equilibrium.E.pdfUploaded byReina
- 4.3Uploaded byapi-3727090
- H2 Econs Essay 1Uploaded byjjnaseem
- 237406_633882287082580595Uploaded byLester de Leon
- Managerial Economics1Uploaded byvijay_paliwal9903
- 155040101111062_DIAN DWI UTAMIUploaded bysola
- Managerial EconomicsUploaded byHema Sakthidharan
- {955a4502-ea04-4b1d-9e86-fd60b129cb3c}_WP7_-_Do_You_See_A_Price_WarUploaded byiqbalj
- ECON 1000 Midterm PracticeUploaded byexamkiller
- 10622513Uploaded byHusna Rosli
- PricingUploaded byUrjit Jhaveri
- Idea Generation Sample AssignmentUploaded byJonelle Morris-Dawkins
- Adv PlanningUploaded byKrishna Chandran

- Lecture 7: More on product differentiation and Introduction to the economics of advertisingUploaded byTom Holden
- Answers to 2012 Mock mid-term testUploaded byTom Holden
- Lecture 5: Price discrimination and nonlinear pricingUploaded byTom Holden
- Lecture 6: Profit reducing price discrimination, and product differentiationUploaded byTom Holden
- Lecture 3: Further static oligopolyUploaded byTom Holden
- Lecture 2: Basic oligopolyUploaded byTom Holden
- Mock mid-term multiple choice testUploaded byTom Holden
- Lecture 8 AuctionsUploaded byTom Holden
- Problem Set 5Uploaded byTom Holden
- Lecture 4: Will profits in reality be higher or lower than under Cournot?Uploaded byTom Holden
- Lecture 9: More on advertisingUploaded byTom Holden
- Lecture 1Uploaded byTom Holden
- Lecture 8: Vertical differentiation and advertisingUploaded byTom Holden
- Lecture 7 Asymmetric Information 2Uploaded byTom Holden
- Problem Set 3Uploaded byTom Holden
- Problem Set 2Uploaded byTom Holden
- Lecture 3.2 ExternalitiesUploaded byTom Holden
- Problem Set 1Uploaded byTom Holden
- Lecture 8Uploaded byTom Holden
- Lecture 3.1 WelfareUploaded byTom Holden
- Lecture 2 Uncertainty 2Uploaded byTom Holden
- Lecture 5 Intertemporal ChoiceUploaded byTom Holden
- Lecture 1 Uncertainty 1Uploaded byTom Holden
- Lecture 6 Asymmetric Information 1Uploaded byTom Holden
- Mock End of Term Test 2012Uploaded byTom Holden
- Lecture 4 Public GoodsUploaded byTom Holden