4 Microeconomics Questions
Fokkusu
LectureNotes.pdf
Slides 10
(Chapter 13)
2
Predatory conduct
We studied entry deterrence via capacity
investment
We will now study predatory pricing
Set up a low price to avoid entry (asymmetric info)
Set us a low price to force exit
3
Predation with asymmetric info
Let entry decision depend on incumbent’s costs
If the incumbent is low cost, do not enter
If the incumbent is high cost, enter
Potential entrant doesn’t know incumbent cost
Does a high-cost incumbent have an incentive to
pretend to be low-cost to prevent entry? Such as
price like a low-cost firm?
4
An example of predation via
asymmetric information Incumbent has a monopoly in period 1
Threat of entry in period 2
Market closes at the end of period 2
Entrant observes incumbent’s actions in period 1
These actions determine whether to enter in period 2
Incumbent might be high-cost or low-cost
no direct information on incumbent’s costs
entrant knows with probability α incumbent is low-cost
Need to specify payoffs in different situations
5
Incumbent profits in period 1 (in $million)
low-cost firm acting as low-cost monopolist: 10
high-cost firm acting as high-cost monopolist: 6
high-cost adopting low-cost monopoly price: 4
Incumbent profits in period 2
if no entry, profits according to true type (10 or 6)
if entry occurs:
low-cost incumbent: 5
high-cost incumbent: 2
Entrant profits in period 2
competing against a low-cost incumbent: -2
competing against a high-cost incumbent: 2
An example of predation via
asymmetric information
6
What does the structure tell us?
What if there is no uncertainty—i.e. entrant knows
whether incumbent is high cost or low cost?
If the incumbent has high costs, enter
If the incumbent has low costs, don’t enter
How does the entrant make a decision with
uncertainty?
Should the incumbent change what she does when there is
uncertainty?
Games like this really underline the importance of “thinking
like an economist”
7
Game tree for our example
Nature
High-
Cost
Low-Cost
I
I
High Price
Low Price
E
E
Enter
Stay
Out
Incumbent: 6 + 2 = 8
Entrant: 2
Incumbent: 6 + 6 = 12
Entrant: 0
Enter
Stay
Out
Incumbent: 4 + 2 = 6
Entrant: 2
Incumbent: 4 + 6 = 10
Entrant: 0
Low Price Enter
Stay Out
Incumbent: 10 + 5 = 15
Entrant: -2
Incumbent: 10 + 10 = 20
Entrant: 0
8
What if a high-cost incumbent
pretends to be low-cost? Consider a high-cost incumbent. It can
Price high in period 1
entry will certainly occur, total profits are 8
price low in period 1 if no entry occurs, total profits are 10
if entry occurs, total profits are 6
High-cost incumbent has incentives to pretend to be low-cost if by so doing it can deter entry
But, will the entrant really stay out if it observes the incumbent price like a low-cost firm?
9
What if a high-cost incumbent
pretends to be low-cost?
If the entrant is a clever, rational firm, it knows that the incumbent may try to deceive it
Issue: What can the entrant infer from observing a low price knowing this may be a deception?
It depends on the probability that observing a low-price means the incumbent is a low-cost firm
10
If the entrant observes the incumbent setting a low-price in period 1, it cannot tell whether the incumbent is high cost or low cost
Thus, the entrant must rely on the unconditional probability that the incumbent is low-cost
Suppose the incumbent is low-cost with probability , in which case entry will lead to a profit of –2
The incumbent is high-cost with probability 1 - , in which case entry will lead to a profit of 2
So expected profit is -2 + 2(1 - ) = 2 - 4
What if a high-cost incumbent
pretends to be low-cost?
11
Expected profit is 2(1 - ) - 2 = 2 - 4
This is negative if > ½
So, if the probability that the incumbent is low cost is “sufficiently high”, an incumbent can deter entry by setting a low price in period 1
What if a high-cost incumbent
pretends to be low-cost?
12
More on predatory pricing
We discussed setting low prices to deter entry
How about setting low prices to force exit?
It has been the subject of numerous legal cases
Much more so than the previous one!
High price firms claim low price firms are predating
International “dumping” cases
13
Classic case:
Matsushita vs Zenith (1986) Allegation in Matsushita vs Zenith
Group of Japanese firms engaged in predatory
pricing against U.S. competitors
The goal was to drive U.S. firms out of the market
Zenith’s evidence noted low prices by Japanese
firms (especially relative to their prices in Japan)
and lost market share by U.S. firms
Penalty if found guilty of predatory pricing
pay triple damages
Civil penalty only; no jail time
14
Classic legal identification of
predatory pricing: P<AVC
Classic standard: pricing is predatory if P < AVC
Why AVC?
Short-run “shut-down” price of profit-maximizing firm
Could use MC, but too difficult to measure
Charging prices lower than competitors’ costs is
not predatory
Does not dissuade inefficient competitors from suing
Nor is charging P < (own ATC)
15
Classic economic idea behind
predatory pricing
$/unit
Quantity
Demand
MR QM
PM
AVC=MC
QP
ATCMonopoly
Profit
Firm needs to drive out competitors so it
can charge the monopoly price Pm
16
Classic economic idea behind
predatory pricing
$/unit
Quantity
Demand
MR QM
PM
AVC=MC
PP
QP
ATC
Losses during predation
Monopoly
Profit
Charge price PP during predatory
phase. This is below your AVC and that
of your rivals. Rivals will have to exit.
17
What are some requirements for
successful predatory pricing?
The predator must be successful in two ways
forcing out competitors with a low price AND
making money as a monopolist
What does the firm needs?
Ability to satisfy demand at lower prices
Otherwise competitors capture “residual demand” and
make money
Need excess capacity or low capacity costs to do this
Sufficient financing to withstand losses
Market power in the monopolist phase
18
Successful predatory pricing
means you must recover losses Is this a profitable strategy?
Predation requires sustaining current losses with
the expectation of recouping monopoly profits later
It is just like an investment problem
Worthwhile if NPV(losses) < NPV(future gains)
Losses are incurred in present, while gains are
less valuable because they come in the future
Need to keep firms from entering in the future
So, you need exit barriers to be low, and entry
barriers to be large
This seems tough!!
19
Back to Matsushita vs Zenith
(1986) Arguments of plaintiffs (U.S. TV manufacturers)
If market is competitive, televisions should sell for
similar prices in different countries
Prices were lower in the U.S. than in Japan
U.S. firms losing market share in the U.S. (over period
of about 10 years)
This must be predation!
Does the plaintiffs’ argument hold water?
20
The NPV of Predation:
Matsushita v. Zenith
Variable Level
Predatory price (% of “competitive” price) 62%
Years of predation 20
Growth in demand 5%
Japanese beginning mkt. share 0.05
Japanese ending mkt. share 0.42
Recoupment price (% of “competitive” price) 119-138%
Alleged predation wasn’t successful even after 20 years—U.S. firms had not exited
NPV calculation is negative even if recoupment period is infinite!
21
From the Supreme Court
decision “In order to recoup their losses, petitioners [Japanese]
must obtain enough market power to set higher than
competitive prices, and then must sustain those prices
long enough to earn in excess profits what they earlier
gave up in below-cost prices. Two decades after their
conspiracy is alleged to have commenced, petitioners
appear to be far from achieving this goal: the two largest
shares of the retail market in television sets are held by
RCA and respondent Zenith, not by any of petitioners.”
22
From the Supreme Court
decision
“The alleged conspiracy's failure to achieve its ends in
the two decades of its asserted operation is strong
evidence that the conspiracy does not in fact exist.
Since the losses in such a conspiracy accrue before
the gains, they must be "repaid" with interest. And
because the alleged losses have accrued over the
course of two decades, the conspirators could well
require a correspondingly long time to
recoup…petitioners would most likely have to sustain
their cartel for years simply to break even.”
23
Legal tests for predatory pricing
There are now two tests that must be passed for
a legal finding of predatory pricing
1. Are prices below a reasonable measure of seller’s
costs?
2. Is there a reasonable chance that the seller will
recoup its investment?
Conservative standard, but it’s better to err by
allowing some predatory pricing than to condemn
some competitive pricing
setting prices low is a hallmark of competition
successful predation is rare (so that the likelihood
of false acquittals is low)
24
Another type of behavior that
looks like predation
Microsoft sold its new premium Xbox 360 games
console at around a third less than the price of its
components - and that is not counting the cost of
assembly, testing, distribution and marketing,
according to research firm supply
Is this predation or are there other stories?
25
Other stories
Sell Xbox console for cheap so people buy more
games. Make a big margin on the games
Probably the main reason
Network effects / word-of-mouth demand: Set low
price initially so lot of people buy. Demand increases
later once people see that the Xbox is lots of fun
Learning-by-doing: Cost of making an Xbox
decreases as you make more of them
Spread the Microsoft brand so people buy other
Microsoft products
26
Why so many suits then?
Current system may make suits a good gamble
Plaintiffs can hire lawyers on contingency, while
defendants typically pay by the hour
Large defendants have extremely high legal costs (per
hour)
In the 1980s AT&T spent $100 million per year
defending against predation claims
To illustrate, MCI won a jury verdict against AT&T
that, with tripling, would have amounted to a $1.8
billion judgment in 1980 dollars
27
Why so many suits then?
Quotes from interviews with jurors in major predatory pricing
trials
"the jurors were overwhelmed, frustrated, and confused by
testimony well beyond their comprehension....”
“At no time did any juror grasp--even at the margins--the law,
the economics, or any other testimony related to the
allegations or defense."
“At no time have I encountered a juror who had the foggiest
notion of what oligopoly, market power, or average variable
cost meant, much less how they applied to the case....
Typical is the response I received when I asked a juror
whether he remembered average variable cost. The juror
replied, ‘Yes, explain it to me. I still don't know what it
means.’”
28
Suing to collude?
Now imagine how a predatory pricing lawsuit could
be used to help organize a tacit collusion scheme:
Plaintiff firm, unhappy with price cuts by a rival, files a
complaint detailing what is wrong with defendant's
price cuts and what reasonable prices would be
Then follows a prolonged period of discovery (often
years) in which
The firms exchange reams of sensitive competitive documents,
senior executives testify about their pricing strategies, business
plans, productive capacity, costs, and many more pieces of
business information that will come very handy to both sides
when considering future pricing and output decisions
29
Suing to collude?
While discovery and trial takes place, a judge
closely scrutinizes the parties' pricing behavior
Judges have issued injunctions against
defendants, prohibiting them from lowering their
prices until a final adjudication of the case
So we have several ingredients of a successful
tacit collusion scheme price signaling
information exchange
policing mechanisms
sanctions for deviating from collusive prices!
30
Suing to collude?
American announced its “value pricing” plan on April
9, 1992, and other major carriers quickly matched or
beat American's price cut. Between April and June,
fares remained relatively flat. Then Continental filed
its predatory pricing lawsuit in early June of 1992
and Northwest filed its parallel suit a few days later
A few days after filing suit, Northwest announced a
10 percent price increase
American and Continental soon followed with price
increases of their own
Between July and the end of the year, while the predatory
pricing case progressed, the major airlines reportedly raised
prices seven times
Slides 11
(Chapters 5 and 6)
2
Price discrimination
First degree (“perfect”): the firm is able to charge
each consumer their willingness to pay
Monopolist observes the WTP of every consumer
No DWL!!!
Second degree: The firm knows there are different
types of consumers, but can’t distinguish them
Typically implemented with quantity discounting or
“block” pricing (selling multiple units of the good)
Third degree: The firm can identify different groups
of consumers and charge them different prices
3
3rd degree price discrimination
4
3rd degree price discrimination
5
2nd degree price discrimination
Stata can 3rd degree price discriminate because
it can find out whether its buyers are students or
faculty
But what if Stata can’t tell?
Motivating example: a club selling drinks
Club has high-demand sector (students), and low-
demand sector (faculty)
But can’t tell them apart
6
2nd degree price discrimination
Idea: use “non-linear pricing” to get the “types” of
consumers to sort themselves
“Non-linear pricing” often involves charging
different consumers different prices based on the
quantity or quality purchased
Example: Air travel
7
Simple example of quantity discount:
two-part tariff
Start with simple model of a club: only one type of
customer ---students
Each student’s demand is P = 16 - 2Q
Marginal cost is $4
Benchmark: standard linear pricing
DMR
10
4 MC
8
Simple example of quantity discount:
two-part tariff
Per-student demand is P = 16 - 2Q; MC = 4
Now, what about a two-part tariff? Charge a cover
charge F plus a per-drink price p
Per-drink price p?
Quantity of drinks sold?
Fixed fee?
Profits?
D
16
4
6
MC
9
Simple example of quantity discount:
two-part tariff
Per-student demand is P = 16 - 2Q; MC = 4
Now, what about a two-part tariff? Charge a cover
charge F plus a per-drink price p
Per-drink price p?
p = 4
Quantity of drinks sold?
Fixed fee?
Profits?
D
16
4
6
MC
10
Simple example of quantity discount:
two-part tariff
Per-student demand is P = 16 - 2Q; MC = 4
Now, what about a two-part tariff? Charge a cover
charge F plus a per-drink price p
Per-drink price p?
p = 4
Quantity of drinks sold?
Q = 6
Fixed fee?
Profits? D
16
4
6
MC
11
Simple example of quantity discount:
two-part tariff
Per-student demand is P = 16 - 2Q; MC = 4
Now, what about a two-part tariff? Charge a cover
charge F plus a per-drink price p
Per-drink price p?
p = 4
Quantity of drinks sold?
Q = 6
Fixed fee?
F = 36
Profits? D
16
4
6
MC
12
Simple example of quantity discount:
two-part tariff
Per-student demand is P = 16 - 2Q; MC = 4
Now, what about a two-part tariff? Charge a cover charge F plus a per-drink price p
Per-drink price p?
p = 4
Quantity of drinks sold?
Q = 6
Fixed fee?
F = 36
Profits?
π = 36 D
16
4
6
MC
13
With one “type” of consumer, two-part
tariff or “block tariff” are like 1st degree PD
With one type: monopolist can extract all surplus
Socially efficient outcome
Alternative, equivalent approach to two-part tariff:
block tariff
Like a “package deal”
Sell consumers 6 drink tickets for total charge $60
14
More interesting: two types of consumers
Let demand for beer differ between two groups
Faculty: P = 12 - 2Q
Students: P = 16 - 2Q
Marginal cost is still $4
There are twice as many faculty as students
15
Suppose we can distinguish the two types
If you can distinguish the types, then you can still 3rd
degree price discriminate
Charge a different two part-tariff to each group
16
8
12
6
44 MC MC
64
$36 $16
StudentsFaculty
Low fixed price = $16
Per unit price = MC
High fixed price = $36
Per unit price = MC
16
More interesting: we can’t distinguish the
two types
Suppose you still offer the same two-part tariffs
Which option will faculty take? Students?
16
8
12
6
44 MC MC
64
$36 $16
StudentsFaculty
Low fixed price = $16
High fixed price = $36
Per unit price = MC Per unit price = MC
17
More interesting: you can’t distinguish the
two types. Block pricing case.
Now suppose you still offer the same block tariffs
Which option will faculty take? Students?
16
8
12
6
44 MC MC
64
$36 $16
StudentsFaculty
Block for faculty:
$32 for 4 tickets
Block for students:
$60 for 6 tickets
$16 $24
18
More interesting: you can’t distinguish the
two types. Block pricing case.
If a student buys the faculty block, valuation of 4
drinks is $48
$48 > $32 charge, so choose the faculty block
12
6
4 MC
4
$16
Faculty
Block for faculty:
$32 for 4 tickets
16
8
4 MC
64
$48
Students
19
When you can’t distinguish types, can’t
perfectly discriminate anymore
Club owner wants to charge a higher cover charge to students
But students will “pretend” to be faculty rather than pay the high cover
Even if they don’t get as many drinks under block tariff
Some options for the owner:
Single two-part tariff for all customers
Two block tariffs under which students still want the student tariff
20
Option 1: Charge a single cover price F
plus per-drink price p to everyone
Idea: students can still select to buy more drinks than
adults. But won’t pay higher cover charge
Our earlier two-part tariff examples suggest that
setting p = MC = 4 is a good starting point
Can the firm do better?
21
Single two-part tariff with two types
6
C =4
4
A =12
8
A =16
B B’
22
Option 2: Offer two entry / drink packages
Have a package that students (“high types”) will choose, and a package that faculty (“low types”) will choose
Low type package will include a smaller number of drinks, so that students won’t want it
23
First think about the low
demand types (faculty)
First package from before: costs $32
and includes four drinks
Producer surplus is:
$32-$16 = $16
12
6
4 MC
4
$16
$16
Recall where these numbers come from
Four drinks because that is what they
demand when P=MC
Then, add triangle ($16) and cost ($16)
24
High demand types (students)
First thought: charge $60 for entry + six drinks
16
8
4 MC
6
$36
$24
THIS WON’T WORK!
25
Students buying the faculty
package
16
8
4 MC
6
Package 1: $32
Includes: entry and 4 drinks
If a high demand consumer
buys package 1, the area under
demand is $48 and they pay $32.
So, they get $16 in surplus.
4
$48
Any package must bring high
demand consumers at least $16
in surplus.
High demand Can the firm do anything?
26
What is needed to make this work?
When will the high demand customer tell the firm
that they are indeed a high demand customer?
When it is in their interest to do so!
They will tell truth when they get more utility from
revealing that they are a high demand customer
So, the key for the firm is to design another
package such that students will want to buy it
That is, students will get more utility from this new
package
27
Constructing a package for the
high types
16
8
4 MC
6
Package 2: $????
Includes: entry and 6 drinks
Why should package have 6 drinks?
$60
28
Constructing a package for the
high types
16
8
4 MC
6
Package 2: $????
Includes: entry and 6 drinks
Why should package have 6 drinks?
If a high demand consumer
buys package 2, the area under
demand is $60 $60
What is the most you can charge
them?
29
Constructing a package for the
high types
16
8
4 MC
6
Package 2: $????
Includes: entry and 6 drinks
Why should package have 6 drinks?
If a high demand consumer
buys package 2, the area under
demand is $60 $60
What is the most you can charge
them?
$60-$16 = $44
Producer surplus is:
$44-24=$20
30
Recap so far: 2nd degree PD
with block pricing “High types” are students; “low types” are faculty
If we can observe type, 3rd degree PD leads to:
Faculty: $32 for 4 drinks
Students: $60 for 6 drinks
If we can’t observe type, we must charge the
students less. Otherwise, they will take the faculty
package.
New prices:
Faculty: $32 for 4 drinks
Students: $44 for 6 drinks Profits = $36
31
Can we do even better?
Let’s think about what just happened…
We must lower the price of the high demand
package because we are worried the high
demand consumers will buy the low demand
package
We lower the high demand package price by the
amount of CS a high demand consumer would
get from buying the low demand package
What would allow us to increase the price of the
high demand package and make more money?
Must make the low package less desirable
32
Options for increasing the price of
the high demand package 1. We could increase the price of the low demand
package
Need to include more drinks to compensate the
low types
Would this make the low package more or less
attractive to the high types?
2. We could “degrade” the quality of the low
demand package by offering fewer drinks
Suppose the low demand package only had 3
drinks, but lower cover charge…
33
Degrading the low demand
package… Let’s work this out on the board
Suppose the low demand package only has 3
drinks, but lower cover charge. What happens?
34
Recap of degrading the “low type”
package What happens when we only offer the faculty 3
drinks?
New faculty package: $27 for 3 drinks
New student package: $48 for 6 drinks
New profits: $39. Higher than before!
Note: prices reflect a bulk discount (one student
package is cheaper than two faculty packages)
Why does degrading the low package work?
35
Degrading the low demand
package… Degrading the low demand package improves
profits because:
Low type has low valuation of quantity on the
margin. So don’t have to compensate the low type
too much for this
High type still has high valuation of quantity at the
low type’s margin. Degrading the low-type
package makes it much less desirable to the high
type
Can charge more for the more attractive high-type
package
36
Degrading the low demand
package is common Lots of examples of degrading the low package
Airlines: make coach seats bunched and
uncomfortable. Serve gross food
Trains in India (may not be true today): no roof for
section of train with the cheapest tickets
IBM (may not be true today): intentionally slowed
its cheaper laser printers
Slides 12
(Chapter 7)
2
Multi-product Monopolies:
A bit of gray area… Some of our examples of 2nd degree price
discrimination can also be thought of as differentiated products models
First-class versus economy airfare
Large vs. small coffee at Starbucks
Economists usually say that price discrimination exists in these cases when:
The per unit profits differ across products
That is, price differences are not entirely explained by cost differences
3
Why is Microsoft “evil”?
It is fairly clear that Microsoft has a monopoly, or near-monopoly in operating systems (MS Windows)
Many antitrust cases about Microsoft haven’t exclusively focused on Windows. Other products of Microsoft have always been central to the cases
Internet browsers, media players, server software
We’re not ready to talk about Microsoft yet, but multi-product issues will play a role
4
Multiple products changes
things… Monopolist offers goods of different varieties
The “big” issues are:
pricing
product variety (how many cars do I offer?)
product bundling (how to bundle? price?)
whether to tie the sales of one product to sales
of another
“quality” discrimination
5
Two types of product differentiation:
vertical vs horizontal
Vertical product differentiation
Products differ in quality
We all agree which product has the highest quality
Horizontal product differentiation
Consumers differ in their tastes: not everyone
agrees which product is best
The firm has to decide how best to serve different
types of consumer
6
Product variety and “quality”
discrimination A natural phenomenon with multiproduct firms
Different varieties of the same car: Honda Accord,
EX, LX, XX, TX, etc…
Want some consumers to buy the high-quality
version, others to buy low-quality
Constraints are similar to price discrimination
Try to make sure that each type of consumers buys
the good that was designed for him/her
7
Quality and price discrimination
How to price goods of different quality?
Similar to price discrimination:
Extract all consumer surplus from low quality good
and as much as possible from high quality good
Use schemes to let the customers sort themselves
Set the prices of higher quality goods so that high
demand customers buy high quality product
8
Monopoly with multiple horizontally
differentiated products
A return to Hotelling!
9
Example: Breakfast cereal
Some basic facts about the cereal industry
Economists call it the “ready-to-eat” cereal industry
Highly concentrated industry
Kellogg’s and General Mills are dominant
Do the firms have high margins (price minus
marginal cost)?
10
Some puzzles
There is little entry into cereal despite high
margins. An exception is recent entry of
healthy brands, e.g., Kashi
Despite lack of entry, incumbent firms
frequently introduce new brands (there are at
least 12 types of Cheerios)
11
Does the world need this many
types of cheerios?
12
Entry puzzle in breakfast cereal: no
new firms, but lots of new brands
A. Lack of entry by other firms suggests presence of barriers to entry
B. Incumbent firms introduce lots of new brands
Two potential stories as to why this might be:
1. High costs of advertising / promotion
2. All brands themselves are a barrier to entry
13
What about story 1: High up-front
advertising?
A. Lack of entry by other firms suggests presence of barriers to entry
B. Incumbent firms introduce lots of new brands
How would an advertising story work given (A) and (B) above?
Need that incumbent firms don’t have to spend as much on ads as an entrant
Possible reason: established firms already have substantial brand recognition
Example: Apple-Cinnamon Cheerios in 1989
We’ll talk more about advertising in two weeks
14
What about story 2: Brand
proliferation?
Idea behind the brand proliferation theory of entry
deterrence
Suppose that, without entry threat, it is optimal to
have N brands of cereal ---we will model this!
But now face threat of entry
Why might it be a good idea to introduce more
brands?
15
Idea behind entry deterrence via
brand proliferation
Incumbent is worried about profitable entry if it has N
brands
Entrant could introduce a new brand that is “different
enough” so that it makes positive profits
Suppose the incumbent introduces the new brand
instead
Now there is no “space” for the entrant to insert its new
product
Incumbent makes less profit (too many brands), but is
still competing only with itself
Prices stay high
16
We need a model to help us
answer some questions here
How to model the concept of multiple products in a
horizontally differentiated market?
We’ve seen this already when we considered
differentiated product Bertrand competition
Use the Hotelling line! Except now the products are
all owned by the same firm
We will use this model to answer questions:
What is the optimal number of brands for a
monopolist?
How does this compare to the socially optimal
number of brands?
17
Return to the Hotelling model of
horizontal differentiation
Early in the course, we used the Hotelling line
model to figure out what happens in
differentiated Bertrand competition
Basic idea: the line was an abstract way to
model differing preferences of consumers
e.g., Coke vs Pepsi
People on one side of the line like Pepsi,
people on other side like Coke
18
Return to the Hotelling model of
horizontal differentiation
“Spatial model”: each product sold has a location
on the Hotelling line
Interpretation of “location” can be rich:
“location” can be thought of in
space (geography)
time (departure times of planes, buses, trains)
product characteristics (design and variety)
19
Applying the Hotelling line to a
multiproduct monopoly
Questions we can ask using the Hotelling line
Pricing: What price to charge for each product?
Design: How differentiated should the products be?
Variety: How many varieties?
20
Basic setup of the model is the
same as in multi-firm case
Model the line as Main Street: 1 mile long
N consumers are evenly distributed on the line
Consumers only buy one item from one store
Suppose a consumer located at x buys from the
store located at y
Then utility is U = V – t|y – x| - py
Where t is the “travel cost”
V is the value from consuming the “ideal” product
py is the price set at store y
21
What will the monopolist do?
Monopolist must decide how best to supply these consumers
That is, what price to charge
And, how many stores (brands) to offer
Suppose MC = c
The fixed cost of operating a store is F
How does this compare to the socially optimal number of brands?
We will study the full model in class as it involves many pictures! PRN 7.2 does a similar example, though graph setup is different
