MONOPOLY + PRICE DISCRIMINATION

(DISPENSE +Capitoli 1-2-5-10 CABRAL + pp. 65-75 e 79-86 e 133-140 TIROLE)

(Riassunti delle parti dei libri non contenute nelle dispense)
CAPITOLO I CABRAL
Market Power ability to set prices above cost ( specifically above marginal cost) for the Chicago School no
market power if there is free entry
Creating market power
- legal protection from competition through patents Xerox and photocopies
- firm strategy (free decoder for cable TV price war, but the forerunner can maintain an advantage)
Maintaning market power
- Predatory prices to keep entrants out of the market
- (the opposite) Legal steps against predatory prices of entrants

Implications of Market Power
Transfer from Consumers to Producers
Inefficient Allocation of Resources
Consumers willing to pay more than the marginal cost, but less than the price are not considered
Inefficient Production (Attenzione Cabral dice che i monopoli non sono efficienti, le dispense dicono di
s: Mi sembra che Cabral intenda dynamic efficiency)
Rent seeking When a company, organization or individual uses their resources to obtain an economic gain
from others without reciprocating any benefits back to society through wealth creation. Furthermore large
amount of resources are wasted
An example of rent-seeking is when a company lobbies the government for loan subsidies, grants or tariff
protection. These activities don't create any benefit for society, they just redistribute resources from the
taxpayers to the special-interest group
Roles of the government
- Regulation prices under a regulators oversight
- Antitrust policy (or competition policy) has a broader range of activities prevent firms from taking
actions that increase market power in a detrimental way
- Industrial policy opposite: government tries to strengthen the market position of a firm or an
industry (eg: national firms)

Structure-Conduct-Performance Paradigm: starting from the structure (number of buyers and sellers,
product differentiation and examining the conduct of firms (prices, advertising etc.), one tries to estimate
how competitive the market is.

RICAPITOLAZIONE CONCETTI BASE (CAPITOLO II CABRAL)
Demand curve direct D(p) , inverse P(d)
- Consumers surplus difference between willingness to pay and price paid for all units
- Price elasticiy
dq p
dp q
c =
Costs average, marginal,fixed, variable, total
Competive markets
A perfectly competitive firm's supply curve S
i
(p) is that portion of its marginal cost curve that lies
above the minimum of the average variable cost curve. A perfectly competitive firm maximizes profit by
producing the quantity of output that equates price and marginal cost. As such, the firm moves along
it's marginal cost curve in response to alternative prices. Because the marginal cost curve is positively
sloped due to the law of diminishing marginal returns, the firm's supply curve is also positively sloped.
The short run market supply curve is given by
( )
1
N
i
i
S p
=

where N is the fixed number of firms (in

the short run this number is fixed)
Because the equilibrium price level is the same for all firms in the long run, the long-run market supply
curve is generally horizontal. This may change based on whether the industry has economies or
diseconomies of scale.

Costs and expenditures, opportunity costs and sunk costs
Economic costs differ from expenditures, because they include opportunity costs that do not correspond
to actual expenditures and exclude expenditures that correspond to sunk costs
- Opportunity costs
- Sunk costs
Fixed costs vs. sunk costs
Fixed costs do not vary with the quantity of output produced. Sunk costs are unrecoverable costs. The two
concepts are independent of each other, but are usually related.
An example of a fixed cost that is not sunk: auto insurance paid monthly. It is fixed in that the amount paid does
not depend on miles driven per month (let's suppose this). But it is not sunk, as the insurance can be canceled any
time.
An example of a sunk cost that is not fixed: money you paid for three rolls of film on a final sale. As you cannot
return the film, the money is sunk. But it is obviously a variable cost.
Sunk costs are costs that have been incurred in the past, and are not relevant to the future, and not relevant to a
future decision. A decision has already been made, in the past, and that specific decision cannot be changed. A
new decision can be made regarding the past decision, but, that is a new decision. Sunk costs cannot be changed.
Sunk costs are never relevant to a future decision.
Reading this, it implies then that whenever you want to analyze relevant cost for the future MOST of FIXED COST
(your building, machine & its depreciation expenses) are then irrelevant - thus SUNK COST.

Economies of scale average cost declines with output (diseconomies if cost increases with output) Empirically
economies of scale for q<q -- constant q<q<q diseconomies q>q (eg because raw materials costs increase)
q minimum efficient scale (lowest output level at which minimum average cost is attained
Economies of purpose the cost of producing outputs q
1
and q
2
together is less than producing them separately
C(q
1
, q
2
) < C(q
1
, 0) + C(0 , q
2
)

(CAPITOLO V CABRAL)
REGULATION OF MONOPOLIES
Monopoly with costs C=F+cq absent regulation P
M
>
P
c
Loss of efficiency E
Profit= q
M
(P
M
-c) F =-F

- If the regulation sets the price at c (compettiitive market price) the firm has a negative profit -F
If the regulator gives the firm a subsidy equal to F, it can do this only by increasing taxes elsewhere in
the economy and this can give rise to an inefficiency E >E. The discretion of the regulator to give
funds to the firm regulatory capture (expenses of the firm to capture the regulators
benevolence corruption)
- If the price is set equal to the average cost (average-cost-pricing) the overall profit is zero (or in the
US rate-of-return regulation the price is set so as to guarantee a fair return)


Natural monopolies for networks (Telefoni, elettricit, gas,etc)
One company has the network (upstream monopoly) and provides services (downstream
competition). The upstreammonopoly can try to exploit its position to avoid the entrance of new
firms
The firm has no incentive to reduce costs.
Low incentives even if there is an interval between
cost reduction and price setting price cap for a
period long enough how long? typically 5 or 10
years
- Upstream monopolies can not carry out downstream activities BUT vertical integration
can increase efficiency
- Efficient component pricing rule New entrants must be at least as efficient as the
vertically integrated company. Fix the fee a for using the network
Suppose the marginal cost for the integrated firm is c and the marginal cost of the entrant is
C
E.
p is the price

The new entrant has a positive return only if p-a- C
E
>0
If a=p-c then p-a- C
E
= c - C
E
>0 entrant is more efficient than the integrated firm

CABRAL Capitolo 10 + Tirole 133-140
PRICE DISCRIMINATION
First degree price discrimination


Trade-off between efficiency 1) and consumers welfare 2)
Trade-off between fairness 3) and making the good accessible to as many users as possible 4)

BUT
- there cases in which total efficiency decreases as a result of price discrimination ( price
discrimination costly)
- there are cases in which the Pareto frontier is improved (both firm and consumers are better off)
First degree price discrimination is very difficult and rare (doctor in a small town knows how much each
customer can pay, airplanes are sold at different prices to different airlines (Boeing knows exactly how
much each is willing to pay)
2
nd
and 3
rd
degree price discrimination

Sellers can price-discriminate either based on observable buyer characteristics (3
rd
degree) or
by inducing buyers to self-select among different product offerings (2
nd
degree)

- 3
rd
degree price discrimination
Let
1 2
, . , . ,
m
p p p
be the prices at m discriminated markets and
( ) ( ) ( )
1 1 2 2
,.., ,
m m
D p D p D p
the corresponding demand curves and
( )
1
m
i i
i
q D p
=
=

the aggregate demand p
i
is chosen so as to maximise

If only one price
Consumers surplus B Firms profit A Total Surplus A+B

If the firm could sell the good at different prices corresponding to each
consumers willingness to pay FIRST DEGREE DISCRIMINATION , it would
produce q=q
D
and would collect the whole surplus A+B+C
Differences:
1) Welfare increase (A+B+C)-(A+B)=C>0
2) Consumers are worse off from B 0
3) Different consumers pay different prices
4) More consumers are served (willing to pay less than p
M
but more than c
( ) | | ( ) ( )
1 1 1
m m m
i i i i i i i i
i i i
p D p C q p D p C D p
= = =
(
=
(



1
i
i i
dC
p
dq
p

=
c


Welfare in case of 3
rd
degree price discrimination
Consumers in high elasticity markets prefer discrimination, in low elasticity markets prefer a
uniform price
Let ( ) ( )
, , , ,
i i i i i i
p q S p p q S p be the prices, outputs and consumers surplus under
discrimination and price uniformity respectively.
i i i
q q q A =

The welfare difference (discrimination-uniformity) is given by
( ) ( ) ( ) ( )
1 1 1 1
,
m m m m
i i i i i i
i i i i
W W S p p c q S p p c q
= = = =
= +
(*)

W is convex with respect to price (per quanto riguarda la parte con S(p)
S(p)=-D(p) S(p)=-D(p)>0 )

Se una funzione convessa
( )
'
( ) ( ) ( )
i i i i i
S p S p S p p p > and
( )
'
( ) ( ) ( )
i i i i i i
S p S p S p p p >
e poich
( ) ( )
'
i i
S p D p = dallequazione (*) seguono i due limiti superiori e inferiori
( )
1
m
i i
i
W W p c q
=
> A
and
( )
1
m
i
i
W W p c q
=
s A

Price discrimination reduces the welfare if
1
m
i
i
q
=
A

(differenza di output complessivo) is zero (or

negative). If the price discrimination does not increase the output, it reduces the welfare.
Example: linear demand curves
( ) all markets are served
i i i i i
q a b p a cb = >
Under discrimination the monopolist maximises
( )( )
i i i i
p c a b p
( ) / 2
i i i i
p a cb b = +
( ) / 2
i i i
q a cb =
/ 2
i i i
i i i
q a c b
( | |
=
( |
\ .


If forced to set a uniform price the monopolist maximises
( )
i i
i i
p c a b p
( | |

( |
\ .

/ 2
i i i
i i i
p a c b b
( | | | |
= +
( | |
\ . \ .


/ 2
i i i
i i i
q a c b p
( | |
=
( |
\ .



e quindi
i
i
q

=
i
i
q

cio non c un aumento di output c una diminuzione di welfare

2nd degree discrimination

Two-part tariff (da Wikipedia)
the price of a product or service is composed of two parts - a lump-sum fee as well as a per-unit charge
(ATTENZIONE: su alcuni testi definita first degree (perch nel caso pi semplice annulla il deadweight loss e
recupera tutto il consumers supply, su altri definita second degree, because it does not depend on the identity of
the buyer, but only on the quantity purchased prevale questa seconda definizione)
The unit charge is above or equal to the marginal cost of production, and below or equal to the price the firm
would charge in a perfect monopoly F+cq

The price charged varies not due to different costs borne by the firm. credit cards which charge an annual
fee plus a per-transaction fee is a good example of a two-part tariff, a fixed fee charged by a car rental
company in addition to a per-kilometre fuel fee is not so good, because the fixed fee may reflect fixed costs
such as registration and insurance which the firm must recoup in this manner. This can make the
identification of two-part tariffs difficult.







The variable price c is set to p
c
(competitive price) =marginal cost so that the
maximum amount of q still profitable for the firm is sold. Output equal to
competitive market output.
The fixed cost is equal to A+B+C (deadweight loss + consumers surplus).
Setting F=A+B+C forces the consumer to buy q
c
. Suppose he buys a
reduced quantity q
r
the price paid (A+B+C)+cq
r
would be greater than his
willingness (A+B+C)+cq
r
, because C>C

Single consumers demand
curve (willingness to pay)
If there are n consumers with the same willingness to pay the firms strategy does not change. CS(n)=n*CS(1)






However, it is possible for the firm to earn even greater profits. Assume it sets the unit price equal to Pm, and
imposes a lump-sum fee equal to area A. Both consumers again remain in the market, except now the firm is
making a profit on each unit sold - total market profit from the sale of Qm units at price Pm is equal to area
CDE. Profit from the lump-sum fee is 2 x A = AB. Total profit is therefore area ABCDE.
Thus, by charging a higher per unit price and a lower lump-sum fee, the firm has generated area E more profit
than if it had charged a lower per-unit price and a higher lump-sum fee. Note that the firm is no longer producing
the allocatively efficient output, and there is a deadweight loss experienced by society equal to area F - this is a
result of the exercise of monopoly power.
Consumer X is left with no consumer surplus, while Consumer Y is left with area B.
Bundling
Bundling is packaging two or more products to gain a pricing advantage
Conditions necessary for bundling
Heterogeneous customers
Price discrimination is not possible
Demands must be negatively correlated

Example film company leased Blockbuster, it required theaters to also lease Failure
Blockbuster Failure
Theater A $12,000 $3,000
Theater B $10,000 $4,000
Renting the movies separately would result in each theater paying the lowest reservation price for each
movie:
Maximum price Wind = $10,000
Maximum price Gertie = $3,000
Total Revenue = $26,000
If the movies are bundled:
Theater A will pay $15,000 for both
Theater B will pay $14,000 for both
If each were charged the lower of the two prices, total revenue will be $28,000
The movie company will gain more revenue ($2000) by bundling the movie

If the demand curve (willingness to pay) is different, the price structure
is different: EXAMPLE Y demand curve is twice X demand curve.
The firm cannot identify each consumer ( 2
nd
degree)

The firm would like to follow the same logic as before and charge a per-
unit price of Pc while imposing a lump sum fee equal to area ABCD - the
largest consumer surplus of the two consumers. In so doing, however,
the firm will be pricing consumer X out of the market, because the
lump-sum fee far exceeds his own consumer surplus of area AC.
Nevertheless, this would still yield profit equal ABCD from consumer Y.
A solution to pricing consumer X out of the market is to instead charge a
lump-sum fee equal to area AC, and continue to charge Pc per unit.
Profit in this instance equals twice the area AC (two consumers): 2 x
AC. As it turns out, since consumer Y's demand is twice consumer X's,
ABCD = 2 x AC. The profit is the same and the producer is indifferent to
either of these pricing possibilities, although consumer Y is better off
this way since she gets consumer surplus BD.
More profitable to bundle because relative valuation of two films are reversed
Demands are negatively correlated
A pays more for Blockbuster ($12,000) than B ($10,000)
B pays more for Failure ($4,000) than A ($3,000)
If the demands were positively correlated (Theater A would pay more for both films as shown) bundling
would not result in an increase in revenue, for instance

Blockbuster Failure
Theater A $12,000 $4,000
Theater B $10,000 $3,000

If the movies are bundled:
Theater A will pay $16,000 for both
Theater B will pay $13,000 for both
If each were charged the lower of the two prices, total revenue will be $26,000, the same as by selling the
films separately

Versioning
Second degree quality discrimination

Examples:
1st class/2nd class train tickets
Business class/economy class airfares
Hardcover/paperback books
Damaged goods
Principles
Offer versions tailored to the needs of different customers
Accentuate the differences between the versions

Dimensions to Exploit
- Delay [book publishers, post services, stock quotes]
- User interface [Dialogweb, DataStar]
- Image resolution [Photodisk]
- Speed of operation [Mathematica]
- Capability [Kurzweil]
- Features and functions [Quicken]
- Customer support

It can occasionally reduce the overall welfare

Intertemporal discrimination
A simple two period model due to Bulow (J. Political Economy, 1982):

- t = 1, 2.
- Good bought in period 1 lasts for two periods with no depreciation.
- After period 2, it becomes obsolete.
- Assume production cost = 0.
- Monopolist and consumers have the discount factor
1/ (1 ) r = + o

- Demand for use of the good each period D(p)=1-p


Leasing
No link between periods.

Monopolist sets p
t
so as
max (1 )
t
t t
p
p p
1 2
1/ 2 p p = =
The total discounted sum of profits is (1 )
l
1 1 1
4 4 4
o o H = + = +


Selling with pre-commitment to not lower price in period 2:

Consumers who buy do so in period 1.

If quantity q is sold then the marginal consumer is one whose per period use value is (1 q) and so
the price charged must be:

p = (1+)(1 q)

The monopolist solves
max q(1 + )(1 q)
q
which yields, q
1
= 1/2 , p
1
= (1+)/2 and the total discounted sum of profits is (1 )
1
4
o +
Same market outcome as leasing.

Selling with no pre-commitment

If q
1
( 1) is the quantity of goods sold by the monopolist in period 1 and q
2
is the quantity of new goods
sold in period 2, then the price in period 2 must satisfy:
p
2
= 1 (q
1
+ q
2
):


So, for any q
1
, in period 2, the monopolist will set q
2
to maximize
[1 (q
1
+ q
2
)] q
2

which yields: q
2
(q
1
) = (1 q
1
)/2= p
2
and the profit in period 2
2 1 1
2
( ) ( ) 1
1
q q
4
t =
Now, we go to period 1.

If monopolist sells q
1
in period 1, then the marginal consumer (the lowest valuation consumer that buys) is
the one with utilization valuation equal to (1 q
1
).
The highest price at which he can sell q
1
is a price p
1
(q
1
) such that this marginal consumer is indifferent
between buying the good in period 1 (and enjoying it for two periods) or waiting to buy it in period 2 at a
price p
2
(q
1
) and using it for one period:

(1 q
1
)(1 + ) p
1
(q
1
) = [(1 q
1
) p
2
(q
1
)] = (1 q
1
)/2







so that
1 1 1
1 1
2
( ) ( ) ( ) p q q
o
= +

The intertemporal discounted sum of profits by selling q
1
in period 1:

q
1
p
1
(q
1
) + 2(q
1
)


and maximizing this yield



The total discounted sum of profits:


2
2
(1 )
4 4
o o
o
+ | |

|
+
\ .

which is strictly less than (1 )
1
4
o + , the total discounted sum of profits when the monopolist leases or sells
with pre-commitment to not lower prices.

Discounted Price
value of paid in
use in 2 period 1
periods
Discounted value
of use in period 2
only minus price
paid in period 2
Durable goods, Coase conjecture and evading the Coase problem(a form of price discrimination)

Formal version of the Coase Conjecture:

- Infinitely lived monopolist with unit cost of production c 0.
- Assume there is a continuum of infinitely lived consumers with unit demand whose
valuation of the durable good (the infinite horizon discounted sum of use value) is
distributed on [c, )
- Prices revised at points of time of interval > 0.
- Discount factor: = e
r
where r is the interest rate.

Conjecture: As 0, the intertemporal (discounted sum of) profit of the monopolist 0
and all prices converge to c. In the limit, almost all trades take place at the initial instant.

Main arguments:
Continuum of consumers implies consumers are price taking.
They decide according to their expectation of future prices which they take as given.
Rational expectations equilibrium:
- consumers anticipate the price path to be chosen by monopolist, take this as given
and buy accordingly;
- given this buying behavior, there should no incentive for the monopolist to deviate
from the price path at any point.

Incentive to wait is lower for higher valuation consumers.
For example, for a consumer whose infinite horizon value of using the good is v the gain
from buying today rather than tomorrow:
(v p
t
) (v p
t+1
) = (1 )v p
t
+ p
t+1

is increasing in v.

If consumer with a certain valuation buys today, all consumers with higher valuation must
have bought by the end of today.

For any fixed > 0, equilibrium price path is nonincreasing over time.
If price increases tomorrow, no one will buy tomorrow.

As becomes extremely small, buyers will wait even if the price declines very slightly and so
the price decline must be extremely small if you want some consumers to buy today rather
than wait.

As 0, if current p >> c, then people anticipate price decline in very near future and wait.
So p c.
- Mitigating factors of Coase conjecture:

- Leasing (see above)
- Pre-commitment (see above)
Make your commitment well known, thus putting your reputation on the line.
Make a monetary contract with someone to increase the benefit of staying on course.
Software programs that block internet access for a predetermined period of time
Limited number of items with destruction of moulds (paintings, sculptures, etc.)
- Asymmetric Information: buyers may not know MC of seller.
Low cost seller can pretend to be a high cost seller, charge a price equal to MC of high cost seller
and earn positive profit that is bounded away from zero (even if interval between price revisions are
close to zero).
- Inflow of new cohorts of buyers
The stock of consumers waiting to buy are on the average of lower valuation than new cohort.
Monopolist has incentive to not lower price too much and sell to higher valuation new consumers,
until the stock of old lower valuation consumers become large and at that point to have a big sale
which, in turn, reduces the stock of consumers sharply and then return to high prices again.
Price cycles.
- Planned Obsolescence:
Reducing durability reduces the quantity of goods carried over and thus convinces buyers that the
price wont fall too much.
Textbook producers frequently bring out new editions to kill used books.
- Money back guarantee - compensate consumers for all future price declines.
Effectively, monopolist pre-commits to not cut price.
Problems in enforcement (secret price cuts, quality improvement etc).




PROBLEM SET N. 1
PROBLEMS ON MONOPOLY AND PRICE DISCRIMINATION: 1-2-3-7-8-9-10-14-15

a) Assegnate le curve della domanda (diretta o inversa) e dei costi calcolare la curva della marginal
revenue, prezzo e output del bene
Problemi 1 -2 -10

b) Calcolare (graficamente e numericamente) CS (Consumer surplus) and Deadweight loss

c) Price discrimination
(i) Diversi mercati: ottimizzare la somma dei profitti per trovare prezzi, output e profitti
Problemi 2- 8
(ii) Discriminazione intertemporale
Problema 3-14 : ottimizzazione semplificata rispetto alla teoria del Tirole (ottimizzare la
somma dei profitti: la dipendenza di q2 e p1 da q1 e p2 assicurata dalla funzione di q2)
(Comportamento naive il consumatore non tiene conto del fatto che i prezzi
diminuirannonel periodo 2)

d) Impresa multi-product
Problema 7 -9

e) Calcolo welfare ricordare la formula W=+CS (=profitto, CS area compresa tra curva della
domanda e il prezzo) se p=a-bq e il prezzo p
m
CS= (a- p
m
) q
m
(dove

q
m
=a- p
m
/b)
Problemi 1- 8-10 (Calcolo anche per mercato competitivo =0