Advanced Microeconomics

Yu Gao
Lecture 6, 2024
Refresh: firm theories

Firms are special “consumers”. They purchase “inputs”. Their “utility”

comes from profits.

Consumers Firms
Utility function Production function
Indifference curve Iso-quant curves
Utility maximization problem Profit maximization problem
(UMP) (PMP)
Expenditure minimization problem Cost minimization problem
(EMP) (CMP)
Table of contents

1. Market
Competitive market

2. Monopoly

3. Monopoly behaviors

4. General equilibrium

1
Market
Competitive market
Competitive market

A competitive firm takes the market price of output as given.

How to understand that?

Let p̄ be the market price. Then the market demand curve facing a
competitive firm takes the form


 0 if p > p̄
D(p) = any amount if p = p̄

 ∞ if p < p̄
The competitive firm is free to set whatever price it wants and produce
whatever quantity it is able to produce. However...
• If sell above market price, no demand.
• If sell below the market price, not wise, because customers are
willing to buy at higher.
• So it must sell at the market price.
2
Firm’s supply

Shutdown vs. Exit

• Shutdown: A short-run decision not to produce anything because of

market conditions
• Exit: A long-run decision to leave the market
• A key difference:
• If shut down in the short-run, must still pay fixed cost.

Example: a restaurant that shuts down during holidays.

3
A firm’s short run decision to shut down

The firm has two concerns:

1. The cost faced by a firm in the short run:

c(q) = cv (q) + cf
The firm will produce when the revenue of producing exceeds the
variable cost:
pq − cv (q) ≥ 0
cv (q)
p≥ = AVC
q
2. Remember from solving for PMP and CMP we know that at the
optimum, price equals marginal cost:

∂c(w, q∗ )
p= = MC
∂q

4
A firm’s short run supply curve

{
0 if p̄ < minAVC(q)
q=
MC−1 (p) if p̄ ≥ minAVC(q)

5
Identify loss and gain

The left figure is a loss situation in case the firm still produces at the
level MC−1 (p̄) while the right figure is a gain situation.

6
A firm’s long run decision to exit

In the long-run, fixed cost should be taken into consideration. The firm’s
concern becomes:
pq − c( q) ≥ 0
c(q)
p≥ = ATC
q

• The firm will exit if p < ATC.

• New firm will enter if p ≥ ATC.

7
The industry supply function

Remember that the industry supply function is simply the sum of the
individual firm supply function.
Example: consider 100 identical firms (in the short run)

8
Long-run supply curve

Long-run equilibrium: the process of entry or exit is complete – remaining

firms earn zero economic profits.

The opposite (new entry) is similar.

9
Long-run supply curve

The opposite (new entry) is similar.

10
Long-run supply curve

The long-run market (industry) supply curve is horizontal at

p = minimumATC.

Think
What determines the min AC?

11
Zero profit

Think
If in the long run all firms have zero profits, then how could the market
exist?

12
Zero profit

Think
If in the long run all firms have zero profits, then how could the market
exist?

• Economic profits include salary, interest and rent (of one’s own
labor, money, and land).
• Entrepreneurial puzzle: entrepreneurs earn less and bear more risk
than salaried workers with otherwise similar characteristics.

12
Zero profit

Think
If in the long run all firms have zero profits, then how could the market
exist?

• Firms may have different costs (technologies). Then for the average
firm, profit=0. Others are either earning or losing money.
• In some industries, the supply of a key input is limited (e.g., the
number of trained workers, the amount of land suitable for farming).
So no extra entry.

13
Monopoly
Monopoly

A monopoly is a firm that is the sole seller of a product without close

substitutes.
Key difference between perfect competition and monopoly:

• A monopoly firm has market power: the ability to influence the

market price of the product it sells
• A competitive firm has no market power

14
Why monopolies arise? The main cause of monopolies is barriers to
entry – other firms cannot enter the market.
• A single firm owns a key resource: De Beers Group sells
approximately 35% of the world’s rough diamond production.
• The government gives a single firm the exclusive right to produce
the good or service: e.g., railways.
• Natural monopoly: a single firm can produce the entire market
demand at lower cost than could several firms: e.g., electricity.

15
Profit maximization

• Consider a demand function x(p), which is continuous and strictly

decreasing in p, i.e., x′ (p) < 0.
• We assume that there is price p̄ < ∞ such that x(p) = 0 for all
p > p̄.
• Also, consider a general cost function c(q), which is increasing and
convex in q.
• In addition, we assume that p(0) ≥ c′ (0). That is, the (inverse)
demand curve originates above the marginal cost curve. Hence, the
consumer with the highest willingness to pay for the good is willing
to pay more than the variable costs of producting the first unit.

16
Profit maximization

A monopolist’s profit maximization problem:

max p(q)q − c(q)

q

F.O.C.,
d(p(q)q) dc(q)
=
dq dq

marginal revenue = marginal cost

Think
In PMP, how to see the difference between a monopolist and a
competitive firm?

17
Profit maximization

Marginal revenue in monopoly

d(p(q)q)
MR = = p (q ) + p′ (q )q
dq
MR describes two effects:

• A direct (positive) effect: an additional unit can be sold at p(q),

thus increasing revenue by p(q).
• An indirect (negative) effect: selling an additional unit can only
be done by reducing the market price of all units (the new and all
previous units), ultimately reducing revenue by p′ (q)q.

18
Profit maximization: an example

Using the example of a linear demand curve: p(q) = a − bq.

• The revenue function should be p(q)q = aq − bq2 .
• The marginal revenue (MR) function is MR(q) = a − 2bq.

19
• A competitive firm takes p as given, therefore as a “price-taker”, has
a supply curve that shows how its output depends on a given p.
• A monopolist is a “price-maker”. This power gives profits>0. There
is no supply curve for monopoly.
20
Is monopoly “bad”?

Competition vs. monopoly

• A competitive industry operates at a point where pc = MC(qc ). A

monopolized industry operates where pm > MC(qm ).
• With monopoly, price is higher (pm > pc ) and output is lower
(qm < qc ). The producer is better off (profit >0) and the consumer
is worse off.
• One must take a value judgement about the relative
welfare of consumers and the owners of firms.
• However, we can have an argument against monopoly on grounds of
efficiency!

21
Consumer surplus (CS)

22
How does CS change when price changes?

There are two areas in the changes in CS:

• Area R measures the loss in surplus due to the fact that the
consumer is now paying more for all the units he continue to
′′ ′ ′′
consume: (p − p )x .
• Area T measures the lost in surplus due to the lost consumption of
the x-good.
23
Producer’s surplus (PS)

The supply curve measures the amount that will be supplied at each
price.

• The area above the supply curve measures the surplus enjoyed by
the suppliers of a good.

24
Producer’s surplus (PS)

′ ′′
Price change from p to p . There are two areas in the changes in PS:

• Area R measures the gain from selling the units previously sold
′ ′′
anyway at p at the higher price p .
′′
• Area T measures the gain from selling the extra units at the price p .

25
Welfare effect: deadweight loss

26
Welfare loss of monopoly

27
Welfare loss of monopoly

Consumer’s surplus:

• Perfect competition: A+B+C

• Monopoly: A

Producer’s surplus:

• Perfect competition: D+E

• Monopoly: D+B

Deadweight loss of monopoly: C+E

∫ q∗
DWL = [p(s) − c′ (s)]ds
qm

28
Welfare loss of monopoly

Consider demand with infinite

elasticity p′ (q) = 0.
• The (inverse) demand curve
becomes totally flat.
• Marginal revenue coincides with
inverse demand.
• Hence, qm = q∗ and DWL=0.

29
Monopoly behaviors
Price discrimination

Selling output at different prices is called price discrimination.

The characteristic used in price discrimination is willingness to pay
(WTP):

• A firm can increase profit by charging a higher price to buyers with

higher WTP.
• Different degrees of discrimination are concerned with different ways
to sort the consumers.

30
First-degree price discrimination

First-degree price discrimination : discriminate consumers

one-by-one accordingly to each consumer’s WTP.

31
First-degree price discrimination

First-degree price discrimination : perfect price

discrimination

• Examples: bargaining at the market.

• Inexperienced consumers like places that have a fixed price to avoid
“been discriminated”.
• A market very close to perfect discrimination: auto repair market
• Gneezy, List and Price (2012) found that disabled receive offers that
are 30 percent higher than the offers received by the abled.
• One possible reason is search cost differences one would expect
search to be more costly for the disabled.
• Magic words: I am getting a few price quotes.

32
First-degree price discrimination

Disadvantages:

• This is ideal for producers, however very costly!

• Experienced consumers can “pretend” to have low WTP.

33
First-degree price discrimination

Think
What could be included in the algorithem?

34
Second-degree price discrimination

Second-degree price discrimination : segregate consumers

by “self-selection” (consumer type is private).

• Experienced consumers can pretend to have low WTP.

• The monopolist cannot observe the type of each consumer (e.g., his
willingness to pay).
• The monopolist can encourage “self-selection” by adjusting
the quantity or/and quality of the good.
• Consumers have other concerns beyond price (e.g., space, comfort,
time, etc.)
• Examples: “bulk discounts”, “Double 11”.

35
Second-degree price discrimination

36
Second-degree price discrimination

37
Second-degree price discrimination

The monopolist offers a menu of two-part tariffs, (FL , qL ) and (FH , qH )

with the property that the consumer with type i = {L, H} has the
incentive to self-select the two-part tariff (Fi , qi ) meant for him.
Assume the utility function of type i consumer

Ui (qi , Fi ) = θi u(qi ) − Fi

where

• qi is the quantity of a good consumed.

• Fi is the fixed fee paid to the monopolist for qi .
• θi measures the valuation consumer assigns to the good, where
θH > θL , with corresponding probabilities p and 1 − p.

The monopolist’s constant marginal cost c satisfies θi > c for all

i = {L, H}.

38
Second-degree price discrimination

The monopolist must guarantee that

1. both types of customers are willing to participate (“participation

constraint (PC)”)
• the two-part tariff meant for each type of consumer provides him
with a weakly positive utility level
2. customers do not have incentives to choose the two-part tariff meant
for the other type of customer (“incentive compatibility (IC)”)
• type i customer prefers (Fi , qi ) over (Fj , qj ) where j ̸= i.

39
Second-degree price discrimination

Considering the PC and IC, the monopolist’s profit maximization problem

becomes
max p[FH − cqH ] + (1 − p)[FL − cqL ]
FL ,qL ,FH ,qH

PCL : θL u(qL ) ≥ FL (1)

PCH : θH u(qH ) ≥ FH (2)
ICL : θL [u(qL ) − u(qH )] + FH ≥ FL (3)
ICH : θH [u(qH ) − u(qL )] + FL ≥ FH (4)

40
Second-degree price discrimination

Both PCH (2) and ICH (4) are expressed in terms of ≥ FH .

• Ideally the monopolist wants to charge a FH as high as possible.

• The monopolist increases FH until such fee coincides with the lowest
of θH u(qH ) and θH [u(qH ) − u(qL )] + FL for all i = {L, H}.
• Otherwise, one (or both) constraints will be violated, leading the
high-demand customer to not participate.

41
Second-degree price discrimination

High-demand customer:
Let us show that ICH is binding while PCH is not.

• Proving by contradiction, assume that FH = θH u(qH ).

• Then, ICH can be written as

θH [u(qH ) − u(qL )] + FL ≥ FH
FH − θH u(qL ) + FL ≥ FH
⇒ FL ≥ θ H u ( qL )

• Combining this result with the fact that θH > θL ,

FL ≥ θH u(qL ) > θL u(qL )

• Which contradicts PCL .

We then reached contradiction. Thus, FH < θH u(qH ). ICH is binding but

PCH is not.
42
Second-degree price discrimination

Similarly, both PCL (1) and ICL (3) are expressed in terms of ≥ FL .

• Ideally the monopolist wants to charge a FL as high as possible.

• The monopolist increases FL until such fee coincides with the lowest
of θL u(qL ) and θL [u(qL ) − u(qH )] + FH for all i = {L, H}.
• Otherwise, one (or both) constraints will be violated, leading the
low-demand customer to not participate.

43
Second-degree price discrimination

Low-demand customer:
Let us show that PCL is binding while ICL is not.

• Proving by contradiction, assume that

θL [u(qL ) − u(qH )] + FH = FL .
• Then, ICH (which has proved to be binding in the previous slide) can
be written as

θH [u(qH ) − u(qL )] + FL = FH
θH [u(qH ) − u(qL )] + θL [u(qL ) − u(qH )] + FH = FH
⇒ θH [u(qH ) − u(qL )] = θL [u(qH ) − u(qL )]
⇒ θH = θL
• which violates the initial assumption θH > θL .

We then reached contradiction. Thus, FL < θL [u(qL ) − u(qH )] + FH .

PCL is binding but ICL is not.
44
Second-degree price discrimination

In summary,

• From PCL binding we have

θL u(qL ) = FL

• From ICH binding we have

θH [u(qH ) − u(qL )] + FL = FH

45
Second-degree price discrimination

The momopolist’s expected PMP can then be written as

max p[FH − cqH ] + (1 − p)[FL − cqL ]

qH ,qL ≥0

= p[θH u(qH ) − (θH − θL )u(qL ) − cqH ] + (1 − p)[θL u(qL ) − cqL ]

• F.O.C. wrt qH :

p[θH u′ (qH ) − c] = 0 ⇒ θH u′ (qH ) = c

• which coincides with that under complete information (if the

monopolist can distinguish H and L type).
• That is, there is no output distortion for high-demand buyer.
• Informally, we say that there is “no distortion at the top”.

46
Second-degree price discrimination

• F.O.C. wrt qL :

p[−(θH − θL )u′ (qL )] + (1 − p)[θL u′ (qL ) − c] = 0

which can be rewritten as

u′ (qL )[θL − pθH ] = (1 − p)c

• Dividing both sides by 1 − p, we obtain

θL − θH p
u′ (qL )[ ]=c
1−p
• the above expression can alternatively be written as
p
u′ (qL )[θL − (θH − θL )] = c
1−p
The output offered to low-demand customers entails a distortion, i.e.,
qL < qSo So
L where qL is the socially efficient quantity solved by
c = u ′ ( qL ) θ L .
47
Second-degree price discrimination

• Since constraint PCL binds while PCH does not, then only the
high-demand customer retains a positive utility level, i.e.,
θH u(qH ) − FH > 0.
• The firm’s lack of information provides the high-demand customer
with an “information rent”.
• Intuitively, the information rent emerges from the seller’s attempt to
reduce the incentive of the high-type customer to select the contract
meant for the low type.

48
Third-degree price discrimination

Third-degree price discrimination : two separate markets,

where firms can easily enforce the division.

• Examples: youth discounts at the movies, restaurants.

49
General equilibrium
General equilibrium

• all prices are variable.

• equilibrium requires that all markets clear.
• at such an equilibrium, no agent would desire to change his or her
actions.
• it takes into account of all of the interactions between markets.

50
Economies without Production: pure exchange

Basic concern: how goods are allocated?

Here we explore the basic problem of distribution in a simplified setting:

pure exchange.

• There is no production.
• Several consumers, each described by their preferences (utilities) and
the goods that they possess (endowments).
• The agents trade the goods among themselves according to certain
rules and only care about his or her individual well-being.

51
Pure exchange: set up

Two individuals A and B, with two products 1 and 2. No production.

• Initial endowment: wA = (w1A , w2A ), wB = (w1B , w2B )

• Consumption bundle (allocations): xA = (x1A , x2A ), xB = (x1B , x2B )
• Utility function: uA (xA ) = uA (x1A , x2A ),uB (xB ) = uB (x1B , x2B )

52
A convenient tool: Edgeworth box

For simplicity, assume two agents. We can represent allocations,

preferences, and endowments in a two-dimensional form.

53
A convenient tool: Edgeworth box

• ICi is the indifference curve of consumer i, which passes through his

endowment point.
• The shaded area represents the set of bundles (x1i , x2i ) for consumer i
satisfying
• ui (x1i , x2i ) ≥ ui (w1i , w2i )
• Consumer A does not exchange w for A.

54
A convenient tool: Edgeworth box

• Bundle B lies inside the lens-shaped area. Thus, it yields a higher

utility level than the initial endowment e for both consumers.
• Bundle D, however, makes both consumers better off than bundle B.
• It lies on “Contract curve”, in which indifference curves are tangent
to one another.
• It is an equilibrium, and Pareto improvements are no longer possible.

55
Pure exchange: Pareto efficiency

Pareto-efficient allocations maximize one agent’s utility while holding the

other agent’s utility constant:
max uA (x1A , x2A )
x1A ,x2A ,x1B ,x2B

s.t. uB (x1B , x2B ) = ū

x1A + x1B = w1
x2A + x2B = w2
Lagrangian procedure:
L = uA + λ(ū − uB ) + µ1 (w1 − x1A − x1B ) + µ2 (w2 − x2A − x2B )
F.O.C.,
∂uA /∂x1A µ
MRSA = = 1
2
∂uA /∂xA µ2
∂uB /∂x1B µ
MRSB = = 1
∂uB /∂xB2 µ 2
Indifference curves of two agents are tangent to each other. 56
Pure exchange: Pareto efficiency

• The set of all Pareto optimal allocations is known as the Pareto set.
• The contract curve is the part of the Pareto set where
both consumers do at least as well at their initial endowments.
• We might expect any bargaining between the two consumers to
result in an agreement to trade to some point on the contract curve.
57
Pure exchange: Walrasian equilibrium

Now we explore the equilibrium in economics where we allow prices to

emerge.
• The budget line go
through endowment
• For each price vection p = (p1 , p2 ), point w with slope
consumer i′ s wealth equals the market −(p1 /p2 ).
value of his endowments of • Only allocations on the
commodities, p · wi = p1 w1i + p2 w2i . budget line are
affordable to both
consumers
simultaneously at prices
(p1 , p2 ).

58
Pure exchange: Walrasian equilibrium

Throughout, we assume that two consumers act as price takers. Although

this may not seem reasonable with only two traders, we could assume
that each consumer (perhaps better called a consumer type) stands, not
for an individual, but for a large number of identical consumers.
Definition
A Walrasian (or competitive) equilibrium for an Edgeworth box
economy is a price vector p∗ and an allocation x∗ = (x1∗ , x2∗ ) in the
Edgeworth box such that for i = A, B,

xi∗ ⪰i xi′ for all xi′ ∈ Bi (p∗ )

At a market equilibrium, markets should clear.

59
Pure exchange: Walrasian equilibrium

• A Walrasian equilibrium
(market is clear)

• A price vector with excess demand

for good 2 and excess supply for
good 1.

60
Pure exchange: Walrasian equilibrium

1. Each individual maximizes own utility

max uA (x1A , x2A ) s.t. x1A p1 + x2A p2 = w1A p1 + w2A p2
x1A ,x2A

max uB (x1B , x2B ) s.t. x1B p1 + x2B p2 = w1B p1 + w2B p2

x1B ,x2B
F.O.C.,
∂uA (x1A ,x2A ) ∂uB (x1B ,x2B )
∂x1A ∂x1B p1
= =
∂uA (x1A ,x2A ) ∂uB (x1B ,x2B ) p2
∂x2A ∂x2B
2. Market clearing (Feasibility)
x1A + x1B = w1A + w1B
2
x2A + x2B = wA + w2B
Equilibrium occurs at a point where the two indifference curves are
tangent.
61
Pure exchange: Walrasian equilibrium

• As the price p varies, the budget line pivots around w, and the
demanded consumptions trace out a curve, denoted by OC (offer
curve).
• Any intersection of the consumers’ offer curves at an allocation
different from the endowment point w corresponds to an equilibrium.

62
Pure exchange: Walrasian equilibrium

• An example of Nonexistence of
Walrasian equilibrium • Multiple Walrasian equilibrium

63
First theorem of welfare economics

First Theorem of Welfare Economics

If (x, p) is a Walrasian equilibrium, then x is Pareto efficient.

64
First theorem of welfare economics

Some implicit assumptions:

• Each agent only cares about his/her own utility (no externalities).
• Each consumer is sufficiently small relative to the size of the market
(price-taker).

With millions of agents, the only thing a consumer needs to know is the
prices of the goods. With “invisible hand”, an efficient outcome is
guaranteed.
• Compare to a social planner (who needs to know the preferences of
every agent to allocate), market can already achieve efficiency with
minimal cost.
• Pareto efficiency does not care about equity/equality / .
• The only possible welfare justification for intervention in the
economy is the fufillment of distributional objectives.

65
Second theorem of welfare economics

The second theorem provides a (partial) converse result.

Second Theorem of Welfare Economics

• Under convexity assumptions, a planner can achieve any desired
Pareto optimal allocation by appropriately redistributing wealth in a
lump-sum fashion and then “letting the market work”. 66
Second theorem of welfare economics

• The second welfare theorem implies that the problem of distribution

and efficiency can be separated.
• The market mechanism is distributionally neutral; whatever your
criteria for a good or a just distribution of welfare, you can use
competitive markets to achieve it.
• However, this might not be realistic because of information and
practical reason. It is difficult to tax on endowments.
• If you tax on choices, there will be distortions.
• We can only achieve the “second-best” (the situation when one or
more optimality conditions cannot be satisfied).

67
Questions?

67