OLIGOPOLY DYNAMICS:

COLLUSION AND PRICE WARS

Overview
• Context: Firms interact over time — possibly using
history-dependent strategies
• Concepts: repeated games, grim strategies, collusion, price wars
• Economic principle: repetition helps enforcing otherwise
unenforceable agreements; greater punishment, greater rewards
(topsy-turvy)
Competition externality
• Competition implies an externality: each firm
maximizes own profit, not joint profit
• Incentive to internalize externality — collusion:

− Organized, public cartel agreements

− Secret agreements
− Tacit agreements — focal points
• Types of agreement

− Increase price
− Reduce supply
− Set levels of service quality, advertising, etc
− Territory restrictions
Stability of collusive agreements
• Homogeneous-product duopoly; firms simultaneously set prices;
constant marginal cost (i.e., no capacity constraints)
• If game is played once, Bertrand equilibrium
• What if game is played repeatedly at t = 1, 2, ... (indefinitely)
• Claim: there may exist grim strategy equilibria whereby p > MC

− set p = p M if p = p M in the past

− set p = MC otherwise
Stability of collusive agreements
• Equilibrium profit NPV

1 M 1 1 1 1
V = π + δ π M + δ 2 π M + ... = π M
2 2 2 2 1−δ

• Expected NPV from undercutting rival

0
V 0 = πM +
1−δ

• Condition for Nash equilibrium

1
V ≥ V0 or simply δ≥
2
Digression: relational contracts
• Many (most) economic relations are based on informal
contracts
• Ditto for most international agreements (e.g. WTO,
Kyoto, etc)
• Agreements are self-enforcing if they form a Nash
equilibrium like to one considered above
Stability of collusive agreements
• Collusion is possible if discount factor is sufficiently high, that is,
if future is important enough
• Discount factor: how is one dollar next year worth now?

1
δ=
1+r

• Equilibrium condition δ ≥ 1/2 is equivalent to r ≤ 100%

• Equilibrium condition rather weak; why isn’t there more collusion?
 Why is collusion not more frequent?
• Determinants of discount factor

(1 + g ) (1 − h)
δ=
1 + r /f
where
− r : annual interest rate
− f : frequency of interaction (rounds per year)
− g : industry growth rate
− h: hazard rate (probability of industry or firm end)
• Antitrust policy
• Credibility of infinite price war (topsy-turvy principle)∗
• Observability of prices

∗ The harsher the punishment, the greater the payoff that can be sustained in equilibrium.
US cars in the 1950s

Production (million units) Price (relative auto CPI)

8 1.02

Price Production
7 0.98

6 0.94

5 Year
1953 1954 1955 1956 1957
Price wars

Grain rate

0.4

0.3

0.2

0.1 Time in weeks

from January
0.0 1, 1880
0 40 80 120 160 200 240 280 320 360
Price wars
• Industry where demand fluctuates
• Each firm observes price, own demand, not market demand
• Low firm demand implies guessing problem:
is it low market demand or price cutting by rival?
• If firms assume it’s market fluctuation and keep setting high price,
then incentive is to cut price
• In order for collusion to be an equilibrium, firm must set low prices
when firm demand is low — even though, in equilibrium, low
demand always results from market fluctuations
• If price cuts are difficult to observe, then occasional price wars
may be necessary to discipline collusive agreements
Demand fluctuations
• Observable demand fluctuations (e.g., seasonality)
• Incentive to undercut greater during periods of high demand
(more to be gained)
• Collusive equilibrium may require that prices be lower during
periods of high demand
Demand fluctuations and price wars
• Unobservable demand shocks ⇒ price ↓ demand ↓
• Observable demand shocks ⇒ price ↓ demand ↑
• Which model is right? It depends!

− Freight shipments 19th century: cyclical price wars

− Cement, retail: counter-cyclical price wars
Price wars in asymmetric industries
• In the preceding models, price wars are a coordinated
effort by all firms as part of a collusive equilibrium
• Price wars initiated by weak firms

− Fares are dictated not by the strongest, but by the

financially troubled. —CEO of Alaska Airlines
• Price wars initiated by strong firms

− Murdoch’s 1993 acquisition of the London Times

Market structure and collusion
• Collusion is more likely in concentrated industries than in
fragmented ones:
− Easier to establish a collusive agreement
− Easier to maintain a collusive agreement
− Example: repeated Bertrand with n firms
• Easier to maintain collusion among similar firms

− Example: duopoly where one firm has a cost advantage over the
other one; efficient equilibrium not stable; and symmetric
equilibrium also likely to be unstable
− Example: diamond industry
− Example: bromide cartels
• Bottom line: collusion easier to maintain among few and similar
firms
Bromide cartel
• Six price wars between 1885 and 1914, two of which right after
publicly announced cartel agreement violations
• Price wars result from disagreement among cartel members
regarding profit distribution; not equilibrium price wars in the
sense explained earlier
• If all firms were symmetric, such disagreements would be less likely
Multimarket contact
• Irrelevance result: if all markets are identical, multi market
contact makes no difference
• Relevance result: firm i has an advantage in market i (cost c
versus rival’s c = c + t)
− Efficient equilibrium: firm i takes over market i
− Taken in isolation, these equilibria are not stable
− Taken together, they may be stable
• Bottom line: Collusion is normally easier to maintain when firms
compete in more than one market
Multimarket contact: US airlines
• Market: flight connection between two different cities
• Average contact in market i: average number of other markets
where competing airlines face each other
• Positively correlated with airfares
• Possible explanation: airlines use competition in other routes as a
means to collude in a given route
Multimarket contact: Dog Food Industry
• In 1980s, US dog food sales greater than $3bn/year
• Market segments: dry, moist, snack, canned, and soft-dry
• In 1986, Quaker Oats (dominant in moist) acquired Anderson
Clayton, increasing share in dry market
• Ralston Purina (dominant in dry) responded by acquiring Benco
Pet Food’s Inc., Quaker’s main rival in the moist market,
Hey, we can come at you in your strong area if you
come after us in our strong area
• The war continued: Quaker launched Moist ‘n Beefy, a clear
attack on Ralston’s Moist & Meaty brand. Ralston Purina
introduced Grrravy, a clear attack on Quaker’s Gravy Train brand.
Information sharing
• Perfect information, one of the conditions for perfect competition
• Natural presumption: the more transparent a market is, the more
competitive it is
• Not necessarily so: information sharing and transparency may
enhance collusion
GE and Westinghouse
• Market: large turbine generators
• Produced to order, mostly to electrical utilities. Sellers (GE,
Westinghouse) submit bids or negotiate with buyers
• 1950s: secret bid ring results in jail sentences
• May 1963: GE announces new pricing policy:

− price book made public

− most-favored customer clause created
− accounting firm to audit
• Within less than a year, Westinghouse copies GE’s policy
• Prices remained stable and identical until 1975, when US
Department of Justice forced a consent decree
Danish ready-mixed concrete market
• Concrete market: regional submarkets, typically oligopolies with 3
o4 4 competitors
• Until 1993: list prices subject to individual, confidential discounts
• October 1993: Danish Competition Council decided gathers and
publishes actual transaction prices (weekly basis)
• Result: higher prices
Danish ready-mixed concrete market

Average 10-MPa Concrete Prices in Århus

550

500

450

400

350

300 Time
Jan-94 Apr Jul Oct Mar-95 Jun Nov
Empirical analysis of cartels and collusion
• Information sources: legal cartels, dismantled cartels, indirect
evidence
• Sugar cartel

− Increase transparency (code of ethics)

• Lysine and citric acid cartels

− sales quotas and buy-backs (individualized punishments)

• Collusion in the US auto industry

− Use data to distinguish collusion from alternative story

US auto industry

Production (106 units) Price (relative auto CPI)

8 1.02

Price Production
7 0.98

6 0.94

5 Year
1953 1954 1955 1956 1957
Identifying the oligopoly solution

Price and cost by model and year

• 1954 price
• 1955 price
•
 cost • •
•
•
•
• • 
• 
• 
• •
 

Quality
A1 B2 C1 D2 E1 F1
THE LAW AND ECONOMICS
OF COLLUSION
Overview
• Context: At an industry convention, a competitor complains that
“competition is much too tough.” How should you reply?
• Concepts: antitrust, competition policy, price fixing, civil and
criminal offense, leniency
• Economic principle: crime does not pay
Outline
• Collusion: US and EU law
• Leniency programs
• Facilitating practices
• Allowed horizontal agreements
• Cases
Outline
• Collusion: US and EU law
• Leniency programs
• Facilitating practices
• Allowed horizontal agreements
• Cases
Motivating example: lysine
• How the cartel worked
• How the DoJ found out about it
• Economic, legal, and personal ramifications
• Watch video from Fair Fight in the Marketplace
Price fixing and related illegal practices
• Price fixing is illegal

− Criminal offense in US (Sherman Act)

− Civil offense in Europe (Articles 101, 102 of Treaty of the EU)
• There may be exceptions and exemptions, e.g.:

− Some sports leagues

− European airlines until early 1990s
• Related forms of collusion

− Territorial or no-customer-poaching agreements

− Capacity reduction
History of anti-price-fixing law: US
• US, late 1800s: price instability led to formation of cartels
• In addition to price stability, this also brought about high prices
• Sherman Act, 1890. Sections 1 and 2 prohibit such contracts.
Felony (prison sentences).
• Some practices per-se illegal. Otherwise, rule of reason is applied.
• Important precedents: GE and Westinghouse, Airline tariff
publishing (non-explicit collusion)
 Top recent DOJ fines

Sherman Act Violations Yielding a Corporate Fine of $10 Million or More

Defendant (FY) Product Fine ($ M) Country
F. Hoffmann-La Roche, Ltd. (1999) Vitamins $500 Switzerland
LG Display Co., Ltd LCD panels $400 Korea
LG Display America (2009)
Société Air France and Koninklijke Air Transportation $350 France (SAF)
Luchtvaart Maatschappij, N.V. (2008) (Cargo) Netherlands (KLM)
Korean Air Lines Co., Ltd. (2007) Air Transportation $300 Korea
(Cargo & Passenger)
British Airways PLC (2007) Air Transportation $300 UK
(Cargo & Passenger)
Samsung Electronics Company, Ltd. DRAM $300 Korea
Samsung Semiconductor, Inc. (2006)
BASF AG (1999) Vitamins $225 Germany
CHI MEI Optoelectronics Corporation LCD panels $220 Taiwan
(2010)
Hynix Semiconductor Inc. (2005) DRAM $185 Korea
Infineon Technologies AG (2004) DRAM $160 Germany

Source: http://www.justice.gov/atr/public/criminal/sherman10.html
History of anti-price-fixing law: EU
• EU Treaty Article 101 (formerly Article 81, formerly Article 85 of
Treaty of Rome). Initial precursor: Treaty of Paris (1951).
• Important precedent: Wood pulp (jurisdiction)
• National regulations, esp. UK (civil and criminal offense)
Public and private litigation
• Almost 90% of antitrust enforcement in US through private civil
suits; almost none in the European Union.
• Clayton Act: treble damages (private litigation quite attractive)
• Example: Sun Microsystems 2002 private federal antitrust lawsuit
against Microsoft: using monopoly in PC operating systems
market to undermine the success of Sun’s Java technology.
• Later Sun also filed a complaint with the European Commission
• Former EC Commissioner Mario Monti: Europe should consider
allowing US-style private lawsuits
Outline
• Collusion: US and EU law
• Leniency programs
• Facilitating practices
• Allowed horizontal agreements
• Cases
Leniency programs
• First introduced by DOJ in 1978
• Major revision in 1993

− Less legal uncertainty

− Automatic amnesty (still discretionary if report received after an
investigation under way)
− Significant increase in activity: from 1/year to 2/month reports
• Similar programs in UK, EU, Japan, etc
What determines fines?
• Study by Connor based on US plea bargains.
• Average fine discount w.r.t. maximum fine: 70%
• Fine discounts lower for 2nd, 3rd, etc. plea bargain in each case
• No relation between fine and harm produced by firm
• Fines depend greatly on cartel characteristics: higher fines for
international cartels, bid-rigging schemes
• Leniency higher during Bush II than during Clinton
Evaluating leniency programs
• Performance measures

− Fines
− Number of cases
• Fairness

− Example: Diana Brooks and Christopher Davidge

• Over-enforcement

− Private suits
− No double jeopardy (why so few in Portugal?)
Mini-case: BA and VA
• Why did BA call VA?
• Are fuel surcharges illegal? Is it illegal for both airlines
to set the same surcharge?
• How would you structure a leniency program?
• Is it fair that two firms are treated differently for the
same crime?
• Are you in favor or against leniency programs?
Outline
• Collusion: US and EU law
• Leniency programs
• Facilitating practices
• Allowed horizontal agreements
• Cases
Facilitating practices
Def: institutional features that make collusion easier
• Price transparency (published prices)
• Most favored customer clause
• Meet the competition clauses
• Examples:

− GE and Westinghouse and DuPont and Ethyl

− Ready-mixed concrete in Denmark
− Federal Election Campaign Act and Medicaid
reimbursement rules
Ready-mixed concrete in Denmark

Average 10-MPa Concrete Prices in Århus

550

500

450

400

350

300 Time
Jan-94 Apr Jul Oct Mar-95 Jun Nov
Mini-case: GE and Westinghouse
• What was GE thinking in 1963?
How did Westinghouse react and why?
• What was the DOJ case in 1974?
How would they have fared in court?
• How does the current legal doctrine treat facilitating
practices?
Outline
• Collusion: US and EU law
• Leniency programs
• Facilitating practices
• Allowed horizontal agreements
• Cases
Non-price agreements
• Research related agreements

− Common standards
− Joint R&D
− Patent pools
• Information exchange

− Industry associations
− Demand and cost data
Exemptions and exceptions
• Major League Baseball; other sports in US
• English Premier League
• European airlines until 1990s
• Export cartels
• Joint marketing programs
Outline
• Collusion: US and EU law
• Leniency programs
• Facilitating practices
• Allowed horizontal agreements
• Cases
Wood pulp
• Wood pulp prices are announced at regular intervals;
competitor prices tend to move in tandem.
• European Economic Community (EEC) initiates
investigation (1977)
• Non-EEC defendants (Finland, US and Canada)
challenge to jurisdiction
• European Courts side with Commission on jurisdiction:
effects doctrine (1988)
• Courts strike down collusion case: price parallelism can
be viewed as the result of collusion only if there is no
other explanation for it
Sotheby’s and Christie’s
• The companies: both founded in England, offices in
many countries. Close to 100% art auction market
− Sotheby’s: based in NY; $3bn turnover
− Christie’s: based in London; private since 1999
• Revenue sources: fees over “hammer” price

− Seller’s (or vendor’s) commission

− Buyer’s premium
− Vary by amount, type of auction; special deals
Sotheby’s and Christie’s
• After late 1980s boom, market collapsed in early 1990s
(Japan economic crisis → drop in demand)
• Fierce competition, low profits
• November 1992: Sotheby’s announces increase in
commissions; Christie’s does the same one month later
• March 1995: Christie’s announces increase in sellers’
fees; Sotheby’s does the same one month later
• June 1996: UK OFT initiates informal inquiries
• June 1997: DoJ subpoenas files from the two houses
Investigation, leniency, conviction
• 1998: François Pinault buys Christie’s for $1.2 billion.
Christopher Davidge replaced as CEO
• Late 1999: Christie’s lawyers prepare for government
investigation, uncover evidence of conspiracy
• January 2000: Christie’s agrees to amnesty conditional
on Davidge cooperating with DOJ. Davidge paid large
sum for doing so
• Early 2000: Ms. Brooks receives permission from
Sotheby’s to do the same but is too late
• September 2001: Ms. Brooks and Sotheby’s plead
guilty (on seller’s commissions)
Court findings
• Mr. Taubman and Sir Anthony met on 12 occasions at
Taubman’s London flat, New York residence
• Ms. Brooks and Mr. Davidge met repeatedly, reporting
to Taubman and Tennant. JFK meeting
• Agreement was established on common fees, no
customer poaching, no special deals to customers (with
exceptions)
Court sentences
• April 22, 2002: Mr. Taubman sentenced to a year and
a day in prison and fined $7.5 million
• April 29, 2002: Diana Brooks sentenced to three years
of probation, including six months of house arrest;
fined $350,000 and ordered to perform 1,000 hours of
community service
• Mr. Davidge marries former Chritie’s employee in
India. Guests paid first-class tickets
• Sir Anthony Tennant refused to testify in US, could
not be extradited from UK. Died in 2011
• Also: civil suit
Sotheby’s and Christie’s: takeaways
• Price fixing is not limited to commodities
• Top management liability conditional on awareness
• Communication isn’t always necessary to maintain high
price equilibrium
Chocolate price fixing
• Class-action suit in Canada (2008)

− Chocolate majors conspired to increase prices

− Discussions at trade shows and association events
− Total settlements C$23.2m
(from Mars’ C$3.2m to Nestlé’s C$9m)
• Mars and Nestlé still face criminal charges
(Canada Competition Bureau).
Hershey off the hook through leniency program
Chocolate price fixing
• Class-action suit in US

Nothing scandalous or improper has been

discovered within our borders, and no evidence
permits a reasonable inference of a price-fixing
agreement.
Chocolate price fixing
• German Cartel Commission (2013)

− Fines over e60m ($82m)

− Mars escaped financial penalty as one of the initial
whistleblowers
Chocolate price fixing: takeaways
• Transporting evidence across jurisdictions
• Leniency is jurisdiction specific
• Private suits
• Paying damages in class action suits
eBooks
• Wholesale model (Amazon): publisher sets price w ,
retailer sets price p and pays w to publisher.
• Agency model (Apple): publisher sets retail price p,
retailer keeps α p. Apple’s α = 30%
• Amazon’s ebook price point: $9.9
• Apple’s proposal to publishers: tiers up to $14.99
• Also: Apple forces MFN clause: if publisher sells for
less than $14.99, Apple has the right to lower price
accordingly
eBooks
• Steve Jobs’ idea: help publishers push Amazon to
agency model. In fact, require publishers to move
e-tailers to agency.
− Apple’s benefit avoid retail price competition: iPad will
beat Kindle on device quality
− Publisher’s benefit: fix Amazon’s low pricing problem
• But α = 30% is very high!
• Hachette, Penguin, Random House, HarperCollins,
Macmillan and S&S accept; Random House declines
eBooks
• DoJ’s case (April 2012): Apple’s proposal is a scheme
to increase prices in a coordinated manner
• Hachette, HarperCollins, Simon & Schuster, Macmillan
and Penguin deny wrongdoing but agree to settle
(December 2013)
• Pay $166 million to customers (call 1-866-621-4153)
• Federal judge finds (July 2013) Apple violated antitrust
law in helping raise the retail price of e-books
• Apple loses appeal, may ask Supreme Court
• Note: decision does not imply that agency model or
MFC clauses are illegal
eBooks: takeaways
• Collusion is not limited to smoke-filled room price
fixing: facilitating practices can be deemed
anti-competitive (cf GE and Westinghouse)
• Decision does not imply that agency model or MFC
clauses are illegal per se (difference w.r.t. rule of
reason)
eBay hiring
• DoJ files a series of civil antitrust suits against Adobe,
Apple, eBay, Google, Intel, Intuit, Lucasfilm, Pixar.
• Claim: “no cold call” agreements. Specifically, senior
executives of eBay and Intuit agreed not to recruit
employees from each other
• District court finds agreement, if proven, constitutes
naked horizontal market allocation agreement:
manifestly anticompetitive, per se unlawful
• Settlements reached with various companies, most
recently eBay (May 2014)
eBay hiring: takeaways
• Illegal horizontal practices are not restricted to sellers’
actions (monopsony power)
• Illegal horizontal practices are not restricted to setting
prices (e.g., limit competition clauses)
Synthetic rubber
• EC fines ENI and Versalis e272.25 million for
participating in the synthetic rubber cartel (Nov 2006)
• Fine includes a 50% increase due to recidivism: firms
had already participated in Polypropylene (1986) and
PVC II (1994) cartels
• In January 2008, Bayer and Zeon are also fined for
cartelization.
− Bayer’s and Zeon’s fines are reduced by 30% and 20%,
respectively, under the EC leniency program
− However Bayer’s fine is increased by 50% because it had
been fined for cartel activity in a previous Commission
decision.
Synthetic rubber: takeaway
• Wishful thinking

This is the fourth cartel decision in the synthetic

rubber industry in just over 3 years. I hope that
this is the last. Buyers of synthetic rubber
should be concerned about how much these
cartels have cost them. And shareholders should
be concerned about how much the fines have
cost them. — Competition Commissioner Neelie
Kroes
• Cartels on commodities abound, will probably never
end (cf Adam Smith)
The Libor scandal
• To set Libor and Euribor rates, banks submit rates at
which they would be prepared to lend money to one
another. Then British Bankers’ Association (BBA)
drops the four highest and four lowest, averages the
remaining into one rate — LIBOR
• Evidence that traders benefited from falsely reported
rates. Net lenders ⇒ tendency to push rates up
• In December 2013, EU fines group of global financial
institutions a combined e1.7 billion ($2.3 billion) to
settle charges (largest combined fine in EC history)
The Libor scandal
Overall takeaways
• Big temptation to agree with rivals on price increases
• Crime does not pay: fines and jail sentences are common
• Leniency programs have made it worse for firms
• Explicit and implicit collusion; per se illegality and rule of reason
MARKET STRUCTURE
Overview
• Context: You’re analyzing a given industry. How would you
expect market structure to depend on type of industry, type of
product, country size, etc
• Concepts: exogenous and endogenous entry costs, scale
economics and minimum efficient scale, business stealing
• Economic principle: market structure depends on many factors,
but there are some regularities
 Industry concentration

C4 (%) in Germany C4 (%) in Belgium

100 100

80 80

60 60

40 40

20 20

C4 (%) C4 (%)
0 in France 0 in France
0 20 40 60 80 100 0 20 40 60 80 100

C4 measures the combined market shares of the 4 largest firms. More on this later.
Market size and market structure
• Cournot competition. Cost: C = F + c qi
Demand Q = (a − P) S, where S is measure of market size
 2
• Equilibrium profit level: Π(n) = S a−c −F
n+1

• Free entry equilibrium:

n) ≥ 0 ≥ Π(b
Π(b n + 1)
r
S
Π(b
n) = 0 ⇒ n = (a − c) −1
F
" r #
S
n = (a − c)
b −1
F

where [x] denotes highest integer lower than x

Market size and market structure
• Equilibrium value of n

− increasing in a, S
− decreasing in F , c
• Relation between S and n is increasing but less than proportional
• Intuition: higher S leads to higher n, but higher n leads to lower
p, which in turn limits the increase in n
 Market size and market structure

n (number of entrants)

20

15

10

5
S (market
size)
0
0 5 10

Parameter values: a = 100, c = 30, F = 100

Evolution of market structure
• Number of firms and firm size changes over time:

− firm growth
− firm entry and exit
− mergers and divestitures
• Evolution determined by changes in

− demand
− technology
− regulation
 Production capacity of US brewers

Capacity (103 barrels) 1959 1967 1975 1983 1989 1998 2001 2006
10 - 100 68 36 10 15 8 77 81 83
101 - 500 91 44 19 12 7 19 19 19
501 - 1000 30 35 13 2 3 1 1 4
1001 - 2000 18 18 13 13 5 4 2 2
2001 - 4000 8 10 12 9 6 7 5 3
4001+ 2 4 15 23 20 20 20 22

Source: Elzinga, in Adams and Brock

History matters
• Same industry, similar countries, different market
structures. Why?
• Branding (e.g., Heinz and Campbell’s in UK and US)
• Learning curves and other self-reinforcing processes
• Multiple technologies and multiple free-entry equilibria
Evolution of new industries

# firms in hard drive industry

30
active firms

20

10
entries exits

0 Year
1975 1980 1985 1990 1995 2000 2005 2010 2015
Exogenous and endogenous entry costs
• Compare Portugal (S = 1) and the US (S = 50)
• Basic model predicts
r r
SUS 50 SP √
nUS ≈ (a − c)
b − 1 ≈ (a − c) − 1 ≈ 50 b
nP
F F

• This is about right in the cement industry

• Beer industry: b
nP = 2, whereas b
nUS = 3. Why does model fail?
Endogenous entry costs
• Much of the entry cost into (national brand) beer is advertising
• Advertising expenditures are roughly proportional to sales (when
comparing various countries)
• Hence, as S increases, so does F
• In the limit, if F is proportional to S, then b
n remains flat w.r.t. S.
" r # " r # " r #
S S 1
n = (a − c)
b − 1 = (a − c) − 1 = (a − c) −1
F kS k

• Prediction: C4 decreasing (b
n increasing) in S. However, relation is
flatter in advertising-intensive industries
Advertising intensity

Low-adv industries A/RS (%) High-adv industries A/RS (%)

Salt 0.26–0.45 Frozen food 1.2–7.1
Sugar 0.06–0.24 Soup 2.7–6.0
Flour 0.17–0.96 Margarine 2.3–10.2
Bread 0.02–0.42 Soft drinks 1.2–5.4
Processed meat 0.30–0.70 RTE cereals 8.34–12.9
Canned vegetables 0.29–0.71 Mineral water 1.5–5.0
Sugar confectionery 1.4–6.0
Chocolate confectionery 2.9–6.5
R&G coffee 1.9–16.7
Instant coffee 2.2–11.1
Biscuits 1.9–8.0
Pet foods 4.0–8.4
Baby foods 0.9–4.2
Beer 1.0–5.43
 Low- and high-advertising industries

C4 (%) C4 (%)
100 100

50 50

0
ln(size/MES) 0
ln(size/MES)
2 4 6 8 10 2 4 6 8 10

MES denotes minimum efficient scale. Used to compare market size across industries.
MARKET STRUCTURE
AND MARKET POWER
Measuring market power
• One firm: margin
p − MC
m=
p
• Many firms: average margin, a.k.a. Lerner index

n
X p − MC i
L≡ si
p
i=1
Measuring market structure
• n (number of firms), or 1/n
• Ci index 4
X
C4 ≡ si
i=1

• Herfindahl-Hirsh index (a.k.a. HHI)

n
X
H≡ si2
i=1

• C4 ∈ [0, 100] (or [0,1]), H ∈ [0, 10, 000] (or [0,1])

What do the extremes mean?
• Pros and cons of each measure
Intensity of competition and market structure
• Exhibit 1: pharmaceutical industry in Sweden

− Each month, government auctions right to be main or

sole supplier
− Bertrand competition
− Very high concentration: leading supplier takes from 50
to 70% market
• Exhibit 2: banking industry in Kenya

− Interest rates around 12%, no variation with respect to

# competitors
− Very little price price competition
− High entry rates; currently more than 40 banks
• Summary: higher degree of competition, more
concentrated market
Intensity of competition and market structure
• Suppose different countries only differ on market equilibrium price
(e.g., regulation)
• Cost function C = F + c q, entry cost E
• Free entry equilibrium

1
(p − c) D(p) − F = E
n

• Solving for equilibrium b

n
 
1
n=
b (p − c) D(p)
E +F

• Hence, lower p (more intense competition) implies lower n

(more concentrated market)
Market concentration and market power
• In previous exercise, vary intensity of competition and find free
entry equilibrium
• Alternatively, fix degree of competition (specifically, Cournot) and
vary market structure
• It can be shown that
H
L=
−
where  is price elasticity of market demand
Example
• Consider two markets with identical demands

− Market 1: two firms with identical market shares

− Market 2: one firm with a 70% market share, two small firms with
15% each
• Assuming Cournot competition in both markets, where is market
power the greatest?
• Answer: whichever market has higher HHI

− Market 1: H = 502 + 502 = 5, 000

− Market 2: H = 702 + 152 + 152 = 5, 350
Structure-Conduct-Performance paradigm
• Structure determines conduct

− e.g.: collusion easier with fewer firms

• Conduct determines performance

− e.g.: Bertrand implies lower margins than Cournot

• Structure determines performance

− e.g.: fixing Cournot, L is proportional to H

Structure-Conduct-Performance hypothesis
• Positive correlation between structure (e.g., H) and performance
(e.g., L), where L is proxied by average profit rate

Ri − VC i p qi − ci qi p − ci
ri ≡ = = = mi
Ri p qi p

• Weak relation in empirical studies

− poor data
− cross industry comparisons: apples and oranges
− reverse causality implies opposite correlation
Collusion v efficiency hypotheses
• Suppose that feedback effects are unimportant: structure
determines performance
• Suppose higher concentration is correlated with higher profits
(some evidence). What is source of correlation?
• Collusion hypothesis: variation in degree of collusion
• Efficiency hypothesis: variation in firm efficiency
• Either hypothesis is consistent with aggregate correlation,
but narratives and policy implications are very different
ENTRY REGULATION
Free entry and welfare
• Suppose number of (identical) firms increases from n to n + 1
• Price drops from pn to pn+1
Total output increases from n qn to (n + 1) qn+1
• Impact of entry in social welfare

− Consumer surplus increases by approx 21 (pn − pn+1 )(Qn+1 − Qn )

− Entrant gains πn+1 − E
− Incumbent firms lose n (πn+1 − πn )
• Presumably, if entry takes place, then πn+1 − E > 0
What about social welfare?
 Free entry and welfare

pn ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ... MC
...
... C...... ...... ...... ...... ......
pn+1 ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ...... ....
....... ...... ...... ...... ......
... ... ...
... ... ...
... ... ...
...
...
A ...
...
B ...
...
... ... ...
... ... ...
... ... ...
...
...
...
.
..
...
... D
... .
.. ...
... .
.. ...
... .
.. ...
... .
.. ...
... .
.. ...
... . ...
...
..
.
.. ... q

n qn+1 n qn (n + 1) qn+1

Marginal entrant’s variable profit: A + B. Increase in total surplus: B + C .

Free entry and welfare
• Business stealing effect: part of the entrant’s variable profit (area
A) is a transfer from incumbents
• Negative entry externality: A is a gain for entrant but not a gain
for society as a whole
• Consumer surplus effect: entry implies an increase in consumer
surplus (area C ) that is not captured by firms
• Positive entry externality: C is a gain for society as a whole gain
but not for entrant
• If A > C , free entry ⇒ excess entry
If A < C , free entry ⇒ insufficient entry
Free entry and welfare
• Product differentiation unimportant, competition soft:
business-stealing effect dominates, free-entry ⇒ excessive entry
• Product differentiation very important, competition very fierce:
consumer surplus effect dominates, free entry ⇒ insufficient entry
• Examples

− banking
− radio stations
− real estate brokerage
Example: US radio stations
• Radio stations make money from advertising
• Important entry decision: location in product space
• Entry externalities:

− entertainment offered to listeners beyond cost of

listening to ads
− stealing listeners from similar stations
• Free entry result in too many rock stations, two few
classical music stations
• Relative so social optimum, loss to firms and
advertisers is 45% of revenue
Firm turnover and allocative efficiency
• Above analysis assumes identical firms and focuses on allocative
efficiency (how much to produce)
• If different firms have different costs, selecting the lowest cost
firms improves productive efficiency (lower total costs)
• Additional benefit from entry and exit: selection and asset
reallocation
− Lower cost firms enter, higher cost firms exit (extensive margin)
− Among incumbents, lower cost firms have higher output (intensive
margin)
Example: telecom equipment industry
• Deregulation of US telecommunications equipment
industry implied significant turnover
− Entry and exit
− Capital reallocation from low productivity to
high-productivity firms
• One way to measure overall effect is to compute
correlation between productivity and output
• Additional benefit from entry and exit: selection and
asset reallocation
− High correlation: capital concentrated on higher
productivity firms
− Low (maybe negative) correlation: capital not
concentrated on higher productivity firms
 Deregulation, turnover and productivity

Correlation (output,productivity)
0.15

0.10

0.05

0.00

−0.05

−0.10

−0.15 Year
75 80 85

Correlation between output and productivity increases as deregulation progresses

 Regulation to encourage entry
• Orphan drugs: developed specifically to treat a rare
medical condition (orphan disease)
• Examples: glioma, multiple myeloma, cystic fibrosis,
phenylketonuria, snake venom poisoning, and
idiopathic thrombocytopenic purpura
• Easier to gain marketing approval; financial incentives
(e.g., extended exclusivity periods)
• Currently more than 400 orphan designated drugs in
clinical trial process in US and EU
• Generally acknowledged as a successful policy measure.
In US, number of orphan drugs increased from 38 in
1983 (Orphan Drug Act) to 1,129 in 2004

Related problem: poor countries’ diseases (malaria, ebola?)

Regulation to discourage entry
• Suppose that p = p̄, no product differentiation, free
entry. Then consumer surplus effect is zero, whereas
business stealing effect is positive: excess entry
• Example: banks and branches. Regulated interest rates
and/or tacit collusion imply p = p̄ > c
• Portugal in 1980s: entry deregulation precedes interest
rate deregulation
• Huge concentration bank branches is urban areas
• Partial solution: entry fees for new branches
Entry regulation as capture
• Real estate brokerage

− In 2001, US Treasury and Fed propose that banks be

allowed as real estate brokers
− National Realtors Association lobbies against
− In 2009, Obama signs Act that permanently bars banks
from real estate brokerage
• Pharmacies

− In 2009, EU court ruled that pharmacy ownership

restrictions are OK
− German legislation frustrate multinational chains
− Many other countries (e.g., Australia) have similar
restrictions
NYC taxi cabs

$ 000
700 14

600 12
Number of medallions issued
500 10

400 8

300 6

200
Medallion price
100

0 Year
1950 1969 1970 1980 1990 2000 2010
 New York taxi medallions
• What determines the price of NYC taxi medallions, in
particular, what factors lie behind the recent increase?
• What do you think will happen to medallion prices in
the near future?
• What determines the number of medallions issued by
the city?
• What policy changes (if any) would you recommend
Mayor Bloomberg?

Alternative investments; E/P ratio

Entry regulation as capture (cont)
• Uber

− Mobile-phone app to link passengers and car drivers

− Brussels, Seattle and Miami have banned / limited
− New York’s Attorney General objects to practice of
pricing more when demand is heavy
• Airbnb

− Online service that lets users easily rent homes or

apartments for short-term stays
− Hotel industry is lobbying hard to kill the upstart
Regulation in the Internet era

What has made the Internet revolutionary is that it’s

permissionless. No one had to get approval from
Washington or city hall to offer Google searches,
Facebook profiles or Apple apps.
Adam Thierer, author of Permissionless Innovation

The only question is how long it will take for these cyber
cowboys to realize that working with the sheriffs is both
good business and the right thing to do.
New York Attorney General Eric Schneiderman
MERGERS
Overview
• Context: market structure changes by means of entry, exit — and
mergers and acquisitions
• Concepts: efficiency and market power; unilateral and coordinated
effects; merger waves
• Economic principle: as always, there is the good, the bad and the
ugly
Why mergers?
• Synergies

− Sony and Columbia Pictures

• Entry into new market

− Nestlé and Rowntree

• Supplier power

− Philip Morris and Kraft

• Distribution efficiencies

− Nestlé and General Mills

Types of mergers
• Horizontal: between competitors
• Vertical and complementary: firms in different or
complementary stages of the value chain
• Conglomerate: industry-unrelated firms
Largest M&A transactions as of 2014

Rank Year Acquiror Target $b eb

1 1999 Vodafone AirTouch PLC Mannesmann AG 202.8 204.8
2 2000 America Online Inc Time Warner 164.7 160.7
3 2007 Shareholders Philip Morris Intl Inc 107.6 68.1
4 2007 RFS Holdings BV ABN-AMRO Holding NV 98.2 71.3
5 1999 Pfizer Inc Warner-Lambert Co 89.2 84.9
6 1998 Exxon Corp Mobil Corp 78.9 68.4
7 2000 Glaxo Wellcome PLC SmithKline Beecham PLC 76.0 74.9
8 2004 Royal Dutch Petroleum Co Shell Transport & Trading Co 74.6 58.5
9 2000 AT&T Inc BellSouth Corp 72.7 60.2
10 1998 Travelers Group Inc Citicorp 72.6 67.2
Effects of horizontal mergers
• Efficiencies

− Common cost savings

− Synergies, complementarities
• Market power

− Unilateral effects (closer to monopoly)

− Coordination effects (closer to collusion)
Merger effects: Hyundai and KIA

# ratio
30 3

# models
20

10
# platforms

# models/platform
0 Year
1
1996 1998 2000 2002 2004 2006 2008 2010
Merger effects: Hyundai and KIA

Prices (1000 KW) Exports (million units)

20 2.0

domestic price 1.6

15

1.2
10
exports
0.8

5
0.4

0 Year
1996 1998 2000 2002 20004 2006 2008 2010
 Merger in Cournot triopoly
• Initially n = 3, all firms with C = F + c q
• Firms 2 and 3 merge, forming 2&3 with C = F 0 + c 0 q
• Merger efficiencies:∗ F < F 0 < 2 F and c 0 < c
• Determine effect on

− merging parties
− non-merging parties
− consumers

∗ Merger efficiencies could also take place at the product level.

Merger effect on merging parties
• Equilibrium profits under Cournot
P 2
a − n ci + cj

j6=i
π
bi = − Fi
n+1

• Firm profit before merger

 2
a−c
π1 = π2 = π3 = −F
4

• Firm 2&3 profit after merger

2
a + c − 2 c0

0
π2&3 = − F0
3
Merger effect on merging parties
• Taking differences
2 2
a + c − 2 c0
 
0 a−c
π2&3 − (π2 + π3 ) = (2 F − F 0 ) + −2
3 4

• Effects

− Fixed cost savings: F 0 < 2 F (positive effect)

− Marginal cost savings: a + c − 2c 0 > a − c (positive effect)
− Market power: n0 < n (positive effect)
− Exit: from 2 to 1 π (negative effect)
Merger effect on outsiders
• Taking differences
2 2
a + c0 − 2 c
 
a−c
π10 − π1 = −
3 4

• Effects

− Rival’s marginal cost savings: a + c 0 − 2 c < a − c (negative effect)

− Market power: n0 < n (positive effect)
Merger effect on outsiders: examples
• BP acquires Amoco (oil industry). Mobil’s stock up $2.625
• Western Digital announced Hitachi acquisition (hard drive
industry). Seagate’s stock up by 9%
• BA and AA announce plans to merge (air travel industry). Virgin
Atlantic paints its aircraft with “BA/AA No Way”
• GKN plc and Alvis plc merge (European defense industry), putting
pressure on Vickers, the third competitor
Merger effect on consumers
• Consumer surplus under Cournot

Pn 2 !2 n
!2
n

1 n a − i=1 ci 1 1 X
CS = 2 = 2 a− ci
n+1 n+1 n
i=1

• Merger effect
 2  2  2
2 1 1 3 2
CS 0 − CS = 1
2 a − (c + c 0 ) − (a − c)
3 2 2 4

• Effects

− Cost savings passed through: a − 12 c 0 + c < a − c (positive effect)

− Market power: n0 /(n0 + 1) < n/(n + 1) (negative effect)

Merger effects: Miller and Coors

Change in price (%)

0.02
change in p
due to change in H
0.01

0.00
change in p
due to change in d
-0.01

-0.02 Time
2007 2008 2009 2010 2011 2012
Merger dynamics
• Preemptive mergers

− Google and Waze

• Merger waves

− US supermarket chains
− Advertising agencies
• Mergers and entry

− US radio stations 2
Merger waves: US radio stations

Number of mergers and acquisitions

800
700
600
500
400
300
200
100
0 Year
1990 1995 2000 2005
 Entry in US radio station markets

Post-merger
0.03

0.02

0.01

0.00

-0.01

-0.02

Pre-merger
-0.03
−0.03 −0.02 −0.01 0.00 0.01 0.02 0.03

Monthly net entry rates

MERGER POLICY
Overview
• Context: you are helping a firm that wants to acquire a competitor
• Concepts: merger review process; efficiency and failing firm
defense; unilateral and coordinated effects
• Economic principle: it’s not just the economics, stupid!
Outline
• Merger policy: general principles
• US and EU merger review processes
• Cases
Outline
• Merger policy: general principles
• US and EU merger review processes
• Cases
Merger policy
• If there are economies of scale, merger leads to lower
cost; various synergies may create additional value
• But: greater concentration leads to greater market
power
− Unilateral effects
− Collusion effects
• Merger policy is an attempt at measuring the pros and
cons of each merger
− Benefits and costs for firms
− Benefits and costs for consumers
Economics analysis
• Relevant market definition
• Lower fixed cost: efficiency gained by firms
• Lower marginal cost: typically shared between firms, consumers
• Fewer competitors (lessening of competition):

− Unilateral effects
− Collusion
• Equilibrium readjustment

− Number of firms (entry or exit)

− Number of locations / varieties
− Prices
Outline
• Merger policy: general principles
• US and EU merger review processes
• Cases
Process (US)
• Hart-Scott-Rodino Antitrust Improvement Act of 1976
• Notification system implemented in 1978
• If merger size is “big” ($200m+) then merging parties must notify
FTC and DOJ. Wait for 30 days
• Either party may request additional information (“second
request”). Clayton Act, Section 7A(e)
• If so, merging parties must provide information; agency then has
another 30 days to respond
• Agency may negotiate remedies; may seek injunction in Federal
District Court to prohibit merger; agency may also challenge
merger ex-post, but that is rare
Historical data
• More than 95% of the filings are cleared during initial period
• More than 75% of the “second request” cases are “enforced”
 FTC approved mergers 1996–2007 (%)

As a function of post merger HHI

..
...
..
0 - 1,799
... 59.4
...
...
...
1,800 - 1,999 ... 38.2
...
...
...
2,000 - 2,399 ... 38.9
...
...
...
2,400 - 2,999 ... 26.1
...
...
...
3,000 - 3,999 ... 30.1
...
...
...
...
4,000 - 4,999 ... 16.0
...
...
...
5,000 - 6,999 ... 13.0
...
...
...
average
7,000 + ... 1.3
...
..
0 10 20 30 40 50 60

Source: http://www.ftc.gov/os/2008/12/081201hsrmergerdata.pdf
 FTC approved mergers 1996–2007 (%)

As a function of change in HHI

....
...
0- 99 .. 92.3
...
...
...
100 - 199 ... 56.4
...
...
...
...
200 - 299 ... 47.2
...
...
...
300 - 499 ... 33.3
...
...
...
500 - 799 ... 27.2
...
...
...
800 - 1,199 ... 22.5
...
...
...
...
1,200 - 2,499 ... 14.5
... average
...
...
2,500 + ... 2.5
...
.
0 10 20 30 40 50 60 70 80 90 100

Source: http://www.ftc.gov/os/2008/12/081201hsrmergerdata.pdf
New Horizontal Merger Guidelines (HMG)
• Merger guidelines are not law, but are highly cited
• Old guidelines (1983, 1992) sought precise, step-by-step
framework around “relevant market” definition (SSNIP test) and
market concentration measurement (HHI index)
• New guidelines (August 2010) envision more flexible approach

− Greater emphasis on economic effects, lesser emphasis on formulas

− Direct evidence of price effect
− Merger simulation models (which may not rely on market definition)
− Statistical analysis of “natural experiments”
New HMG (cont)
• Thresholds and safe harbors:

− Higher HHI thresholds (more on this below)

− 35% market share “safe harbor” deleted
• Greater transparency in agencies’ analytical process

− Much of what HMG 2010 include was already in use

• Greater scrutiny in differentiated product and R&D intensive
industries (new section on innovation)
New HMG (cont)
• But courts will likely continue placing weight on market definition
and market shares
− Will agencies use old HMG in court?
− Will agencies’ court performance suffer?
− Will legal uncertainty increase?
HHI thresholds in HMG
• HHI increases by less than 100: no problem
• Post-merger HHI < 1,500: no problem (old threshold 1,000)
• Post-merger HHI between 1,500 and 2,500: if increase in HHI is
over 100, significant concerns
• Post-merger HHI over 2,500 (old threshold 1,800): if HHI
increases by 100–200, concerns; if HHI increases by more than
200, presumed anticompetitive
EU regulations
• Roughly similar to US regulations
• In addition, separate non-horizontal merger guidelines

− More likely efficiency effects (e.g., double marginalization)

− Input foreclosure
− Customer foreclosure
− Coordinated effects
Outline
• Merger policy: general principles
• US and EU merger review processes
• Cases
Mini-case: The ill-fated GE-Honeywell merger
• What was the strategic logic for GE to purchase
Honeywell?
• Do you find it compelling?
• What was the European Commission’s logic in blocking
the merger?
• Do you think it made the right decision?
• What should Jack Welch have done?
• Should he have pressed ahead for a deal?
• What are the main learning points from this case?
Staples and Office Depot: the merger
• Staples wants to merger with Office Depot (1996)
• FTC asks for several store divestitures as remedy
• Staples disagrees, case taken to court (1997)
Staples and Office Depot: the case
• Discussion on market definition

− Stores selling office supplies

− Office supplies super stores
• Discussion on cost efficiencies

− FTC claims many efficiencies would have taken place

through internal growth
− FTC claims pass-through rate small
• Discussion effects on prices

− Prices 11.6% lower in markets with Staples and Office

Depot than Staples only
• Court sides with FTC
Staples and Office Depot: takeaways
• FTC signals willingness to fight till the end
• Court signals importance of economic analysis
• Unilateral effects important; not just coordination
effects (as suggested by 1992 merger guidelines)
Oracle and PeopleSoft (2004)
• SAP, Oracle and PeopleSoft largest firms in ERP
• Oracle makes bid for PeopleSoft. DOJ challenges in
Court
• DOJ uses simulation model
• Court rules in favor of defendants (DOJ did not
provide sufficient evidence)
HP-Compaq merger (2002)
• Given HMG and any other information you think is relevant, do
you agree with the FTC’s decision not to oppose merger?

Company Desktop PCs Servers

Dell 13 7
Compaq 12 16
HP 8 14
Gateway 4 –0
IBM 6 26
Source: Bank of America report, October 2001
Data for 2001Q2
The Nestlé Perrier acquisition
• In February 1992, Nestlé makes a bid for Perrier S.A.
• Market shares prior to merger: Perrier, 35.9%; BSN,
23%; Nestlé, 17.1%; others, 24%
• The Nestlé/Perrier merger ⇒ leading producer with
53% second with 23%
• Nestlé agrees to sell Volvic to rival BSN
• Expected post-merger market shares: Nestlé/Perrier,
38%; BSN (with Volvic), 38%; others, 24%
• Some analysts argued this made things worse: collusion
more likely among similar competitors
AT&T and T-Mobile
• March 2011: AT&T announces $39 billion acquisition
of T-Mobile USA
• August 2011: DOJ sues to block merger
• Within weeks, Sprint Nextel and C Spire Wireless file
private antitrust suits to block merger
• November 2011: FCC announces it is considering order
to oppose merger
• December 2011: parties abandon transaction; AT&T
pays T-Mobile $4 billion breakup fee (largest in US
antitrust history)
AT&T and T-Mobile: what were they thinking?
• AT&T argues it does not compete with T-Mobile while
the latter runs adds stating “sometimes you have to
pay more to be slower”
• AT&T claims no growth in isolation, although
T-Mobile had announced plans to expand unilaterally
• Combining operations would have saved $3 billion per
year in duplicative resource costs
AT&T and T-Mobile: takeaways
• Cost efficiencies count for DOJ, but greater weight is
placed on competition effects
• Sectoral regulators such as the FCC have broader
mandate, frequently tighter standards
• Business strategy

− Mergers and acquisitions are expensive operations

− Failed mergers and acquisitions have additional
reputational costs
− Business strategy should be complemented by in-house
economic and legal analysis
Live Nation and Ticketmaster
• Ticketmaster and Live Nation announce merger plans
in February 2009
• Live Nation had recently entered into ticketing, so
merger is both vertical and horizontal
• Strong opposition, including Bruce Springsteen

The one thing that would make the current

ticket situation even worse for the fan than it is
now would be Ticketmaster and live Nation
coming up with a single system, thereby
returning us to a near monopoly situation in
music ticketing
Live Nation and Ticketmaster
• UK Competition Commission

− October 2009: provisionally block

− December 2009: approve
− January 2010: rival ticket agency Eventim challenges
− May 2010: re-approve merger
• In January 2010, DOJ accepts merger conditionally on
consent decree
− Divestiture of Ticketmaster’s Paciolan division
− Prohibition of certain discriminatory practices (e.g.,
must use Ticketmaster if hosting Live Nation artists)
Live Nation and Ticketmaster
• Post-mortem

− Divestitures had little effect

− Prices have remained high
Live Nation and Ticketmaster: takeaways
• Increasingly, the distinction between vertical and
horizontal mergers becomes blurred
• Broadly speaking, two types of concessions:

− divestitures
− behavioral
• Regulators incur Type I and Type II errors;
this one looked like a Type II error
XM and Sirius satellite radio
• FCC awards two satellite radio licenses
• Sirius and XM begin broadcasting in 2001
• Subscription based ($10/m), commercial free service
• Mostly used by drivers
• Music, talk, sports, etc; some premium content
• By the end of 2006, 17 million subscribers
• No profit. February 2007, Sirius and XM propose to
merge
XM and Sirius satellite radio
• Merger is approved by DoJ in March 2008; and by FCC
in July 2008 (longest investigation in history)
− alternative audio services
− no competition in key input markets
(nearly inelastic supply)
− beneficial merger specific efficiencies
(penetration pricing externality)
• Subject to FCC-imposed behavioral conditions

− price freeze (for three years)

− a la carte pricing
− noncommercial and diversity programming
XM and Sirius satellite radio
• Post-mortem as of 2012 (vd 2009)

− nearly bankrupt in 2009, currently profitable

− monthly fee $14.99 (up by 11.5%)
− 21.9 million subscribers (up 26.5%)
XM and Sirius satellite radio: takeaways
• Failing firm can be a valid argument for merger
• Countervailing power can be a valid argument too
• Efficiencies can be more than saving fixed costs
Comcast and NBC
• In December 2009, Comcast announces purchase of
controlling stake in NBC
• Primarily a vertical merger
• Concerns: upstream foreclosure and downstream
foreclosure
− U: other content suppliers
− D1: competing multi-channel video programming
distributers (MVPDs), e.g., TimeWarner and Verizon
− D2: online video programming distributors (OVDs),
e.g., Hulu and Netflix
• Benefits: greater coordination in content creation and
distribution (in a rapidly changing industry)
Comcast and NBC
• Potential threats greater than benefits
• Approval subject to conduct-oriented conditions

− Allow MVPDs and OVDs the option of binding

arbitration over rates
− Commit as ISP not to discriminate against OVDs
• Almost no complaints since then (exception:
Bloomberg successfully requests that its channel be
placed in a news neighborhood)
• DoJ investigation on ISP discrimination
Comcast and NBC
• In Fall 2009, Comcast planned to launch an Internet
service for the poor that was sure to impress federal
regulators
• David Cohen, Comcast’s chief of lobbying, told the
staff to wait: Comcast was planning a controversial
$30 billion bid to take over NBC Universal, and Cohen
needed a bargaining chip for government negotiations
• The strategy was quintessential Cohen. The
hard-charging 56-year-old veteran of Philadelphia
politics and Democratic campaign bundler: the “wonk
rock star”
Comcast and NBC
• Al Franken opposes merger, Alec Baldwin makes fun of
it
• 97 Congressmen signed a letter urging FCC to approve
the merger without conditions; 84 of these received
donations from Comcast (460 out of 535 Members of
Congress received donations from Comcast)
• FCC approves merger on January 18, 2011
• Four months later, Meredith Attwell Baker, one of the
commissioners who voted for the deal, joined
Comcast’s Washington lobbying office
Comcast and NBC: takeaways
• Creative use of concessions

− binding arbitration over rates (increasingly common)

− use merger as quid pro quo (e.g., no discrimination as
ISP, diversity in hiring)
• It’s the politics, stupid!
Comcast and TimeWarner
• February 2014: Comcast and TimeWarner announce
plans to merge
• Non-overlapping cable providers
• Together, 40% as Internet Service Providers (ISP)

− Netflix and others oppose merger

• Greater monopsony power w.r.t. content providers

− lower prices (partly passed on to consumers)

− lower investment incentives (upstream)
Comcast and TimeWarner: things to look for
• Behavioral remedies to ease OVDs’ concerns
(e.g., Netflix); will arbitration do?
• Monopsony power: lower prices but also lower
incentives for content creation (cf Sirius XM)
• Expect a lot of politics
• Update: meeting scheduled for April 22, 2015
(first meeting since merger was announced)
MARKET FORECLOSURE:
ENTRY DETERRENCE
AND PREDATION
Overview
• Context: How can you discourage rivals from entering a market
you are in or plan to enter? If you’re the entrant, how can you get
past the incumbent’s defenses?
• Concepts: commitment, cheap talk, entry deterrence, preemption
• Economic principle: sometimes, lack of flexibility is more valuable
than flexibility
Commitment
• For many decisions, it’s useful to have lots of options
• In games, it’s sometimes useful to have fewer options:
to eliminate moves that lead to unattractive equilibria
• We refer to this limitation of your options as
commitment (as in, you’re committed to a particular
course of action)
 Dupont and titanium dioxide
• TiO2 : white chemical pigment employed in the
manufacture of paint, etc.
• Primary raw material: ilmenite ore (DuPont) or rutile
ore (six smaller rivals)
• Sharp increase in rutile ore price (1970): DuPont cost
advantage up to 44% from 22%; also, ilmenite better
compliant with stricter environmental regulation
• Window of opportunity: expand capacity faster than
demand, discourage expansion (or entry) by rival firms
• Goal: increase DuPont’s market share from 30% in
1972 to 56% in 1980 and perhaps 65% in 1985
• Market share did increase; by 1985, five domestic
competitors had exited

Ghemawat, Pankaj (1984), “Capacity Expansion in the Titanium Dioxide Industry,” Journal of Industrial Economics 33, 145–163; Hall, Elizabeth A
(1990), “An Analysis of Preemptive Behavior in the Titanium Dioxide Industry,” International Journal of Industrial Organization 8, 469–484.
 Incumbent-entrant capacity game
• Incumbent chooses capacity
• Entrant observes incumbent’s choice and chooses capacity
(0 = no entry). If entry, pay cost F

Entrant
0 24 26 28
0 768 780 784
40
1920 960 880 800
0 672 676 672
Incumbent 44
1936 880 792 704
0 576 572 560
48
1920 768 672 576

Demand curve: Q = 100 p. Capacity costs: c1 = 12, c2 = 4.

Case 0: no entrant

40 1920

Incumbent 44 1936

48 1920

• If there is no entrant, incumbent’s best choice is 44

• Denote this as monopoly capacity
Case 1: F = 500
Entrant
0 24 26 28
0 268 280 284
40
1920 960 880 800
0 172 176 172
Incumbent 44
1936 880 792 704
0 76 72 60
48
1920 768 672 576

• Given Entrant’s strategy, Incumbent is better o↵ by choosing 40

• We call this a strategy of entry accommodation
Case 2: F = 600
Entrant
0 24 26 28
0 168 180 184
40
1920 960 880 800
0 72 76 72
Incumbent 44
1936 880 792 704
0 -24 -28 -40
48
1920 768 672 576

• Given Entrant’s strategy, Incumbent is better o↵ by choosing 48

• We call this a strategy of entry deterrence
Case 3: F = 700
Entrant
0 24 26 28
0 68 80 84
40
1920 960 880 800
0 -28 -24 -28
Incumbent 44
1936 880 792 704
0 -124 -128 -140
48
1920 768 672 576

• Given Entrant’s strategy, Incumbent is better o↵ by choosing 44

• We refer to this situation as blockaded entry
Incumbent-entrant capacity game
• F = 500. Entrant is better o↵ by entering. Incumbent’s optimal
capacity: 40 (entry accommodation)
• F = 600. Entrant stays out if and only if incumbent chooses
capacity 48, which is optimal choice (entry deterrence)
• F = 700. Incumbent chooses optimal monopoly capacity (44)
and entrant stays out (blockaded entry)
• Notice that incumbents optimal capacity is non-monotonic with
respect to F : from 40 to 48, then down to 44
The miracles of science

TiO2 Capacity (000 tons)

600

rival firms

400 DuPont

200

0 Year
72 73 74 75 76 77
 Games hospitals play
• Previous theory suggests relation between likelihood of
entry and investment is non-monotonic:
Low entry probability: blockaded entry
Medium entry probability: entry deterrence
High entry probability: accommodated entry
• Possible test: period after Medicare announces likely
reimbursement increase but before it takes e↵ect
• Incumbent’s strategy: sales volume increase (market
for electrophysiological studies, procedure to identify
cardiac arrhythmias)
• Measure of entry probability: # of potential entrants

Dafny, Leemore (2005), “Games Hospitals Play: Entry Deterrence in Hospital Procedure Markets,” Journal of Economics and Management Strategy
14 (3), 513–542
 Games hospitals play

Volume increase (%)

0.5

0.4

0.3

0.2

0.1 Number of
potential
0.0 entrants
0 1 2 3 4

Dafny, Leemore (2005), “Games Hospitals Play: Entry Deterrence in Hospital Procedure Markets,” Journal of Economics and Management Strategy
14 (3), 513–542
Product proliferation
• Example: Breakfast cereals

Highly profitable industry

Low technological barriers to entry
Same small number of incumbents for decades
• Idea: increase number of varieties so as to leave no
room for potential entrants
 Preemption by product proliferation

Enter
......................................................................................... 0, 20
......
......
......
..... ......
2 varieties ......
....................................................................................................
.....

...
... Entrant ......
......
..
. ......
... ......
....
.
... ......
Do not enter
......
........................................................................................ 60, 0
....
.
...
...
.
.
...
....
Incumbent ...
...
...
...
...
...
...
Enter
.......................................................................................... 23, -3
... ......
... ......
... ......
... .
. ..
. ......
...
... 3 varieties
..................................................................................................
......
..
Entrant ......
......
......
......
......
......Do not enter
...........................................................................................
40, 0

Assumptions: entrant creates 4 varieties when incumbent has 2; entrant creates 1 variety when incumbent has 3.
Proliferation
• As with capacity expansion, product proliferation
sacrifices short-run monopoly profits for the value of
entry preemption.
• Ditto for store location. Example: Staples

Staples was trying to build a critical mass of

stores in the Northeast to shut out competitors
. . . By building these networks [of stores] in the
big markets like New York and Boston, we have
kept competitors out for a very, very long period.
— Thomas Stemberg, in Staples for Success
• Other examples?
Limit pricing
• If Southwest Airlines is present in airports A and B,
then the likelihood of o↵ering A-B is 70 times higher
than if it is not present in A or B
• Suppose SA is present in A; what do A-B incumbents
do if SA enters B
fares are cut (less so in “guaranteed-entry” routes)
capacity remains the same
number of passengers increases
• Could be adjustment to entry, could be deterrence
(e.g., signaling low cost)
Contracts and bundling
• Long-term contracts: Nutrasweet
• Exclusive dealing: Unilever
• Per processor fees: Microsoft
• All-unit discounts: Intel
• Bundling: Microsoft IE, airlines
• MFC clauses: insurer-hospital contracts
IE vs Netscape
• Windows (only operating system) WTP: $50
• Computer user requires at most one Internet browser
• Three groups of users (1 million each)

IE fans: IE WTP: $25, Netscape WTP: $10

Netscape fans: IE WTP: $10, Netscape WTP: $25
Die-hard Netscape fans: IE WTP: $10,
Netscape WTP: $40
No bundling equilibrium
• Windows price: $50
• IE price: $25
• Netscape price: $25
• Microsoft’s profit: 3 ⇥ 50 + 1 ⇥ 25 = $175 million
• Netscape’s profit: 2 ⇥ 25 = $50 million
Bundling equilibrium
• Windows+IE price: $60
• Netscape price: $30
• Microsoft’s profit: 3 ⇥ 60 = $180 million
• Netscape’s profit: 1 ⇥ 30 = $30 million
• Intuition: bundling commits Microsoft to price IE
aggressively
The aspartame market
• Aspartame: low-calorie sweetener discovered (by
accident) in 1965
• Searle (owner) extends original patent to 1987
(Europe), 1992 (US)
• Monsanto acquires Searle in 1985, sells branded version
of aspartame: Nutrasweet (main market: diet soda)
• Holland Sweetener Company (HSC) creates plant in
1986 in anticipation of aspartame’s patent expiry.
• Monsanto’s preemptive moves:

Sign long-term exclusive contracts with Coke and Pepsi

Slash prices in Europe (first patent to expire)
Naked exclusion
• Entrant requires 30% market share to cover entry costs
• Incumbent “bribes” (70 + ✏)% customers with price
they’d get under duopoly (minus ✏)
• Entrant stays out; (30 ✏)% customers pay high price
 Extracting “entrant’s surplus”
• Potential entrant of uncertain cost efficiency. Absent
exclusive contracts, entry if and only if entry costs are
low (“entry surplus”)
• Incumbent and buyer sign exclusive contract: akin to
“colluding” and setting entry “price” (fee that buyer
must pay incumbent to break exclusive contract)
• Given breach penalty, efficient entrant forced to set
lower price. Good for the buyer and for the incumbent
• If entrant is more but not much more efficient than
incumbent, contract deters efficient entry (analogy
with monopoly pricing)

Exercise 12.4 formalizes this intuition

Raising rivals’ costs
• Most favored nations clause: hospital agrees to “fully
and promptly inform” insurer about lower rates o↵ered
to other insurance companies
• Cases:

Reazin v. Blue Cross & Blue Shield of Kansas

Ocean State Physicians Health Plan, Inc. v. Blue Cross
and Blue Shield of Rhode Island
Controlling essential assets
• Sleeping patents: Xerox plain-paper copying
• Synthetic insulin: Eli Lilly and Genentech
• Patent portfolios: Nortel
• Airport landing slots: Compass II
• Incumbent’s and entrant’s incentives:

Incumbent stands to lose ⇡ m ⇡ d from rival entry

Entrant stands to gain ⇡ d from entry
Typically ⇡ m ⇡ d > ⇡ d
Pay for delay
• When a medical drug patent expires, generics
manufacturers are willing to enter
• Branded drug companies stand to lose ⇡bm ⇡bd
from entry by generics
• Generics entrants stand to gain ⇡gd

• Typically ⇡bm > ⇡bd + ⇡gd , that is, ⇡bm ⇡bd > ⇡gd
• Gains from trade: there is a price p such that
⇡bm ⇡bd > p > ⇡gd such that entrant agrees to be
“bribed” to delay entry
Predatory pricing
• The “dual” of entry deterrence as foreclosure strategy
• Common definition: pricing below cost with the intent
of driving rival out of the market
• Related to concept of dumping (international trade)
Spirit Airlines

$ Detroit-Boston
300

250

200

150

100 Northwest Airlines

Spirit Airlines
50

0
Jan 1995 Jan 1996 Jan 1997 Jan 1998
Predatory pricing: Chicago argument

prey
...
...
.................................................................. ⇡p , ⇡p
...
....
.
...
...
stay in ...
...
......................................................................
..
...
..
L ...
...
... ...
... ...
.
.... ...
...
... ...
prey ...
...
.
. ...
... don’t prey
..
.
.
...
.
.. ..
. ..
.. .
.. ..
. ..
.. .
.. ..
. ..
.. ..
. ..
.. .
.. ..
. ..
.. .
.. ..
.. .
.. ..
. .
S
.
.. .
. ...
....
...................................................................
⇡d , ⇡d
.. ....
.
... ..
. ...
... ...
... ...
... ...
...
.....
... exit
...................................................................
⇡ ,0 m
L ...
...
...
...
...
...
...
...
... don’t prey
................................................................... d d
⇡ ,⇡
Predatory pricing with deep purses
prey
...................................................................
...
...
⇡p , ⇡p
...
..
...
....
.
...

y
bank loan ...
.......................................................................
...
...
L ...
...
..... ...
...
.
... ...
... ...
... ...
stay in .....
...
... don’t prey
....
. N .
.........................................................................
...
.
...................................................................
⇡d , ⇡d
. . ...
...
. ...
...
..... ...
.. ...
prey ...
...
......................................................................
...
no loan
... m
...
... S ...
...
...................................................................
⇡ ,0
.... ...
...
..... ...
.. ...
... ...
... ...
...
.....
... exit
....................................................................
⇡ ,0 m
L ...
...
...
...
...
...
...
...
... don’t prey
................................................................... d d
⇡ ,⇡
Predation as an equilibrium
• The Chicago School approach
• Asymmetric information: reputation for toughness
• Asymmetric information: lending contracts and deep pockets
• Dynamics: learning curves, network e↵ects, etc
Takeaways
• Commitment is the art of limiting your options for strategic
advantage. The commitment must be known and believed by
others
• Incumbent firms have a variety of ways to discourage entry:
long-term contracts, aggressive pricing, excess capacity, product
proliferation, reputation for aggressive response
PUBLIC POLICY
TOWARD ABUSE
OF FIRM DOMINANCE
Outline
• Public policy: false positives and false negatives
• Cases
Outline
• Public policy: false positives and false negatives
• Cases
Basic trade-off: Type I and Type II errors
• Errors in decision making can be classified as

− False positive: competitive practice labeled anti-competitive

− False negative: anti-competitive practice labeled competitive
• Trade-off increasingly difficult as we move down the list:

− Horizontal agreements
− Mergers
− Dominant position (e.g., exclusion)
Observable conditions and economic effects

Horizontal agreements

•
Price fixing
Common standard
•

Anti-competitive Pro-competitive
Observable conditions and economic effects

Aggressive pricing

•
Below cost pricing
Quantity discounts
•

Anti-competitive Pro-competitive
Exercise: pro-competitive or anticompetitive?
• Below cost pricing
• Bundling, quantity forcing, etc
• Exclusive contracts
• Pay for delay
Beware of false positives (dominant firm)
• Competitive justification frequently correct
• Preemptive effect on firm’s competitive strategies
• The extreme statistics phenomenon

− Multiple jurisdictions with global impact

− Jurisdiction balance expected costs from false positives, negatives
− Binding decision is toughest
− In expected terms, it will likely be a false positive
 US and EU on predatory pricing

US E.U.
Market Necessary condition Necessary condition
dominance
Price test p<AVC p<AAC per se illegal
necessary condition p>LRAIC OK (?)
Recoupment Reasonable prospect Not necessary
Intent Talk is cheap Helpful in proof
Key quote Predatory pricing schemes A dominant undertaking
are rarely tried, and even has no interest in applying
more rarely successful such prices except that of
eliminating competitors

AVC: Average Variable Cost. AAC: Average Avoidable Cost. LRAIC: Long Run Average Incremental Cost.
Quotes from Justice L. Powell on Matsushita, 1986; and European Court of Justice on AKZO.
Competition policy and football

The competition authority must be like

a good referee: easy enough to let the
game flow but not too easy to the point
of letting the game become a mess.
Outline
• Public policy: false positives and false negatives
• Cases
Microsoft
• What is the idea of the “fee for processor” strategy?
• What was the thrust of the DOJ 1998 case against
Microsoft?
• What arguments would you present against and in
favor of Microsoft?
• Would you answer be different in 2004 than 2014?
AMD vs Intel
• Intel offers rebates to customers (PC manufacturers)
who agree to limit purchases from AMD
• Compaq’s CEO Michael Capellas tells AMD (in 2000)
“I had a gun to my head” and had to stop buying
AMD chips
• AMD engages in a global litigation strategy

− files complaint with competition authorities in EU,

Korea, Japan
− files lawsuit in US for discovery of certain Intel docs
− files lawsuit in US re Intel’s practices (2004)
• Various government initiated cases: Japan; Korea; EU
DG Comp (2004); and NY AG (2009)
AMD vs Intel
• US Supreme Court approves legal strategy of suing
abroad and pursuing discovery at home
• Japan and Korea competition authorities fine Intel in
the $ millions
• Intel and AMD settle US dispute in 2009:

− Intel to pay AMD $1.25 billion and abide by a long list

of prohibitions;
− AMD to drop its litigation efforts worldwide (suits and
complaints)
• DG Comp fines Intel e1.06 ($1.44 billion at the time)
EU court to rule June 12 on Intel appeal
AMD vs Intel: takeaways
• Global antitrust

− Suing in parallel fronts across the world

− Choosing more favorable jurisdiciton
− Spillovers across cases
• No “double jeopardy” protection
• Rule of reason implies legal uncertainty

It’s as easy to find legal certainty in this case as

it is for scientists to discover the Higgs boson —
Intel lawyer Nicholas Green
Solvay and Actavis
• Solvay receives 2003 patent for brand-name drug,
Androgel (low testosterone levels).
• Actavis files 2003 patent for generic modeled after
Androgel.
• Solvay sues Actavis for patent infringement under
“paragraph IV” litigation (Hatch-Waxman Act)
• Actavis enters a reverse payment settlement agreement
with Solvay (2006), whereby Actavis
− keeps its generic drug off the market
(for a “specified number of years”)
− agrees “to promote Androgel to doctors”
− receives monetary compensation
Solvay and Actavis
• In 2013, US Supreme Court rules 5-3 that FTC cannot
be prevented from bringing antitrust action against
defendants
• Stopped short of declaring pay-for-delay illegal
• Previously Supreme Court had refused to hear similar
cases and FTC had lost
Solvay and Actavis: takeways
• Relation (and possibly conflict) between patent law
and antitrust law
Spirit Airlines vs Northwest Airlines
• What are the case facts regarding pricing?
• What are the points of factual disagreement?
• What are the legal standards for predation?
• How would you have advised Spirit Airlines?
Spirit Airlines

$ Detroit-Boston
300

250

200

150

100 Northwest Airlines

Spirit Airlines
50

0
Jan 1995 Jan 1996 Jan 1997 Jan 1998
Spirit Airlines: takeaways
• In theory, we agree; in practice we don’t
• US standards to prove predation are very high