ECON 3440: Regulation and Competition Policy
4. Monopolisation Practices
Dr Heiko Gerlach
Associate Professor
The University of Queensland
Semester 1, 2023
Regulation and Competition Policy 1
Our Plan
1 What is Monopolisation?
2 Predatory Pricing
3 Exclusionary Contracts
4 Bundling and Tying
Regulation and Competition Policy 2
1. What is Monopolisation?
in practice few pure monopolists but many dominant firms
in particular relevant for (i) deregulated industries and (ii) industries
with network effects
monopolisation=pursuit of anticompetitive practices to either sustain
or extend market power
examples: exclusive contracts, tying/bundling, capacity preemption,
price predation, strategic incompatibility, input foreclosure, raising
rivals costs, refusal to supply
monopolisation is prohibited by law
Regulation and Competition Policy 1. What is Monopolisation? 3
US Sherman Act Section 2
Every person who shall monopolize, or attempt to monopolize, or combine
or conspire with any other person or persons, to monopolize any part of the
trade or commerce among the several States, or with foreign nations, shall be
deemed guilty of a felony, and, on conviction thereof, shall be punished by fine
not exceeding $100 million if a corporation, or, if any other person, $1 million,
or by imprisonment not exceeding ten years, or by both said punishments, in
the discretion of the court.
AUS Competition and Consumer Act 2010 Section 46
[Misuse of market power] (1) A person (the first person) who has a substantial
degree of power in a market shall not take advantage of that power in that or
any other market for the purpose of:
(a) eliminating or substantially damaging a competitor of the first person
or of a body corporate that is related to the first person in that or any
other market;
(b) preventing the entry of a person into that or any other market; or
(c) deterring or preventing a person from engaging in competitive conduct
in that or any other market.
Regulation and Competition Policy 1. What is Monopolisation? 4
EU-Article 82: Abuse of dominant position
Any abuse by one or more undertakings of a dominant position shall be pro-
hibited. Such abuse may, in particular consist in
(a) directly or indirectly imposing unfair purchase or selling or other unfair
trading conditions
(b) limiting production, markets or technical development to the prejudice of
consumers
(c) applying dissimilar conditions to equivalent transactions with other parties
(d) making the conclusion of contracts subject to acceptance of supplementary
obligation which have no connection with the subject of such contracts
- minor differences in these laws with respect to whether purpose and/or
effects test is required (in AUS currently only purpose required)
Regulation and Competition Policy 1. What is Monopolisation? 5
Prosecution of Monopolisation
- laws require following steps for prosecution of monopolisation:
1 show that firm was in a dominant position, i.e. had a substantial
degree of market power in the relevant market
2 show a motive/mechanism by which monopolisation practice could be
rational for defendant
3 show evidence that firm engaged in anticompetitive behaviour which
outweighs any procompetitive effects of this business practice
- rule of reason approach: weigh costs against benefits of practices
- no hard evidence of anticompetitive intent (like internal documents)
necessary
- creation of dominant position is not illegal but its abuse is!!!
- building market power through innovation, investment, marketing is
perfectly legal
Regulation and Competition Policy 1. What is Monopolisation? 6
Judge Kaufman in Eastman Kodak Case (1979):
One must comprehend the fundamental tension - one might almost
say the paradox - that is at the heart of Section 2...The successful
competitor, having been urged to compete, must not be turned
upon when he wins.
- law should not work against monopolies built up by superior efficiency
- two-tiers approach to companies (big versus small)
- courts make difference between dominant companies and non-dominant
companies: dominant firm not allowed to engage in some practices
- aggressive conduct allowed to competitors but not to dominant firm
which has “special responsibility”
- in practice judging dominance and abusive behavior is not an easy task
Regulation and Competition Policy 1. What is Monopolisation? 7
When is a firm in a dominant position?
supra-normal profits or market share thresholds are not sufficient for
dominance
dominant firm has unique access to instruments that makes
competition on even playing field hard, such as patents, economies of
scale, network effects or essential facilities
definition of relevant market is important for this judgment
we discuss market definition in more detail in merger analysis
market includes all firms and products that a hypothetical cartel
would need to control in order to raise price in a permanent way
disputes over market boundaries occupy much time and effort in
monopolisation cases
Regulation and Competition Policy 1. What is Monopolisation? 8
2. Predatory Pricing
Definition
Predatory pricing are practices of one or more firms that aim at
1 deterring entry of competitors or
2 driving competitors out of the market
Economic mechanism:
- predator is taking strategic actions to reduce (actual or expected) current
or future profitability of (potential) competitor (=prey)
- motivated by gaining market power in the long run
- predator deviates from his short-run (non-predatory) optimal strategy
- once target of predation has left industry, predator enjoys higher profits
through maintained market power
- price predation is very controversial competition policy issue
Regulation and Competition Policy 2. Predatory Pricing 9
CASE: Entry in Airline Market
Regulation and Competition Policy 2. Predatory Pricing 10
Regulation and Competition Policy 2. Predatory Pricing 11
- court in Matsushita vs Brooke (1986):
“that predatory pricing schemes are rarely tried, and even more
rarely successful” ...
“mistaken inferences in cases like this one are especially costly,
because they chill the very conduct the antitrust laws are designed
to protect”
- antitrust task: search for motives and evidence
Questions we want to address in this section:
1 What are ingredients for predation to work?
2 How to distinguish predation from “innocent” competition?
3 What should and can competition policy do?
Regulation and Competition Policy 2. Predatory Pricing 12
2.1 The Chicago School’s view on Predation
McGee (1958): Predatory pricing is unlikely to happen because
1 a large firm (=predator) suffers more from price cuts than small
(=prey)
not the case if price discrimination is possible
2 recouping predation cost impossible because assets of prey still in
industry
however, there might be sunk cost; reputational effects
3 no predation in the presence of perfect capital markets
why does the incumbent have deeper pocket?
4 predation has to be more profitable than other options like mergers
merger might not be allowed: can attract more entrants
Regulation and Competition Policy 2. Predatory Pricing 13
2.2 Economic Theories to rationalise price predation
two lines of arguments are used to explain profitable predation
(a) predation due to imperfect capital markets (“deep pockets”)
(b) predation due to incomplete information and reputation building
(a) Predation and Imperfect Capital Markets
firm with more financial resources wages price war to drive financially
weaker rival out of market
firms cannot make losses for a long time and weak firms will exit
market
but Chicago School argument: if weak firm is at least as efficient as
incumbent and capital markets are perfect, why shouldn’t it get a
loan and stay in the market?
Regulation and Competition Policy 2. Predatory Pricing 14
predation theory has to take into account relationship between firms’
finance and entrant
in following: simple version of deep pocket predation
Deep Pocket Predation
two periods and two firms (incumbent I and entrant E)
in first period incumbent can fight or accommodate, entrant then
decides to stay or exit
in second period there is either accommodation (if E stays) or a
monopoly for I (if E exits)
profits for both firms satisfy:
P < 0 < A < M
both firms discount second period profits by δ ∈ [0, 1]
Regulation and Competition Policy 2. Predatory Pricing 15
Extensive Form Game:
E
enter out
E
I
fight accommodate
E
(P+δM, P<0)
out stay
(A+δA, A+δA)
(M+δM, 0)
- predation occurs if
P + δM ≥ (1 + δ)A
Regulation and Competition Policy 2. Predatory Pricing 16
An Asset-based Model of Predation
I and E produce homogenous good, MC = 0
compete in quantities with infinite horizon, predation only in period 1
weight of future profits is δˆ = δ/(1− δ) > 0, δ discount factor
market with linear per-period demand D(p) = 1− p
E has to pay amount D = 1/36 to stay in the market after period 1;
if not, it has to exit (imperfect capital market)
two interpretations possible:
1 E has to pay back debt D; if not, bankruptcy
2 E has assets A smaller than fixed costs D = F − A
sunk costs make re-entry cost not profitable
backward induction: second period profits after exit
qm = 1/2,Πm = 1/4
Regulation and Competition Policy 2. Predatory Pricing 17
if E stays on, firms play Cournot in second period with reaction
functions
Ri (qj) =
1
2
(1− qj)
and second period quantities and profits of
qd = 1/3,Πd = 1/9
at the end of period 1, E has to exit if
(1− qI − qE )qE ≤ D
or
qI ≥ qI (qE ) =
qE (1− qE )− D
qE
two types of scenarios for period 2 as a function of qI in period 1:
competition (when E stays on) and predation (when E leaves)
Regulation and Competition Policy 2. Predatory Pricing 18
1. Competitive equilibrium
- incumbent sets qI such that E stays in market
- firms get Cournot duopoly profits in second period
- firms maximise first period profits and play Cournot quantities in period 1
- entrant E has no incentive to deviate as it would strictly lower first
period profits and weakly lower second period profits
- firm I could deviate to quantity qI (q
d) such that E leaves market
- this is incentive-compatible if
Πd + δˆΠd ≥ (1− qI (qd)− qd)qI (qd) + δˆΠm
or
δˆ ≤ δC = 4(1− 9D)
2
5
=
9
20
Regulation and Competition Policy 2. Predatory Pricing 19
Graph: Competitive Equilibrium
RI(qE)
RE(qI)
qE
qI
firm E exits
qd
qI(qE)
firm E stays
qd
Regulation and Competition Policy 2. Predatory Pricing 20
2. Predation equilibrium
- for predation: firm I sets minimum quantity for which there is no qE with
which firm E could make a profit of at least D
- this is the quantity at the maximum of qI (qE ):
∂qI (qE )
∂qE
= (D − q2E )/q2E = 0 or qPE =
√
D
which gives qPI = 1− 2
√
D
- there exists an equilibrium in which firms play (qPI , q
P
E ) and firm E exits
after period 1, if for I it holds that
(1− qPI − qPE )qPI + δˆΠm ≥ (1− RI (qPE )− qPE )RI (qPE ) + δˆΠd
or
δˆ ≥ δP = 9(1− 3
√
D)2
5
=
9
20
Regulation and Competition Policy 2. Predatory Pricing 21
Graph: Predation Equilibrium
RI(qE)
RE(qI)
qE
qI
firm E exits
qIP
qEP
qI(qE)
firm E stays
Regulation and Competition Policy 2. Predatory Pricing 22
Do consumers prefer the competitive or the predation equilibrium?
- consumer surplus in each period S(Q) = Q2/2
- welfare in competitive equilibrium:
CSC = (1 + δˆ)S(2qD) = (1 + δˆ)
(2/3)2
2
=
2(1 + δˆ)
9
- welfare in predation equilibrium:
CSP = S(qPI + q
P
E ) + δˆS(q
m) =
(1−√D)2
2
+ δˆ
(1/2)2
2
=
25
72
+ δˆ
1
8
- in first period consumers prefer predation (25/72 > 2/9), in second
period consumers prefer competition (2/9 > 1/8)
Regulation and Competition Policy 2. Predatory Pricing 23
- competition dominates predation if CSC ≥ CSP or
(
2
9
− 1
8
)δˆ ≥ 25
72
− 2
9
or δˆ ≥ δcs = 9
7
- if future less important for consumers: predation is better than
competition!
- overall we get three regimes
9/20
δ/1-δ
9/7
predation eqmcompetition eqm
predation optimal competition optimal
Regulation and Competition Policy 2. Predatory Pricing 24
- whenever competitive equilibrium occurs it is inefficient since
δC = 9/20 < δcs = 9/7
- predation can be efficient when it occurs (δC < δˆ ≤ δcs)
- predation can be inefficient when it occurs (δˆ > δcs)
Comments:
- firm with deeper pocket can predate rival
- short run loss through price war is weighed against long term monopoly
profits
- consumers can be better off with predation if price war phase is long
relative to recoupment
- arguments holds for more sophisticated entrant’s credit contracts
Regulation and Competition Policy 2. Predatory Pricing 25
(b) Predation with incomplete information
- with incomplete information, incumbent might be able to build up
reputation for fighting entry (Kreps and Wilson, 1982)
- model of sequential entry with two entrants
- entrants don’t know whether incumbent is “tough” and fights entry or
“soft” and accommodates
- behaviour of incumbent with first entrant is observed by second entrant
- 2nd entrant updates his belief whether he faces soft or tough I
- as tough I always fights first entrant, soft I might pretend to be tough by
fighting first entrant
- second entrant cannot learn from first entrant and does not enter if
initial probability of facing tough incumbent is high
- soft incumbent can deter entry that would otherwise occur with
complete information!!!
Regulation and Competition Policy 2. Predatory Pricing 26
2.3 Evidence for Predatory Pricing
- two necessary conditions for predation:
1 predating firm sacrifices short-run profits
2 expected ability to increase profits in the long run
Areeda-Turner test (1974)
- price below short run MC should be unlawful
- MC is hard to estimate; short run MC can be approximated by AVC
Areeda-Turner test: A price below reasonably anticipated average
variable cost should be conclusively presumed unlawful
- not in accordance with theory:
predation price and cost not systematically related
predation price can be above and below MC
Regulation and Competition Policy 2. Predatory Pricing 27
- there are many circumstances under which non-predatory price can be
optimally below MC
1 learning effects, increasing returns to scale (e.g. aircraft
manufacturing)
2 introductory offers and imperfect information (e.g. free samples to
create future demand)
3 consumer switching costs
4 network externalities
5 producer of complementary products
- high risk of type-I error: declaring competitive firm behaviour as
predation
- high risk of type-II error: predation with p > MC remains undetected
Regulation and Competition Policy 2. Predatory Pricing 28
Predation Claims in Practice
1 analysis of the industry to determine market power of alleged firm; if
firm dominant, continue; if not, stop investigation.
dominant firm less likely to charge low price; reduce risk of type-I error
in Brooke vs Brown and Williamson Tobacco (1993) court ruled that alleged
predator’s market share of 12% is too low to allow recoupment
2 analyze relationship of price and cost; if price below pre-determined
threshold continue, if not stop
price threshold increases legal certainty and reduces legal cost
3 identify possible predation mechanism and establish chances to
recoup losses from predatory pricing
in Matsushita vs Zenith (1986) court ruled that a lifetime monopoly profits
would not be enough to compensate alleged 10-year predation phase
Regulation and Competition Policy 2. Predatory Pricing 29
CASE: Spirit vs Northwest Airlines (2005)
- 1996: price war starts as Spirit (SA, regional airline) enters two domestic
routes served by Northwest (NA) with 70% market share
- when SA starts operating second aircraft, NA responds by increasing its
own capacity
- prices drop to unexpectedly low levels
- load factor of SA fell to 30% and by September 1996 canceled both
flights
- after SA’s exit, NA raised it fares sharply and reduced capacity
- SA lodged complaint of anticompetitive behaviour and case went to
district court
Regulation and Competition Policy 2. Predatory Pricing 30
Graph: Spirit vs Northwest Airlines
Regulation and Competition Policy 2. Predatory Pricing 31
Competition or Predation?
- three steps to prove predation:
1 define relevant market, show predator has market power
2 determine whether predation prices were below appropriate cost
measure
3 show that predator could charge price high and long enough to recoup
short run losses
Ad 1. Relevant market and market power.
- parties agreed on relevant market being the city pair connection
- NA argued that connecting passengers should be part of relevant market
- SA claimed only local passengers as NA had different internal fares for
local vs connecting passengers
Regulation and Competition Policy 2. Predatory Pricing 32
- NA had high market share before and after price war (not sufficient for
market power)
- several barriers to entry (frequent flyer program, limited ground-based
facilities, gate lease from incumbents)
- NA invoked US Airways as large enough competitor
Ad 2. Prices below average variable costs.
- parties differ on what to includes as variable what to exclude as fixed cost
- main difference in submissions: NA included all connecting passengers
which reduced their average variable cost
- NA maintained in their analysis that they made profits in the short run
- SA disagreed: NA’s revenue/passenger: $61.98 < costs/passenger:
$65.87− 85.24
Regulation and Competition Policy 2. Predatory Pricing 33
Regulation and Competition Policy 2. Predatory Pricing 34
Ad 3. Recoupment Test.
- predatory pricing like investment; was investment profitable from ex ante
point of view?
- weigh expected cost of predation against expected future profits
- SA calculated recoupment profits with two methods: cost-based and
fares in other markets NA serves
- necessary recoupment period for 3 months predation: 3-6 months
- actual length of recoupment period depends on barriers to entry
- in fact, there was not entry for the next 17 months in this market, NA
had enough time to recoup
- NA refuted the basis of the calculations
- NA asked why SA left market if they thought NA’s fares were not
profitable; they should have waited until NA increased fares to sustainable
levels...
Regulation and Competition Policy 2. Predatory Pricing 35
Conclusion
- in District Court NA won on summary judgment
- Sixth circuit court of appeals found SA’s case credible and remanded the
case back to District Court for full trial
a reasonable trier of fact could find that at the time of predation, North-
westO˜s prices were below its relevant costs for these routes, the market in the
two relevant geographic routes was highly concentrated, Northwest possessed
overwhelming market share, and the barriers to entry were high. Accordingly,
a reasonable trier of fact could conclude that Northwest engaged in predatory
pricingE´in order to force Spirit out of the business.
- in 2007 Northwest went bankrupt and case was closed thereafter
Regulation and Competition Policy 2. Predatory Pricing 36
3. Exclusionary Contracts
Definition
Exclusive dealing describes an arrangement whereby one party’s willingness
to deal with another is contingent upon that other party (1) dealing with
it exclusively or (2) purchasing a large share of its requirements from it.
- vertical (as opposed to horizontal agreements) contractual agreements
Incumbent
Buyer
Entrant
exclusive  contract  
- potential monopolisation of upstream and downstream markets
- strategic use of contracts to gain and protect market power
Regulation and Competition Policy 3. Exclusionary Contracts 37
Several recent high-profile antitrust cases:
- US/EU vs Microsoft (2001-07): MS accused of requiring manufacturers,
ISP and software producers to exclude Netscape’s browser in favour of its
own Explorer; same argument in recent case on windows media player vs
real player
- US vs Visa (2003): Visa had to abandon agreements with banks that
prohibited them from distributing rival credit cards like American Express
and Discover
- US vs Dentsply (2001-06): dominant maker of artificial teeth accused of
illegally excluding rival manufacturers through exclusive agreements with
dental wholesalers selling to dental labs; DOJ alleged exclusion but district
courts did not condemn Dentsply; however Court of Appeals reversed
decision and deemed conduct anti-competitive
Regulation and Competition Policy 3. Exclusionary Contracts 38
- well-documented classic cases:
- Standard Fashion vs Magrane-Houston (1922): leading manufacturer of
dress patterns contracted with prominent Boston retailer to sell patterns
under condition that Magrane not sell patterns of any other manufacturer;
all over US 40% of the 52.000 pattern outlets had exclusive contracts with
Standard; court condemned contract:
“The restriction of each merchant to one pattern manufacturer ...
must amount to giving such a single manufacturer a monopoly of
the business...”
- US vs United Shoe Machinery (1922): dominant manufacturer of shoe
machinery (95% market share) used exclusive clauses in lease contracts
with shoe manufacturers
Regulation and Competition Policy 3. Exclusionary Contracts 39
- courts apply rule of reason in these cases acknowledging both the
potential for monopolisation but also possible efficiency-enhancing effects
of exclusive contracts
- in the following we consider
1 Pro-competitive Effects of Exclusive Contracts
2 Chicago School View
3 Contracts as Barrier to Entry
4 Naked Exclusion with Buyer Externalities
Regulation and Competition Policy 3. Exclusionary Contracts 40
3.1 Pro-competitive Effect of Exclusive Contracts
- exclusive contracts encourage parties to make non-contractible,
relation-specific investment that enhances value of partnership
- without exclusive relationship these investments might not be undertaken
can  invest  
to  reduce  cR  
Manufacturer  1  
Retailer  1   cR  
cM  
can  invest  
to  reduce  cM  
Manufacturer  2  
Retailer  2  
- examples include manufacturer advertising, training of dealer staff,
sharing of trade secrets with retailers, and promotional investments
Regulation and Competition Policy 3. Exclusionary Contracts 41
- risk of expropriation through free riding (e.g. manufacturer advertising
and retailer selling product of other manufacturer)
- GM and Fisher signed exclusive contract in which GM agreed to buy only
Fisher autobodies (GM/Fisher 1919); protect Fisher’s investment in
specialised equipment for GM cars
- United Shoe Machinery Co. required shoe manufacturers only to buy
United’s machines (United Shoe Machinery 1922)
- United argued they needed to protect their investment in training shoe
manufacturers how to efficiently organize their production processes
- without exclusive contract, manufacturer could use knowledge with other
firms’ shoe machine
- potential efficiency gains of exclusive contracts recognised by courts who
treat exclusive dealing cases with a rule-of-reason standard
Regulation and Competition Policy 3. Exclusionary Contracts 42
3.2 Chicago School View
- throughout most of the 20th century courts treated exclusive dealing
harshly for fear of exclusion of competitors and monopolization
- Chicago School economists attacked this pessimistic view with simple
price theoretic/monopoly models
- rational firms would not engage in exclusion for anticompetitive reasons
- exclusive contract only signed if seller and buyer benefit from it:
[Bork (1975) on Standard Fashion]: “...exclusivity has necessarily been
purchased which means the store has balanced the inducement offered
by Standard against the disadvantage of handling only Standard’s pat-
terns... The store’s decision made entirely in its own interest, necessarily
reflects the balance of competing considerations that determine con-
sumer welfare... the reason is not the barring of entry but some more
sensible goal such as obtaining the special selling effort of the outlet.”
Regulation and Competition Policy 3. Exclusionary Contracts 43
- exclusion of a more efficient competitor is not profitable since buyer
would hurt himself by restricting competition when accepting an exclusive
contract
- exclusive contract only signed if it induces investment that enhances
overall efficiency
- in fact Standard Fashion had explicitly contracted with its retailers to
actively promote its patterns; required its retailers to provide a pattern
department at “a prominent position on the ground floor in the store,” a
designated “lady attendant” to give “proper attention to the sale” of
patterns, and a minimum inventory level
- Chicago School proposed following simple model to show exclusive
cannot be anticompetitive
Regulation and Competition Policy 3. Exclusionary Contracts 44
Chicago School model of exclusive contracting:
- incumbent seller I faces buyer B and a potential entrant E
- incumbent can sign exclusive contracts with B before E ’s entry
- buyer’s demand is D(p) and his surplus is CS(p)
- entrant is more efficient than incumbent: cE < c
- let pm(c) denote monopoly price at cost c
- suppose efficiency advantage of E is non-drastic, i.e. pm(cE ) > c
- potential entrant has to pay entry cost f to enter the market
timing of the game as follows:
1 I offers B exclusive contract along with a payment t
2 B decides whether to accept or reject
3 E decides to enter or not
4 active firms name price to B who chooses from whom to purchase
Regulation and Competition Policy 3. Exclusionary Contracts 45
Price equilibrium after entry:
- E wins market at price just below c and makes profits that cover entry
cost since
(c − cE )D(c) > f
- buyer B gets CS(p = c)
Price without entry:
- I charges pm(c) and makes monopoly profits πm(c) and buyer gets
CS(p = pm(c))
Exclusive contract approval stage:
- B accepts contract if and only if
t + CS(p = pm(c)) ≥ CS(p = c)
t ≥ CS(p = c)− CS(p = pm(c)) (IC-B)
Regulation and Competition Policy 3. Exclusionary Contracts 46
- if t sufficiently high, exclusion is possible, but it is only profitable for I if
πm(c)− t ≥ 0 (IC-S)
- exclusion occurs if the (IC-B) and (IC-S) jointly hold, i.e. if exclusion
maximizes the joint surplus of B and S
- no t > 0 exists such that both constraints are jointly satisfied if
CS(p = c) > CS(p = pm(c)) + πm(c)
which always holds because of the deadweight loss created by the
monopoly of the incumbent
- exclusive contracts never arise in equilibrium!
- specific assumption about competition, bargaining process and
bargaining power but conclusion is more general
Regulation and Competition Policy 3. Exclusionary Contracts 47
Graph: Chicago School on exclusive contracts
q
pm(c)
p
c
qm(c)
CS(pm(c))+pm(c)
Regulation and Competition Policy 3. Exclusionary Contracts 48
Bilateral contracting principle: if two parties (i) contract in isolation, (ii)
have complete information about their payoffs and (iii) lump-sum transfers
are possible, then they will reach an agreement that maximizes their joint
payoff.
- the split of this maximized joint payoff is determined by bargaining power
and process, timing of bids etc.
- Chicago School conclusion: inefficient exclusion is in principle feasible,
but not profitable for incumbent-buyer pair
- the gains for the incumbent from exclusion are less than what the buyer
would lose from such contracts
- therefore, if we observe exclusive contracts, it must be due to efficiency
gains from exclusive contracts!
Regulation and Competition Policy 3. Exclusionary Contracts 49
3.3 Contracts as Barriers to Entry
- first model to show anti-competitive effects of exclusive contracts was
Aghion and Bolton (AER, 1987)
- consider partial exclusion contracts: buyer and incumbent sign contract
that specifies damage payment in case buyer switches and buys from
entrant
- this type of stipulated damage contract allow I -B pair to extract surplus
from E
- this rent extraction argument, however, means that I does not want to
exclude E but cash in on the damage payment as the buyer switches
- inefficient exclusion only optimal with incomplete information about
entrants cost (“inefficient exclusion by lack of information”)
Regulation and Competition Policy 3. Exclusionary Contracts 50
3.4 Naked Exclusion with Buyer Externalities
Lit: Rasmusen, Ramseyer & Wiley (1991), Naked Exclusion, AER, and
Segal & Whinston (2000), Naked Exclusion: Comment, AER.
Naked exclusion=incumbent benefits only from full exclusion of entrant,
not through any stipulated payments and partial exclusion
Argument: negative externality and lack of coordination between multiple
buyers can be used by incumbent to deter efficient entry
- negative externality arises from entry cost (or scale economy) of entrant
- incumbent has incentive to sign subset of buyers to make entry
unprofitable and thereby monopolize all buyers
- tied down buyers (who might receive better conditions) ignore that they
make entry impossible and increase cost of purchase to other buyers
- naked exclusion of efficient E does not require incomplete information
Regulation and Competition Policy 3. Exclusionary Contracts 51
Simple Model with two buyers:
- incumbent I and potential entrant E produce homogenous good
- two buyers B1 and B2 value good at v and buy at most one unit from
cheapest supplier
- I has unit cost of c ; entrant is more efficient, produces at cE < c but has
fixed cost of entry F
- two buyers B1 and B2 can sign exclusive contract with I before E ’s entry
- “naked exclusion” contracts (=agreement not to buy from other
supplier, i.e. penalty payment →∞)
- minimum scale requirement for E:
entrant needs to get sales from both buyers to make entry prof-
itable (parameter assumption (MS) below)
Regulation and Competition Policy 3. Exclusionary Contracts 52
Timing of the generic game:
1 I makes offers of fully, exclusionary contracts to both buyers, offers
payment ti to Bi [with/without price discrimination]
2 buyers accept or not [simultaneous or sequential]
3 entrant observes outcome and decides whether to enter or not
4 active firms name prices to “free” buyers (pI , pE ); only I names price
to “locked” buyers pLI
Stage 4: price competition
1. free buyers after entry:
- E serves free buyers at price c making per-customer profits of c − cE
- buyer gets v − c
2. locked-in buyers and free buyers if there is no entry
- incumbent charges monopoly price pLI = v
Regulation and Competition Policy 3. Exclusionary Contracts 53
Stage 3: entry decision
- entry only profitable if E serves both customers:
c − cE < F < 2(c − cE ) (MS)
- if one or two buyers sign the contract, entrant does not enter
Stage 2: buyers’ acceptance
- we consider three different scenarios:
1 non discriminatory offers (t1 = t2 = t)
2 discriminatory offers
3 sequential offers
Regulation and Competition Policy 3. Exclusionary Contracts 54
Scenario 1: Non-discriminatory offers
- I offers t1 = t2 = t to both buyers who simultaneously decide whether to
accept or not:
- two pure-strategy equilibria:
1 if t ≥ 0: (accept, accept) NE
2 if 0 ≤ t ≤ v − c: (reject,reject) NE
- exclusionary and non-exclusionary equilibria possible
Regulation and Competition Policy 3. Exclusionary Contracts 55
- both buyers accepting (and E not entering) is NE as unilateral rejection
does not change entry decision
- minimum scale of entry for entrant requires both buyers to be free
Stage 1: optimal contract proposal of I
- monopoly profit of I per locked buyer: v − c
- to avoid the (reject, reject) NE, I would need to pay t > v − c to each
buyer which is not profitable
- at any profitable t: two Nash equilibria sustainable
- buyers prefer (reject, reject) but might fail to coordinate on it
- externalities among buyers: acceptance increases cost of rejection of
other buyer
- I ’s profit with (accept, accept) and exclusion: 2(v − c)− 2t
Regulation and Competition Policy 3. Exclusionary Contracts 56
Scenario 2: Simultaneous, discriminatory offers
- suppose I can price discriminate and offers exclusive contracts in
exchange for payment t1 ̸= t2 to buyers
- buyer decide simultaneously whether to accept or not
- additionally assume: buyers can coordinate in case of multiple equilibria
- which offers (t1, t2) destroy non-exclusion equilibrium?
Regulation and Competition Policy 3. Exclusionary Contracts 57
- to destroy (reject, reject) NE: any ti > v − c and tj ≥ 0 enough to
capture one buyer
- I ’s profit with exclusion: 2(v − c)− t1 − t2
- I sets lowest (t1, t2) that ensures that at least one buyer accepts
exclusive contract
- profit-maximizing offer:
ti = v − c + ϵ and tj = 0
- makes (accept, accept) unique NE, coordination of buyers does not help
- negative externality between buyers when accepting exclusive contract
- “divide and conquer” strategy : attract one buyer is enough to
monopolize both buyers
- equilibrium profits of I are equivalent to monopolising one buyer (v − c)
Regulation and Competition Policy 3. Exclusionary Contracts 58
Scenario 3: Sequential, discriminatory offers
- I first makes an observable offer t1 to B1, then t2 to B2
- what is subgame-perfect equilibrium?
B1
B2
accept
reject
accept reject
t1
t2
0
t2
t1,
0
v-c
v-c
accept
B2
I
t1
t2 t2
reject
I I
Regulation and Competition Policy 3. Exclusionary Contracts 59
- if B1 accepts, there is no entry; I sets t2 = 0 and B2 accepts
- if B1 rejects, I offers t2 = v − c + ϵ and B2 accepts
- B2 always accepts and entry never occurs, no matter what B1 does!
- B1 accepts any offer t1 = ϵ ≥ 0 rather than to reject
- B1 knows that B2 only considers his own profits in second stage and
accepts
- this acceptance imposes cost of externality on B2 who gets 0 instead of
v − c
- I extracts all the rent from his technology and makes profits of
ΠI = 2(v − c)
- in unique SPE: I can achieve naked exclusion at (almost) no cost!
- sequential offers very effective form of divide and conquer strategy
Regulation and Competition Policy 3. Exclusionary Contracts 60
Comments on naked exclusion with buyer externalities
- rational, inefficient exclusion mechanism without incomplete information
requirement
- critical for inefficient exclusion is existence of externality, here: economies
of scale for entrant; similar analysis/result with network externalities
- symmetry or number of buyers not crucial
- with continuous entry threat I should try to keep number of free buyers
low at any time: staggering expiration date of contracts, longer contracts
- partially exclusionary contracts: penalty contracts like in Aghion and
Bolton (1987) imply trade-off for I : either exclude using externality or
extract rents from entrant
- both could emerge as most profitable strategy
Regulation and Competition Policy 3. Exclusionary Contracts 61
Case: US vs Visa-MasterCard (2003)
- Visa’s by-law 2.10(3) and MasterCard’s competitive program policy
prohibited member banks from issuing the cards of certain other
competitors, including AmEx and Discover
- plaintiffs claimed violation of Sherman Act by restraining competition in
general purpose card network products and excluding competitors
- steps of procedure for courts:
1 Establish market power in relevant market
2 Provide rationale/mechanism for anticompetitive behaviour
3 Give evidence that practice reduces welfare taking into account
possible procompetitive justifications
Ad 1. Court established general purpose card networks as relevant market
excluding the cash/debit card market
- Visa (46%) and Master (27%) card held joint market shares of 73%
Regulation and Competition Policy 3. Exclusionary Contracts 62
- entry in network markets is difficult as it requires to find sufficient
distribution among banks and merchants
- Court argued that Visa-Mastercard recently raised interchange rates
charged to merchants a number of times without losing merchants
Ad 2. Naked exclusions might work particularly well in network industries
as returns from network size increase rapidly (externality more important)
- individual banks are not taking into account that by signing exclusive
contract they reduce relative value of signing AmEx for other banks
Ad 3. Exclusion leads to reduction in access to different card systems for
consumers and loss of choice (match bank and card)
- defendants tried to argue that exclusive contracts protected
Visa-Mastercard’s investment in marketing and development of cards but
court was not convinced
- in 1999 District Court ordered Visa and Mastercard to repeal those
contract features; the decision was affirmed by the US Court for Appeals
Regulation and Competition Policy 3. Exclusionary Contracts 63
4. Bundling and Tying
bundling=products are offered and priced by same company
tying=customers that purchase one good (the tying good) are required to
also purchase another good from producer (the tied good)
- tying/bundling as horizontal foreclosure instrument
Leverage Theory: multi-product firm with market power in one market can
use leverage of its market power to monopolise a second market
- historically received harsh treatment in US courts
- Chicago School heavily criticised this argument and courts more lenient
since 1980s
- Whinston (1990) and others thereafter demonstrated flaws in the
Chicago school arguments
- bundling no longer per se illegal, judged on case-by-case basis, e.g.
Microsoft case
Regulation and Competition Policy 4. Bundling and Tying 64
4.1 Tying with Complementary Products
- Chicago School argued that a component monopolist of a system good
would not have an incentive to use tying in order to extend his monopoly
power to other component market
The Single Monopoly Profit Theory:
Even without tying, a monopolist of one component of a system
good can achieve the monopoly profit of the whole system. Tying
is not profitable as it cannot achieve more than the one monopoly
profit of the system.
- component pricing interdependent: monopolist of one component can
earn system monopoly profits without bundling
- we first present Chicago School argument
- we then show that if there are economies of scope between entry
decisions, tying can be anti-competitive
Regulation and Competition Policy 4. Bundling and Tying 65
Chicago School model
- system good with components A and B used in a one-to-one ratio:
buyers only derive utility v if they have both products
- firm 1 produces goods A and B, firm 2 only produces good B
- mass 1 of consumers want one unit of system good and value it at v
- unit cost of production for good A is cA
- in market B firm i ∈ {1, 2} has production cost cBi where
cB2 < cB1
- demand for system depends on total price Pi = pA + pBi ≤ v for both
products
- pricing in market A and B interdependent with complementary products
Regulation and Competition Policy 4. Bundling and Tying 66
Graph: Tying with complementary products
component  A   component  B  
Firm  1  
Firm  2  
Firm  1  
v  
cA   cB1  
cB2  
- timing:
1 firm 1 decides whether to bundle components or not
2 both firms set their prices given products on offer
3 buyer chooses from which firm to buy
Regulation and Competition Policy 4. Bundling and Tying 67
- first, consider tying: with tying firm 2’s product has no value and makes
no sale
- firm 1 makes monopoly profits of
Πm = v − cA − cB1
- second, consider no tying:
- best response of pB1 and pB2 is to undercut each other as long as price
above MC: cB2 ≤ pB2 ≤ cB1
- best responses of pA and pB2 to each other satisfy: pA + pB2 = v
- set of NE prices such that firm 1 sells A and firm 2 sells component B:
pA = v − pB2 pB1 = pB2 + ϵ
pB2 = λcB1 + (1− λ)cB2, λ ∈ [0, 1]
Regulation and Competition Policy 4. Bundling and Tying 68
- total system price is v , that is, no incentive to increase price for the
firms, price decrease would lower revenue
- firm 2 sells component B and makes profit
Π2 = λcB1 + (1− λ)cB2 − cB2 = λ(cB1 − cB2)
- 1− λ share of firm 2’s rent in market B that is appropriated by firm 1 in
price competition as firm 1 makes profits of
Π1 = v − λcB1 − (1− λ)cB2 − cA
= v − cA − cB1 + (1− λ)(cB1 − cB2) ≥ Πm = v − cA − cB1
- tying complementary products is dominated by selling components
separately
- no entry deterrence motive either as result holds even when tying
prevents entry while selling components induces entry in adjacent market
Regulation and Competition Policy 4. Bundling and Tying 69
Economies of scale between entry decisions
Main idea: Bundling can be an effective entry barrier if entry requires
economies of scale; then, by bundling, the incumbent can prevent the
entrant from realising these economies and deter entry (Carlton and
Waldman, 2002)
Case: US vs Microsoft:
- one of the main argument against Microsoft was the “applications barrier
to entry”: it is difficult to enter OS market because of the large number of
existing applications that run on Windows
- a rival browser has potential to challenge this monopoly in the long run if
sufficient applications for this rival browser would be written
- economies of scope and scale in browser market might challenge
Microsoft’s market power in OS market
- therefore, Microsoft might have an incentive to monopolize browser
market to prevent erosion of market power in OS market
Regulation and Competition Policy 4. Bundling and Tying 70
4.2 Tying with Independent Products
- Whinston (1990) argues that Chicago school argument does not extend
to case with independent products when there are economies of scale in
production or sunk cost of entry
- use same two-product model as Chicago School with independent
valuations vA and vB for the two goods
- multi-product firm 1 faces potential competition if firm 2 enters in
adjacent market B; entry cost K > 0
- argument: bundling sale of good A and B makes firm 1 more aggressive
competitor in market B and discourages entry
- bundling as commitment to compete harder as bundling entails the risk
of losing sales in both markets
Regulation and Competition Policy 4. Bundling and Tying 71
Model set-up: tying with multi-market firm
Product  A   Product  B  
Firm  1  
Firm  2  
Firm  1  
vA   vB  
cA   cB1  
cB2  
- mass 1 of consumers want one unit of each good: value of good A is vA,
value of good B is vB
- assume firm 2 more efficient in producing good: cB1 > cB2
- firms compete in prices: firm with higher net value (valuation minus
cost) sells to consumer at a price that makes consumer indifferent between
firms when losing firm sells at marginal cost price
Regulation and Competition Policy 4. Bundling and Tying 72
1. Entry and pricing without bundling
- firm 1 sells good A at price vA making profits of vA − cA
- after entry: firm 2 supplies the whole market B if cB2 < cB1 making
profits of cB1 − cB2
- thus, entry is profitable for firm 2 if
K ≤ cB1 − cB2 or cB1 ≥ cB2 + K (E-without)
- in this case entry is also socially efficient
- if this condition holds and the entrant wins over market B, firm 1’s
profits are
vA − cA
- if there is no entry, firm 1 earns monopoly profits in both markets
vA − cA + vB − cB1
Regulation and Competition Policy 4. Bundling and Tying 73
2. Entry and pricing in the presence of bundling
- if there is no entry, firm 1 makes monopoly profits vA + vB − cA − cB1
- if there is entry, consumers can buy bundle or product B from firm 2
- firms can cut prices down to their respective marginal costs
- who wins consumers? firm offering (at cost prices) more net value
- firm 1 wins if
vA + vB − cA − cB1 ≥ vB − cB2
or, if the value of consuming A is larger than the cost advantage of firm 2
in market B
vA − cA ≥ cB1 − cB2 (ADV)
- if (ADV) holds, firm 1 wins over both markets and E does not enter
- if (ADV) does not hold, firm 2 might still not find it profitable to enter
Regulation and Competition Policy 4. Bundling and Tying 74
- firm 2’s maximum price (when winning the sale) is defined by
vA + vB − cA − cB1 = vB − p2 or p2 = cB1 − (vA − cA)
and entry is profitable for firm 2 if p2 − cB2 > K or
cB1 > cB2 + K + vA − cA (E-with)
- if (E-with) holds, firm 1 is not selling at and making a profit of 0
- if (E-with) does not hold, firm 1 is making monopoly profits across the
two markets
3. Comparison
- (E-without) easier to satisfy than (E-with): more entry without bundling
as firm 1 is pricing more aggressively when products are tied
- difference are monopoly profits in market A: when competing with
bundle, firm 2 needs to lower by additional vA − cA to make sale
- this reduces incentives to enter the industry in the first place
Regulation and Competition Policy 4. Bundling and Tying 75
Graph: Profits without/with bundling as function of firm 1’s cost cB1:
cB1  cB2   cB2+vA-­‐cA  
cB2+vA-­‐cA-­‐cB1  
vA-­‐cA  
bundling  no  bundling  
vA-­‐cA+vB-­‐cB1  
cB2+k   cB2+vA-­‐cA+k  
entry  without  
bundling  
entry  with  
bundling  
Regulation and Competition Policy 4. Bundling and Tying 76
- there is more entry without bundling as firm 1 is pricing more
aggressively with a bundle
- if (E-without) holds while (E-with) does not hold, then bundling leads to
deterrence of efficient entry in the adjacent market
- bundling can be anti-competitive with independent products
Necessary conditions for tying to be profitable
1 Firm 1 must commit to tie-in sales
▶ once firm 2 enters, firm 1 would be better off selling good separately
▶ firm 1’s profits with bundling after entry are
vA − cA − (cB1 − cB2) < vA − cA
▶ tying requires strong commitment, e.g. through technological choices
(selling integrated systems) rather than purely commercial or
contractual bundling
2 strategy must indeed deter entry
Regulation and Competition Policy 4. Bundling and Tying 77
Case: 3M versus LePage (2003)
in early 1990s, 3M had a near-monopoly in the market for Scotch
brand transparent tape, with a market share greater than 90%
LePage (LP) (an office supply company) had 88% of the private label
tape business in the U.S
in 1992, 3M started making its own private label tape and offered
bundle rebate program across six of its product lines (including
transparent tape)
LP suffered huge losses during 3M’s rebate program and a smaller
manufacturer, Tesa Tuck, exited in 1997
LP complained to DOJ that 3M intended to monopolise the private
label market
Regulation and Competition Policy 4. Bundling and Tying 78
1. Market power in relevant market. Relevant market was the manufacture
and sale of transparent tape for home and office use in the US
- LePage argued that 3M had market power pointing at (i) high market
share, (ii) high profits, (iii) economies of scale and difficulty of small scale
entry
- 3M conceded that it possessed market power in transparent tape brands
2. Motive for exclusionary tying. Tying products can make firm more
aggressive competitor in adjacent markets and exclude rival firms from the
market
- 3M was aware of the threat of LePage’s entry in the adjacent market as
it put pressure on prices for its transparent tape brands
- the reduced profits from tying to exclude LePage from the private label
market could be recouped through the monopoly profits from the
transparent tape market
Regulation and Competition Policy 4. Bundling and Tying 79
3. Anti and procompetitive effects of tying rebate program.
- short run effect for consumers might have been positive, the long run
negative
- 3M had no solid efficiency rationale for its business practice: it argued
bundled rebates offered customers convenience in terms of invoicing
- 3M argued that prices were always above cost and that should be safe
harbour
- even LePage admitted that 3M was more cost efficient than itself
- District court judged 3M guilty of unlawful maintenance of monopoly
power and ordered treble damage exceeding $68M
- Court of Appeal affirmed the decision in 2003, Supreme Court decided
not to take case
- special responsibility for dominant firm: tougher criteria for
anticompetitive practices
Regulation and Competition Policy 4. Bundling and Tying 80
4.3 Efficiency Defense of Tying
- five arguments why bundling can increase efficiency
1. economies of scale and scope
more efficient that computer manufacturer assembles pieces than
individual consumer
2. informational asymmetry
- between consumers and producer, or between producers
- this avoids inefficient component matching
- however: other solutions to information problem possible:
e.g. IBM required from customers to use their punch cards with their
mainframe computers; Supreme court forced them to publishing that cards
with minimum quality from other manufacturers work as well
Regulation and Competition Policy 4. Bundling and Tying 81
3. tying can flatten demand function and avoid inefficient exclusion
- consider monopolist selling two products to two types of consumers
- valuations of different types are as follows
product A product B
type 1 14 6
type 2 5 15
- without tying: firm sells A at $14 to type 1 and B at$15 to type 2
- with tying: firms sells bundle of A and B to both types at $20
- bundling is profitable and increases total welfare
- bundling can avoid inefficient exclusion of consumers
- total welfare argument, consumer surplus decreases
Regulation and Competition Policy 4. Bundling and Tying 82
4. tying as ex post metering device to price-discriminate
- for example copy machine and toner
- to price discriminate between low- and high-intensity users, firm could
bundle sale of copy machine and toner
- charge low price for machine and high price for toner
- total welfare argument
5. bundling can increase incentives for R&D
- bundling allows more surplus extraction in price competition
- more incentives to invest in R&D
Regulation and Competition Policy 4. Bundling and Tying 83