Transcript In Search of Predatory Pricing

In Search of Predatory
Pricing
R. Mark Isaac and Vernon L.
Smith
Is Predatory Pricing Observable in a
Laboratory Environment?
Outline
 Overview
of Research Procedures and
Results
 Predatory Pricing from Literature to
Experimental Design
 The Plato Posted Offer Procedure
 Alternative Hypotheses
 Experimental Results
 Conclusions
Overview of Research Procedures
and Results

Assumptions
1.
2.
3.
4.
5.
6.
7.
Two firms- one large, one small
Scale economies- larger firm has a relative cost
advantage
“Deep pocket” possessed by the advantaged firm
Sunk entry costs
Complete information regarding competitors‘ costs
Induced Rivalry
Antitrust Rules
- Semi permanent price reduction rule
- Quantity expansion limit
Overview of Research Procedures
and Results
 Predatory




pricing distinct from
Competitive prices
Shared monopoly pricing
Dominant firm price
Edgeworth- style price cycles
Overview of Research Procedures
and Results






First series of experiments conducted with
conditions 1-3 (3 experiments)
Second series conducted with conditions 1-4
Third series of experiments conducted with
conditions 1-5
The fourth series induced rivalry amongst
competitors
In the fifth series the large firm acted as
confederate of the experimenters
Two additional antitrust restrictions imposed in
addition to conditions 1-4.
Overview of Research Procedures
and Results
Predatory Pricing from Literature to
Experimental Design
 Distinction
between low prices due to
“good competition” and “bad predation”
exists in literature
 Based
on literature the trading
environment consists of firms producing a
homogenous product in a posted offer
market with full demand revelation
Predatory Pricing from Literature to
Experimental Design
1.
Number of firms
- Single Predator
- Singular Prey (Salop 1981, p.11)
- Plural Prey (Scherer 1980, p.335; Kreps and Wilson
1982; Milgrom and Roberts 1982; Selton 1978)
- Two firm market (Coursey, Isaac and Smith (CIS)
1984; Isaac and Smith 1984)
Predatory Pricing from Literature to
Experimental Design
2. Costs of the Firms
Predator and Prey distinguished by costs
(Mcgee 1958, p.40)
-
Predator and Prey “may” or “may not” have
equal costs (Ordover and Willig 1981 p.308; Salop 1981, p.19)
-
Predator and Prey “do not” have identical costs
(Geskins in Scherer 1980, p.338)
-
To create conditions favorable to predation
predator given an important cost advantage
Predatory Pricing from Literature to
Experimental Design
Predatory Pricing from Literature to
Experimental Design
3. Deep Pocket
- Predator has capital market advantage
(Wall Street Journal, 1993)
- “The length of its purse assures it of
victory.”
(Edwards in Scherer, 1980)
- Predator given double the endowment of
Prey
Predatory Pricing from Literature to
Experimental Design
 4.



Sunk Cost entry and reentry barriers
Small firms face barriers to entry
Economies of scale not the only barrier to
entry (CIS, 1984)
Additional Sunk cost of entry and exit
(Ordover and Willig, 1981, p.305)

Large firm also has incumbent knowledge advantage
(Salop, 1981)

Each seller obtains permit before entering the market
(CILS, 1984)
Predatory Pricing from Literature to
Experimental Design
 5.
Information
- Firms have complete information about
rivals’ costs
(Salops, 1981; Kreps and Wilson,1982 and Milgrom and
Roberts, 1982)
-
Large and Small firm swapped position for
there to be full information of rivals cost
structures.
Predatory Pricing from Literature to
Experimental Design
6. Rivalry
- Predation dependent on the intent of the
predator
- In one series rival intent to exclude small
firm
Predatory Pricing from Literature to
Experimental Design
7. Predatory Pricing Antitrust Program.
- Semi permanent price reduction regulation
on large firm
- All of large firm's price reductions had to be
maintained for five consecutive periods
-
Output expansion limit
- Small firm bought permit to produce
- Large firm could not expand its output for two
periods
The Plato Posted Offer Procedure
 Posted

Offer Institution (Plott and Smith 1978)
Seller sets “take it or leave it” price
Quantities selected by buyers subject to
sellers’ capacity limits
Used the posted offer mechanism
programmed for the Plato computer
system
(Ketcham, Smith and Williams, 1984)
The Plato Posted Offer Procedure

Buyers and sellers sit on separate terminals and trade for 25 periods
 Display screen shows subject’s record sheet

Maximum units that can be purchased (sold)

Each units marginal valuation (cost)

Buyers and sellers can cash rewards equal to the difference between
the marginal value (selling price) and the purchase price (marginal
cost)

No penalties for carryover orders

Each period begins with sellers setting a price and corresponding quantity
and Plato calculates the resulting profit
 Each seller sees the other sellers only after both have entered their offers
 The buyers were then randomly ordered by Plato using a computerized
subroutine under the assumption that demand was fully revealed
 Sellers were not told what the final period of the experiment would be
Alternative Hypothesis
 Alternative
 Any
oligopolistic pricing behaviors
behavioral hypothesis might apply as
long as one or both firms chooses prices
Alternative Hypothesis

Predatory Pricing
-
Price lower than short-run optimal price

Price that deters entry or drives out competitors
Firm A is the Predator when the following conditions hold

 P(Q) A  MC (Q) A
 P(Q) A  MinAC (Q) B
 P(Q) A  [$2.60,$2.65] & QA  8
 P(Q) A  AVC (Q) A
Alternative Hypothesis

Competitive Equilibrium

Price cutting less severe than predation
QA  7 & QB  3
PA & PB  [$2.66,$2.76]

Dominant Firm Equilibrium
PDF  $2.84
QA  6 & QB  3
Alternative Hypothesis

Edgeworth Price cycles



Price cuts of A matched by B until all incentives
are wiped off
Prices cut till PA  PB  $2.79
Shared Monopoly

Tacit collusion
PM  $3.21, QM  5 & Pr ofitM  $3.43
Experimental Results
Experimental Results
Experimental Results
Experimental Results
Experimental Results
 The



absence of Predatory Pricing
In series 1 no seller posted predatory prices
In instances when prices were in the
predatory range the quantities were not
predatory
Such behavior only in period 1
• Early period price experimentation
• Super-sophisticated signal of predation in future
Experimental Results

Series 2



With the introduction of sunk costs no predatory
pricing price quantity pairs chosen in the 69
observations
In 54 cases the smaller firms contested the market
with permits
Series 3


Despite perfect knowledge regarding competitors
costs no predatory behavior
Small firms remained in the market in all 54 instances
Experimental Results
 Series


4
The introduction of $1.00 reward for rivalistic
behavior failed to generate any predatory
behavior
Small firm did not exit the market
Generally there are no price quantity pairs
that lie in the predatory range, however
there are instances when the large firms
behavior was suggestive of predatory
behavior
Experimental Results
 The
dominant firm price setting model is
the most observed price setting behavior

Of the 9 experiments 6.5 can be described by
the dominant firm model
 Only
two of the 10 experiments were
edgeworthian

Signaling a move towards monopoly prices
rather than competitive behavior
Experimental Results
Conclusion
 Unable
to produce predatory pricing in
structural environment
 Dominant
firm equilibrium is predominant