Testing profit maximization

1
Testing profit maximization?

• Monopolies

2
Natural Monopoly and Contested Markets

• Coursey, Isaac, and Smith JLE 1984
• Natural Monopoly
• Decreasing average cost
• Entry barrier if already sunk
• If the fixed cost is recurring decreasing AC is
  not by itself an entry barrier
• Contested Market hypothesis
• Even in the presence of decreasing average cost
• Potential entrants provide discipline on the
  pricing behavior of the incumbent due to threat
  of entry

3
The contestable market hypothesis is falsifiable

• The drop in price when adding potential entrants
  with zero entry barriers is a refutable
  hypothesis
• Page 94

4
A-C-E

• Assertion monopoly firms are maximizing profits,
  buyers choices depend only on price and marginal
  value
• No strategic considerations like demand
  withholding
• Test Condition/Assumptions manipulable and
  observable entry barriers, firms have identical
  technologies, ceteris paribus
• Event Predicted When entry barrier is reduced to
  zero the incumbent will set a price at zero
  profits so entry will not occur

5
Refutable hypothesis

• Nul hypothesis based on observable exogenous
  manipulation of the entry barrier, pricing will
  change according to prediction
• Alternative hypothesis price may stay close to
  monopoly price if firms signal a market sharing
  interest
• The price charged in nul hypothesis is
  significantly lower than the price charged in the
  alternative hypothesis

6
Another refutable hypothesis

• Nul hypothesis if the incumbent firm prices
  above AC, the contesting firm will immediately
  enter and undercut pricing
• Alternative hypothesis the contesting firm may
  avoid entering the market altogether fearing a
  price war
• The entry behavior of the contesting firm is
  significantly different under the two hypotheses

7
Two firms profit maximize

• Max1 P1Q1(P1, P2)-WL1-RK1
• Max2 P2(Q2(P1, P2)-Ce-WL2-RK2
• Ce is the entry cost for firm 2
• Q1a-bP1 if P1ltP2
• Q10 if P1gt P2
• Choice variables Pi, (Qi), Li, Ki
• Exogenous variable W, R, Ce

8
Comparative statics

• Change in entry barrier Ce
• Infinite entry cost single seller
• Zero entry cost duopoly
• Because a decrease in Ce will decrease P2
• The predicted response for firm 1 is to lower P1
  to below P2
• Nash Equilibrium is P1P2MC
• Only one firm will remain then, two would be
  unprofitable

9
Experimental Test

• Implement the comparative statics
• Control condition and Treatment condition
• Within firm or Between firm
• Control single seller
• Treatment two potential sellers and zero entry
  cost
• Assertions must be stationary between control and
  treatment conditions or else we cannot claim that
  change in pricing is due to change in entry cost

10
Alternative testing method

• One condition test only the treatment condition
• Is the pricing in the treatment condition
  significantly different from the theoretical
  prediction of monopoly?
• Is the pricing in the treatment condition
  significantly different from the theoretical
  prediction of zero profits?

11
Risk of confounds

• The pricing decision may be influenced by other
  (un-modeled) variables
• Ceteris paribus is not automatically fulfilled in
  the data whether from field or laboratory
• This can only be tested in a control condition
• Thus the need for comparative statics

12
(No Transcript)
13
Do sellers maximize profits?

• Monopoly may be confounded by buyer strategies
• Monopolist may expect buyer strategizing and act
  to avoid it
• Demand would no longer be a simple linear
  function of price different assertion of buyer
  responses
• Duopoly
• Presence of trust may result in market sharing
  arrangement
• Aversion to losing a pricing war
• Changes the assertion of profit maximization to a
  utility function including trust or war aversion
• Stationary preferences must still be maintained

14
The Effects of Market Organization on
Conspiracies in Restraint of Trade

• Isaac, Ramey and Williams, JEBO, 1984
• Motivation for using PO markets
• Implies a price commitment that cannot be
  negotiated during trades
• Costly negotiations
• Retail markets
• Experimental sessions po92, po94, po102, and
  po107

15
Predictions

• Po competitive price
• Qo 7
• Pm Po0.60
• Qm3
• Price at which 4 units are demanded (allowing one
  per seller) P4Po0.25

16
Po102 Monopoly

• High efficiency
• Low monopoly effectiveness
• Mean prices close to Po
• Appeared to have been satisficing?

17
PO94 Monopoly

• Low efficiency
• High monopoly effectiveness
• Demand withholding disappeared

18
Seller conspiracy

• Under which market form is seller conspiracy to
  raise prices most likely to be successful?
• 4 buyers and 4 sellers in each multi-seller
  market
• Maxi PiQi(Pi, Pnoti)-WLi-RKi
• Non-cooperative Equilibrium PiMC

19

• Nul hypothesis
• When conspiracy is costly (no communications are
  possible) PPo
• When conspiracy is low cost (communications are
  possible) PPm or PP4gtPo

20
Comparative statics of changing the cost of
conspiracy
21
Posted Offers allow conspiracies to be upheld
more easily
22
PO conspiracy is similar to monopoly, both lower
than monopoly predictions
23
Profit maximization?

• Monopoly with human buyers
• Demand with holding possibilities although less
  so in Posted Offer than it would be in Double
  Auction
• Oligopoly
• Nash Equilibrium prediction of PMC is confirmed
  over time, thus no implicit cooperation
  consistent with profit max
• Conspiracy has a tension between risk/trust and
  profit maximization risk attitudes?

24
Experimental Evaluation of Institutions of
Monopoly Restraint

• Harrison, McKee and Rutstrom
• Constant, Increasing and Decreasing marginal cost
  conditions
• No human buyers no strategic demand withholding
• Pre-tested for risk neutrality
• Unregulated, subsidized, franchised and
  subsidized, contested

25
Monopoly effectiveness

• Unregulated Constant MC Experienced
• 78 of monopoly profits
• Unregulated Increasing MC Experienced
• 23 of monopoly profits
• Unregulated Decreasing MC Experienced
• 71 of monopoly profits
• Inexperienced UD
• 38 of monopoly profits

26
Profit maximization?

• Cleanest test of these three papers
• No strategic buyer behavior
• No changes in demand conditions over time
• Learning still appears necessary experience
  level changes behavior
• Increasing MC appears more difficult than
  constant or decreasing
• Cognitive resources?

27
Testing theories

• Comparative statics suffer less from confounds
  than testing level predictions
• Difficult to test profit maximization assertion
  due to confounds
• Buyer withholding
• Cognitive costs
• Field data or experimental data
• Experience level of traders
• Control of confounds

28
Implications for applied use of profit
maximization models

• Trading rules and institutions matter
• Double Auction vs. Posted Offer and demand
  with-holding
• Contesting firms may not be visible in data
• Conspiracies affect market power and are affected
  by information conditions, communications
  possibilities, and trading rules