Exclusionary Price Cuts Massimo Motta University of Bologna

1
ExclusionaryPrice Cuts Massimo
MottaUniversity of Bologna

• ACE Meeting - Milan,
• 17 October 2008

2
Acknowledgment

• This builds on a series of papers, with
• Chiara Fumagalli (AER, 2006 EJ, 2007 several
  projects)
• Liliane Karlinger (network industries)
• Helder Vasconcelos (two-sided markets)

3
Economics of exclusionary abuses

• Traditionally, economists skeptical about
  exclusionary practices Chicago-school critique
• More recent economic theory exclusion (of as- or
  more- efficient rivals) may be profitable
• Often (but not always), a common mechanism behind
  different exclusionary practices
• In industries where (demand- or supply-side)
  scale economies matter, if an incumbent firm
  deprives the more efficient entrant of key buyers
  (or markets, or profits) the latter will not be
  able to operate profitably.

4
Exclusionary price cuts an example

• Assume two buyers (or market segments), B1
  and B2, who buy in sequence.
• The incumbent, I an entrant, E, still has to pay
  entry costs, F E is more efficient (cEltcI) but
  can cover F only if it serves both buyers.
• If I induces B1 to buy from it, E will not enter,
  and I will remain monopolist.

5
Example, contd

• Entrant and Incumbent will fight to get B1. Two
  possible outcomes
• If efficiency gap cI-cE small enough ? Exclusion
  B1 will buy from I at price below cI and B2 at
  monopoly price (predation)
• If efficiency gap cI-cE large enough ?
• Entry B1 will buy from E at price below cI and
  B2 at price equal to cI
• Note exclusion is bad for consumers and
  welfare (I is productively inefficient)

6
Beyond the example

• The same results (aggressive pricing may deter
  efficient entry, but for given market structure
  they raise consumer welfare) hold
• If the game is simultaneous
• With N buyers (possibly asymmetric)
• With demand-side scale economies I has customer
  base, E has not (and needs both buyers to reach
  viable customer base)
• If some buyers are captive (e.g. high switching
  costs)
• If I resorts to other practices (e.g., exclusive
  dealing) to subtract scale to E

7
Policy implications, I (rules)

• If price discrimination (or below-cost
  pricing, or rebates) was banned, exclusion
  would not take place when efficiency gap cI-cE is
  small.
• But this ban would reduce consumer welfare when
  cI-cE is high entry occurs anyway, but since I
  cannot price aggressively, E will not need to
  sell below cI to get B1 (chilling competition
  effect).
• Thus, per se prohibition of price discrimination
  would not be a sensible policy (unless there is
  strong belief that in the economy cI-cE is
  small)
• Note a case-by-case policy contingent on
  realised costs makes no sense either

8
Policy implications, II (abuse)

• The more dominant the alleged violator the more
  likely that the practice is bad
• ? Focus on strongly dominant firms
• Counterfactual on prices, market shares, firms
• ? How likely that the victim will exit?
• Price below incremental costs as a necessary
  condition for abuse
• Analysis of market and documents does a
  rationale for exclusion exist? (Overlook intent
  considerations)
• No form-based rules different practices may
  achieve the same objective
• Efficiency defenses (e.g., investment promotion)

9
Napp Pharmaceuticals (UK)

• Sustained release morphine is sold in two
  inter-related segments
• Hospital segment specialist doctors can better
  assess the value of the products high price
  elasticity (hospital pays)
• Community segment patients are prescribed the
  drug by GPs, who mimic hospitals (reputation
  effect, follow-on effect), and for whom price is
  not important (NHS pays)
• Hospital segment only 10-14 of the total market,
  but necessary condition for entrants success.

10
Napps abuse(s)

• Hospital segment excessively low prices where
  facing competition
• Community segment excessively high prices
• Assessment
• Rationale for exclusion (see above- B1 hospital
  B2, community important fixed sunk costs for
  promoting brand to GPs)
• Napp superdominant (and its market share
  increases with the allegedly abusive practice)
• Napps prices below direct costs
• Likely exit (? Oramorph withdrawn)

11
Possible justification?

• Napps defence two-sided market, or
  externalities (the more hospital sales the higher
  community demand).
• ? Pricing below costs efficient in two-sided
  markets (to have both sides on board), or
  normal behaviour with externalities
• But why did Napp do it only when it faced
  competition? When uncontested, Napp did not
  charge lower prices to hospitals!