Transcript Competition Policy

Competition Policy
Market definition and the Assesment
of Market Power
Step 1: Market Definition
 Market definition is done with the aim of
assessing market power
 The RELEVANT market is the set of products (&
geographical areas ) that exercise some
competitive constraints on each other
 The test that guides the analysis of market
definition is the SSNIP Test (or “Hypotethical
monopolist test)
 SSNIP: Small but Significant Non-transitory
Increase in Prices (originally introduced by the
US Dept. of Justice)
SSNIP







Ex.:merger between two sellers of bananasfocus on the
definition of the product market
Suppose an hypothetical monopolist: Would he find it profitable
to raise the price of bananas above the current level by 510%?
Answer: YES, then no significant competitive constraints from
other (substitute) products  bananas are a separate market
Answer: NO, because part of the demand will be redirected to
kiwi, and to other exotic fruits then bananas could not be
considered a separate market
The test continue by considering a wider market: banana and kiwi
Would the H. Monopolist find it profitable to raise the price of
babanans and kiwi by 5-10% ANSWER Yesbananas and kiwi
are a separate market
Answer: No the test continue until we have identified a separate
market (exotic fruits?)
Demand & Supply substituability
 Do not consider just substitutes from the
point of view of demand
 There could be supply substitution if
producers can switch to a new product if its
price increases
 Switching must be easy-rapid and feasible
 No considerable sunk costs and entry
barriers (it should be easy and cheap to
overcome entry-barriers)
 Example of civil aviation: no supply
substituability because of airport
congestion.
Problem in non merger cases:
abuse of dominant positions
 Using SSNIP test in non merger investigations may raise
problems
 Ex.: abuse of dominant position: the question is: did the
firm increase prices above the competitive level?
 Applying SSNIP to current prices may distort results: the
firm, if dominant, has already increased prices it won’t
find it profitable to further raise prices according to
the SSNIP test we must reply the test for a wider market
 A too wide market could be identified and the calculation
of small market shares follows  No dominance is found
 Caution in applying SSNIP in non merger cases
Implementing the SSNIP test
 Own price elasticity: not sufficient
 Cross price elasticity useful to rank
substitutes (some closer some not): % change
in the demand for B with a 1% increase of
product A
 When own price elasticity of A is high ( a
monopolist is not likely to charge higher
prices) look at cross elasticities
 If estimates of cross elasticities are low,
products are not close substitutes and suggest
a separate market
Geographic market definition
 The same considerations hold
 EX: merger between mineral water
producers in Italy SSNIP test: would a h.
monopolist increase the price by 5-10%?
YESthe geographic market is Italy
 NoImports from France are expected and
rising prices is unprofitabletest to be
repeated on a H. Monopolist in Italy &
France
Step 2: assessment of market
power
 Perfect competition is an “ideal” modelin
reality we expect each firm to enjoy some
market power (some fixed cost and some
substitutes in most cases do exist)
 1.Which measure of market power?
 2.Which treshold to call the attention of
competition agencies?
Which measure
 A theoretical measure: the Lerner Index: L =
(P – MC)/ P it increases with the mark up
charged by the firm
 The direct application can create problems: 1)
Estimating MC is not easy 2) High cost can be
due to the productive inefficiency caused by
market power paradox: one obtains high
costs and lower margins lower market power
 Alternative approach: L = 1 /εP  εP easier to
assess with modern econometrics
 Traditional approach look at market shares
 Other variables: potential entrants/
countervailing power of buyers
Traditional approach: market
shares
 To assess market power let us measure
market shares
 A firm with a tiny market share would be
unable to excercise much market power
restraints on the ability to increase prices come
from competitors
 BUT a firm with high market share is not
necessarily dominant If entry was easy the
firm would not be able to increase price OR if
the buyer has countervailing power
Market shares as screening devices
 Market shares could be used as a screening
device: a market share below a treshold
(40%..) leads to presume low market power (it
cannot be considered dominant..) – above the
treshold dominance is presumed and the
burden of proof may fall on the defendant
 In practice the EC and the European Court of
justice follow this approach (useful to increase
legal certainty and reduce the cost of
investigations)
Market shares and beyond
 Market shares both in units and in values might be
available
 In certain industries reserves might be more informative
 If one firm is supposed to be not a crucial player in the
future (cause it uses an old inefficient technology)
current market shares may overestimate competitive
constraints
 Excess capacity by rival firms may be important: if it is
just enough to satisfy current demand their supply
elasticity is low – If they have huge excess capacity the
market power of the investigated firm is low
 It is also the persistence of market shares over time that
informs on market power if the distribution varies over
a short time we presume no dominance and greater
competition