BSc Economics and related programmes

Economics of Competition
and Regulation
Lecture 5
Market definition and market
power
Overview
• Why we need market definition
• Key ideas:
– substitutes
– constraints on pricing
• Hypothetical monopolist /SSNIP approach
• Practical methods and examples
2
Market definition...
is a preliminary stage in the assessment of
competition, such as:
• the existing state of competition in a market (the
existence of market power, although remember
that competition policy is against the abuse of
market power);
• the effect of a merger on competition;
3
Reasons for specifying a market
definition
• The market is the context in which
competition is assessed, e.g.:
– Significant lessening of competition (mergers)
– Significant market power (Abuse of mkt power)
• Practical: impossible to consider every
possible way of specifying the market
• Choice may be contentious: => must be
– clear
– based on evidence
4
Key ideas
Example: Vue/Ster cinema merger
• OFT referred a merger between 2 cinema
chains. Concern about SLCs in:
– Romford/Upminster/Thurrock area
– Leeds
– Edinburgh
– Basingstoke
• Questions:
– what is the product market?
– how to define the geographic market?
5
The product market
• Candidates
– cinema exhibition of films/movies
– films including DVD
– the leisure market including restaurants
6
Two Key ideas
• Substitution:
– if two goods are good substitutes they are
effectively in the same market
• Constraining pricing power
– if the presence of another product “B” limits the
ability of “A” to raise prices above competitive
levels then B is in the same market as A
These ideas are obviously related
7
Key ideas Hypothetical monopolist test
Hypothetical monopolist test
X
Y
Z
1. Start with the narrowest definition
(X)
2. Could a hypothetical monopolist of X
raise price above the competitive
level (and get profits)?
If yes, keep this as the definition and stop
Or does the presence of the extra products in Y (in the ex.
DVDs) make this impossible?
If so, new definition is Y. Repeat the procedure with Y and Z
8
Reasons for market definition Key ideas Hypothetical monopolist test
Hypothetical monopolist test (3)
X
Y
Z
What do we mean by
“raise price above the competitive level” ?
A
S mall but
CC
Usually
mergers
5-10%
S ignificant
5%
Non-transitory
Typically around a year
I ncrease in
P rice
- if the increase in prices delivers an increase in profits then the
product does not face significant ‘external ‘ competition
9
Supply substitutability
The reference to non-transitory increases means
that we should look at long-term elasticities
(usually larger than short-term elasticity) and
also at supply side. So one need to take into
account:
• Possibility of entry in the market;
• Contestable markets;
The cellophane fallacy
• Warning: if the industry is already
monopolistic, the SSNIP test fails
automatically. Increasing the price may be
profit-reducing in the limit or convey an
ambiguous result anyway;
• Hence there is an interaction with the
assessment of market power, less worrying for
merger assessment.
Some practical methods
•
•
•
•
•
•
•
•
Price level comparisons
Price correlations
Demand elasticities
Critical loss analysis
Diversion ratio analysis
Price-concentration studies
Transport costs
Journey time analysis (“isochrones”)
12
Practical methods
Price level comparisons
Price level comparisons
Price correlations
Demand elasticities
Critical loss analysis
Diversion ratio analysis
Price-concentration studies
Transport costs
Isochrones
Law of one price:
• Homogeneous products in same market will
have the same price because of arbitrage;
• Large price differentials (if unexplained by
quality differences) tend to indicate separate
markets;
• However price difference with quality differential
do not necessarily mean different markets;
13
Practical methods
Price correlations
Price level comparisons
Price correlations
Demand elasticities
Critical loss analysis
Diversion ratio analysis
Price-concentration
studies
Transport costs
Isochrones
• Similarly, if prices move closely together
for differentiated products they may be
in the same market- differences in levels may be due to
transport cost or other factors
Issues:
– How close is “closely?” Can we get a benchmark
– General inflation tends to bias results. When inflation is high
correlation tend to be large anyway. Deflate by RPI or CPI first?
– There may be some other, independent of market, cause of
correlation. A common input price for example
14
Practical methods
Price correlations
Price level comparisons
Price correlations
Demand elasticities
Critical loss analysis
Diversion ratio analysis
Price-concentration studies
Transport costs
Isochrones
Examples mentioned in Motta:
1. “Stigler and Sherwin (1985) used correlation analysis to test
whether the cities of Chicago, Detroit and New Orleans are
in the same market for wholesale petrol. They correlate
monthly fuel prices in the three cities during the period
1980-82 inclusive. Their results indicate that the correlation
coefficients are very high: the coefficient between New
Orleans and Chicago is 0.792; that between New Orleans
and Detroit is 0.967; and that between Chicago and Detroit
is 0.77, hence the three cities are in the same market.”
15
Practical methods
Demand elasticities
Price level comparisons
Price correlations
Demand elasticities
Critical loss analysis
Diversion ratio analysis
Price-concentration studies
Transport costs
Isochrones
• Own elasticities:
if “low” elasticity there are no close substitutes=>
relevant market. EX: if raise the price by 10% quantity will
drop less than 10%, and assume therefore that SNIPP test
is satisfied. Actually ε<1 means that a price increase,
increases revenues not profits.
SO how low is low enough for the elasticity? Observation
is of course elusive
• Cross elasticities: εxy=(dx/x)/(dpy/py). High cross
elasticity suggests goods are close substitutes and we
should add another product/location
16
Practical methods
Diversion ratio analysis
(usage of products)
Price level comparisons
Price correlations
Demand elasticities
Critical loss analysis
Diversion ratio analysis
Price-concentration
studies
Transport costs
Isochrones
Response to question:
If you could not buy A , what would you buy
instead? B? C? nothing?
It is a hot summer’s day and you fancy a Coke.
But the Coke is all gone...What will you have
instead?
Survey data (not very reliable)
17
Geographic markets
• Import data are relevant when looking at
national or regional markets
• Check both:
– Little in from outside
– Little out from inside
Only If both Ok then market is self-contained
Practical methods
Transport costs
Price level comparisons
Price correlations
Demand elasticities
Critical loss analysis
Diversion ratio analysis
Price-concentration
studies
Transport costs
Isochrones
Useful for geographic definition
If transport cost are low between P & Q
suggests P & Q are in same geographic market
(can be checked by comparing prices and price
trends)
19
Isochrones
Useful for geographic definition
Simplified case:
Store
Price level comparisons
Price correlations
Demand elasticities
Critical loss analysis
Diversion ratio analysis
Price-concentration
studies
Transport costs
Isochrones
max distance
travelled by most
consumers
20
Isochrones
Simplified case:
Overlap area
Price level comparisons
Price correlations
Demand elasticities
Critical loss analysis
Diversion ratio analysis
Price-concentration
studies
Transport costs
Isochrones
max distance
travelled by most
consumers
Store A
Store B
Are there enough consumers in the overlap area to
constrain the pricing of the other store? If so they are
in the same market
21
Mkt power
• The main index of market power is the Lerner Index, that is
the price-cost margin.
• Problems:
– How to measure marginal costs? At best we can estimate average
variable cost;
– A high cost may be due to monopoly situation and inefficiency;
– Of course we could exploit the fact that, under monopoly,
it is the inverse of the elasticity (and estimate the elasticity
instead). But little practical importance in the assessment
of oligopoly mkt power.
– An old fashioned practical approach is to rely on market
shares and concentration measures
Power: Thresholds
• Which market share is worrying?
• The choice of thresholds may increase legal
certainty. Lower threshold (under which a
position is presumed to be non-dominant) and
higher over which there is a positive
presumption and the burden of proof is on the
firm;
• OFT has 40% and 50% thresholds, the EU
possibly set the negative threshold at 25%.
Power: other factors
• Input reserves. May significantly limit the ability
of competitors to constraint the behaviour of one
firm in the future;
• Excess capacity of competitors- a large market
share may be irrelevant;
• Stability of market share; infrequent orders. Mkt
power and shares should be evaluated in the
medium term;
Power: other factors
• Entry: see Cont. Mkts. Note that barriers to
entry may be endogenous (R&D, Advert.);
• Must pay attention to all factors that may
decrease competition and deter entry:
switching costs, part. when artificial, network
externalities; How likely is entry?
• History of industry; reputation for thoughness;
• Buyers’power may limit seller’s power;
however may have some adverse effect on
competition in downstream markets;
Mkt. Power in a Cournot Oligopoly
MKt power
In other words an aggregate index of market power
depends on the index of market concentration,
HHI=∑m2. That explains why HHI is commonly used in
assessment of market power.