About SAFLII
Databases
Search
Terms of Use
RSS Feeds
South Africa: Competition Tribunal
SAFLII
>>
Databases
>>
South Africa: Competition Tribunal
>>
2018
>>
[2018] ZACT 121
|
|
Shoprite Checkers Proprietary Limited and Other v Massmart Holdings Limited (CRP034Jun15/EXC241Mar17, CRP034Jun15/EXC240Mar17, CRP034Jun15/EXC239Mar17) [2018] ZACT 121; [2018] 1 CPLR 155 (CT) (13 February 2018)
COMPETITION
TRIBUNAL OF SOUTH AFRICA
Case
No: CRP034Jun15/EXC239Mar17
CRP034Jun15/EXC240Mar17
CRP034Jun15/EXC241Mar17
In
the exception applications of:
SHOPRITE CHECKERS PROPRIETARY
LIMITED
First Excipient
PICK 'N PAY RETAILERS
PROPRIETARY LIMITED
Second Excipient
SPAR GROUP
LIMITED
Third Excipient
and
MASSMART
HOLDINGS LIMITED
Respondent
In
re
the matter between:
MASSMART
HOLDINGS LIMITED
Applicant
and
SHOPRITE
CHECKERS PROPRIETARY LIMITED
First Respondent
PICK
'N PAY RETAILERS PROPRIETARY LIMITED
Second
Respondent
SPAR
GROUP
LIMITED
Third Respondent
SOUTH
AFRICAN PROPERTY
OWNERS
ASSOCIATION
Fourth
Respondent
Panel
:
Norman Manoim (Presiding Member)
: Andiswa Ndoni (Tribunal Member)
: Anton Roskam (Tribunal Member)
Heard
on
: 19 September 2017
Reasons
Issued on : 13 February 2018
Reasons
for Decision
Introduction
[1]
In
a complaint referral pending before us, the applicant, Massmart
Holdings Ltd ("Massmart"), a retailer of fresh groceries,
alleges that three of its rivals have exclusive leases with the
landlords of shopping malls nationwide. The applicant alleges that
these leases constitute a restrictive vertical practice prohibited in
terms of section 5(1) of the Act, because they exclude it
from
competing in the relevant markets. The respondent retailers allege
that the complaint is excipiable on several grounds and
should be
dismissed. It is the latter objection of the respondents that we have
to decide in the present matter.
The parties
[2]
Massmart
is the holding company of a group of companies whose activities range
from wholesaling to retail of general merchandise.
The group trades
through various divisions, but only one, Game, is pertinent to this
case. Game is a retailer of general merchandise
and groceries and has
stores located nationwide.
[1]
[3]
Three
of the respondents in this case, Shoprite Checkers Proprietary
Limited ("Shoprite"), the first respondent, Pick
n Pay
Retailers Proprietary Limited ("Pick n Pay"), the second
respondent, and Spar Group Limited ("Spar"),
the third
respondent, are retailers of,
inter
alia,
groceries. They each have a
national presence, and it is common cause, are rivals of Massmart in
the grocery retail market. Each
has filed an exception. The fourth
respondent is the South African Property Owners Association ("SAPOA")
a non-profit
association, representing its members, who are the
landlords in the commercial property market. SAPOA is cited according
to Massmart
because its members have aninterest in the matter and the
relief sought in the proceeding. SAPOA briefed attorneys to keep a
watching
brief but did not otherwise participate in the proceedings.
[4]
This
is not the first time that that Massmart has had to respond to
exceptions in this case. Massmart first brought a referral in
June
2015. That referral was the subject of a number of exceptions and an
application to stay the proceedings. The exceptions were
mostly
upheld, whereas the application to stay proceedings was dismissed by
us on 01 September 2016.
[2]
The present referral is a complete rewrite of the first referral and
seeks to be responsive to the exceptions that were upheld.
In these
reasons we have to consider whether Massmart has met these
objections. It asserts it has; the respondents assert otherwise
and
seek the dismissal of the referral.
Background
[5]
On
31 October 2014, Massmart brought a complaint to the Competition
Commission ("Commission") alleging that exclusive
leases
between the respective respondents and their landlords, who own
shopping malls, were anticompetitive. The Commission non
referred the complaint on 12 May 2015, citing as its reason for this
that it had taken a decision to conduct a market inquiry into
the
grocery retail sector. Massmart then brought a referral itself to the
Tribunal in June 2015 ("the first referral").
[6]
The
respondents all brought exceptions. The referral and the exceptions
are all described in our earlier decision and need not be
repeated
again here. In the first referral Massmart alleged that the conduct
of the respondents amounted to both an abuse of dominance
in terms of
section 8 and a contravention of section 5(1) of the Competition Act,
No. 89 of 1998 ("the Act'').
[3]
The present referral is confined to section 5(1).
The Referral
[7]
In
the referral Massmart seeks an order interdicting the respondents
from:
"..
.enforcing,
or
invoking
with respect to the Applicant, and whether directly against the
Applicant or against the relevant landlord, any provision
in any
lease agreement in terms
of
which any
of
the First to
Third Respondents acquires or is granted the exclusive entitlement ta
trade in fresh grocery products, including fresh
fruit, vegetables,
meat, dairy, and bakery products [“fresh groceries”]
now
or
in the future
at
a
store
located in
a
shopping
mall."
[4]
[8]
Massmart also seeks a declaratory order that the respondents have
contravened section
5(1) of the Act.
[9]
Section 5(1) states as follows:
''An agreement between parties
in
a
vertical
relationship is prohibited if it has the effect of substantially
preventing or lessening competition in
a
market, unless
a
party to the
agreement can prove that any technological or other pro-competitive,
gain resulting from that agreement outweighs that
effect."
[10] The
agreements that form the centrepiece of Massmart's case are the
leases between the respondents
and their landlords in the respective
shopping malls. These leases, whilst varying in their terms, have as
an essential feature,
a term preventing or restricting the landlord
from allowing any other tenant in the mall from selling fresh grocery
products. Included
in the definition of fresh grocery products are
fresh fruit and vegetables, meat, dairy and bakery products.
[5]
(As a shorthand from now on we will refer to these products as
"fresh").
[11]
Sometimes
there is an exception to the restriction mapped out for one of the
other respondents, another competitor or for a smaller
player.
[6]
For instance Cape Gate's lease with Shoprite carves out an exception
to the restriction in favour of Pick 'n Pay and Woolworths.
[7]
Others have a term requiring the landlord to obtain the consent of
the respondent retailer to allow that rival to sell fresh in
the
mall.
[12]
However,
Massmart alleges they have as a common feature that they grant a
measure of protection from competition to the respondents
in the
supply of groceries in these shopping malls.
[8]
Thus, the vertical feature of this arrangement comprises the
landlords in an upstream market for the supply of retail space in
shopping malls suitable for the sale of both fresh and non-perishable
groceries and the downstream market for retail of the same
products
by stores located in shopping malls.
[9]
[13]
Massmart
locates Game as a customer of the landlord suppliers in the upstream
market and a competitor of the other respondents in
the downstream
market.
[14]
The
agreement, Massmart alleges, leads to anticompetitive effects in this
downstream market. The effects are not only their impact
on price
competition because of the exclusion of competitors, but also on
consumer choice; what it terms an
"alternative
shopping experience".
These
anticompetitive effects are felt by consumers.
[10]
However, Massmart also claims the anticompetitive effect is felt by
suppliers. With restrictions removed, Massmart claims that
the growth
of Game would reduce the respondents buying power and incentivise
suppliers to invest in new products and innovation.
[11]
[15]
Game
puts forward the following facts as to the extent of its exclusion.
It has 119 stores of which 102 are in malls in which at
least one of
the respondents is present. Of these 102 stores, it seems that they
know of 53 that are subject to exclusive leases.
There is also one it
concedes does not contain a restrictive lease.
[12]
This leaves 48 stores for which the position is unknown to Massmart.
However, it says that Game's experience here is that the respondents
have sought to enforce exclusivity even where there is no lease
provision that they could rely on to assert this right.
[13]
[16]
Massmart,
however, is unable to present a clear picture of how many such leases
have been enforced or invoked against it and by
whom. Indeed, it
readily concedes that it does not presently have all the information:
"Game is
currently
unable conclusively to confirm the existence, nature or duration of
all the relevant exclusivities in the respondents'
leases since it
is
not privy
to
the lease
agreements held by the respondent retailers.
”
[14]
[17]
Game,
it says, is unable to obtain this information prior to the referral.
[18]
But
it states that even when exclusionary lease terms have not been
enforced, the threat of enforcement in the future discentivises
Game
from a further roll-out of its fresh offering.
[15]
[19] Massmart also
asserts that the leases are often of" ...
extreme
longevity."
As examples, it
cites Pick 'n Pay's lease in the Midlands Mall where the lease has a
duration of 30 years from 2003 and in Cape
Gate, the Shoprite lease
has a duration of 12 years with options to renew for two periods of
five years each. Massmart appears
to contend that the exclusivity
provisions run concurrently with the terms of the leases.
[16]
[20] The leases in this
case are between the respective landlords and the three retailer
respondents. Enforcing
the terms of exclusivity, according to
Massmart, could come either from the retailer respondents directly or
from the landlords
of their own accord or from pressure on them from
the respondents to do so. Hence the relief is framed in the terms it
is, interdicting
the respondents from the "..
.enforcement,
threatened enforcement and invocation"
of
the exclusionary lease terms.
[21] Since all three
respondents are cited in the same matter, despite this being a
section 5(1), and not
a section 4(1) case, Massmart explains that the
conduct has a cumulative effect. But sensitive to the fact that this
cumulative
effect was a subject that was controversial in respect of
the first referral, it tries to explain that the cumulative effect is
not relied on to make out the case for anticompetitive effects. Its
section 5(1) case is an individual case against each respondent.
The
cumulative effect is relied on only because it serves to exacerbate
the anticompetitive effect.
[22] Massmart in its
concluding paragraph explains it in this way:
“
Every
instance of such enforcement, threatened enforcement
or
invocation, as
against
Game,
by
any respondent retailer of any such exclusivity
has
this
anticompetitive effect. However, without derogating
from
the assertion
that the enforcement
of
exclusivities by
any one
of
the
respondents is sufficient
to
render the
anticompetitive harm complained of it
is
worth noting that
the exclusivities in the instant
case
also operate
cumulatively, such that the effect is all the
more
acute given that
Game
faces
the problem
of
exclusivity
enforcement not in relation
to
one competitor
only but each
of
the three largest
national retail chains. This
of
course compounds
the effect
as
Game
and Massmart
thus
has
very
little
room
to
manoeuver. In the instant
case,
the extent
of
this inhibition
is substantial, and extends
across
a
material
portion
of
Game's footprint-
exacerbating the lessening and prevention of competition that single
instances
of
hampering this
peculiar threat would entail.”
[17]
The
exceptions
[23]
All three retailer respondents filed
exceptions. Shoprite, whilst filing an exception, also filed an
answering affidavit. For the
purpose of deciding this matter, we will
not consider that answer but approach the exceptions in the normal
manner and decide the
matter on the facts set out in the referral on
the assumption that they can be proven.
[24] There
is a degree of overlap in the exceptions so we can set them out
briefly as follows:
(a)
Massmart's referral
is not sufficiently aligned to its complaint and thus constitutes a
new referral in violation of the so-called
referral rule.
[25]
This exception, raised by two of the
respondents (Shoprite and Spar) can be dealt with shortly. Deciding
this point requires us
to compare the terms of the complaint to the
new referral. We do not have the complaint in the papers so we cannot
decide this
issue without substantial unfairness to Massmart.
(b)
Failure to join the
relevant landlords
[26]
All the respondents have taken the point
that the relief is not competent if the landlords are not joined
since the relief impacts
on them. Massmart argues that the way in
which the relief is formulated impacts on the respondents, not the
landlords and hence
joinder is not necessary. In any event it argues
that the landlords' industry association, the South African Property
Association
(SAPA), has been cited in the proceedings as the fourth
respondent and it has not opposed, opting to abide.
[27]
We do not need to decide this point
before we conclude on the others. If the point is good, we would
allow Massmart to join the
landlords. It does not on its own
constitute a basis for dismissing the referral.
(c)
Relief not competent in law
[28]
Pick 'n Pay argued that the relief
sought was not competent in law as the respondents could not enforce
the terms of a lease against
Massmart, only the respective landlords
could. This defence was based on a recent Constitutional Court
decision
Masstores (Ply) Limited v
Pick n Pay Retailers (Pty) Limited
(CCT242/15)
[2016] ZACC 42
;
2017 (1) SA 613
(CC);
2017 (2) BCLR 152
(CC) where
the court refused to grant an interdict that Pick 'n Pay had sought
to enforce against Massmart and Game in respect
of the Cape Gate
shopping centre on the basis of the delict of unfair competition. The
Court held that this remedy was not available
to Pick 'n Pay, as it
was not a party to the lease with Game.
[29]
Massmart's response was that the Constitutional Court had decided a
narrow point of unfair competition
that did not resolve the
apprehension it held regarding enforcement of the exclusivity
provisions for competition law purposes.
[30]
Again,
this is not a point we need to decide because Massmart still has to
deal with the hurdles raised by the remaining exceptions.
(d)
Failure to make out a cause of
action
[31]
Although
articulated in different ways, this was the central argument in all
the exceptions. The argument here is that the referral
does not make
out a cause of action in respect of section 5(1), the only section of
the Act that Massmart now relies on.
(e)
Lacks pleading of sufficient
material facts
[32]
The
respondents allege that the upstream and downstream markets lack
specificity. Further, that allegations of an anticompetitive
effect
are insufficiently pleaded.
[33] We
will deal with both these latter points (d) and (e) in our analysis
that follows.
Analysis
[34]
In
our decision in relation to the first referral, we identified several
problems facing Massmart. The first was that market definitions
were
insufficiently clear. The second was that for theory of harm to exist
under section 5(1), it needed to disclose an infringement
of section
5(1).
[18]
[35]
The
respondents argue that Massmart has still not cured these
deficiencies. We go on to consider whether this is the case. First,
we consider if Massmart has made out a case in terms of section 5(1)
following what we term a conventional approach to vertical
restrictive practice analysis. Then we go on to consider whether
Massmart makes out a case following what it terms its unique
competitive threat approach. This approach, as we understand
Massmart's argument, requires us to consider the subjective
competitive
threat posed by Game as an entrant as a counterfactual to
the status quo.
Conventional
approach to section 5(1) analysis
[36]
In
order to bring its case under section 5(1) that makes out a cause of
action, Massmart needs to allege the following:
•
Who the parties to the vertical
relationship are;
•
What the challenged agreement is; and
•
That the agreement has an
anticompetitive effect in a market.
[37]
Massmart
succeeds in establishing the first two elements. The question is
whether it has pleaded sufficient facts to establish a
case against
each respondent in respect of the third.
[38]
Although we have not considered the
issue of what a substantial prevention or lessening of competition
means for the purpose of
section 5(1), the approach taken in the
SAA
[19]
case is apposite here, even though that matter concerned a section 8
contravention. Textually, the language of section 8 differs
from that
in section 5(1)- the test under section 8 is an
'anticompetitive
effect'.
The language of section
5(1) refers to '..
.substantially
preventing or lessening of competition'.
From
an economic perspective the tests used to identify these effects
would be the same. Both are premised on effects and allow
a defence
of pro-competitive justification that is textually identical. It
would be incongruous if the tests for harm differed,
but the tests
for justification are identical. Hence, the approach taken in SAA is
instructive here.
[39]
In SAA, we held that there were two ways
in which an anticompetitive effect could be demonstrated:
“
If
the conduct meets the requirements of the definition, we then enquire
whether the exclusionary act has an anti-competitive effect.
This
question will be answered in the affirmative if there is (i) evidence
of actual harm to consumer welfare or (ii) if the exclusionary
act is
substantial or significant in terms of its effect in foreclosing the
market to rivals. This latter conclusion is partly
factual and partly
based on reasonable inferences drawn from proven facts. If the answer
to that question is yes, we conclude that
the conduct will have an
anti-competitive effect. Whichever species
of
anti
competitive effect we have, consumer welfare or likely foreclosure,
we
have
evidence of a quantitative nature and hence we can return to the
scales with a concept capable of being measured against the
alleged
efficiency gain.
”
[20]
[40] We
examine the referral to see if either form of anticompetitive effect
is made out.
[41] The
theory of harm advanced in this case is that the challenged lease
terms exclude rivals of the respondent
from the downstream market for
the retail of fresh in shopping malls. Both upstream and downstream
markets are premised on the
mall as the subject of the contestation.
In the upstream, landlords are suppliers of malls that are suitable
for the retail of
fresh. In the downstream, the respondents provide
goods in the form of fresh to consumers from malls. In the upstream,
the geographic
market is national. Nationally landlords supply mall
space to those who wish to compete in the downstream. In the upstream
market
the mall is the product being supplied to those who wish to
compete in the downstream. But in the downstream market the mall
constitutes
the boundary of the geographic market where the product
-fresh - can be supplied to consumers.
[42] The
US scholar Phillip Areeda, in his well-known treatise, points out
that it is not necessary for
the upstream and downstream market to
have the same geographic dimension. He accepts that an upstream
market may be national but
the effects may be felt downstream in
local markets.
[21]
[43]
If the mall constitutes the downstream
market, then an agreement that excludes competitors from competing in
it to sell fresh, would
seem,
prima
facie,
exclusionary.
[44]
If this is the case, one might conclude
that Massmart has made out a cause of action for the respondents to
answer.
[45]
But although Massmart asserts that the
mall is its candidate for the geographic market, it fails to make the
necessary factual allegations
to justify this. (Note that in our
first Massmart decision we indicated that a complainant must do more
than allege what the market
is. It must allege at least why it
is.)
[22]
[46]
At best, Massmart relies for this
proposition on generalised contentions made by some of the
respondents in submissions to an ongoing
retail market inquiry being
undertaken at the time of this hearing by the Competition
Commission.
[23]
It quotes passages from submissions made to the enquiry by the
respondents and other retailers, where these retailers emphasise
the
strategic need to have access to retail space in malls.
[47]
However, none of the extracts relied on
are pertinent to supporting the mall as the relevant market
definition or at best, are peripheral
to this issue. The extract from
Spar is simply descriptive of where its stores are located. The
extract from Pick 'n Pay's submission
is a defence to why spaza shops
are not realistic competitors to national chains and hence
exclusionary lease terms do not impact
on them. The extract from
Woolworths in Massmart's own summation identifies "...
the
'type of shopping centre in which the store is located'
as
having an important influence on the
catchment area of stores''.
[24]
[48]
The most that can be said for these
contentions is that they do not negate the idea that location in
malls is commercially advantageous
for large retailers of fresh.
However, they do nothing to help advance a case for why they should
be regarded as the relevant market
[49]
Indeed, Massmart appears to concede as
much, as appears from the following paragraph in the referral:
“
At this
stage
of
the
proceedings, Massmart does not have access to all the information
that would enable it
to
confirm
or
finesse
the precise boundaries
of
the
relevant product market, such as store-level information
on
the respondent retailers' sales over
time (which would
allow
interrogation of, among other things,
the influence of product range on substitution),
or
the extent to which the respondent
retailers' real estate departments consider locations outside
shopping malls to be suitable.
”
[25]
[50]
But
why does Massmart need to rely on what the respondents' retail estate
departments' state, in order to make its own case? If
it has similar
people available to it in its real estate department, why weren't
these views put up?
[26]
[51]
Were
Massmart's ambitions more modest, it would have taken a single mall
or a smaller universe of malls with an exclusionary lease
and
explained why it was not practical for competitors to find suitable
alternative retail space to compete for those consumers.
But Massmart
has not done that. It has made the same allegation against all the
respondents in respect of each space it regards
as a mall, wherever
situated, from small towns to large cities.
[52]
All
in all there may be 102 of them, or maybe more, (as Game is not
presumably in the sum total of all malls for which such leases
exist)
or maybe less, as it is only presently certain of exclusionary lease
terms in 43 of them. What the referral seems to take
for granted is
that all these malls in respect of all the respondents must have the
same characteristic as the single mall example
we consider above. Why
this is so is not explained beyond the generalised submissions made
to the retail market enquiry.
[53]
Thus,
to the extent that the case for Massmart rests on there being a
primary downstream market candidate constituted by the boundaries
of
the mall, this case has not been sufficiently pleaded.
[54] But
justification of its choice of market definition is not the only
problem with the case advanced
by Massmart. Even if its market
definition is considered to be adequately pleaded (which we do not
consider it has), the complainant
needs to establish how much of the
candidate market or markets are foreclosed by the restraint in issue.
[55] The
conventional approach to analysing evidence of foreclosure by an
exclusionary vertical arrangement
is succinctly expressed by Areeda
as follows:
“
In
sum, the preconditions for competitive harm are (a) exclusive dealing
or similar arrangements
covering
a
significant
portion
of
the downstream
market (b) entry barriers or equivalent impediments making it
difficult for rivals
or
potential rivals
in the upstream market
to
obtain efficient
access to the downstream market and (c) resulting prolongation
of
the dominant
firm's ability of earn monopoly profits in the downstream
market.
”
[27]
(Our
emphasis)
[56] The
pre-condition suggested in (a) viz. that the exclusive dealing covers
a "significant portion
of the downstream market" is also
well accepted in the European Commission's Guidelines on Vertical
restraints. Here it is
noted that the Block Exemption Regulation
exists to cover situations where the relevant market share of the
supplier and buyer,
each, do not exceed a 30% threshold.
[28]
[57] What
both approaches illustrate is that, to make a case for exclusion by a
restraint, some dimension
of foreclosure should be alleged. Whilst
our Act is not prescriptive of what the extent of foreclosure needs
to be or conversely,
as in Europe, what a safe harbour would be, that
does not exempt the pleader from making some case out on extent.
[58] Since
the anticompetitive effect occurs in the downstream market, we must
ask what case Massmart has
made for the extent of foreclosure in that
market.
[59] Let
us start first with what Massmart offered as its alternative
candidate for the downstream market.
Offering an alternative market
definition is not in itself problematic, as a market definition is a
conclusion based on certain
facts and conclusions might differ.
However, the alternative is described as a market for fresh that is
"..
.not limited to stores in shopping malls."
[60]
Defining a market not by what it is, but what it is not, constitutes
vague and embarrassing pleading.
If respondents do not know where the
complainant maintains the boundaries lie, how do they meet a case
based on exclusion? Depending
on what radius from the mall is
selected, some may have rival stores located within them, others may
not. The exception taken to
this alternative market definition is
well- founded.
[61] This
leaves us to consider the primary candidate for the downstream local
market which is defined as
the shopping mall. Here there can be no
complaint about the precision of the boundaries.
[62] But
Massmart then goes on to say that:
“
In
either case, the downstream market has both local and national
dimensions. However, Massmart submits that it is the national
dimension of competition that is
of
the most
relevance for this referral.
”
[29]
[63] Quite
what Massmart means by this dual set of dimensions is confusing.
[64] If
the market definition is the mall, it seems what Massmart is saying
is that if there is exclusivity,
then, in that local market, there is
absolute foreclosure - only the respondent firm is able to compete.
However it does not say
so expressly nor why the market should be
limited in every case of a mall in this way. Certainly no facts are
put up. The most,
as we noted earlier, is its reliance on the
submissions to the market enquiry.
[30]
[65]
Second, annexure FAS to the Referral which lists all those malls in
which Game is located, indicates
that in 31 of these malls at least
two of the respondents are present. If the case is that there is
total foreclosure, this thesis
would seem to be undermined by the
presence of at least two rivals in the same mall. Nor does it allege
that even where only one
of the respondents is present that there is
not another non-respondent competitor having a presence. We know from
the referral
that Woolworths, which is not a respondent in this
complaint, is alleged to have a presence in "shopping
centres".
[31]
The fact that Game is excluded from malls does not equate to an
exclusion of competition if another rival is present. Section 5(1)
requires proof of a substantial prevention or lessening of
competition in a market. Mere proof of exclusion of a particular
competitor
does not suffice.
[66]
Thus, to the extent that Massmart relies
on two candidate markets for its downstream market, both are
deficient. The alternative
market definition lacks precision and
there is no allegation as to the extent of foreclosure. The primary
candidate, the mall,
may have geographical precision, but why it
should be so is not sufficiently motivated, and it too, lacks any
allegations as to
the extent of foreclosure, unless one presumes that
each respondent in a mall is a monopoly, but that case too, is not
made out.
[67]
That leaves us to consider, still
following the conventional approach, the
"national
dimension case':
which Massmart
emphasises as the "more
relevant".
We understand this case as follows.
[68]
Although customers for fresh may shop in
some local market, as they presumably do not travel great distances
to shop for fresh,
this demand at the local level does not constitute
the significant competitive constraint. Rather, because the
respondents are
national chains, which have national chains of supply
and advertise prices on a national basis, the constraint each poses
on the
other comes about from this national competition.
[32]
By extension, to the extent that Game is foreclosed nationally from
having an equivalent footprint to offer its fresh, because
it cannot
use its presence in malls to do so, the competitive threat it would
pose is thus constrained and a
fortiori
there is a substantial prevention or
lessening of competition.
[69]
However, if this national dimension of
competition exists, then it must follow that all three respondents
(unless acting collusively)
must constrain one another. Nor may they
be the only firms with a national footprint in fresh.
[33]
On this national market definition Game does not explain why the
challenged lease provisions a
fortiori
create a substantial prevention or
lessening of competition.
[70]
It needs to explain why, notwithstanding
what may be its exclusion from, or the attenuation of its presence in
this national market,
there is still a substantial prevention or
lessening of competition. Nor does Massmart at the national level
give any indication
of the extent of market foreclosure. We know only
of the extent of Game's presence in malls (102) and, with
considerably less certainty,
from how many it may have been excluded
(43). We thus don't know how many malls suitable for fresh are
available nationally and
hence the salience of Game's exclusion from
43 of them.
Conclusion
[71]
Thus
on a conventional approach, Massmart's referral fails in two respects
raised in the exceptions. The requirement to plead a
clear market
definition has not been met in respect of the alternative candidate
market. In respect of the two primary markets
the locally defined
mall market and the national market for the retail of fresh in malls,
no allegations as to why these constitute
the boundaries of the
relevant market have been adequately made out. Finally, as far as a
cause of action is concerned, in none
of the three markets, has the
requirement of the extent of foreclosure been pleaded and hence,
based on SAA, the inference of an
anticompetitive effect.
[72]
A
complainant needs to allege more than the existence of a contractual
restraint. As the Competition Appeal Court noted recently
in the
SAB
[34]
case, where one of the aspects of
the Commission's case concerned an exclusive territory division a
supplier had with its distributors
which it alleged contravened
section 5(1):
“
The
core problem with the evidence provided by the Commission is that it
was based on the argument that any restraint necessarily
constrains
competition. There was no significant attempt to illustrate how the
effects of this alleged restraint would substantially
lessen or
prevent competition in that market.
”
[35]
The
unique case analysis
[73]
The
two tests for consumer harm set out in the SAA case were harm to
consumer welfare and foreclosure to the market of rivals. The
conventional analysis is premised on the latter of these two. The
unique case focuses on the former i.e.
effect
on consumer welfare.
[74]
Massmart
placed most of the emphasis of its case in pleading this unique
threat theory of harm. Translated into the SAA case test,
this means
a case premised on harm to consumer welfare.
[75]
We
have previously held that actual evidence of harm to consumer welfare
would suffice even where market definition may be imprecise.
[36]
This is because market definition serves as a proxy for market power,
not evidence of it. Thus, direct evidence of harm to consumers
is
more probative than proxy evidence.
[76] Thus,
in approaching the manner in which the unique case is pleaded, we ask
whether Massmart has sufficiently
pleaded consumer harm. If it has,
this would meet the criticisms made earlier of its failure to
adequately plead market definition
and extent of foreclosure.
[77] let
us examine what it puts up as the unique competitive threat case.
[78]
Massmart wants to be able to turn its Game stores, with their present
favourable locations in malls,
into vendors of fresh. lt wants to be
able to do this at all its stores so that it acquires a national
footprint and a reputation
for selling fresh at every outlet. If it
can, it says it will constitute a formidable competitor to the three
respondents.
[79]
Absent these restrictions, it will have the benefit of both economies
of scale and scope.
[37]
The strength of its brand in general merchandise and the backing of
the financial muscle of the Massmart group add to its competitive
threat. This exemplifies its uniqueness. It also alleges it has, what
it terms,
"
...substantial ubiquity''.
This
is because it is already located in 102 malls and has a national
footprint on a par with the three respondents.
[80]
Massmart could have demonstrated consumer harm by adopting the
following approach: ln malls where it
already sells without
restriction, consumer welfare has been enhanced by way of improved
competitive outcomes - price and non-price.
This could be by way of
reference to its own offering, as compared to that of the respective
respondent located in that mall, or
a comparison of the effects on
that respondent in a contested mall, compared to one where a
respondent is not contested by a Game
fresh offering.
[38]
[81] But
Massmart has not done so. In the only evidence of this in the record,
a judgement in respect of
an interdict brought against Massmart by
Spar in 2014 in the KwaZulu-Natal High Court, the court papers
suggest that Massmart was
anxious to show that it did not constitute
a threat to the incumbent, as its offering was going to be
modest.
[39]
[82] In
the referral in this case the most that Massmart suggests, by way of
specifics, is what its future
aim is:
"Game's
aim
is
to
be at least
2-5%
cheaper than
comparative retailers for food (perishable or non-perishable)".
[40]
[83]
But that is not evidence that the restraint presently harms consumer
welfare. Rather, it is evidence
of future ambition to improve
consumer welfare.
[84] Put
differently, it is a case of what 'might be', not 'what is'. But a
case under section 5(1) requires
a showing of an
existing
substantial
lessening of competition. Hypothetically, any market may experience
improved competitive outcomes by the addition of
more competitors,
but that does not obviate the necessity for a complainant to plead
facts to show that there exists in a market,
by virtue of a
challenged agreement, an existing substantial prevention or lessening
of competition. Massmart's confident assertion
of the first
proposition does not, a
fortiori,
establish the
existence of the second.
[85]
Furthermore, competition law is premised
on the exclusion of competition as an objective fact, not the
exclusion of a particular
competitor. Massmart's focus and its relief
in this case focuses on Game as a specific competitor. To quote
Areeda again:
"The concern of antitrust
laws is with injury to competition which generally means injury
resulting in lower output and higher
prices in
a
properly defined
market. A foreclosure injury to
a
private firm
occurs when that firm is denied access
to
a
market that
could presumably be open absent the challenged restraint. But
...
this private
injury might not result even from an exclusive dealing covering
a
significant share
of the market. Further even when this private injury does occur, it
may not be injury to competition."
[41]
[86]
Competition Jaw does not look at effects
subjectively from the vantage point of a particular firm, but
objectively from the vantage
point of a hypothetical competitor.
Certainly the former helps inform the latter, but it is not a
substitute for it.
Conclusion
[87]
We thus find that Massmart has failed to
make out a case following either approach. Massmart does acknowledge
that it had difficulties
in getting facts. Massmart's solution to the
problem is to assert that after discovery has taken place, it will be
in a better
position to expand its case. While later discovery might
provide some additional evidence to Massmart to prove its case, it
cannot
be relied on to make a case not made out in the referral.
[88]
The other problem for Massmart is that
its business case has driven its competition case and not
vice
versa.
[89] A
case under section 4(1) against all three respondents or against one
of them in terms of section
8 might have been better suited to the
macroscopic approach to the facts that Massmart had adopted in the
referral. A microscopic
one may have served it better in terms of the
allegations it needed to plead to make out a case in terms of section
5(1).
[90] We
find that the new referral does not deal with the core problems
associated with the first referral.
The referral makes out no cause
of action in terms of section 5(1) and is vague and embarrassing.
[91]
Accordingly, the exceptions are upheld.
Should
Massmart be given another opportunity to amend its referral?
[92]
Massmart has been given an extensive amount of time to remedy the
problems identified with the first
referral. In our order following
the first round of exceptions, we gave Massmart 40 business days to
file a supplementary affidavit
and we then extended this period
twice, by 10 business days and then by a further 3 business days. It
mentioned in its application
for extension that its legal team needed
more time to consult with managers on the ground.
[42]
No doubt they did so. If the information was there to be had, it
would have used it. There is therefore little point in prolonging
this matter further and burdening the respondents with costs.
[93]
Therefore, the referral is dismissed.
Costs
[94]
Costs will follow cause, so accordingly
the first to third respondents are each entitled to their costs. Spar
had sought the costs
of three counsel and pointed out that Massmart
had used three counsel. This is correct, but Massmart had to take on
three contenders,
whilst Spar was faced with one. We do not think
that this case merits the respondents getting more than the costs of
two counsel
each.
ORDER
1.
The exceptions of the first to third
respondents are upheld;
2.
The referral is dismissed; and
3.
Massmart is to pay the costs of the
first to the third respondents, including the costs occasioned by the
employment of two counsel.
13 February 2018
DATE
Mr
Norman Manoim
Ms
Adiswa Ndoni and Mr Anton Roskam concurring
Case
Manager:
Kameel Pancham
For the
Applicant:
F Snyckers SC, MM Le Roux, VS Bruinders instructed by
Cliffe Dekker Hofmeyr
For
the First Respondent:
L Kuschke SC, MJ Engelbrecht instructed
by Werksmans
Attorneys
For
the Second Respondent: DN
Unterhalter SC, GD Marriot instructed by Nortons Inc.
For the
Third Respondent:
A Annandale SC, M du Plessis and A Coutsoudis instructed
by
Garlicke and Bousfield
[1]
We will for the purpose of this case refer to Massmart, since it is
the applicant, but where it is relevant to refer to Game,
we do so.
This is also the approach adopted by all the parties in this case.
See Massmart Complaint Referral at paragraph 4.3.
Note as well we
refer to the parties as they are described in the complaint referral
and not as the excipients and Massmart as
the respondent as they are
described In the several exception applications.
[2]
Shoprite Checkers and Two Others and Massmart Holdings
LimitedCRP034Jun15/EXCOBBJul15;
CRP034Junl5/EXC107Aug15;CRP034Junl5/EXC109Augl5;CRP034Junl5/STA204Dec15.
[3]
During the course of argument on the except ions to the first
referral,
Mass
mart through Its counsel abandoned it reliance
on section 8.
[4]
Massmart complaint referral, notice of motion, paragraph 1.
[5]
See Notice of Motion prayer 1. See also Referral paragraph 18.1. In
paragraph 18.1 the term non-perishable products is also used.
[6]
Massmart classifies these lease exclusivities as individual, Joint,
complete or partial. See referral paragraph 53.
[7]
Referral paragraph 52.1. 2. Pick n Pay's lease with the Cape Gate
also has a carve-out for Shoprite , but although concluded
later,
contains no reference to Woolworths. Annexure FA3, Record page 70,
paragraph 10. Also referred to Is Pick n Pay's lease
in the Highveld
Mall which has a carve-out for Woolworths. (52.2.2)
[8]
Referral paragraph 20.
[9]
Referral paragraphs 18.1-18.2
[10]
Referral paragraph 83.Massmart describes this effects as on price,
quality, range and service.
[11]
Referral paragraph 86.
[12]
Referral paragraph 55.5. This Is in Kokstad where Spar is the
landlord and confirms there is no exclusivity.
[13]
It gives as one example of this a letter from the landlord of a mall
in King Williamstown which states that its lease with Shoprite
Checkers does not permit it to let Game introduce fresh on its
premises. Referral paragraph 55.4 read with annexure F 11 Record
page 313.
[14]
Referral paragraph 56
[15]
Referral paragraph 25. Massmart alleges that Pick n Pay's legal
department In it eagerness to enforce it exclusivities wrote
threatening letters to landlords about Game selling fresh including
in two malls where Game does not have a presence. See referral
paragraph 55.2
[16]
Massmart does not say this expressly. It says the leases have this
duration and the leases contain exclusivities. See referral
paragraph 59
[17]
Referral , paragraph 93.
[18]
Shoprite Checkers and Two Others and Massmart Holdings
LimitedCRP034Jun15/EXC088Jul15; CRP034Jun15/EXC107Aug15;
CRP034Jun15/EXC109Aug15;CRP034Jun15/STA2040ec15
at paragraphs
37-47.
[19]
Competition Commission v South African Airways (Pty)Ltd
Case
number 18/ CR/ Mar01.
[20]
Competition Commission v South African Airways (Pty) Ltd
Case
number 18/CR/Mar01l at paragraph 132.
[21]
P.
Areeda and H.Hovenkamp, Fundamental s
of Antitrust Law,
Fourth Edition ,
paragraph 1802d
page 65.
"When
exclusive dealing is considered as a "foreclosure offence",
it ordinarily becomes necessary to examine market
power or share at
both of the two market levels involved. Importantly, the geographic
boundaries of the two markets are not necessarily
the same..."
[22]
Shoprite Checkers and Two Others and Massmart Holdings
LimitedCRP034Jun15/EXC088Jull5; CRP034JunlS/EXC107AuglS;
CRP034Jun15/EXC109Aug15i CRP034JunlS/STA204DeclS at paragraphs 37.
[23]
It is known as the Grocery Retail Sector Market Inquiry. See
referral paragraph 36 onwards.
[24]
Referral paragraph 36.3
[25]
Referral paragraph 37.
[26]
See record
page
250 paragraph 40. In
a
decision on an
Interdict brought In KZN High Court by Spar the judgment quotes an
earlier judgement which refers to a Ms Gounder,
described as a
senior property manager at Massmart.
[27]
P. Areeda and H.Hovenkamp,
Fundamentals of Antitrust Law, Fourth
Edition,
paragraph 1802b page 64" (This is stated in a
section Is dealing with minimum condition for competitive harm ln a
Chapter
on Exclusive dealing and related practices.)
[28]
See European Union Guidelines on Vertical Restraints paragraph
110(2)
[29]
Referral paragraph 18.1.
[30]
Referral paragraph 36.
[31]
Referral paragraph 36.3. Also the Cape Gate lease for Shoprite
Checkers provides for exclusivity for Shoprite Checkers but exempts
Pick n Pay and Woolworths from the exclusivity.
[32]
These allegations are set out in the Referral from paragraph 40.2 to
43.
[33]
The referral notes the presence of Woolworths for Instance. See
paragraph 41.5 to 41.6
[34]
Competition Commission v South African
Breweries
Limited and Others
Case No . 129/CAC/Apr14.
[35]
Competition Commission v South African Breweries limited and
Others
Case No. 129/CAC/Aprl4 at paragraph 59.
[36]
Seeat paragraph 45 of
National Association of Pharmaceutical
Wholesalers
&
Others and Glaxa Wellcome(Pty) Ltd
&
Others
Case No. 68/IR/Jun00
"Section 4 and 5 claims
do not
require a prior identification
of
the relevant
market- that is, the relevant market can be read bock, as it were,
from evidence of the anticompetitive practice,
thus sidestepping the
formalism inherent in efforts at a prior identification
of
the
market"
- referring to
Natal Wholesale Chemists (Pty)
Ltd and Astra Pharmaceuticals (Pty) Ltd
&
Others
Case
No . 98/IR/Dec00
[37]
Economies of scope arise as the seller of general merchandise can
increase its offer and hence its competitiveness by giving
the
consumer the convenience of a one stop shop.
[38]
Thls is the approach that was taken by the U.S Federal Trade
Commission In obtaining an order blocking a merger between two giant
stationery retail chains. The FTC led evidence that showed that In
local markets where both firms had a presence prices were
lower than
they were in markets where only one had a presence. See Gellhorn et
al
Antitrust law and Economics in a Nutshell,
Fifth Edition,
pages 455 and 460, discussing evidence in
FTC v Staples, Inc.
970
F Supp. 1066 (D.D.C. 1997)
[39]
See
Spar v Synergy et al
decision Case number 802/2014. The
relevant extract is at page 252 of the record, where the court
"summarise the contents
of an email from Ms Gounder a Massmart
employee where she stat es: " ... the current allocation of
such goods in the Game
range would be a floor area of 99.2 square
metres, 'hardly a threat one would say to a 3500 square metre to
5500 square meter
supermarket anchor' ".
[40]
Referral paragraph 62.2.
[41]
Areeda, Treatise, paragraph 1802(b),
page
62.
[42]
Massmart Holdings Limited and Shoprite Checkers Proprietary
Limited
CRP034Jun15/CON211Nov16 at paragraph 41.