COMPETITION TRIBUNAL OF SOUTH AFRICA
Case No: 81/IR/APR06
In the matter between:
THE BULB MAN (SA) PTY LTD Applicant/Complainant
And
HADECO (PTY) LTD Respondent
______________________________________________________________
Panel : DH Lewis (Presiding Member), N Manoim (Tribunal
Member), and M.T. K Moerane (Tribunal Member)
Heard on : 24 October 2006
Decided on : 14 November 2006
Reasons Issued: 28 November 2006
REASONS AND ORDER [NON_CONFIDENTIAL VERSION]
INTRODUCTION
[1] The applicant has brought an application for interim relief in terms of
section 49(C) of the Competition Act 89 of 1998(“the Act”). This application
was heard on the 24 th of October 2006. The respondent opposed the
application. On The 14 th of November 2006, we dismissed the application.
Our reasons for dismissing the application follow.
[2] The application concerns a decision by the respondent, a supplier of
flower bulbs, to refuse to supply the applicant, a firm that distributes flower
bulbs to schools for fund raising schemes, on terms that it had previously
supplied it. The applicant contends that this refusal amounts to a
contravention of sections 5(1), 8(c), 8(d)(ii) and 9 of the Competition Act 89 of
1998 (“the Act”). The respondent does not deny making such a decision, but
places in issue whether it has any competitive implications. In particular, it
alleges that its actions were motivated by legitimate business concerns,
because it was experiencing a deteriorating business relationship with the
applicant.
1
THE PARTIES
[3] The applicant is The Bulb Man (SA)(Pty)(Ltd) (“TBM”), a South African
company with its principal place of business at Bellville in Cape Town. The
respondent is Hadeco (Pty) Ltd (“Hadeco”), also a South African company,
with its principal place of business at Parktown, Gauteng.
BACKGROUND
[4] The respondent is a distributor of flower bulbs that can be used in the
garden or to produce cut flowers. The respondent sources its bulbs from its
own production, other domestic growers and imports. The respondent has
developed its eponymous brand name Hadeco, which has strong resonance
with consumers of this product. The respondent distributes its product through
various outlets both retail and wholesale. Recently school fundraising
schemes have become a niche outlet for the sale of bulbs to the public and
this is where the applicant fits in.
[5] Len Ginsberg established the applicant's business in 1970. Some
years later, his son Lee acquired an interest in the business and later, in
1984, a controlling interest. 1 (The present business of the applicant has
undergone several changes of ownership form in that time but in substance,
the business has remained the same and it has always been owned by
members of the Ginsberg family.) 2
[6] The business started off modestly by selling flower bulbs to scout
groups as a means of fundraising. In 1989 the business decided to expand
the fundraising concept to schools and this seems to have led to an increase
in its fortunes as the applicant states that the business experienced
substantial growth in the 1990”s. 3
[7] The Bulb Man business has never produced its own bulbs and has
relied on others for its source of supply. Between 1972 and 1984 it purchased
on a wholesale basis from the respondent. From 1985 until 1993 it purchased
on a wholesale basis from the respondent. From 1985 until 1993 it purchased
its flower bulbs from a firm called Hollandia. According to the respondent, it
1 See page 402 of the record
2 The Bulb Man (Pty) Ltd was formed in 1978 and the Bulb Man Scheme became the business of that
company. The company was converted into a Close Corporation in 1984. Len Ginsberg and his son Lee
Ginsberg held the members interest in the close corporation. A further close corporation was formed in
Cape Town, with the members’ interest held in the same proportions between Len Ginsberg and Lee
Ginsberg. In 1996 Bulbman Limited, a company registered in the Hague and a branch thereof
registered as an external company in the Republic of South African, acquired the business of the two
close corporations. In 2004, the complainant was formed and the business of the Bulbman Limited with
all its assets, liabilities, rights and obligations were transferred to it in that year.
3 See page 403 of the record
2
purchased from Hollandia on a wholesale basis. 4 The applicant is less clear
on this point only describing the nature of services Hollandia performed.
According to the applicant, Hollandia was unable to meet its increased needs,
because Hollandia was in turn being supplied by the respondent, which was
unwilling to extend Hollandia credit. 5 The respondent’s version of the supply
switch is that the applicant did not want to continue an arrangement with
Hollandia where it, the applicant, had to carry the risk of nonsale stock. In
1993 it had purchased two million bulbs too many from Hollandia and
because it no longer wanted to carry this risk itself it switched suppliers so
that it could enjoy an agency relationship with the respondent. 6
[8] Whichever version for the change in supplier is correct, it is at least
common cause that in 1993 the applicant resumed its direct relationship with
the respondent. It is also common cause that there was an agreement
regulating this relationship. The legal nature of the relationship is not common
cause and has been the subject matter of dispute for years. The respondent
states that the applicant became its agent. 7 The applicant denies this and
alleges that it has been the principal throughout. For our purposes not much
turns on how the relationship would be classified by the common law.
Although in this decision we will from now on refer to this as an ‘agency’
relationship this is out of convenience, to distinguish this from the wholesale
relationship – it is not indicative of any legal conclusion we have reached
about the nature of the relationship.
[9] The dispute over whether the applicant was the respondent’s agent was
not a mere conceptual debate, but related to who bore the commercial risk of
not a mere conceptual debate, but related to who bore the commercial risk of
the scheme, and who could hold out to schools as being the principal. The
respondent, both agreed, was responsible for dispatching invoices to the
schools but it was on whose behalf that there was disagreement. 8
Correspondence forming part of the record shows that this dispute dates back
as far as 1996, at least. 9 From what we have on record, this dispute is never
resolved, although the business relationship seems to persist despite the
apparent faultline looming beneath it.
[10] Eventually the inevitable rupture occurred in 2005. It was precipitated
by the departure of Len Ginsberg, the founder of the applicant's business and
at the time its chief executive officer. We are not given any explanation in the
interim relief papers why this happened, but we do know that Len Ginsberg
reentered the bulb market through another entity, known as Len’s Flower
Bulbs CC (“LFB’). LFB entered the market as a direct competitor of the
applicant. Whether LFB already existed prior to his departure or not is in
4 See page 116 of the record at paragraph 11
5 See page 403 of the record paragraph 43
6 See record pages 117 117 A paragraph 15
7 See page 116 of the record at paragraph 12
8 See record page 4534.
9 See record pages 469 472
3
dispute. On the respondent‘s version, the wife of Len Ginsberg had founded
LFB and he had joined on his departure from the applicant. On the applicant’s
version this business, although established in April 2005, could not have
started without the full involvement of Len Ginsberg whose employment with
the applicant had terminated in July 2005. 10
[11] Hostilities between LFB and the applicant commenced swiftly. The
applicant accused LFB of passing off and brought an interdict against it in the
Cape High Court in August 2005. That dispute has taken its own trajectory,
but it is how the respondent has become embroiled in the dispute that is
relevant for the present matter.
[12] In 2005, the respondent was the supplier of both the applicant and
LFB, both it appears being supplied on the “agency” basis. Initially, it seems
that the respondent adopted a neutral stance in the war, supplying both sides
on equal terms. 11 Eventually the respondent ended its neutrality and
appears to have favoured LFB over the applicant. Why this has come about is
not clear, nor important for the purpose of this application. What is important
is the fact of the breach, as this has become the rationale for the respondent’s
refusal to supply the applicant on the “agency” terms.
[13] The respondent then attempted to terminate its contract with the
applicant. The applicant contested its right to do so and the matter too went to
court in the Transvaal Provincial Division of the High Court (TPD) and
culminated in a consent order in November 2005. The terms of the consent
order appear to be a preservation of the status quo in terms of the existing
contract as well as an attempt to plug some gaps in it – its terms are lengthy
and need not be repeated here. The consent order also contained a series of
interdicts against interference in the applicant’s business, inter alia, by either
the respondent or LFB.
the respondent or LFB.
[14] The terms of the consent order only obliged the applicant to supply the
respondent on that basis until June 2006. In November 2005, the
respondent's attorneys indicated that it would no longer supply to the
applicant on the terms contained in the order beyond that period.
[15] In December 2005 the applicant then lodged a complaint with the
Competition Commission, alleging that the respondent was contravening
sections 8 and 9 of the Competition Act by refusing to supply it on the terms
contained in the original supply arrangement what we have termed the
“agency”’ arrangement. In March 2006, the Commission indicated that it
would not refer the complaint and issued the applicant with a certificate of
nonreferral. Thereafter, the applicant referred the dispute directly to the
10 See page 403 of the record
11 See paragraph 17 on page 9 of the record
4
Tribunal. In the complaint referral the applicant seeks an order that it be
supplied on the same terms as set out in the TPD order. At a prehearing of
the complaint matter, held in July 2006, we set the matter down for hearing
from 22 January 2007.
[16] Concerned that even if it succeeds in the main complaint, this may
come to late to save its business, the applicant decided to launch interim relief
proceedings in addition. It is unusual for a party that has filed a complaint
referral and obtained hearing dates, to at the same, time seek interim relief.
The applicant has its reasons for doing so, which we explain later, when we
look at the mechanics of the scheme and its time sensitivity.
[17] The respondent has opposed both the main complaint proceeding and
the interim relief proceeding. It has however indicated that it is prepared to
continue supplying the applicant on the terms it supplies its bulk wholesale
customers, pending determination of the main complaint proceedings. 12 This
offer of wholesale terms is not acceptable to the applicant, hence this
application.
THE NATURE OF THE SCHEME
[18] The bulbselling scheme, as implemented by the applicant, and
presumably by LFB as well, is organised over a long time line. It begins in
June of the previous year when the applicant designs the scheme. This
involves determining how it will market the scheme to schools. Schools are
offered various incentives to participate, including prizes and free bulbs. Since
free bulbs are part of the incentives, this involves consultation with the
respondent.13 From July until November, the applicant markets the scheme
to schools and signs up participants. (This explains why the applicant applied
for interim relief now. If it has to wait until next year to ascertain if it has an
agency contract it fears it will not be able to recruit schools in time. If it recruits
agency contract it fears it will not be able to recruit schools in time. If it recruits
schools without certainty about its conditions of supply, it might assume an
unacceptable risk.)
[19] In November, the respondent informs the applicant at what prices it will
supply it. This becomes the basis on which the applicant sets its selling price
to the schools, and importantly, the price at which the school sells to members
of the public. Under the bulb selling scheme, it is the intermediary, that is, the
applicant or LFB, that sets the price that the school sells to the public.
[20] In January the following year, the applicant distributes promotional
material to the schools that have signed up. Included in this material are
12 See paragraph 28 on page 121 of the Record
13 See paragraph 11.1 on page 5 of the Record
5
colour brochures of the flower bulbs, which pupils use to solicit orders from
the public. They are given a window period to do this, usually it is from
February to March. At the same time as taking the order the pupil is supposed
to get payment, which is then remitted to the school. 14 The applicant then
collates the orders and forwards them to the respondent. The respondent
packs the bulbs in pack sizes suitable for the schools.
[21] The respondent then delivers the bulbs to the schools, invoices them
and collects payment. Once the respondent has collected payment from the
schools, it deducts the amounts owing to it for the cost of the bulbs and pays
the balance to the applicant.
[22] We have attempted to describe the scheme, as best as we can, on a
common cause set of facts. It has not been easy for us to do so as the parties
have differed in the papers over detail, no doubt because the agreement
between them has always been the subject of dispute, as we stated earlier. 15
[23] What is relevant to us from the scheme are the following features. The
school adopts a relatively passive role. It takes its purchase price and selling
price from the applicant. The school relies entirely on the applicant to run the
scheme. It appears to write draft letters from the school to parents on how the
scheme works, advises on whom to sell to etc. 16
[24] Secondly, the scheme whether considered in law an agency contract or
not, involves several interactions between the applicant and the respondent
over the life of the scheme and no doubt a high degree of cooperation
between respective staff to coordinate deliveries, collection of monies and
dividing up the spoils. Add to this the appearance of joint marketing by way of
carrier bags and branding, and one appreciates that if the relationship
between the two firms breaks down it may create havoc for both businesses
and bemusement for the third party, the school.
and bemusement for the third party, the school.
[25] By contrast the wholesale arrangement, which the respondent offers
the applicant, at least until the complaint proceeding has run its course,
involves far less contact and interaction. 17 As we understand it the applicant
would then be treated on the same terms as the standard bulk wholesale
client, that is, a nursery, for instance, and there is no special regard for the
special requirements of the fund raising scheme. This means the applicant
would have to take the risk that the schools would procure what it has
14 According to the applicant’s forms, “ all orders must be cash on order not cash on delivery.”
15 See pages 288293 of the Respondent’s answering affidavit and pages 415421 of the Applicant’s
replying affidavit
16 See record page 476 – document from applicant entitled “suggested letter to parents”
which suggest inter alia selling to Dad’s golf and Mom’s tennis friends.)
17 Under the wholesale arrangement it appears the Bulb Man would be supplied “unbranded” bulbs.
Under the agency arrangement the bulb packages contain the Hadeco logo.
6
ordered, pack the bulbs and deliver them. The applicant also claims that the
respondent will not guarantee that it will supply it with sufficient bulbs to serve
its needs. Thus as we understand the applicant, the wholesale model may
serve the general retailer well, but not a firm like it which has a niche market
depending on timing and a precise match between orders and supply. In the
agency model the risk of matching orders and supply rests with the
respondent. In the wholesale model, the risk is with the applicant; it has to
order before it knows its demand from the schools and so it faces the risk that
it will have too much or too little. Recall that on the respondent’s version the
applicant was left with two million unsold bulbs when it had a wholesaler
relationship with Hollandia.
[26] Rather surprisingly having made this fact known in its papers the
respondent's economist has still attempted to suggest that the applicant is
commercially better off as a wholesale customer than it is as a customer with
an “agency” relationship. 18 This is not a factual dispute that we need to
resolve now – the fact is that the applicant does not share this sanguine view
of the wholesaling relationship, hence the present application. Of course, one
could argue that the applicant is only at risk for one season, until the main
complaint has been heard. The applicant states, however, that the wholesale
model is so unattractive that it may force it out of the market for this season
and that if this happens it will be hard to return as an effective competitor. In
the meantime, it contends, its competitor LFB continues to be supplied on the
agency terms and hence will be able to take its customers from it.
RELEVANT MARKETS
[27] Much of the dispute in the papers concerns definition of the relevant
markets, as the competition problem manifests itself in an upstream market,
markets, as the competition problem manifests itself in an upstream market,
where the respondent supplies product to intermediaries, and a downstream
one, where the applicant sells product to schools. The applicant and the
respondent do not agree on the definition of either upstream or downstream
market. As a matter of practice, since this is an interim relief case, we would
want to avoid making a decision on market definition, which might at the main
hearing prove premature, unless it is necessary for us to do so to determine
this application. We find it is not necessary for us to do so, because even if we
adopt those market definitions most favourable to the applicant’s case, it still
fails to establish the existence of a prohibited practice.
[28] The applicant defines the upstream market as one for the production
and supply of dry flower bulbs and that this would make the respondent the
only major supplier in the market. The applicant is not able to provide market
share figures for its version of the market so we understand its case to be that
18 See report by Price Metrics, the respondent’s economist, page 207.
7
the respondent is the only viable supplier for its business model. The
respondent, whilst conceding the market share figures are hard to ascertain,
contends that the market is one for all bulbs and includes imports among its
competitors. The scope of this debate involves more botany than economics
and we can spare the reader the full details now. 19 Briefly, however, dry
bulbs are bulbs produced for growing in a garden. The other type of bulb is a
bulb used for growing what will become cut flowers. Lest the horticulturally
challenged assume that a bulb is still a bulb by any other name and hence
interchangeable, the applicant’s expert Keith Kirsten contends that they are
not. Bulbs to be used for cut flowers are “programmed” and “forced” during
their growth period to force or delay flowering and are not suitable for garden
use. The respondent has both a horticultural, and hence it follows, an
economic disagreement on this point. It contends that cut flower bulbs can be
sold to the domestic consumer and hence should be included in the relevant
market. Once that exercise is done, its economist concludes that it would
have no more than [2025 %CONFIDENTIAL] of the relevant market. 20 Note
that this is a derived figure and involves making certain assumptions about
land under cultivation and the respondent’s share of exports and imports.
[29] It will be seen from this very brief summary of the differences that a
definitive answer to the market definition is not a task for a panel considering
interim relief. For the benefit of the applicant, we will assume that this is an
upstream market in which the respondent is a dominant firm expressed
differently, that it has some market power in relation to the applicant. We will
assume as well that, given the niche nature of the bulb fundraising scheme,
the respondent is the only viable supplier that can supply the amounts needed
the respondent is the only viable supplier that can supply the amounts needed
on a national basis to retain the applicant at its present level of supply to
schools.
[30] The applicant defines the downstream market as the market for the
sale of dry bulbs to school fundraising schemes. The respondent denies that
this is a market at all and suggests that if schools are in some fundraising
market, then all forms of fundraising schemes must be considered as viable
substitutes, not just those that involve the sale of bulbs. Again, we take no
view on this debate at this stage. As the downstream market is the one where
the anticompetitive effect is alleged to manifest itself, we will examine that
contention from the perspective of both contending downstream market
definitions. Prior to doing so, we consider the relief sought and the legal
standard to be applied
RELIEF SOUGHT
19 In its replying affidavit the applicant has included an affidavit by a horticulturist Keith Kirsten, This
was not left unanswered by the respondent which filed a supplementary affidavit inter alia engaging
with Kirsten on these matters
20 See record of main complaint application Price Metrics report May 2006, paragraph 64.
8
[31] The applicant seeks the following relief
(i) Ordering the respondent to supply the applicant with bulbs
on the same terms set out in the order of the Transvaal
Provincial Division of the High Court of South Africa dated 10
November 2005
(ii) Directing that the order set out in paragraph (i) above shall
remain in force for six months after the date of its issue or
until the matter before the Tribunal has been finally
determined, whichever is earlier
(iii) Ordering the respondent to pay costs of the application in so
far as it opposes it
REQUIREMENTS FOR INTERIM RELIEF
[32] The requirements for the interim relief are set out in section 49C (2)(b)
of the Act, which states that, the Competition Tribunal:
“…may grant an interim order if it is reasonable and just to do so
having regard to the following factors:
i) The evidence relating to the alleged prohibited practice;
ii) The need to prevent serious or irreparable damage to the
applicant; and
iii) The balance of convenience
[33] As the applicant correctly argues, in our case law, following the
practice of our civil courts in interdicts, these factors are considered as
interrelated and not individually decisive. In Natal Wholesale Chemists (Pty)
Ltd v Astra Pharmaceuticals Distributors (Pty)Ltd 21, we approved the
following dictum from Eriksen Motors ( Welkom)Ltd v Protea Motors
Warrenton22 a decision of the Appellate Division that the factors in the
section:
“…are not individually decisive, but are interrelated, for
example, the stronger the applicant’s prospect for success the
less the need to rely on prejudice to himself. Conversely, the
more the element of “some doubt”, the greater the need for the
other factors to favour him .” 23
[34] That notwithstanding, the applicant must at least make some showing
[34] That notwithstanding, the applicant must at least make some showing
that there is a prohibited practice. As we put it in Nuco Chrome (Pty) Ltd and
Xstrata South Africa (Pty) Ltd, Rand York Minerals 24:
21 [20012002]CPLR 363 CT
22 1973 3 SA 685 A
23 Tribunal Case Number: 98/IR/Dec00
24 Tribunal Case Number: 31/IR/Apr04 page 6
9
“ While the requirement of section 49(2)(b) that we have regard to the three
factors listed above suggests that a strong positive finding on two factors
might outweigh a lesser or possibly negative finding on the third, we would, as
we have observed elsewhere, be extremely reluctant to uphold an application
for interim relief in the absence of evidence confirming the restrictive practice
alleged.25
[35] In this case, for reasons that follow, we have found that the applicant
has failed to make out a case that a prohibited practice has occurred. For this
reason we do not need to consider the remaining factors set out in section
49C(2)(b)(ii) and (iii).
THE CONTRAVENTIONS
[36] The applicant initially alleged that the applicant had, inter alia,
contravened section 5(1) of the Act in that the respondent had an exclusive
supply arrangement with its competitor LFB. The respondent has denied that
this is an exclusive arrangement and this claim has not been persisted with.
26
[37] The applicant thus relies on three provisions of the Act. The first is the
complaint under section 8(c) of the Act, the second is under section 8(d)(ii)
and the third is under section 9 of the Act.
[38] Section 8(c), provides that it is a prohibited practice for a dominant firm
to
“ engage in an exclusionary act, other than an act listed in
paragraph (d), if the anticompetitive effect of that act
outweighs its technological, efficiency or other pro
competitive gain; or ”
[39] Section 8(d)(ii) states that it is a prohibited practice for a dominant firm
to –
“ engage in any of the following exclusionary acts, unless the firm
concerned can show technological, efficiency or other procompetitive
gains which outweigh the anticompetitive effect of its act
(ii) refusing to supply scarce goods to a competitor when
supplying those goods is economically feasible;”
supplying those goods is economically feasible;”
25 Tribunal Case Number: 31/IR/APR04 .
26 See applicants heads of argument paragraph 103 and counsel’s opening address record
page 7
10
[40] Section 9(1) states that:
“ An action by a dominant firm, as the seller of goods
or services is prohibited price discrimination, if
a) it is likely to have the effect of substantially preventing or
lessening competition ;
b) it relates to the sale, in equivalent transactions , of goods or
services of the like grade and quality to different purchasers;
and
c) it involves discriminating between those purchasers in terms
of
(i) the price charged for the goods or services;
(ii) any discount, allowance, rebate or credit given or
allowed in relation to the supply of goods or
services
(iii) the provision of services in respect of the goods or
services; or
(iv) payment for services provided in respect of the
goods or services.
(Our emphasis. We explain the reasons for this later when we analyse the
anticompetitive effect)
EVIDENCE OF A PROHIBITED PRACTICE
[41] Proof of an anticompetitive effect is a necessary element of all the
sections that the applicant alleges that the respondent has contravened as
can be seen from the underlined phrases in the Act we have cited. 27
[42] Accordingly, we have first approached the contraventions alleged
holistically, that is, we considered as a point of departure, whether an anti
competitive effect is established. Without some evidence, even if weak, of this
effect, the application must necessarily fail. Our conclusion is that the
applicant has failed to establish any evidence of an anticompetitive effect,
hence its case in respect of both the alleged section 8 contraventions fails.
[43] Because price discrimination may require a lower threshold for anti
competitive effect, we have considered whether the applicant would
nevertheless succeed if it could cross this hurdle. We find nevertheless that it
fails to make out a case for a further necessary element of that contravention.
We discuss this more fully below.
We discuss this more fully below.
27 We use “anticompetitive effect” as shorthand to include the phrase in section 9(1)(a)
“substantially preventing or lessening competition in the market”
11
ANTICOMPETITIVE EFFECT
[44] The applicant contends that the relevant market in which the anti
competitive effect is experienced is what it terms the market for the sale of dry
bulbs for fundraising. This market is downstream to the one for the supply of
bulbs – the market in which the respondent is alleged to be dominant. As we
understand the applicant's case, the respondent leverages its dominance in
the upstream market to exclude the applicant from the downstream market by
engaging in conduct ‘tantamount’ to a refusal to supply it. We use the word
‘tantamount’ advisedly because it is common cause that the respondent's
refusal to supply is not absolute. It is willing to supply the applicant on the
same terms as its wholesale customers, at least until the complaint referral
has been determined. It is not willing to supply it in terms of a previous
arrangement between the applicant and the respondent. The applicant alleges
that the terms of the wholesale arrangement are so unattractive for it that it
would be forced to exit the downstream market, and so the respondent’s
conduct amounts to a refusal to deal by a dominant firm. If the applicant is
excluded from the downstream market, this would have an adverse effect on
school fundraising, as, faced with less competition, those left in the market
would reduce the margin enjoyed by the school. Recall that in terms of both
the applicant and LFB’s business models it is the intermediary, not the school,
which sets the price at which learners sell to the public.
[45] The respondent says the downstream market is for school fundraising
schemes. If bulb schemes do not offer a school an attractive return it will
choose another form of fundraising, since the school’s object is to raise funds
not to engage in the business of bulb selling.
[46] The applicant seeks to counter this argument by contending that its
own experience, which it says is confirmed by its economists who have
own experience, which it says is confirmed by its economists who have
interviewed school staff, is that schools do not make their choice of fund
raising scheme by reference to the returns that they could enjoy in respect of
different types of fundraising scheme, but rather see bulb selling as a market
on its own. Expressed differently, schools will compare the virtues of various
bulbselling schemes, but they would not compare bulbselling, to say cake
sales or a fete, in their choice of scheme. This, argues the applicant, is proof
that the market must be the one for which it contends.
[47] The applicant relies crucially for this on the fact that when LFB entered
the market, it offered schools a margin of 40% instead of the 33% traditionally
offered by the applicant. When it became clear that the applicant might be
excluded from the market, LFB dropped the schools margin back to 33%.
This, says the applicant, is clear evidence that other school fundraising
schemes do not constrain the bulb schemes and that such schemes are not in
the same market.
12
[48] It is not necessary for us to decide which of these two downstream
market definitions is the correct one. Instead of making this determination we
will analyse whether an anticompetitive effect is established on either version
of the downstream market.
ANTICOMPETITIVE EFFECTS RESPONDENT’S MARKET DEFINITION
[49] Recall that the respondent argues that the downstream market is the
market for school fundraising schemes. On this version, if the flower bulb
scheme is unattractive for schools they can substitute it with an infinite variety
of other schemes whose object it is to raise funds for schools. The respondent
does not attempt to describe the extent of this market. Presumably, this is
because there is an infinite variety of schemes in which schools can engage.
The important point here is that if this is the correct downstream market, then
the exit of the applicant would have no anticompetitive effect, as schools can
substitute with other schemes, whether they involve the selling of flower bulbs
or some other goods, be it cakes or hamsters. The respondent notes that
schools do not consistently engage in the flower selling scheme and that
annual repeats by the same schools of the bulb scheme are less frequent
than schools who take a break from the scheme for a year or so. We do not
need to decide what the other schemes are; it would seem clear that schools
that use the bulb scheme substitute with other schemes on a frequent basis. If
the market is therefore one for fundraising, the flower bulb segment is not
sufficiently ubiquitous that the exit of even a major player such as the Bulb
Man would have an anticompetitive effect. The applicant does not appear to
contest the notion that there would be no anticompetitive effect if the market
is for fundraising. Rather, it focussed its efforts on contending for a narrower
market. Therefore, if the respondent has defined the market correctly then the
market. Therefore, if the respondent has defined the market correctly then the
exit of the applicant would have no significant anticompetitive effects as
schools, as buyers of fundraising services, would have other choices.
ANTICOMPETITIVE EFFECTS APPLICANT’S MARKET DEFINITION
[50] It is not the applicant’s case that the ordinary consumer of bulbs
experiences the anticompetitive effect – it concedes that consumers can
purchase bulbs from nurseries and other retailers. 28 Indeed, on the
respondent’s version, these sell to consumers at lower prices than they could
purchase through any schoolbased scheme. The reason school schemes
survive at all according to the respondent is because consumers purchasing
choices are sentimental not rational. 29
[51] The applicant’s case is that the anticompetitive effect is felt in the
28 See page 8 of the transcript of the hearing.
29 See page 123 paragraph 35.2 “ This reflects the fact that the former purchases [school fund
raising sales] are driven by school fundraising considerations rather than by rational market
driven considerations .”
13
downstream market for fundraising for schools through the sale of bulbs. The
major players at present in this market are the applicant and LFB. There is
some disagreement as to who else may be in this market. The respondent
appears to supply some schools directly, who it would appear, purchase from
it in the same way as other wholesale customers would. There is also
reference to the presence of other intermediaries – but their significance as
players is disputed. 30 The applicant suggests that the others are bit part
players, whilst the respondent, not surprisingly, sings their praises. This is not
something we can resolve on these papers. We will assume in the applicant’s
favour that if it were to exit, then the market would comprise LFB and those
schools that can order directly from the respondent.
[52] The question then is whether it is clear that this will have an anti
competitive effect on the schools. In respect of those provisions for which
dominance is a prerequisite, it would not seem that the exit of the applicant,
enhances or establishes the market power of the respondent. What is lacking
in the applicant’s case is a credible theory of competition harm. If the
respondent is intent on excluding the applicant from the downstream market
what anticompetitive object does it have for doing so? The respondent is not
its competitor in these markets – it is thus not apparently motivated, as many
dominant firms who compete with their customers are, to exclude them, take
their customers or squeeze their margins. On the present papers, the
respondent's margins to intermediaries, and its pricing power to consumers,
remain the same whether the applicant is in this market or not. Indeed, if
anything the respondent is worse off without the applicant, if it is as effective
an intermediary it contends it is.
[53] The only evidence of competition harm the applicant advances is to
[53] The only evidence of competition harm the applicant advances is to
drawn an inference from a change in margins offered by LFB. In 2005 when it
entered the market, LFB circulated a pamphlet advising schools that if they
signed up with it they would get a 40% margin. Up until then, the applicant
had only offered schools a 33.3% margin. In 2006, however, LFB has reverted
to the 33,3% margin offered to schools, at least so it appears from a pamphlet
circulated to schools this year. 31 The applicant draws from this the conclusion
that when there was competition in the market for schools in 2005, they
benefited by getting higher margins. At this stage LFB knew the applicant was
its competitor and needed to win market share from it. In 2006, presumably
because it knew that the applicant’s supply arrangement with the respondent
was to end in June, LFB could confidently withdraw the additional margin
offered to the schools in 2005, without fear of an adverse competitive
response from the applicant.
[54] This does not help the applicant in its case against the respondent. At
best it suggests that LFB, not the respondent, is the beneficiary of the
30 Names of the firms are confidential
31 See record page 25 and exhibits JAG 5 and JAG 12 to the founding affidavit of Jackie Ginsberg.
14
applicant's exclusion, as there is no evidence on the papers that the margin
foregone and then retained by LFB, was passed on to the respondent. Nor is
the LFB margin move on its own significant. This may have been no more
than an extravagant marketing promotion from a new entrant, since, on the
applicant’s papers, LFB only entered the market in mid 2005. An aggressive
publicity promotion, in the absence of evidence of actual harm, is not sufficient
to found a case for an anticompetitive effect. If it was, every sale offer once
granted and subsequently withdrawn, might lead to inferences of the insidious
presence of monopoly. In the face of this margin switch by LFB, the
respondent’s market position remains constant. If it has market power, the
absence or presence of the applicant in the downstream market makes no
difference to it.
[55] The crucial question here is : does the respondent’s refusal to deal
advantage its alleged market power? In York Timbers Ltd v SA Forestry
Company Ltd, we cited the following page from Areeda and Hovenkamp’s
treatise, which we suggest is apposite:
“An arbitrary refusal to deal by a monopolist cannot be unlawful unless it
extends, preserves or creates, or threatens to create significant market power in
some market, which could be either the primary market in which the
monopoly firms sells or a vertically related or even collateral market. Refusals
that do not accomplish at least one of these results do not violate section 2 (of
the Sherman Act), no matter how much they might harm the person or class of
persons declined service. Nor are such refusals an ‘abuse’ of monopoly power
in the sense of using power in one market as ‘leverage’ to increase one’s
advantage in another market.” 32
[56] We can look at the anticompetitive effect from another perspective.
[56] We can look at the anticompetitive effect from another perspective.
Why is the dominant firm refusing to deal? 33 As the authorities show, even
dominant firms are entitled to refuse to deal. However, if the dominant firm
lacked a proper explanation for its conduct, this might shift the probabilities in
favour of the applicant.
Faul and Nickpay observe in relation to European jurisprudence that:
“A refusal to deal by a dominant undertaking will not be considered an
abuse under Article 82 of the EC Treaty if it is objectively justified. This
will be the case if the refusal can be justified on business grounds
other than the intention to eliminate a competitor from the market.” 34
[57] In this case, the respondent advances what on the papers appears to
32 See [20012002] CPLR 408 (CT)
33 Note that we have not found the respondent dominant w e simply make this favourable
assumption to the applicant, to test the other aspects of its case.
34 See Faull and Nickpay The EC Law of Competition 3.156
15
be a legitimate business justification for its conduct.
[58] When the respondent gave notice of its intention to terminate its
agency agreement with the applicant it justified doing so on the basis that
their relationship had become strained as a result of the dispute between the
applicant and LFB which had culminated in High Court litigation in which the
respondent had become embroiled. 35
[59] It is common cause that the Ginsberg family have been in the business
of selling flower bulbs for fundraising schemes for some years and that
various members of the family have been involved in some capacity over
time. Historically, they were all involved, trading as the Bulb Man until 2005,
when, for reasons not clear to us, the joint family enterprise fractured into two
rival businesses, with the erstwhile founder of the applicant establishing LFB,
in competition with his son and daughterinlaw who remained with the
applicant. The respondent too appears to be a business where family
relationships figure large. The chairman and chief executive; Floris and Stuart
Barnhoorn are father and son. In the letters exchanged around the agency
debate in 1996, which we referred to earlier, both offspring Ginsberg and
Barnhoorn are the protagonists and refer to understandings they had with the
respective fathers. 36 Unsurprisingly, in business relationships where personal
contact seems of great importance, the Ginsberg fallout was to have
consequences for the respondent, which appears to be considered not merely
the preferred, but only possible supplier, by each of these warring firms. That
the respondent appears to have chosen sides in this dispute seems probable
from the papers, as is the observation that once it had chosen sides, the loser
would inevitably have to pay the commercial price in a business environment
would inevitably have to pay the commercial price in a business environment
where a supplier relationship is a prerequisite to success. It is not for us to
presume on the rights and wrongs of this dispute or whether the respondent
could have avoided becoming embroiled. The fact is that it did, and having
done so, relationships, as the correspondence and the history suggest,
reached a most unhappy state.
[60] In the context of this dispute the respondent’s conduct in refusing to
supply the applicant in terms of an agency arrangement amount not to the
actions of a dominant firm flexing its anticompetitive muscle, but a supplier
divorcing its customer based on irretrievable breakdown. 37 It must be borne in
mind that the agency relationship, unlike the wholesale one, calls for repeated
contacts between the supplier and the agent – a breakdown in trust between
the two is far more problematic to manage in this arrangement than in
35 See the letter from the respondent’s attorneys on page 53 of the record.
36 See letter from Stuart Barnhoon to Jackie Ginsberg page 470, “This entire arrangement was
agreed to three years ago with Len” and letter from Lee Ginsberg to Stuart Barnhoorn page
471, “ The following issues were discussed and agreed at this meeting between Floris, Len
and myself.”
37 See letters emanating from the respondent’s attorney, record page 56 and 90 and 104.
16
wholesale relationship characterised as it is by its impersonal nature and
minimal contact.
[61] On the present papers, the more probable version for the respondent's
conduct in refusing to supply the applicant on an agency basis is the
breakdown in the business relationship, as opposed to an attempt to wield
market power or to exclude the applicant for an anticompetitive end.
[62] Without any evidence that this conduct leads to an enhancement of the
respondent’s market power on the present papers, we cannot see the
likelihood of an anticompetitive effect, even making the most favourable
assumptions for the applicant, on all the other facts. The applicant has
disputed the rights and wrongs of many of the incidents giving rise to the
present parlous state of the business relationship but it does not dispute the
fact of its decline. Nor does the applicant place before us an anticompetitive
motive for the conduct. At best, it suggests the dispute will lead to its exit and
hence an anticompetitive outcome, but it does not explain why the refusal
itself furthers some anticompetitive design. There is no reason then for us not
to accept the respondent’s justification.
[63] It follows, that if on either version of the market there is no anti
competitive effect, assuming that the applicant is excluded from the market,
the application must fail. We examine briefly if this conclusion holds for the
applicant’s case in respect of the remaining contravention, namely section
9(1).
SECTION 9(1)
[64] The applicant’s case under section 9(1) is that the respondent is a
dominant firm and is supplying the applicant's competitor on one set of terms
which it refuses to provide to the applicant. The terms offered to the applicant,
it alleges, are discriminatory as to both service and credit.
it alleges, are discriminatory as to both service and credit.
[65] As we indicated earlier, to found a section 9(1) claim, the applicant
must still establish that the price discrimination has an anticompetitive
effect.38
[66] In Sasol Oil (Pty) Ltd and Nationwide Poles CC the Competition Appeal
Court held that the test for competitive harm in terms of section 9(1)(a) is to
ask is there a reasonable possibility that competition may be adversely
affected by a practice. 39
[67] The CAC test is a more stringent test than the one adopted by this
38 This is because of section 9(1)(a) which states an action by a dominant firm is prohibited price
discrimination if “ it is likely to have the effect of substantially lessening or preventing competition.
39 See page 28 of this Competition Appeal Court decision: Case Number 49/CC/Apr05
17
tribunal.40 However, it is not clear, as the CAC did not need to pronounce
upon this point, whether, because of a slight difference in the language of the
sections, that test is less demanding than the test for section 8.
[68] It would not be appropriate in an interim relief case to explore that
issue further and for that reason we will, out of caution, assume that it is a
lesser test, and that the applicant might have met that threshold under section
9(1) although, not under section 8.
[69] This means we must consider the respondent’s other defence to the
section 9 (1) claim, and, that is, that the applicant fails to make out the next
element of the section 9 claim, which requires that transactions are
equivalent.
[70] The socalled agency agreement is one that involves several different
elements that distinguish it from a wholesale agreement. Thus even if we
accept that the latter is less favourable than the agency agreement they
involve sufficiently different commercial terms, logistics and interaction
between supplier, customer and third parties that justifies them not being
considered equivalent. Most significantly, the commercial risk of a scheme,
which requires someone to bet on whether the number of orders will
correspond to the available supply of product, shifts depending on the choice
of scheme – this factor alone negates their equivalence. The problem for the
applicant is that it has sought to distinguish the commercial imperative of the
two arrangements to justify why it cannot survive under a wholesale
arrangement, but in so doing, it has succeeded in emphasising not only their
difference but also their nonequivalence.
[71] As we stated in Nationwide Poles 41 on the interpretation of this
subsection of section 9:
“Thus transactions may be functionally equal one
subsection of section 9:
“Thus transactions may be functionally equal one
business class seat or one telephone call between Cape
Town and Johannesburg may be functionally equal to
another business class seat or telephone call, but they
may not be equivalent (a call or a flight made in peak
time as opposed to one made during a non peak period)
in the sense that their economic effect is different and
hence the legislature, recognising this, chose to bring
‘nonequivalent’ transactions under the rubric of
prohibited price discrimination despite the fact that the in
other respects they may be regarded as equal”
40 The tribunal’s test is that the complainant must situate the complaint as being one relevant
to competition but does not require proof of some standard of harm
41 See Tribunal case number: 72/CR/Dec03 on page 34 paragraph 132
18
[72] The applicant fails in its claim in terms of this section as well.
CONCLUSION
[73] Since the applicant has failed to establish that the conduct of the
respondent has an anticompetitive effect, it fails to make out a case that
there is evidence of an alleged prohibited practice. Whilst we have observed
in the past that the factors we take into account for a section 49C claim are
interrelated and that a weakness in one respect may be compensated for by
strength in others, this case fails on its most important constituent element,
namely, the competitive effect of the conduct of the respondent. It is not a
case of there being some evidence albeit weak – it is a case of there being no
evidence at all. Even granting the applicant the benefit of many assumptions
on the facts, it fails to rise to a case about competition harm, albeit it may be
about commercial harm.
COSTS
[74] As with our normal practice costs follow cause and there is no reason
on the facts of this case to depart from that practice. The respondent is
entitled to its costs.
ORDER
[75] We make the following order:
1. That the application be dismissed; and
2. That the applicant pays the respondent’s costs in the application on a
party and party scale, and such costs are to include costs of one
counsel.42
__________________ 28 November 2006
N Manoim Date
Tribunal Member
Concurring: D Lewis and M. T.K. Moerane
42 Note that the order was made earlier on 14 November 2006.
19
Tribunal Researcher : J Ngobeni
For the Applicants : Mark Wesley
Instructed by Jowell Glyn & Marais
For the Respondents : Jerome Wilson
Instructed by Edward Nathan Sonnenberg
20