THE COMPETITION TRIBUNAL
REPUBLIC OF SOUTH AFRICA
CASE
NO: 72/CR/Dec03
In the matter between:
Nationwide
Poles
Complainant
And
Sasol (Oil)
Pty Ltd
Respondent
DECISION AND ORDER
Background
1.
The
complainant, Nationwide Poles CC (‘Nationwide’), is a small
producer of treated
wooden poles based in the Eastern
Cape province. It procures
supplies of
untreated pine poles from the sawmills and then impregnates the poles
with a
wood preservative. In the case of
Nationwide the preservative employed is creosote, or, to be more
specific, SAK
K, the brand name of a wax-additive creosote produced by Sasol.
Although the
Nationwide plant is based in the Eastern Cape the bulk of its
customers are vineyards in the Western Cape.
2.
The
respondent, Sasol Oil (Pty) Ltd (‘Sasol’),
a major subsidiary of the Sasol
group of companies, is
responsible for the marketing of Sasol's liquid fuels and lubricants.
The
process of producing synthetic fuel releases a tar by-product which is
then
utilised as the feedstock for the production of a range of other
products
manufactured through Sasol’s carbo-tar division. The carbo-tar
division
comprises a number of business units corresponding to the range of
products
generated from the tar feedstock these being wood preservatives, DIY
and black
disinfectants and surface coatings. The wood
preservative, creosote, produced by Sasol is utilised by its customers
for a
range of different uses including the treatment of poles.
3.
Nationwide Poles was
acquired by Mr. Jim Foot on the 31st May 2002 at
a time when its business was ailing. Mr
Foot brings this complaint on behalf of Nationwide Poles.
4.
In about August 2002 Foot became aware that Sasol was
charging him a higher price for his purchases of creosote than that
charged to
his competitors. Foot approached Sasol for a price list which, after
some
apparent reluctance on Sasol’s part, was furnished. The price
list confirmed
that the price charged by Sasol for creosote supplied to Nationwide was
notably
higher than that levied on Woodline, a very large pole manufacturer in
the
Eastern and Western Cape and
Nationwide’s most
important competitor. It is, indeed,
common cause that Sasol’s price schedule for the sale of creosote
allows for
discounts based on purchase volumes, with its largest customers
receiving the
most preferred prices while its smallest customers, of whom the
complainant is
one, charged the highest price on Sasol’s price schedule.
5.
On 30 April 2003
Nationwide lodged a complaint against the respondent with the
Competition Commission. It alleged
contravention of sections 4(1)(b) and 9(1) of the Competition Act
(‘the Act‘).
It received a Notice of Non-referral from the Commission on the 12 November 2003. Nationwide then elected to
approach the Tribunal directly. In the
present proceedings Nationwide is only pursuing a claim in terms of
Section 9
of the Act, the section that proscribes ‘prohibited price
discrimination’. In
essence, Nationwide alleges that the
discount structure utilised in the pricing of Sasol’s wood
preservative,
creosote, meets the test of prohibited price discrimination and it
requests
that the Tribunal makes a finding to this effect. Nationwide
also asks the Tribunal to order
Sasol to supply it on the same price terms as those available to its
competitors.
Section 9 of
the Competition
Act
6.
Section 9 provides:
(iv)
a sale in
good faith in
discontinuance of business in the goods or services concerned.
7.
Section
9 is contained in Part B of Chapter 2 of the Act, that part dealing
with abuse
of a dominant position. Hence in order
to sustain an allegation of price discrimination the complainant must
first
establish that the respondent is indeed dominant in the relevant market. Having discharged this onus, it is then for
the complainant to establish that the price differentiation complained
of is
indeed ‘prohibited price discrimination’.
The Act requires that in order to make this finding we
must be
persuaded, firstly, that the complained of action ‘is likely to
have the effect
of substantially preventing or lessening competition’. Secondly, we must be satisfied that the
transactions in respect of which price discrimination is alleged are
‘equivalent’ transactions. Thirdly,
the
discriminatory action in question must relate to price, discounts
provided,
services provided or to payment for those services.
8. If
the complainant succeeds in establishing that a dominant firm is indeed
engaging in ‘prohibited price discrimination’ in terms of
Section 9(1), then
Section 9(2) entitles the respondent to mount a defence.
In order to mount a successful defence the
respondent must show that the differentiation in question either
reflects cost
differences that attach to the transaction in question, or that the
differential is an act, in good faith, to meet competition, or is
dictated by
particular market conditions such as obsolescence of the good or
service in
question or the discontinuance of the business.
9.
We will first describe the price
structure for Sasol’s creosote.
Sasol’s
creosote: a description of the price structure
10. Sasol
sets the prices of its products by way of a price list. Sasol reviews
or
adjusts its pricing annually, usually on about the 1st of
July of
each year.
11.
Much evidence was lead about the transparency of the
published
price list. Sasol claims that its price list is made available to all
its
customers in order to allow its customers easy access. However, the
evidence
that the price list is marked “For Internal Use only” would
indicate otherwise.
It appears also from the evidence that Foot only acquired the pricelist
after
asking specifically for it in relation to the Commission’s
investigation.
12.
Sasol
estimates how much creosote the customer is going to buy over the next
year, based on previous months’
sales. This is evaluated
every three months based on purchases
actually
made, that is, Sasol determines how much a customer has ordered in the
previous
three months and then determines a price in terms of its pricing
structure by
reference to the purchasing volumes of those three months.
The three-month purchasing pattern is annualised and Sasol then
evaluates into
which category the customer falls on its price list.
This is then used to determine a price that
will be applicable for the next three months. In other words, every
three
months Sasol sets the customer’s
future price
for three months at least and then re-evaluates the price. Therefore,
Sasol
argues, the price list dictates what price that customer will be
charged for
prospective purchases. Once reaching a
particular price threshold, the customer will continue to get the lower
price
for at least three months.
13.
Sasol
has a range of customers classified as small, medium and large
depending on the
size of their orders in a given month.
According to Mr. Van Wyk, it had about 5 large customers in 2004. Sasol
contends that, although it has several large customers, the spread of
its
customers across size categories is even, with no single customer
having
achieved the maximum price break that is reflected in the pricing
structure. The evidence in Exhibit 1, Creosote
Monthly Sales
Volumes at 2004, would appear to bear
out that
testimony. However the complainant’s
evidence reflects that between 2001 and 2003, Sasol had at least one
customer
in the maximum price category.
Sasol’s Creosote Sales Volumes as at February 2004
|
Category
|
No. of Customers
|
Volume (tons/annum)
|
Percentage of uptake from Sasol accounted for by customers
|
|
1
|
4
|
0-450
|
5%
|
|
2
|
3
|
451-1000
|
13%
|
|
3
|
6
|
1001-2500
|
60%
|
|
4
|
1
|
2501-3600
|
22%
|
|
5
|
|
3601-5500
|
0
|
|
6
|
|
More than 5500
|
0
|
Source: Table
derived from Exhibit 1 , Sasol’s figures as at February 2004
14.
In
other words:
Customers
representing 5% uptake from Sasol fall into category 1.
Customers
representing 13% uptake from Sasol fall into category 2.
Customers
representing 60% uptake from Sasol fall into category 3.
Customers
representing 22% uptake from Sasol fall into category 4.
15.
What
this illustrates is that, based on Sasol’s own figures,
Sasol’s large customers
(those that buy more than 1000 tons per annum) account for more than
80% uptake
of creosote while its small customers account for only 5% of uptake.
Mid-size customers account for 13% of Sasol creosote. This accords
generally
with Mr. Foot’s analysis based on data on page 9 of his
supplementary info
bundle (“CSI”).
He
states that in 2003, 5 plants (or customers) accounted for 81% of
market (by
volume). Similarly, in 2004, 5 plants accounted for 87% of the market
(by
volume). In the above analysis based on Sasol’s own figures, 7
plants accounted
for about 82% in 2004.
16.
On Nationwide’s analysis of creosote
uptake per customer, large firms fall between categories 3-6.
The split between large and small customers is therefore,
approximately, 80/20.
What seems clear from these contentions and the table above is that the
smaller
firms account for a relatively small proportion of Sasol’s total
revenue from
creosote.
17. Genesis,
the firm of consultants retained
by Sasol in this matter, reproduced a table reflecting Sasol
carbo-tar’s price
schedules for creosote between 2003 and 2004. Genesis computed the
cumulative
percentage difference between price bands or categories, which are
reflected in
columns 4 and 5:
18. According
to the Genesis figures, the larger customers enjoyed discounts of
between 10%
and 15% relative to the base price charged to small customers between
2003 and
2004.
The Genesis table also reveals that the difference between the prices
received
by the firms in the first category and those in the last category of
Sasol’s
discount structure, who received the maximum discounts in 2004, is
approximately 15%. The complainant’s figures go further than this
and suggest
that between 2001 and 2003 the differential between the highest and
lowest
price categories ranges from 26% (2001) to 16% (2003). On both sets of figures there is a
significant difference between the lowest and highest levels of price
categories.
19. Evidence
was led that in 2001 Sasol announced to its customers that in the
impending
several years it would switch from utilising the tar feedstock
generated by the
production process at its Sasolburg plant to the feedstock generated by
its
Secunda plant. For various reasons which
are elaborated below, this resulted in a major shift in the way in
which Sasol
priced the creosote feedstock, that is, the tar stream – the
price for the
feedstock charged by Sasol to its carbo-tar unit increased from R350
per ton to
R1800 per ton. This was to be passed on
to Sasol’s customers. Sasol’s
carbo-tar
unit claims to have offered its customers the choice of absorbing a
very
significant price increase once the change in the feedstock had been
effected,
this being some years down the line, or of phasing in the increase over
a
number of years. In Mr. Van Wyk’s
words:
“So the
alternatives, we put them all on
the table. We did our own calculations as well. And we asked the
industry would
they prefer us keeping the price on the normal PPI increases and so on
and then
after 4 years you give them a 500% or whatever increase. Or do you want
us
gradually increasing it to a point where we believe we can start doing
normal
business again. And that happened in the past 3 years. And what they
told us is
they prefer us phasing it in, because it won’t give them a total
shock, because
on their contracts as well they won’t be able to negotiate their
contracts on a
big increase. They can maybe incorporate it in a phased manner.”
20.
The customers consulted apparently
preferred the latter option.
What
we have to determine
21.
We
will examine each of the constituent elements of Section 9. We will commence the analysis by identifying
the relevant market. As commonly occurs in anti-trust litigation, this
requires
us to decide major factual and analytical disputes between the parties.
Nationwide prefers a narrow relevant market – indeed it argues
that the
relevant market is that for the product named SAK K, a particular
wax-additive
creosote produced by Sasol alone. On
this version of the relevant market Sasol is a monopolist.
Sasol, for its part, contends for a
significantly wider market. It insists
that the market is that for wood preservatives, and that this market,
far from
being confined to SAK K, includes not only all creosote but also a
product
called CCA or ‘copper chrome arsenate’, a product that,
alleges Sasol, is
directly substitutable for creosote. This
involves an examination of certain of the technical
characteristics
of the products concerned. Having determined the relevant market we
then ask
whether, in that market, the respondent, Sasol, meets the Act’s
definition of a
dominant firm.
22.
Because,
as we elaborate below, we do find that the respondent is indeed a
dominant
firm, we then go on to ask whether the complainant has successfully
established
that the practice in question conforms to the elements of prohibited
price
discrimination provided for in Sections 9(1)(a), (b) and (c). Sasol has made much of the proper
interpretation of Section 9(1)(a), in particular the nature and extent
of the
evidential burden that the complainant has to discharge to show that
the price
discrimination is ‘likely to have the effect of substantially
preventing or
lessening competition’.
23.
Because we do conclude that Sasol is engaged in the
practice
of prohibited price discrimination, we then proceed to examine whether
or not
the respondent has successfully invoked the defences provided for in
Section
9(2).
24.
The
matter was heard on the 4-6 August , 22nd, 23rd,
31st
August and 1st December
2004. The following witnesses testified:
25. For the
complainant
26.
For
the respondent
The
relevant
market
27.
Three
possible relevant product markets have been proposed.
As already indicated, Nationwide has proposed
that wax-additive creosote be considered the relevant market,
alternatively
creosote. Sasol is the only manufacturer
of wax-additive creosote in South Africa, the relevant geographic
market. There is only one other producer
of creosote
in South Africa, this being Suprachem/ICC, part of the large iron and
steel
producer, ISPAT/ISCOR (Iscor), now named Mittal Steel.
Suprachem distils and
refines crude tar, which is a
by-product of Iscor’s coke production, and manufactures and
markets coke, tar
and related by-products. The company is engaged in the distilling of
tar and
crude benzol into 40 different industrial chemicals.
28.
Sasol,
for its part, insists that the relevant market is that for wood
preservatives. This market, avers Sasol,
essentially comprises two substitutable products, creosote and
copper-chrome-arsenate or CCA. There are
other products employed as wood preservatives but their share is
marginal and
does not affect our conclusions. If CCA
is part of the relevant market then it is common cause that
Sasol’s share would fall below 35%
and, in order to
establish dominance, the complainant would have to prove market power.
29.
We do
not accept the narrower of the market definitions proposed by
Nationwide. This
is the market for wax-additive creosote. It
appears that there is only one such product, that being
SAK K, which
is produced by Sasol. Accepting this
definition of the relevant market would effectively render Sasol a
monopolist
in the market in question. While we have no reason to doubt Mr.
Foot’s stated
preference for SAK K or even the superior quality of his preferred
product –
indeed it seems reasonably clear that the wax additive confers certain
advantages on SAK K - we have not been presented with any evidence that
suggests that it cannot be relatively easily substituted by other
creosote
products or that the addition of a wax additive is beyond the capacity
of
creosote users like Nationwide.
30.
However,
Nationwide is on considerably firmer ground when it argues for a
creosote
market in opposition to Sasol’s insistence that the market be
defined as that
for wood preservatives. As noted,
Sasol’s broader definition would incorporate the second product,
CCA, into the
relevant market. We must then consider
the substitutability of CCA for creosote.
31.
It is
instructive to note at the outset that Sasol did not initially argue
for the
substitutability of CCA for creosote. There
is no mention of CCA in the Commission’s
notice of non-referral.
When this omission was put to Mr. van
Wyk, the Sasol executive who testified at the hearing and who had, in
the
relevant period, headed the Sasol division producing creosote, he could
not
offer an explanation short of insisting that the Commission had been
provided
with a full exposition of the market. More telling is the omission of any reference
to CCA in Sasol’s answering affidavit filed in the present
proceedings. Again van Wyk could offer no
explanation for
this omission. Indeed the first mention of CCA is made in Nationwide’s
replying affidavit.
However the existence of CCA loomed large in the oral evidence
presented by
Sasol’s witnesses at the hearing. In
this belated fashion, the substitutability of CCA for creosote, by
providing
the basis for Sasol’s denial of dominance, emerged as one of the
two main
pillars of Sasol’s defence, the other being Sasol’s
insistence that its
opponent had not established that the complained of price
differentiation had
compromised competition.
32.
Given
the extent to which the alleged substitutability of CCA and creosote
has
subsequently been relied upon by Sasol, its failure even to make
mention of CCA
in its initial pleadings is nothing short of startling.
It certainly tends to support the inferences
sought to be drawn by Mr. Foot from persistent reference in
Sasol’s internal
documents to a ‘creosote market’ as well as to the marked
absence of reference
to CCA in these documents. While
ordinarily we are prepared to accept that the term ‘market’
is frequently used
in everyday commerce in a manner that is not intended to identify a
relevant
market for anti-trust purposes, the fact is that even for
anti-trust
purposes the respondent appears to have decided only belatedly to
incorporate CCA into its own definition of the relevant market. On the other hand, Foot’s testimony
denying
the substitutability of CCA for creosote was consistent with his
earlier filed
affidavits and, in this important matter, certainly, he emerges as a
significantly
more reliable witness than van Wyk.
33.
The
relevant South African Bureau of Standards (SABS) regulations stipulate
the use
of either CCA and creosote for preservation of the wood of products in
contact
with the ground. Vineyard poles have to comply with the H4
SABS specification, which will therefore be fulfilled by the use of
either CCA
or creosote. As far as creosote is concerned, the standard does not
differentiate between SAK 100 and SAK K.
34.
However,
it appears that, notwithstanding the SABS regulation, the actual degree
of
substitutability between creosote and CCA is largely dependent upon the
intended end use of the wood product that is subject to the treatment. It is clear that the possibility of
substituting CCA-treated poles for creosote-treated poles for use in
telephone
or electricity transmission is highly limited – the former are
unable to
withstand veldt fires as successfully as the latter and this is the
major
reason for the well-nigh exclusive use of creosote treated poles by
these
important consumers.
It was suggested that there is some recent evidence of CCA-treated
poles being
used in these applications, but it appears to be common cause that this
remains
limited and that the purchasers of poles for these uses will continue
to insist
on creosote-treated poles.
35.
Sasol
has made rather more of the fact that the complainant does not produce
poles
for use in telephone and electricity transmission, and, hence, that the
lack of
substitutability of CCA for creosote in this use has no bearing on the
selection
of the relevant market. We reject this
argument. This appears to be the largest segment of the poles market
and we
have little doubt that any pole manufacturer wishing to expand its
business
would want to bid for a slice of this market segment.
Nationwide avers that the reason why the
electricity transmission and telephone poles market is effectively
reserved for
the larger pole manufacturers is because the wood suppliers refuse to
provide
the complainant and other smaller pole manufacturers with the wood
input that
would allow them to produce poles for these purposes.
Leaving aside this limitation – itself a
prima facie contravention of the Competition Act
– there seems to be no reason why Nationwide or any other pole
manufacturer
would not wish to contest this important market segment and, should
this
happen, there would be no effective substitute for creosote in the
treatment
process.
36.
Creosote-treated
poles have also been favoured for use in vineyards, the market segment
in which
Nationwide is active. It appears that
the reason why creosote- treated poles have historically been favoured
in this
segment is because the superior moisture retention capacity of creosote
poles
renders them less brittle than CCA- treated poles and so better able to
withstand
the pressure exerted by the mechanical grape-harvesting process. However Sasol avers that this consideration
–
and hence the non-substitutability of CCA for creosote in this
application –
only applies to the limited number of vineyard poles that are at the
end of the
line and that must accordingly bear most of the pressure of mechanical
harvesting. Moreover, insists Sasol,
technological developments have enhanced the moisture retention
capacity of
CCA- treated poles, rendering them less brittle and more suitable for
vineyard
use. Sasol avers that major wine-producing vineyards have switched from
creosote to CCA-treated poles.
37.
However, issues related to the toxicity of the respective
products appear to resolve this debate in favour of the narrower
definition of
the relevant market. Although both
creosote and CCA clearly have toxic qualities, it appears that relevant
EU
regulations are moving decisively in the direction of prohibiting the
importation of wines from vineyards that utilise CCA-treated poles. Several witnesses insisted that this was a
purely protectionist measure, that, in other words, CCA-treated poles
had no
substantive impact on the safety of the vineyard’s product and
that the
regulation prohibiting this wood preservative was cynically designed as
a
protectionist measure. This is certainly the view of Mr. Angus Currie,
the head
of the South African Wood Preservers Association (“SAWPA”)
who testified at the
hearing, but he nevertheless conceded that continued use of CCA-treated
poles
in vineyards are likely to be used as an environmental barrier to the
entry of
South African wines into export markets. He
referred to the case of the Nederberg estate which, he
averred, was
told not to use CCA-treated timber any longer. Another of Sasol’s witnesses, Mr Stears,
from
South African Timber Auditing Services, while insisting that the issue
of CCA
toxicity was based on ‘emotional issues’ conceded that
CCA-treated poles were
likely to be phased out of use in South African vineyards within the
next six
to eight years.
38.
It appears that CCA’s toxic qualities are an issue
in other
areas of treated pole usage as well. Ms.Tammy
Bruno of Botar Enterprises, who also testified at
the hearing,
averred that the World Bank has refused to fund projects that use
CCA-treated
transmission poles because of the arsenic content of the preservative,
a
requirement that had effectively precluded CCA poles from being used in
Zambia.
39.
Certainly
it would be wholly unreasonable to expect a producer in the position of
Nationwide to incur any cost of switching from the use of creosote to
CCA if it
is accepted in the segments of the market that serve the telephone and
electricity providers and also the agricultural sector that
creosote-treated
poles are, for one reason or another, the preferred product,
particularly when
it appears certain that regulatory requirements will protect and extend
creosote use in the immediate future.
40.
We should add here that we heard lengthy submissions
concerning the cost of switching between CCA and creosote in the pole
treatment
process. In essence Nationwide insisted that because it operated a
creosote
treatment plant, the fact that CCA was technically substitutable for
creosote
was of little relevance to it and to the definition of the relevant
product
market with which it engaged. Nationwide,
at any rate, was stuck with creosote - its
reality was that
of a purchaser in a market for creosote. Sasol
argued that switching a pole treatment plant from
creosote to CCA
was a technically simple and relatively costless exercise.
Nationwide, for its part, insisted that
switching involved considerable expense and downtime.
This debate generated significantly more heat
than light. However we are able to
conclude that while the larger firms generally operate parallel CCA and
creosote treatment facilities in their plants, and while there appears
to be
some evidence of firms switching permanently from one wood preserver to
another, there is no evidence of a firm alternating a single treatment
facility
between creosote use and CCA use.
41.
However,
Sasol has, in order to support its contention that creosote and CCA
belong in
the same relevant market, placed considerable reliance on data which,
it
insists, demonstrate that when it increased the price of creosote,
demand for
its product fell off significantly and purchases of CCA increased
concomitantly. However, the data
relied
upon are open to question.
42.
It is
common cause that the price of creosote has increased, relative to the
price of
CCA. Sasol insists that in consequence of this movement in relative
prices it
has lost market share to CCA. Sasol contends that evidence of the two
products being substitutes is found in the SAWPA statistics, which
reflect that
the use of CCA increased relative to that of creosote. Its expert, Mr. Malherbe of Genesis, produced
a table entitled “Changes in Sales” which is reproduced and
discussed below.
Sasol also claims to have lost market share to Suprachem, the other
producer of
creosote. The reliability of these data
is open to question for various reasons:
i.
There is
evidence that the
SAWPA data relied on by Genesis may include export figures, therefore
we do not
know what the extent of local demand actually was.
ii.
We have to
rely on estimates by
Genesis as to Suprachem/ICC’s sales volumes for 2000 and 2001
because no
evidence of this was submitted.
iii.
The CCA
volumes are also
derived estimates and are open to question.
43.
Sasol
produced at the hearing a handout prepared by its experts, Genesis,
based on
SAWPA and Suprachem sales volumes, documenting changes in sales for
creosote,
CCA and a third product, Boron, between 2001 and 2003. This is
reproduced
below:
Sasol’s Changes in Sales Figures (in 000 m3)
|
|
CCA
|
Boron
|
Suprachem
|
Sasol
|
|
2001
|
133
|
2
|
163
|
210
|
|
2003
|
190
|
7
|
184
|
152
|
|
Absolute Change
|
57
|
5
|
21
|
-58
|
|
Percentage Change
|
43%
|
250%
|
13%
|
-28%
|
GenesisTable
produced at hearing sourced from SAWPA data (shaded areas represent
creosote
sales)
44.
Sasol
utilises this in an attempt to show that during the period documented
in the
table, Sasol’s sales losses were taken up by both Suprachem and
CCA. It
contends that over the period 2001 to 2003, there was a rise in the
demand for
CCA of 43%; further that there was a rise in demand for the creosote
offered by
Suprachem of 13%, while Sasol’s creosote product suffered a 28%
decline over
the same period.
45. It
is
common cause that the SAWPA data include pole volumes for domestic and
export
sales. Nationwide contended that insofar
as the SAWPA data included pole volumes for both domestic and export
purposes,
they could not be considered a reliable indicator of local demand for
creosote:
in other words, that the figures were flawed.
46.
At the
second set of hearings Mr. Foot
cross-examined Sasol’s expert, Mr Malherbe, on Sasol’s
sales figures.
He asked whether Sasol had extracted export orders from its analysis.
Malherbe indicated that the figures on which Sasol relied did not
include
export orders.
He confirmed then and later, in response to a question from the
Tribunal, that
Sasol’s figures had extracted export sales which had been
stripped out by his
team.
However, it was later put to him by the Tribunal that in the earlier
hearings,
Mr. Currie, the SAWPA representative, in
response to a
question from the panel as to whether the SAWPA sales figures reflected sales
in South Africa or whether they were sales by South
African
producers for export as well, had confirmed that the SAWPA figures did
in fact
reflect both local and export sales.
This was put to Mr. Malherbe, Sasol’s expert witness, and he
could not confirm
the reliability of either the SAWPA or Sasol sales figures:
“MR
MANOIM: This is, Mr
Currie is in the
witness box and I asked him a question, I said: “Sorry, just as a
point of
clarity on the Sawpa figures that Mr Unterhalter has shown you, are
these
figures of sales in South Africa, or are those figures of sales by
South
African producers either in South Africa or for export as well?”
and he says:
“It’s the latter.” I say: “The latter?”
and he says: “Yes.”
MR MALHERBE: So in other words he said it included export sales.
MR MANOIM: Yes that’s how I would understand that
exchange.”
MR MALHERBE:
Yes let me just confirm that. Okay
I think that the thing to say here is that we believe that our Sasol
numbers do
not include exports, but it’s not exactly the same calculation as
we did for
ICC.
MR
MANOIM:
Where did you get the Sasol numbers from? Were they given to you, are
they part
of the record, or were they given to you under instruction …[end
of tape]…
MR MALHERBE:
Well here is a possible issue. The
way that we derived at the Sasol figures for these purposes was from
the sulpha [this should read
“SAWPA”] figures less our ICC figures for domestic
market and our
understanding from Mr Currie was that the numbers that he provided us
did not
include exports and on that basis we assumed that that number that we
have,
effectively was equivalent to Sasol’s domestic sales. Now it
seems as if our
impression from him and what he said in the record might be
inconsistent and
that might have an impact on the numbers. I’m not sure.”
(Our emphasis)
47.
Sasol
later submitted that while the SAWPA figures reflected sales of
creosote-treated poles, they included sales of creosote-treated poles
destined
for export. However, Sasol argues that this is irrelevant, because even
if the
poles are exported, they are still an accurate reflection of local
demand
for creosote. However, apart from the doubt that this unresolved
debate casts on
the reliability of the data, the question of whether the treated poles
are for
the domestic market or for export markets has implications for
substitutability. For example, we have
no knowledge of the use to which the exported poles are put. It is conceivable that they were for use
in housing
construction where CCA poles may have been favoured for reasons of
creosote’s
odour rather than because of changes in relative price. What is clear
is that
Mr. Currie of SAWPA conceded that the gain in creosoted poles in 2001
could
have been attributed to an increase in exports and that this calls into
question the analysis of substitutability and its relationship to
movements in
the relative prices of CCA and creosote.
48.
There
is similar confusion surrounding the Suprachem/ICC figures. Not all of the Suprachem/ICC
figures were disclosed during discovery, and it seems that Sasol
estimated the
export figures for 2000 and 2001 based on the proportion of creosote
that
Suprachem/ICC exported in 2002. Sasol argued
that
export sales were removed from these “estimated” sales
figures for 2000 and
2001.
We agree with the complainant’s contentions that, because we do not have hard evidence
of what Suprachem’s
2000 and 2001 export creosote figures actually were, there is no way to
deduce exactly what Suprachem’s local sales of creosote were in
2001.
49.
Similarly, we do not know what the CCA
volumes in the market were, therefore cannot accurately compute the
degree to
which creosote sales declines were attributable to rises in sales of
CCA.
50.
In summary then we must approach with
considerable caution the assertion that Sasol’s data in the above
table
indicate substitution from Sasol creosote to CCA or to
Suprachem’s creosote
product, and assertions about the extent by which Sasol’s market
share was reducing,
if at all. Firstly, it is clear that even Sasol’s own expert was
confused as to
what data had been used and on which a fundamental component of
Sasol’s case
was based. Secondly, since the figures
included local and exported poles, we have no way of knowing to what
extent
demand was driven by price or the physical use to which the poles were
put. Dr
Roberts, Nationwide’s expert, pointed out that the demand for the
alternative
product could have been changing for a host of other reasons unrelated
to
price.
Thirdly, Dr Roberts pointed out that the analysis of switching
encompassed a
two year period, which was an inappropriately long period in which to
assess
substitutability, as it would increase the percentage change during
that
period. He argued it would have been better to assess year-by-year
changes over
a longer period of time, to get an accurate picture of substitutability.
51.
Dr
Roberts also pointed out that it is, in this case, particularly
difficult to
determine whether or not the pre-increase price of creosote was set at
the
competitive level. In our discussion of
market power we will show that Sasol’s pricing of creosote has
not responded to
that of its competitors. In these
circumstances it is reasonable to infer that Sasol’s price level prior
to the significant increases was already supra-competitive. An increase from a supra-competitive price
level may well give rise to a sharp decline in demand for the product
in
question and a concomitant increase in the demand for an alternative
without
suggesting that at competitive price levels the two products
are
substitutes. This is the well-documented
operation of the ‘cellophane fallacy’.
52.
The
technical characteristics of the two products – creosote and CCA
– indicate
that substitutability is, at best, limited in key applications and,
because of
regulatory interventions, is being further constrained in favour of
creosote
use. The evidence of substitutability
that Sasol produced based on, inter alia, the SAWPA data is
inconclusive and
clearly unreliable. We conclude then that the relevant market is
that for
creosote.
53.
We
will proceed to examine whether or not Sasol is dominant in that
market. We
will show that Sasol’s market share exceeds 45% and that it is,
therefore,
presumptively dominant in terms of Section 7(a) of the Act.
Dominance
54.
Section 7 of the Act provides:
A
firm is dominant in a market if –
(a)
it has at
least 45% of that
market;
(b)
it has at
least 35% but less
than 45% of that market, unless it can show that it does not have
market power;
or
(c)
it has less
than 35% of that
market, but has market power.
55.
The
Act defines ‘market power’ as ‘..the
power of a firm to control prices, or to exclude competition, or to
behave to
an appreciable extent independently of its competitors, customers or
suppliers.’
The
creosote market – market
share data establish Sasol’s dominance
56. The
evidence clearly establishes that Sasol’s share of the creosote
market exceeds
45% and is therefore presumptively dominant.
i.
SAWPA levies
SAWPA
extracts
levies from the two manufacturers, Suprachem and Sasol, based on a
percentage
of their sales. Therefore Nationwide contends that the levies represent
a
reasonable approximation of what their market shares must be. If we
assume that
the SAWPA levies do represent a reasonable proxy of what the volumes
would have
been then we must conclude that Sasol had 66% of the creosote market in
2001
and 53% in 2004.
ii.
Iscor/ICC figures
:
Nationwide relies
on information
submitted by Iscor, based on creosote tonnages
sold, and computes Sasol’s market share as follows:
2002 total market: 36,543 tons
Sasol share:
18, 251 tons
Sasol % :
50%
2003 total market: 37,644 tons
Sasol share :
19,250 tons
Sasol % :
51%
57. In 2004,
Sasol’s own figures indicate that, as at February
2004, it had 56% of the total creosote market.
Furthermore, its own information - once again forming part of the
Tribunal
record - indicates a South African market share of 56% for the 2003
year.
In the business plan of the Sasol
Carbo-Tar division it places its own share of the creosote market at
53%.
We
are satisfied
then that Sasol is, by virtue of its market share alone, clearly
dominant in
the creosote market because all evidence establishes a share in excess
of 45%
of the market throughout the relevant period up until 2004, that is
from April
2001 until August 2004.
58. Although we are
satisfied that Sasol’s market share
establishes that it is presumptively dominant in terms of Section 7(a)
of the
Act, we will also show that it is has exercised market power in this
market
insofar as it has, in setting the price of its creosote,
‘behave(d) to an
appreciable extent independently of its competitors, customers or
suppliers’.
59.
Sasol
has traditionally manufactured petrol and diesel from coal. This
process
involves converting coal into a gas stream which is converted into
liquid fuel.
This process leaves both ash and tar as by-products. The tar stream is
then
utilised to produce a bouquet of products which are, in turn, utilised
in a
variety of applications. These products make up Sasol’s carbo-tar
business which produces a range of value-added tar and carbon products
at both
its Secunda and Sasolburg plants and is a relatively small business
unit within
the entire Sasol group. As indicated
earlier the product categories in the carbo-tar division are creosote,
a wood
preservative, a product for the raw tar market, DIY and black
disinfectants,
and surface coatings, mainly comprising primers for road bases and the
binder
pitch which is sold to aluminium smelters.
60.
The Sasolburg plant has the capacity to produce
approximately
50 000 tons of the tar feedstock each year. Of
that raw feedstock, between 20% and 40%
could be converted to creosote. In 2002 and 2003, Sasol produced
between 20 000
and 23 000 tons of creosote per annum.
61. Mr.
van Wyk’s testimony revealed an important distinction between the
economic
drivers of the Sasolburg and Secunda plants, a distinction that
critically
influences Sasol’s pricing behaviour. The Sasolburg plant was
designed
to produce petrol and diesel from the gas stream only and not
from the raw tar stream. As
indicated the Sasolburg production process generates some 50 000
tons of
the tar feedstock annually. By contrast, Secunda was later designed so
that the
total tar stream could also be converted to a diesel stream –
hence although
Secunda produces an annual tar stream of 500 000 tons all of it was
intended to
be utilised in the production of liquid fuel. Two
important consequences flow from this:
62.
Firstly,
although there is considerable tar feedstock available at Secunda, the
plant is
not set up to utilise this feedstock in the production of the tar based
products such as creosote. The Secunda
feedstock has to be transported to Sasolburg to produce the various tar
products.
63.
Secondly,
because Secunda was designed, and the capital was invested, to produce
liquid
fuel from its tar stream by-product, the alternative value of the
Secunda tar
stream is the value of petrol and diesel. Therefore the opportunity
cost of
using that supply is the international dollar price of petrol or diesel
referred to as the ‘fuel equivalent price’.
Moreover, a key element of Sasol’s strategic plans
is the importation of
natural gas from Mozambique through a pipeline, the construction of
which is to
be completed 3 or 4 years hence. One of the core business units which
is to
utilise the gas is the Sasolburg plant. The
end result is
to be the elimination of the gasifiers because the plant will no longer
be
coal-based, hence the by-product of raw tar would fall away. Therefore
Sasol
predicted that in 3 or 4 years’ time Secunda would be the only
source of tar
feedstock all of which would be priced at the fuel equivalent price. It
was
this set of factors that underpinned Sasol’s decision to hike
massively the
price of its feedstock and, hence, the price of creosote and the other
products
emanating from this feedstock.
64.
Mr.
Foot avers that there is no semblance of negotiation between Sasol and
its
customers over the price of its creosote. The
price is laid down – for a year at a time
– in a schedule supplied
by Sasol. Customers are then informed of
Sasol’s decision and they either adhere to Sasol’s price or
they purchase their
product elsewhere. There is evidence in Van Wyk’s testimony on
how prices are
determined year-by-year. Sasol
determines an overall price increase, which is then allocated between
the
different price categories, and they are enforced in that manner, with
little
room for negotiation. | 