• Law Reports
  • Harmony Gold Mining Company Limited and Gold Fields …

Harmony Gold Mining Company Limited and Gold Fields Limited (93/LM/Nov04) [2005] ZACT 29; [2005] 2 CPLR 484 (CT) (18 May 2005)

78 Reportability
Competition Law
View PDF SAFLII Source
Share WhatsApp LinkedIn X Email

Brief Summary

Competition Law — Merger Approval — Hostile takeover of Gold Fields by Harmony Gold Mining Company — Competition Tribunal approved the merger subject to conditions limiting retrenchments — No retrenchments below Patterson grade C; maximum of 1000 retrenchments at or above grade C permitted — Merged entity required to submit quarterly reports to the Competition Commission for 24 months — Tribunal found no substantial competition issues in the gold market despite the merger creating the largest gold producer in South Africa and the world.

Comprehensive Summary

Summary of Judgment


Introduction


These reasons concern large merger proceedings before the Competition Tribunal of South Africa in relation to a hostile takeover in which Harmony Gold Mining Company Limited was the primary acquiring firm and Gold Fields Limited was the primary target firm.


The matter followed a contested regulatory history. Harmony launched a takeover bid in October 2004 in two stages described as an “early settlement offer” and a “subsequent offer”. Procedural litigation ensued, culminating in a ruling of the Competition Appeal Court on 26 November 2004 that, in substance, the two offers formed part of a single transaction to acquire control. The Competition Appeal Court interdicted Harmony from voting its shares in Gold Fields pending the final determination of the merger by the Tribunal (under section 16(2)) or by the Competition Appeal Court (under section 17) of the Competition Act.


Harmony notified the merger to the Competition Commission, and Gold Fields confirmed to the Commission that the transaction was a hostile takeover. On 11 February 2005, the Commission recommended approval subject to employment-related conditions. The Tribunal then conducted pre-hearings, heard intervention and confidentiality/discovery applications, and proceeded to an extended merits hearing (with multiple witnesses for both merging parties and one witness for the intervenors). Although mining unions were notified, none ultimately participated in the merits hearing in a substantive way.


The dispute concerned whether the merger raised competition concerns (including alleged buyer power effects in input markets) and whether it raised public interest concerns, principally in relation to employment. A further public interest contention advanced by Gold Fields was that the merger posed “systemic risk” to the gold sector and the South African economy.


Material Facts


It was common cause between the Commission and the merging parties that the merger raised no competition concerns in the gold production market. The Tribunal accepted that, while the merged entity would be the largest gold producer in South Africa and the world by certain measures, its share of world gold production would be about 9.5%, and the market for gold production is highly fragmented. The Tribunal also accepted that the daily gold price is fixed internationally through a relatively transparent mechanism not readily susceptible to direct producer influence. On this basis, the Tribunal treated the gold market as not requiring further analysis for competitive harm.


The main competition dispute raised by Gold Fields related not to gold sales but to alleged effects in markets for the supply of inputs to gold producers. Gold Fields contended that reducing South African “gold producing majors” from three to two would confer buyer power on the merged entity, enabling it to depress input prices and reduce suppliers’ output, with potential dynamic effects on suppliers’ incentives to invest and innovate. Gold Fields attempted to support this by reference to supplier engagement it had convened and a supplier survey, from which it asserted that significant proportions of suppliers expected business contraction or exit.


The Tribunal treated key aspects of Gold Fields’ supplier evidence as unreliable. It found the supplier survey and meeting process vulnerable to bias, noting the alarmist tone of the materials presented to suppliers and the presence of disputed statements framed as fact. The Tribunal also attached significance to the fact that suppliers who had initially made submissions subsequently withdrew them rather than subject their claims to scrutiny and cross-examination before the Tribunal.


By contrast, the Tribunal relied on the Commission’s investigative steps in relation to suppliers. The Commission contacted a material number of Gold Fields’ suppliers (including the ten largest identified in merger notification requirements) and obtained responses indicating that suppliers were generally not wholly dependent on Gold Fields and often served multiple major mining customers, including other mining houses and the platinum sector, with some being branches of larger groups or international firms. The Commission’s summary recorded that many supplier concerns were more about uncertainty in tender-based procurement than about a competition problem.


A further factual dispute concerned an economic expert report for Gold Fields that employed a “dispersed HHI” measure using highly aggregated industrial data to suggest potential buyer power concerns in broad product groupings such as “pumps”, “other rubber products”, and “mining machinery”. The Tribunal found that this approach did not define relevant markets and, in the Tribunal’s view, could not reliably indicate competition harm.


On public interest, the Tribunal recorded that the Commission’s recommendation identified “serious concerns” regarding job losses. Harmony stated that it could not be precise without due diligence but indicated willingness, pragmatically, to accept employment conditions as comfort to stakeholders. Gold Fields argued that Harmony’s claimed efficiencies would necessitate job losses far exceeding the Commission’s proposed ceiling, but it was accepted in argument that Harmony’s claimed savings to shareholders were not, as such, legally binding commitments for the purposes of merger control.


The Tribunal accepted that merger-specific retrenchments were most plausibly linked to rationalisation of corporate and regional management structures rather than mine-level operational consolidation, since the mines themselves were not to be merged. The Tribunal considered evidence about the social consequences of job losses in gold mining and accepted that retrenchments of semi-skilled and unskilled workers may have severe long-term effects. This underpinned the Tribunal’s view that a condition preventing retrenchments below a specified grade was warranted.


Gold Fields also advanced a public interest contention of “systemic risk”, based on the proposition that Harmony’s financial distress and management quality could lead to the collapse of the merged entity with grave macroeconomic consequences. The Tribunal rejected this factual thesis, including on the basis that the witness advancing it lacked relevant expertise and that the alleged chain of causation was inadequately supported.


Legal Issues


The central competition-law question was whether the merger was likely to substantially prevent or lessen competition in any relevant market, including whether it would create or strengthen buyer power (oligopsony power) in markets for inputs supplied to the gold mining sector. This raised issues of application of competition principles to fact, including the probative value of supplier surveys, the necessity of relevant market definition, and the reliability of concentration metrics used without proper market delineation.


The central public interest questions arose under section 12A of the Competition Act. First, the Tribunal had to determine the correct interpretation of section 12A(1)(b) in circumstances where the merger raised no competition concerns: specifically, whether a merger that passes the competition enquiry must nevertheless be prohibited unless the parties show an affirmative, net positive public interest justification, or whether approval is permissible provided the merger does not have a substantial negative effect on the public interest factors in section 12A(3). This was primarily a question of law (statutory interpretation), informed by policy considerations articulated within the judgment itself.


Secondly, the Tribunal had to decide whether, on the facts, the merger would have a substantial negative effect on public interest considerations, particularly employment, and (as alleged by Gold Fields) the stability of a sector or region via the “systemic risk” contention. This required an evaluative assessment of evidence about likely job losses and the appropriateness of conditions to address them.


Court’s Reasoning


On competition, the Tribunal accepted at the outset that the gold production market presented no competition problem, given global fragmentation and international price-setting. It therefore focused on the input-supply theories of harm advanced by Gold Fields.


In assessing buyer power allegations, the Tribunal emphasised that competition analysis requires engagement with actual markets and participants, and that there is no substitute for defining the relevant market before concentration tools such as the Herfindahl-Hirschman Index (HHI) can be meaningfully applied. The Tribunal treated Gold Fields’ supplier survey evidence with caution and ultimately attached no weight to it, because of the manner in which it was gathered, the alarmist framing, and the risk that suppliers would respond in ways that aligned with a major customer’s stance. The Tribunal also regarded the subsequent withdrawal by suppliers of their objections (when faced with open proceedings and cross-examination) as further undermining the reliability of the supplier-based narrative advanced by Gold Fields.


The Tribunal preferred the Commission’s investigative approach as an appropriate initial step. It accepted the Commission’s rationale that a detailed market-by-market inquiry was not warranted given that supplier interviews did not indicate competition concerns sufficient to trigger deeper investigation. The Tribunal also rejected the “dispersed HHI” analysis used by Gold Fields’ expert as rarely used in antitrust analysis and as fundamentally flawed because it sought to infer concentration without defining a relevant market. By illustrating how broad industrial categories like “pumps” collapse multiple non-substitutable products and disparate customer sets into a single aggregation, the Tribunal concluded that the method could generate misleading inferences (both false positives and false negatives).


The Tribunal further observed that buyer power allegations should be approached with circumspection, noting that buyer power can often benefit ultimate consumers and that it would be short-sighted for purchasers to squeeze suppliers to the point of exit and long-run supply harm. It also regarded arguments about procurement philosophies—long-term partnerships versus tendering and price pressure—as not germane to the competition enquiry and, in any event, not demonstrably aligned with competition or public interest outcomes.


Having found no likely substantial lessening or prevention of competition in any market, the Tribunal turned to public interest.


Before applying public interest factors, the Tribunal addressed the interpretation of section 12A(1). Gold Fields contended that even where a merger raises no competition issues and no negative public interest issues, it must still be prohibited unless there is evidence that it can be justified on public interest grounds, meaning an affirmative net benefit. The Tribunal rejected this interpretation as an overly literal and mechanistic reading, unsupported by the structure of section 12A, logic, and sensible public policy.


The Tribunal reasoned that section 12A mandates that the competition inquiry be performed initially, and only then the public interest inquiry, suggesting that the public interest test is not a freestanding requirement that every merger must affirmatively improve the public interest. The Tribunal held that “can or cannot be justified” is best understood as requiring a balancing of potentially positive and negative public interest effects to reach a net conclusion, rather than requiring a net positive public interest gain as a precondition to approval. The Tribunal also rejected the attempt to import a constitutional “culture of justification” approach to private merger transactions, considering it inappropriate and illogical in the statutory context, particularly because the Act does not require an affirmative showing that a merger promotes competition when assessing the competition leg.


Applying this legal conclusion, the Tribunal held that, since the merger did not raise competition concerns, Harmony did not have to prove a positive public interest benefit; it needed to establish that the merger would not have a substantial negative public interest effect.


On the specific public interest grounds, the Tribunal rejected the “systemic risk” argument. It held that the evidence relied upon (accounting ratios and speculative causal assertions about management quality leading to collapse) did not justify the dramatic conclusions advanced, and it criticised the lack of relevant expertise and corroboration. The Tribunal also reasoned that even firm failure would not necessarily entail cessation of mining activity because assets could be sold and operations continued, with shareholders most directly affected.


On employment, the Tribunal accepted that the merger raised serious public interest concerns and that conditions were warranted, especially given the lack of due diligence in a hostile takeover context and the potential long-term harms of retrenchments among less skilled workers. It accepted the Commission’s approach of limiting merger-specific retrenchments largely to higher categories, who were viewed as better positioned to secure alternative employment. The Tribunal imposed a stricter ceiling than the Commission recommended, reducing the number of permissible retrenchments, and sought to reduce ambiguity by specifying grades and by expressly including contract labour within the definition of “employees” for purposes of the order. The Tribunal also imposed detailed monitoring and reporting obligations to enable the Commission to supervise compliance.


Outcome and Relief


The Competition Tribunal approved the hostile takeover by Harmony of Gold Fields subject to conditions.


The conditions limited merger-related retrenchments at the merged entity for 24 months from the Tribunal’s order. The order prohibited merger-related retrenchments of employees below Patterson grade C (or equivalent) and permitted merger-related retrenchments of up to a maximum of 1000 employees at or above Patterson grade C (or equivalent). The term “merged entity” was defined to include Harmony and its subsidiaries including Gold Fields, and “employees” was defined to include contract labour.


The Tribunal ordered that the Competition Commission must monitor compliance. The merged entity was required to provide quarterly reports containing specified information regarding employment levels, retrenchments, reasons, planned retrenchments, status, and processes, with submission deadlines and continuation of reporting if retrenchment processes were not finalised within the 24-month period.


No costs order is recorded in the reasons.


Cases Cited


Harmony Gold Mining Company Ltd and Randfontein Estates, Competition Tribunal Case No: 16/LM/Feb00.


Franco-Nevada Mining Corporation Ltd and Gold Fields Ltd, Competition Tribunal Case No: 77/LM/Jul00.


Randfontein Estates Ltd and Anglogold, Competition Tribunal Case No: 03/LM/Jan01.


Clidet No 383 (Pty) Ltd and The Free State Operations of Anglogold Ltd, Competition Tribunal Case No: 05/LM/Jan02.


Crown Gold Recoveries (Pty) Ltd and IDC of South Africa Ltd and Khumo Bathong Holdings (Pty) Ltd, Competition Tribunal Case No: 31/LM/May02.


Armgold/Harmony Freegold Joint Venture Company (Pty) Ltd and St Helena Gold Mines Ltd, Competition Tribunal Case No: 54/LM/Aug02.


Anglogold Ltd and Driefontein Consolidated (Pty) Ltd, Competition Tribunal Case No: 66/LM/Nov03.


Harmony Gold Mining Company Ltd and African Rainbow Minerals Gold Ltd, Competition Tribunal Case No: 25/LM/May03.


Ubuntu-Ubuntu Commercial Enterprises (Pty) Ltd and Anglovaal Mining Ltd/Avgold Ltd/Harmony Gold Mining Company Ltd, Competition Tribunal Case No: 06/LM/Feb04.


Distillers Corporation (SA) Limited and Stellenbosch Farmers Winery Group Ltd, Competition Tribunal Case No: 08/LM/Feb02.


Anglo/Kumba, Competition Tribunal Case No: 46/LM/Jun02.


Competition Tribunal Case No: 86/FN/Oct04.


Competition Appeal Court Case No: CAC/43/Nov04.


Legislation Cited


Competition Act 89 of 1998 (as referenced), including sections 2, 12A, 13(3), 16(2), and 17.


Labour Relations Act 66 of 1995 (as referenced in the discussion of labour market regulation and collective bargaining).


Rules of Court Cited


No specific rules of court are cited in the reasons.


Held


The Tribunal held that the merger was not likely to substantially prevent or lessen competition in any market. It rejected Gold Fields’ buyer power theory as insufficiently supported, particularly given the absence of defined relevant markets, the unreliability of supplier-survey evidence gathered under conditions the Tribunal considered compromised, and the limited probative value of the “dispersed HHI” approach.


On the public interest, the Tribunal held as a matter of law that where a merger raises no competition concerns, the merging parties are not required to demonstrate an affirmative positive public interest benefit in order to secure approval. Instead, the Tribunal held that approval is permissible provided the merger does not have a substantial negative effect on the public interest factors, understood through a balancing of public interest considerations.


The Tribunal further held that the merger posed a potential adverse public interest effect through employment impacts, and that this warranted conditions. It imposed a time-limited moratorium and ceiling on merger-related retrenchments, confined in effect to higher employment grades, together with Commission monitoring and reporting obligations. It rejected the alleged “systemic risk” contention as inadequately substantiated and advanced by a witness not shown to have the relevant expertise.


LEGAL PRINCIPLES


The Tribunal applied the principle that merger assessment under section 12A is staged: a competition assessment must be undertaken first, followed by a public interest assessment. The public interest evaluation is required even where the merger does not raise competition concerns, but the Tribunal interpreted “can or cannot be justified on substantial public interest grounds” as requiring a net evaluative conclusion after weighing competing public interest effects, rather than requiring proof of a net positive public interest gain in every case.


In evaluating buyer power allegations, the Tribunal applied the principle that meaningful reliance on concentration indices requires prior relevant market definition. The Tribunal treated highly aggregated industrial categories as an insufficient basis for competition conclusions and endorsed an investigative approach that begins with engagement with market participants, with deeper market delineation and measurement to follow only if initial enquiries reveal material concerns.


The Tribunal also applied a principle of evidentiary caution in merger proceedings: evidence gathered through processes likely to bias responses—particularly where materials are presented in an alarmist manner or framed with undue authority—may be afforded little or no weight, especially where affected parties withdraw objections rather than defend them under scrutiny.


On remedies, the Tribunal applied the principle that merger approval may be conditioned to address substantial public interest risks, particularly employment impacts, and that conditions may be supported by monitoring mechanisms (including reporting obligations) to ensure compliance over a defined period.

COMPETITION TRIBUNAL 
REPUBLIC OF SOUTH AFRICA
      
Case no:  93/LM/NOV04
In the large merger between:
HARMONY GOLD MINING COMPANY LIMITED             Primary Acquiring Firm
and
GOLD FIELDS LIMITED           Primary Target Firm
Reasons for decision
Introduction
1. On 10 May 2005 the Tribunal approved the hostile take­over by Harmony Gold  
Mining   Company   Limited   of   Gold   Fields   Limited   subject   to   the   following  
conditions:
1. The following limitations shall be placed on the retrenchments at the merged  
entity – 
a) There shall be no retrenchments of employees at the merged entity  
below the level of Patterson grade C or equivalent as a result of the  
merger;
b) The   merged   entity   may   retrench   up   to   a   maximum   of   1000  
employees at or above the level of Patterson grade C or equivalent  
as a result of the merger.
2. For purposes of paragraph 1 –
a) ‘merged entity’ means Harmony Gold Mining Company Limited  
and its subsidiaries including Gold Fields Limited; and
b) ‘employees’ includes contract labour.
3.   The undertaking in paragraph 1 shall apply for a period of 24 months from  
the date of the Competition Tribunal order.

4. The Competition Commission must monitor the above conditions. 
5. In   order   for   the   Competition   Commission   to   properly   monitor   the   above  
conditions the merged entity must adhere to the following procedures:
a. Provide the Commission with quarterly reports regarding the effects of  
the proposed transaction on employment.
b. Each report must include the following information:
(i.)the   current   levels   of   employment,   per   job   category,   at   the  
merged entity;
(ii.)the   number   of   actual   retrenchments   per   job   category   in   the  
quarter reported on;
(iii.)the reasons, per job category, for retrenchments;
(iv.)the number of planned further retrenchments per job category;
(v.)the status of further retrenchments; and
(vi.)the   process   upon  which   the   retrenchments   will   take   or  have  
taken place.
c. The quarterly reports must be submitted for the period of this order.  
Should the retrenchment process in the merged entity, as a result of the  
proposed transaction, not be finalized within the period of this order,  
the merged entity shall be obliged to submit further quarterly reports  
until the entire retrenchment process has been finalized.
d. The   quarterly   reports   must   be   submitted   to   the   Competition  
Commission no later than one calendar month following the end of  
each quarter.  
Transaction Background
6. In October 2004 Harmony Gold Mining Company Ltd (“Harmony”) launched a  
hostile takeover bid for rival mining house Gold Fields Ltd (“Gold Fields”). The  
bid was made in two stages, the first ‘ the early settlement offer ’ and the second  
stage, ‘ the subsequent offer ’. 
7. The   ‘ early   settlement   offer’ ,   which   Harmony   claimed   was   not   subject   to  
regulatory  approval,  was the subject  of a  whole  set of procedural  applications  
brought against Harmony and which resulted in the Competition Appeal Court  
ruling,   on   26   November   2004,   that   ‘ the   early   settlement   offer ’   and   ‘ the

subsequent   offer ’   in   substance   formed   part   of   a   single   transaction   to   acquire  
control of Gold Fields and, therefore, interdicted and restrained Harmony from  
voting its shares in the share capital of Gold Fields prior to the final determination  
2

of   the   merger   by   the   Competition   Tribunal   in   terms   of   section   16(2)   or   the  
Competition Appeal Court in terms of section 17 of the Act. 1 
8. Subsequent to the Competition Appeal Court hearing Harmony acquired 11.8% of  
Gold   Fields’   shares.   On   8   November   2005,   at   the   same   time   that   Harmony  
announced its ‘ early settlement offer ’ it also filed its merger notification with the  
Competition   Commission.   Gold   Fields   responded   on   15   December   2005,  
informing the Competition Commission that it is a hostile take­over. 
9. On   11   February   2005   the   Competition   Commission   recommended   that   the  
proposed merger be approved subject to the following conditions:
1. The following limitations shall be placed on the retrenchments at the merged  
entity – 
a. There shall be zero retrenchments at the merged entity below the level  
of corporate, management and supervisory positions as a result of the  
merger;
b. The merged entity may retrench up to a maximum of 1500 employees  
in corporate, management and supervisory positions as a result of the  
merger.
2. Corporate, management and supervisory positions shall mean positions from  
shift boss level up to the chief executive.
3. The moratorium mentioned in 1 above shall apply for a period of 24 months  
from the date of the Competition Tribunal order.
Monitoring of the recommended conditional approval 
The following procedures must be adhered to in order for the Commission to properly  
monitor the abovementioned proposed conditions: 
1. The merged entity must:
a) Provide the Commission with quarterly reports regarding the effects of the  
proposed transaction on employment.
b) Each report must include the following information:
i.)the current levels of employment, per job category, at the  
merged entity;
1  The full history of these proceedings are reported in Competition Tribunal Case No: 86/FN/Oct04 and  
Competition Appeal Court Case No: CAC/43/Nov04.
3

ii.)the number of actual retrenchments per job category in  
the quarter reported on;
iii.)the reasons, per job category, for retrenchments;
iv.)the   number   of   planned   further   retrenchments   per   job  
category;
v.)the status of further retrenchments; and
vi.)the  process upon  which  the  retrenchments  will  take   or  
have taken place.
c) The quarterly reports must be submitted for the period of this order. Should  
the retrenchment process in the merged entity, as a result of the proposed  
transaction, not be finalized within the period of this order, the merged entity  
shall   be   obliged   to   submit   further   quarterly   reports   until   the   entire  
retrenchment process has been finalized.
d) The quarterly reports must be submitted to the Competition Commission no  
later than one calendar month following the end of each quarter.
10. A   pre­hearing   was   held   on   25   February   2005   during   which   intervenors   were  
identified, a time­table for filing submissions was agreed on and discovery and  
confidentiality issues addressed. A second pre­hearing date was set for 20 April  
2005 to discuss the final logistics of the case and the hearing dates were set down  
for 3 to 6 May 2005.
11. On 30 March and 8,19 and 20 April 2005 applications to intervene, confidentiality  
applications and discovery were heard. During the pre­hearing, held on 20 April it  
became apparent to the Tribunal that it would need additional hearing dates since  
the since the parties had called 13 witnesses in total. The hearing days were thus  
extended to 7,8 and 9 May.
12. Three parties intervened, Stitch Wise (Pty) Ltd, Paragon Textiles (Pty) Ltd and  
Knee’d’em (Pty) Ltd. The intervenors called one witness, Ms N. Killasy, who is a  
Director of all three intervenors.
13. The following witnesses were called by Harmony and Gold Fields:
Gold Field witnesses:
1) Mr. T. P Goodlace Senior   Vice   President,   Strategic   Planning,   Gold  
Fields
2) Mr. J McLuskie Expert Witness on deep level mining

Fields
2) Mr. J McLuskie Expert Witness on deep level mining
3) Dr N.S. Segal Independent consultant
4) Prof. H Bhorat Director, Development Policy Research Unit, UCT 
5) Mr. J Hodge Engagement Manager: Competition  and regulation  
4

practice, Genesis Analytics
6) Mr. N. Goodwin Gold Analyst, TSEC Securities
7) Mr. M.J. Mitchley Senior Manager Gold Fields
8) Prof. SA du Plessis Associate Professor in macro­economics University  
of Stellenbosch
Harmony witnesses:
1) Dr. C. Caffarra Economist and Director of Lexecon Ltd
2) Mr. B.M Saunders Executive: Investor Relations Harmony
3) Mr. A. Clay Director of Venmyn Rand (Pty) Ltd
4) Prof. S. Roberts Associate   Professor   of   Economics,   University   of  
Witwatersrand
5) Mr. Z.B. Swanepoel Chief Executive Officer of Harmony 
14. Although  all   the  Unions  that  represent  the   mineworkers  were  informed   of the  
hearing none were represented before the Tribunal. UASA and Solidarity attended  
the pre­hearing on 25 February 2005 but did not submit any further submissions  
nor  attended   any further   hearings  in this  regard.  The  National   Union of  Mine  
Workers, on 26 April 2005, requested an opportunity to address the Tribunal at  
the hearing but never showed up.      
COMPETITION EVALUATION
The Gold Market – competition implications
15.  It is common cause between the Commission and the acquiring and target firms  
that this merger presents no competition problems in the gold market.  Although  
the merged entity will, by most relevant measures, be the largest gold producer in  
South Africa and in the world, it’s share of world gold production will still only  
be   9,5%.     In   short   the   structure   of   the   market   for   the   production   of   gold   is  
characterised by its high degree of fragmentation. Moreover daily prices are fixed  
internationally   through   a   relatively   transparent   mechanism,   which   appears  
relatively   impervious  to  direct   producer  influence.    The  gold  market   has been  
analysed   in   previous   transactions   and   this   transaction   does   not   change   the  
conclusions reached in these earlier decisions. 2  Accordingly the gold market will

conclusions reached in these earlier decisions. 2  Accordingly the gold market will  
2  See: Harmony Gold Mining Company Ltd and Randfontein Estates, Case No: 16/LM/Feb00
Franco­Nevada Mining Corporation Ltd and Gold Fields Ltd, Case No 77/LM/Jul00
Randfontein Estates Ltd and Anglogold, Case No: 03/LM/Jan01
Clidet No 383 (Pty) Ltd and The Free state Operations of Anglogold Ltd, Case No: 05/LM/Jan02
Crown Gold Recoveries (Pty) Ltd and IDC of South Africa Ltd and Khumo Bathong Holdings  
(Pty) Ltd Case No: 31/LM/May02
5

not be examined further.
Markets for the supply of inputs to gold producers – competition implications
16. Gold   Fields   alleges   that   the   merger   will   lead   to   a   substantial   lessening   of  
competition in the markets for the supply of inputs to the gold mining sector.  
Gold   fields   argues   that   because   the   merger   will   reduce   the   number   of   South  
African   gold   producing   majors   from   three   to   two,   competition   will   be  
substantially lessened in many of the markets in which inputs are sold to the gold  
mining sector – the merger will, in other words, create oligopsonistic or buyer  
market   power.   The   consequence,   asserts   Gold   Fields,   will   be   manifest   in   the  
ability   of   the   buyer   to   force   input   prices   to   sub­competitive   price   and   output  
levels.   Gold   Fields   also   argued   that,   in   addition   to   these   static   allocative  
inefficiencies,   the   accretion   to   buying   power   will   give   rise   to   dynamic  
consequences   insofar   as   the   suppliers’   incentive   to   invest   and   further   develop  
their products will be dampened.   It is not clear whether Gold Fields contended  
that these dynamic consequences would be generated by a change in the structure  
on the demand side of the market or in consequence of the particular modality  
employed   by   Harmony   to   procure   supplies.     It   was   alleged   that   Harmony’s  
approach to procurement is dictated by considerations of price alone whereas the  
Gold Fields’ approach is allegedly more sensitive to quality considerations.  Gold  
Fields argues that the latter mode of procurement is more conducive to supplier  
investment in product improvement.
17.   Gold Fields attempted to adduce evidence in support of its contentions.   This  
evidence was gathered by a survey conducted of its suppliers and a meeting it  
convened of some 200 suppliers.  On this basis Gold Fields concluded that 39%

convened of some 200 suppliers.  On this basis Gold Fields concluded that 39%  
of their suppliers would be forced out of business and 56% expected a material  
downscaling in business volumes. 3  
18.   We are sceptical  of evidence gathered in the manner.   Gold Fields’ suppliers  
would, under these circumstances, understandably be inclined to provide answers  
supportive of an important customer’s clearly expressed standpoint – indeed the  
survey result may well reflect the power of Gold Fields   vis a vis   its suppliers  
rather   than   genuine   apprehension   of   the   merger.     We   should   add   that   the  
presentation to the suppliers has compromised the value of this evidence.  Highly  
disputed figures of the employment loss predicted by Gold Fields are presented as  
Armgold/Harmony Freegold Joint Venture Company (Pty) Ltd and St Helena Gold Mines Ltd Case No: 54/
LM/Aug02
Anglogold Ltd and Driefontein Consolidated (Pty_) Ltd Case No: 66/LM/Nov03
Harmony Gold Mining Company Ltd and African Rainbow Minerals Gold Ltd Case No: 25/LM/May03
Ubuntu­Ubuntu Commercial Enterprises (Pty) Ltd and Anglovaal Mining Ltd/Avgold Ltd/Harmony Gold  
Mining Company Ltd Case No: 06/LM/Feb04  
3  Transcript p.28
6

fact.     The   tone   of   the   communication   is   decidedly   alarmist   and   manifestly  
designed to strike fear into the hearts of the suppliers and their employees. 4   It  
appears, moreover, that the presentation, on behalf of Gold Fields, was made by  
its legal advisers, Edward Nathan – the presentation slides certainly carry Edward  
Nathan’s   explicit   imprimatur.     Not   only   does   this   provide   a   veneer   of  
independence, but it may have imbued certain important statements with a legal  
authority which they should not have enjoyed. For example the analysis presented  
to the suppliers and the dire impact predicted is predicated on an analysis of the  
‘extensive efficiencies’ promised by Harmony, ‘promises’ which, Edward Nathan  
baldly and quite incorrectly asserts (in bold type), ‘Harmony is not entitled  to  
abandon’.5  The reasonable reader of this document would assume that Harmony  
was legally obliged to achieve these efficiencies and hence been more susceptible  
to come to panic induced conclusions. We attach no weight to evidence gathered  
under these circumstances.
19.   Our scepticism  of the  evidence  presented  by Gold Fields is deepened  by the  
conspicuous failure of any suppliers to make submissions to the Tribunal.  Indeed  
a significant  number of suppliers had initially  made submissions.   However it  
appears   that   when   they   realised   that   they   may   be   obliged   to   repeat   their  
allegations in an open enquiry and subject to cross­examination they all withdrew  
their   statements   and   objections.   This   may   well   indicate   that   they   feared  
victimisation in the event of a successful Harmony acquisition (although as the  
Commission   points   out,   the   mere   fact   that   they   feared   victimisation   from  
Harmony  post­merger indicated  that  they  envisaged competing  for the merged  
entity’s custom – that is, contrary to Gold Fields’ evidence, its suppliers had not

entity’s custom – that is, contrary to Gold Fields’ evidence, its suppliers had not  
resigned themselves to going out of business).  However their reticence to defend  
their   claims   may   also   indicate   that   they   were   largely   designed   to   please   an  
increasingly anxious Gold Fields’ management and they feared that they would  
not stand up to scrutiny. 6  In the event, the only supplier who made submissions  
to the Tribunal was a representative of what is best understood as a corporate  
social investment project and she, indeed, made her representations in terms of the  
public interest.
20.   The   Commission   investigated   the   possibility   of   a   substantial   lessening   of  
4  Note particularly the slide titled ‘effect on suppliers’ on pages 3772­7 of the record, where the suppliers  
are told that, inter alia, many of them will ‘forced out’ of business and that ‘entire communities will be  
harmed extensively’.
5  We say incorrectly because counsel for Gold Fields conceded in final argument that Harmony was not  
under any obligation to achieve its claimed efficiencies.
6  Certainly one of the suppliers interviewed told the Commission that he had signed the document  
presented at the meetings in order to ensure that he retained Goldfields’ business.  Another who indicated  
opposition to the merger at the meeting, indicated to the Commission that he was neutral.  And, as we have  
already observed several – including some very large companies – who made submissions to the  
Commission opposing the merger were not prepared to subject these to the scrutiny of the Tribunal  
proceedings.
7

competition in the supplier markets.   Parties notifying a merger are required to  
identify   their   10   largest   suppliers.     The   Commission   telephonically   contacted  
about   52   Gold   Fields’   suppliers,   including   its   10   largest   suppliers. 7    The  
Commission summarises its efforts and some of its most important findings:
The suppliers contacted ranged suppliers of hydraulic pumps, gum boots,  
bearings, cleaning chemicals, pumps, rolling stock, skips, heat exchange  
for underground use, batteries, motors, valves, mechanical seals, hoses  
and   fittings,   conveyor   belts,   backfill   bags,   knee   and   arm   guards,  
radiators, oil coolers and the construction of underground dams, pump  
stations and pipe installations, inter alia.  Of the suppliers contacted none  
was  wholly   dependent   on  Gold  Fields   as  a  customer  with  many  citing  
Gold   Fields,   Harmony   and   Anglo   American­Ashanti   and   being   their  
major  customers.    Some  of   the   contractors   located   in  the   gold  mining  
areas   were   branches   of   larger   companies   such   as   Rocla   (part   of   the  
Murray and Roberts group), Builder’s Market (part of the Iliad Group  
with   branches   in   Welkom,   Klerksdorp,   Rustenburg,   Gauteng   and  
Polokwane), Conway Johnson (part of the Inmins group listed on the JSE)  
and Alstom (an international company).  Contractors and suppliers in the  
Carletonville area cited that, in addition to the gold mines, the platinum  
mines are important customers. 8
21. It   appears   that   the   Commission’s   enquiries   elicited   a   range   of   responses.  
Predictably, some amongst those Gold Fields’ suppliers that have no established  
relationship   with  Harmony  expressed  disquiet  at   the   prospective   merger.    The  
Commission summarises its findings:
In   summary   it   can   be   said   that   the   majority   of   businesses   who   have  
expressed   concerns   about   the   merger   operate   in   markets   where

expressed   concerns   about   the   merger   operate   in   markets   where  
competition   takes   place.     The   concerns   of   many   of   the   company  
representatives with respect to the merger can be said to be related to  
their uncertainty about being able to secure business in a market where  
most projects go out on tender. 9
22.  Counsel for Gold Fields made much of the Commission’s failure to go behind the  
responses   gleaned   from   the   telephone   interviews   and   to   identify   the   relevant  
markets and then conduct a detailed competition  analysis of the impact  of the  
transaction.   The Commission reasonably responded that because its interviews  
with   a   significant   sample   of   key   suppliers   indicated   little   if   no   competition  
concern with the merger there was no need for a time consuming and resource  
sapping   examination   of   each   of   the   many   product   and   geographic   markets   in  
7  See transcript of 5 May at page 558.
8  Commission’s Recommendations p21
9  Commission’s Recommendations p23
8

which a vast array of inputs are supplied to the merging parties. 10 
23.   Gold   Fields   also   relied   upon  evidence   and   argument   presented   by  one   of  its  
expert economic witnesses, Dr. S. du Plessis.  Dr. du Plessis purported to measure  
concentration in these markets by using a measure dubbed the ‘ dispersed  HHI’.  
The Herfindahl Hirschman Index or HHI is widely and legitimately employed as a  
measure   of   market   concentration   in   anti­trust   analysis.     However,   a   necessary  
prior step to calculating the HHI is the identification of the relevant market. This  
is more than a filter – it is the necessary preliminary step towards the construction  
of a filter and that filter is the HHI, which measures concentration in the relevant  
market.  In other words, once the relevant markets have been identified the market  
shares of the various participants in these relevant markets are calculated and the  
index is applied in order to assess the change in market concentration that will  
arise from the merger of two of the participants.  It is widely used by competition  
authorities, including the South African Competition Commission, as a measure  
of the increase in concentration resulting from a horizontal merger and, hence, as  
a first cut indicator of a possible accrual of market power that may arise from this  
increase in concentration.   However, the HHI is never, on its own, construed as  
sufficient   evidence   of  market   power  –   this   requires   a  detailed   evaluation   of  a  
range of factors including barriers to entry, the prospect of import competition  
and the dynamics of the market in question.
24.   However the   dispersed  HHI utilised by Dr. du Plessis does not serve the same  
useful filtering purpose as the HHI and accordingly is, du Plessis acknowledged  
and the Harmony expert, Dr. C. Caffarra, confirmed, rarely, if ever, utilised in  
anti­trust analysis. 11   There is no substitute for defining the relevant market and

anti­trust analysis. 11   There is no substitute for defining the relevant market and  
the Commission’s interviews constituted precisely the appropriate first step in this  
direction.  Had its interviews revealed any concern amongst the diverse suppliers  
that it contacted then it would have been obliged to conduct a deeper investigation  
in   order   to   identify   the   relevant   market.     Thereafter   it   would   have   calculated  
market shares and then computed the HHI in order to measure the changes in  
concentration that would accrue in consequence of the merger of the two market  
participants, of the two buyers.   
25.  The Gold Fields’ expert witness has made no attempt whatsoever to delineate the  
10  The Commission’s decision not to take this line of investigation any further is powerfully endorsed by  
the fact that these were  Goldfields’  suppliers that were taking a neutral view of a Harmony acquisition of  
their customer.  Indeed had they indicated concern, the Commission would have had to press further partly  
because of the real likelihood that their stated concern lay less with the absence of buyer competition in the  
post­merger market than with the fear that they were about to lose a valued customer.  This latter is not a  
competition problem – it is, if anything, a spur to competition. 
11  Indeed Dr du Plessis himself claims so little for his analysis that its probative content is seriously in  
doubt.  In the end he seems to suggest little more than that his report – and his analytical tool, the dispersed  
HHI – provides pointers for further investigation.  However the validity of even this claim is, as we outline,  
called into question. 
9

relevant markets. What he has done is to utilise highly aggregated (and dated)  
industrial statistics – specifically, the ‘Supply and Use’ tables for South Africa in  
2000 – which have then been employed to identify 9 broadly defined sectors, or,  
more accurately, products, in which the gold mining sector accounts for more than  
4% of total demand.   Then, assuming that each purchasing sector constituted a  
single monopoly – an assumption patently at odds with reality ­ a dispersed HHI  
was calculated which purported, in the fashion of the HHI, to identify the market  
concentration   of   the   sector   as   a   purchaser   of   the   output   in   question   –   he  
effectively insists that he has, in the manner of the HHI, provided an indicative  
measure of market concentration without defining the market.   12   In this manner  
– and after several iterations with the Harmony expert in which he purported to  
modify his results in order to accommodate specific criticisms of his efforts – Dr.  
du   Plessis   identified   three   problematic   groups   of   products,   these   being   ‘other  
rubber products’, ‘pumps’ and ‘mining machinery’.  These are product groups in  
respect of which the   dispersed  HHI exceeds 0.18.   It appears that this threshold  
was   chosen   in   order   to   provide   an   appearance   of   conformity   with   the   HHI  
threshold utilised by the US Department of Justice.
26. Let   us   consider,   by   way   of   example,   the   product   designated   by   the   industrial  
statistics as ‘pumps’, one of the ‘markets’ in which Dr. du Plessis’ dispersed HHI  
indicates cause for concern.  But the most casual observation tells one that this is  
no market at all.  It is an aggregation of several diverse markets.  It incorporates a  
range   of   non­substitutable   products   –   pumps   utilised   in   underground   mines,  
pumps utilised  in farm boreholes, pumps utilised in domestic swimming pools

pumps utilised  in farm boreholes, pumps utilised in domestic swimming pools  
and, for all we and, indeed, Dr. du Plessis, know, pumps utilised in automobile  
engines.   Had he even spoken to a single supplier of pumps to the gold mining  
sector he may have discovered that, as did the Commission on many occasions,  
the producer in question was unconcerned with the merger.   Had he been of a  
mind  to  interrogate   this response  further,  it  may  then  have  been  revealed   that  
these pumps are utilised across the mining sector and, hence, that the merger of  
even   two   large   gold   mining   companies   would,   in   the   face   of   an   attempt   to  
exercise market power, not necessarily depress prices below the competitive level  
– the pump producers would simply turn to their other mining customers.  He may  
also have found that the mining pump producers also actively supply international  
markets.  He may have found that the mining pump supplier, faced by an exercise  
of buyer power on the part of his mining customers, is easily able to switch to  
supplying pumps to the agricultural sector.  Or he may have discovered that the  
pump supplier enjoyed a monopoly in its market and, in consequence, that he felt  
relatively impervious to the change in the structure of his customers’ market.
27.   We   could   go   on   in   this   fashion   ad   nauseam.     The   conclusion,   though,   is  
12  As Counsel for Harmony pointed out by designating each sector a monopoly, the merging of two small  
buyers in a fragmented market would yield the same dispersed HHI as would a merger to monopoly of the  
only two firms in the market.
10

reasonably clear:   anti­trust investigation  does not easily lend itself to desktop  
research utilising highly aggregated industrial data.  It is micro­economic research  
and there is, as the Commission has shown, no substitute for engaging with the  
actual producers and customers.  This was not even attempted by the Gold Fields’  
expert.  Instead reliance was placed on a measure and a data set that yielded, at  
best,   no   results   of   consequence.     At   worst,   the   results   may   be   downright  
misleading   –   just   as   Dr.   du   Plessis’   high   dispersed   HHI’s   may,   on   relatively  
cursory   examination   of   the   relevant   markets   embedded   in   the   product   ranges,  
reveal no cognisable competition problems, so may some of his low dispersed  
HHI’s camouflage, on a proper definition of the relevant market, very definite  
problems.   In other words, the dispersed HHI is susceptible to both Type 1 and  
Type 2 errors – it may signal problems where there are none; and it may signal the  
absence of problems where there are some.   Dr. Caffarra, the Harmony expert,  
characterised the dispersed HHI:
‘..as a screening devise that doesn’t really screen because (it) is subject to  
type 2 errors.  It can tell us that there are some sectors where a problem  
exists, but then we may find there isn’t one.  It may tell us that there are  
some   sectors   where   there   isn’t   a   problem,   because   you   are   below   the  
threshold   and   still   there   is   a   problem   when   you   define   the   market  
correctly.  That is to me the hallmark of a not very useful approach.  You  
want to do other things instead.   You want to look….that’s why surveys  
exist.  You want to look…in a perfect world, you would want to calculate  
the elasticity of supply, but in a world where you don’t have this type of  
data it won’t be possible to do complicated econometrics and calculate  
the elasticity of supply or the supply curve.

the elasticity of supply or the supply curve.
So, you are down to asking nitty­gritty questions to suppliers.  Where are  
your concerns and how dispersed is the group of buyers that you sell to?  
As well as this fact that we are not really defining the relevant market.  
None of the categories that Prof. du Plessis has come up with resembles  
even remotely a relevant market, something that he has conceded but it  
doesn’t seem to worry him.  It worries me very much from the perspective  
of competition economics because I only can make any kind of meaningful  
inferences, however indirect, from market shares of from concentration  
indices to the extent that they pertain to one market. 13 
28. We should add here that those concerned to protect competition – and such is the  
stance adopted by Gold Fields and its expert in this matter – should approach the  
concept of buyer power with considerable circumspection.  Buyer power is most  
frequently alleged in respect of large retailers.   While there can be little doubt  
about the purchasing power of retail giants, there can be equally little doubt that  
13  See page 666 of the transcript dated 6 May 2005.
11

this frequently redounds to the benefit of the ultimate consumer.   Moreover, to  
argue that customers have an interest in squeezing their suppliers’ margins to such  
an extent that they are forced to cut back production and even exit the market is to  
impute  considerable   short­sightedness,  even  irrationality,   to the  purchaser  who  
may,   in   consequence   of   his   exercise   of   market   power,   not   only   be   faced   by  
supplies of compromised quality but even ultimately by a monopoly supplier. 
29. Finally we note that Gold Fields has made much of the fact that it enters into  
long­term partnerships with its suppliers and that it does not privilege price above  
quality   in   its   approach   to   procurement.     This,   it   insists,   contrasts   with   the  
Harmony approach, which is alleged to be aggressively focused on securing price  
concessions from suppliers who are denied the certainty of long­term contracts.  
This has nothing to do with our enquiry.   It also appears to be empty rhetoric.  
Certainly Gold Fields’ appraisals of its own procurement policies indicate – and  
we would expect no less ­ that it is as concerned with price as is Harmony.  Nor  
are we persuaded that long term contracts as opposed to regular tendering and  
regular   pressures   on   suppliers   to   review   their   pricing   is   preferable   from   a  
competition or public interest perspective.  Certainly markets in which suppliers  
are reviewed at relatively short intervals are subject to keen competitive pressures  
and may also be more conducive to new entry.
30. We find accordingly that the merger is not likely to substantially lessen or prevent  
competition in any market.    
PUBLIC INTEREST
31. This   matter   then   turns   on  argument   and   evidence   regarding   the   impact   of  the  
transaction on the public interest.  A range of public interest concerns have been  
submitted   for   our   consideration   and   we   will   examine   each   of   these   in   turn.

However we will  first examine  the  legal  regime  that  the  Act provides for the  
regulation of the public interest. 
The legal position
32. Section 12A sets out the relationship between the competition inquiry we have  
just performed and the public interest.
12

Section 12A. Consideration of Mergers 
(1) Whenever required to consider a merger, the Competition Commission or  
Competition Tribunal must initially determine whether or not the merger is  
likely to substantially prevent or lessen competition, by assessing the factors  
set out in subsection (2), and­
a. if   it   appears   that   the   merger   is   likely   to   substantially   prevent   or  
lessen competition, then determine­
i.)whether   or   not   the   merger   is   likely   to   result   in   any  
technological,   efficiency   or   other   pro­competitive   gain  
which will be greater than, and offset, the effects  of any  
prevention or lessening of competition, that may result or  
is likely to result from the merger, and would not likely be  
obtained if the merger is prevented; and
ii.)whether   the   merger   can   or   cannot   be   justified   on  
substantial public interest grounds by assessing the factors  
set out in subsection (3); or
b) otherwise, determine whether the merger   can or cannot   be justified  
on substantial public interest grounds by assessing the factors set out  
in subsection (3).   
33. Note that we have underlined the words can and cannot that appear twice in  
this section. Gold Fields argues that if a merger raises no competition problems  
and no negative public interest issues, it must still be prohibited if there is no  
evidence that it  can be justified on public interest grounds. 
34. Thus when making a public interest enquiry we first ask if the merger  cannot  be 
justified on public interest grounds. If we come to the conclusion that there is no  
basis for concluding that there is not, in the words of Mr Cockrell who argued for  
Gold Fields on this point, we cannot close the books and walk away.   The Act  
then   requires   us   to   consider   if   the   merger   can  be   justified   on   public   interest  
grounds. If there is no evidence that the merger can be justified on public interest

grounds. If there is no evidence that the merger can be justified on public interest  
grounds it must be prohibited. Expressed differently, unless there is a net positive  
public interest gain from a merger, it must be prohibited.
35. This   is   a   far­reaching   conclusion   and,   as   we   suggested   in   our   hearing   to   Mr  
Cockrell,  would render a good measure of the  mergers which come  before us  
daily, susceptible to prohibition. 
36. Let us see how Gold Fields comes to this conclusion and then examine if there is  
any basis for it.
37. Gold Fields starts by noting that the Act requires the Tribunal to have regard to  
13

public interest criteria even when no competition issue is implicated. This is  
because of the word ‘otherwise’ found in section 12 A(b). This interpretation  
thus   far   accords   with   our   finding   in   Anglo/Kumba   Tribunal   Case   No:  
46/LM/Jun02  where we held:
As   the   IDC   has   contended,   and   in   our   view   correctly,   the   use   of   the   word  
“otherwise” in section 12A(1)(b) means that the public interest evaluation must  
still be undertaken by the Tribunal, regardless of the outcome of the section  
12A(2) ‘competition’ analysis. As we have previously stated the public interest  
can operate either to sanitise an anticompetitive merger or to impugn a merger  
found not be anticompetitive. 14
38. Gold Fields goes on to note that in the purpose section of the Act, Section 2,  
public interest criteria are reflected in some sections such as (c), (e) and (f) which  
it argues, are not intuitively competition concerns.
39. This then gives Gold Fields the platform to support its reading of section 12A  
because it is a reading of the Act as we shall see that requires divorcing the public  
interest from any relationship to competition.
40. Let   us   first   examine   how   the   procedural   aspects   of   section   12   A   work,   as   it  
involves a set of stages and conclusions that are necessary to appreciate in order  
to understand the textual argument made by Gold Fields. 
41. All   mergers  must  first  be  subject   to  a  competition   evaluation. 15  This  is  not  
because of some administrative preference but because this is what the section  
requires.
42. Two   possible   outcomes  flow   from   this.  If  the  merger  is  not  found  to  lessen  
competition we follow the path set out in sub­section (b). This merger can thus  
be considered as having passed the competition inquiry. If the merger is found  
to be anti­competitive (we use this term again simply as shorthand) then the

to be anti­competitive (we use this term again simply as shorthand) then the  
next   stages   enumerated   in   sub­section   (a)   follow.   First   one   performs   the  
efficiency   trade   –off   required   by   (a)(i). 16  This   inquiry   again   can   have   two  
outcomes.   The   efficiency   trade   off   can   be   greater   than   and   offset   the   anti­
competitive effects in which case the merger can again be considered to have  
passed   the   competition   inquiry   and   be   on   all   fours   with   the   merger   that  
followed the path of subsection (b). If the efficiency trade off does not redeem  
14  See the large merger between Distillers Corporation (SA) Limited and Stellenbosch Farmers Winery  
Group Ltd, Tribunal Case No: 08/LM/Feb02, para 210.
15  Shorthand for whether the merger is likely to substantially prevent or lessen competition.
16 ‘ Efficiency’ is shorthand for  “technological, efficiency or other pro­competitive gain”  .
14

the merger then the merger emerges from the competition enquiry as having a  
net harm to competition. 
43. What   is   evident   then   is   that   under   path   (b)   a   merger   emerges   always   having  
passed the competition enquiry, but a merger under path (a) may, depending on  
the verdict of the efficiency trade off, pass or fail the competition test. 
44. Mergers following either path are then subject to the public interest inquiry.  
The (a) merger does so in terms of (a)(ii) and the (b) merger in terms of that  
sub­paragraph. The language of the public interest test however is identical for  
both  namely,  whether  the   merger  ‘can  or  cannot   be  justified   on  substantial  
public interest grounds by assessing the factors set out in subsection (3).’
45. Now   the   words   can   or   cannot   indicate   that   a   merger   that   has   failed   the  
competition   test  can still   be  passed  on  the  public   interest  test  and hence  be  
approved. Conversely, that a merger that has passed the competition test could  
still fail the public interest test and hence be prohibited.
46. Under path (a) can and cannot could arguably be interpreted as signifying the  
existence of these two outcomes, because the possibility of transforming both a  
‘failed’ and ‘passed’ merger exist. But this interpretation does not explain why  
under   option   (b)   where   we   only   have   ‘passed’   mergers   the   legislature   still  
provides for a “can”, when logically, it seems the only aspect of transformation  
that can take place is for the merger to ‘fail’ on public interest grounds and  
hence ‘cannot’ would have sufficed.
47. This gets to the nub of the Gold Fields argument, which is premised on the  
notion that the word ‘can’ is otherwise superfluous in paragraph (b) unless it  
can be given the interpretation for which they contend. It follows that if Gold  
Fields is correct about the interpretation given to (b) the same interpretation

Fields is correct about the interpretation given to (b) the same interpretation  
must be given to (a)(ii) as the language is identical. Hence Gold Fields argues  
that one not only has to determine whether the merger ‘cannot’ be justified on  
public   interest   grounds,   but   also   assuming   the   absence   of   a   rationale   for  
‘cannot’, still prove that it can i.e. the existence of a positive  impact on the  
public interest. Only by adopting this approach, so it appears to be contended,  
are   the   can   and   cannot   sensibly   accounted   for   as   alternatives   in   the   same  
phrase.
48. As a matter of pure interpretation the Gold Fields approach is a possible, albeit  
a very mechanistic reading. On this reading, finding the absence of a negative  
public interest does not end the Tribunal’s task. It must then make a “can”  
finding, which translated, means a finding that the merger benefits the public  
15

interest.
49. But, at best for Gold Fields, only an adherence to a sterile literalism commends  
this approach. Neither logic, the manner in which section 12A is constructed or  
sensible public policy, support this interpretation, as we go on to show. Nor, if we  
can   confine   ourselves   only   to   a   purely   textual   analysis   is   the   Gold   Fields  
interpretation the only possible one.
50. It appears that the only policy rationale beyond the semantic that Gold Fields  
can   enlist   to   its   cause   is   the   approach   to   the   culture   of   justification   in  
constitutional theory. Gold Fields argues that the use of the word justification  
is   a   signal   that   the   approach   to   the   public   interest   must   be   capable   of  
justification in the sense that this term is understood in constitutional law.  In  
constitutional law we ask whether state action is capable of being justified. If a  
decision   is   made   it   must   be   capable   of   being   rationally   explained.   Because  
section   12A   has   used   the   word   justification,   says   Gold   Fields,   a   similar  
approach must be adopted. Thus it is not good enough to determine whether the  
merger may not be unjustifiable on public interest grounds, the constitutionally  
derived   culture   of   justification   requires   us   to   determine   affirmatively   that   a  
merger  is justifiable.
51. This   approach   is   doubtful   for   a   number   of   reasons.   There   is   no   reason   to  
import  the  constitutional   approach to  justification   simply  because  there  is a  
congruence   in   language   in   the   Competition   Act.   Secondly,   mergers   are  
transactions   of   private   players.   To   equate   them   to   a   doctrine   requiring  
justification of state action is wholly inappropriate nor can it be justified by  
public policy. Thirdly, it seems to defy logic. Why should the legislature make a

public policy. Thirdly, it seems to defy logic. Why should the legislature make a  
firm justify its merger affirmatively on public interest grounds, but not require  
that   affirmative   justification   on   competition   grounds,   when   this   is   a   statute  
primarily   concerned   with   competition,   and   hence   presumably   elevates   that  
object more highly than any other subsidiary object. 17
52. There are a number of responses to Gold Fields’ interpretation. The first and  
perhaps most obvious is that having to show a negation of the reasons for why  
something cannot be justified seems to lead to the obvious default conclusion  
that one should be treated as if one can; as opposed to Gold Fields’ default  
position which is that absent an assertion of a positive, the default is that you  
17  The word justification is only used in the public interest leg of the inquiry not the competition leg. In the  
competition leg one is required to simply consider whether the merged will adversely effect competition  
not whether it will promote competition. Even if the efficiency defence is invoked as a gesture to  
‘justification’ it needs to be noted that it operates to negate a negative conclusion and not to assert a  
positive conclusion over a neutral one.
16

are treated as if you cannot.  18
53. Thus far we  have  approached the semantic  argument by cautioning  against  
attaching too much significance to the presence of a “possibly” superfluous  
can   in   sub­section   (b).   (Can   as   we   have   indicated   above,   is   not   “possibly”  
superfluous   in   (a)(ii)   as   there   are   two   outcomes   here,   recall   the   discussion  
above.)
54. But, there is an obvious approach to the textual interpretation that Gold Fields  
has not considered and which more credibly explains what the legislature is  
signalling through the explicit choice of the words can and cannot. The public  
interest   grounds   once   evaluated,   do   not   always   point   to   the   same   net  
conclusion.   Indeed,   in   certain   cases,   as   we   pointed   out   in   our   decision   in  
Distillers Corporation (SA) Limited and Stellenbosch Farmers Winery Group  
Ltd,  Tribunal  Case  No:  08LM/Feb02 ,  (“the  Distell  case”),  they  may  lead  to  
opposing conclusions which requires an internal weighing up to lead to some  
net conclusion on the public interest. 19 For instance a merger may lead to no  
competition problems and hence the merger follows the path of sub­section (b).  
If in examining the public interest and we find that it leads to some employment  
loss it will be public interest negative (Section 12 A (3)(b)). However the merger  
could  also lead to  the creation of a national  champion and hence is public  
interest positive (Section 12 A (3)(d)). Thus the Tribunal is required to perform  
an internal balancing of two conflicting public interest considerations before  
coming   to   net   conclusion.   The   words   “can   or   cannot”   are   then   instructive.  
They tell us that the public interest can have both adverse and benign effects.  
Secondly, they indicate that the competition authority is required to balance the  
positive  and negative  outcomes  and come  to a   net  conclusion on the public

positive  and negative  outcomes  and come  to a   net  conclusion on the public  
interest as opposed to a net  positive conclusion as Gold Fields would have it.
55. Thus far we have only toyed with the textual semantics of the argument. If Gold  
Fields’ interpretation were correct it would have led to a construction of the  
section, which is wholly at variance with what it is at present. Since Gold Fields  
argues that every merger must be shown to have a positive effect on the public  
interest in order to be approved, it follows then that this ought to be the first  
inquiry, because absent a showing of a positive impact on the public interest the  
merger must fail – why bother with the complexities of inquiring into market  
definition and efficiencies until this showing has been made. If it is to be the  
first inquiry one would expect the section to be drafted to reflect this. Instead  
section 12A directs the competition authorities to do the exact opposite of what  
18  Gold Fields argues that there can be no default position and hence its position is the correct one.  
However its position leads, as we have shown, inevitably to a default position.
19  See  Distell paragraph 214
17

on   the   Gold   Fields’   interpretation   they   should   logically   be   doing   first.   It  
expressly states in 12 A (1) that the Tribunal “must initially  determine” the  
competition question and “then determine” the public interest.
56. This   prioritisation   of   the   competition   inquiry   explains   the   use   of   the   word  
justification in the public interest test. The public interest inquiry may lead to a  
conclusion that is the opposite of the competition one, but it is a conclusion that  
is   justified   not   in   and   of   itself,   but   with   regard   to   the   conclusion   on   the  
competition  section.  It is not a blinkered approach, which makes the  public  
interest inquiry separate and distinctive from the outcome of the prior inquiry.  
Yes, it is possible that a merger that will not be anti­competitive can be turned  
down on public interest grounds, but that does not mean that in coming to the  
conclusion   on   the   latter,   one   will   have   no   regard   to   the   conclusion   on   the  
first.20  Hence section 12 A makes use of the term “justified” in conjunction  
with the public interest inquiry. It is not used in the sense that the merger must  
be justified independently on public interest grounds. Rather it means that the  
public   interest   conclusion   is   justified   in   relation   to   prior   competition  
conclusion.
57. Gold  Fields   argument   that  we   referred  to   earlier  that   section  2  supports  its  
interpretation of the autonomy of the public interest concerns in the Act is also  
not correct. Although some sub­paragraphs of section 2 refer to purely public  
interest concerns they do so in competition context. Section 2 commences with  
the   injunction   that   “the   purpose   of   this   Act   is   to   promote   and   maintain  
competition in the Republic in order ­….”
58. The   public   interest   in   the   purpose   section   is   seen   clearly   as   dependent   not  
independent of competition.

independent of competition.
59. Finally,   we   consider   whether   the   Gold   Fields   approach   is   consistent   with   the  
policy of the Act. Beyond its culture of justification argument, Gold Fields has  
not tried to do so. While many already consider our public interest requirements  
an anathema to merger control policy, few would argue for a position that mergers  
are so inherently harmful, that absent a positive contribution to the public interest,  
a merger that raises no competition concerns must be stopped. 
60. On the other hand the contrary position is compelling. That is, that a merger  
that raises no competition concerns and no negative public interest concerns  
should be permitted. There is no public policy that can be advanced to suggest  
that the species of contract that we define as a merger is so inherently harmful  
that if it cannot be shown to make the public interest better should be stopped.  
20  This is what we held in Anglo/Kumba. See citation supra.
18

Recall for a moment that the public interest showing must be substantial. Since  
substantial is not a relative term, it is either a substantial effect or it is not, this  
means that mergers involving large firms are more likely to be able to make a  
positive   showing   than   ones   involving   small   firms   simply   because   they   are  
bigger players economically. If that is so then small firms will always struggle  
to find a positive substantial public interest and hence be more likely on the  
Gold   Fields   approach   to   be   prohibited.   Recall   that   even   parties   to   what   is  
defined as a ‘small merger’ can be required to notify to the Commission and  
once so required their merger is subject to the section 12A regime. 21 How can  
such   parties   ever   show   that   their   merger   is   justified   on   substantial   public  
interest grounds? Even if they can show that a few more people will have jobs  
post merger that is hardly sufficient to meet the test of substantiality – on the  
Gold Fields test they must fail.
61. We find that as a matter of law it is not necessary for Harmony to show that the  
merger can be justified on public interest grounds. All that it needs to establish,  
having found as we have earlier  that it will not have a likely anti­competitive  
effect, is that the merger will not have a substantial negative effect on the public  
interest.
62. We   now   go   on   to   examine   the   facts   to   see   whether   the   merger   will   have   a  
substantial negative effect on the public interest.
The effect of the merger on a particular industrial sector or region a sector
63. A brief section of the report submitted by one of Gold Fields’ economic experts,  
Dr.   du   Plessis,   alleged   that   a   successful   acquisition   of   his   client   by   Harmony  
portended ‘systemic risk.’  The low watermark of his argument is that the merged  
entity   might   collapse,   allegedly   under   the   combined   weight   of   Harmony’s

entity   might   collapse,   allegedly   under   the   combined   weight   of   Harmony’s  
distressed   finances   and   its   poor   management   (which   management,   alleged   du  
Plessis, had been responsible for Harmony’s financial distress). In other words, a  
contagion originating in Harmony would infect the merged entity resulting in its  
descent into ruin and bankruptcy.
64.   We should say at once that even if this threat were to be realised, it is by no  
means clear that this would impact negatively on the public interest.  The scenario  
implicitly   sketched   by   du   Plessis   is   one   where   firm   failure   is   manifest   in   the  
sudden cessation of mining activities in the merged entity.  This is, of course, an  
extremely unlikely proposition.  Were Harmony or, indeed, the merged entity, to  
fail, then the company or its individual mining assets would simply be sold off to  
others   at   a   price   sufficiently   discounted   to   enable   the   continuation   of   mining  
21  See section 13(3).
19

activities.     Under   this   scenario   the   only   stakeholders   likely   to   suffer   adverse  
consequences would be the shareholders of the failed concern. And then there is,  
of course, no apparent reason why we should assume that the merging of two  
firms – one sound, the other relatively distressed – will result in the inevitable  
demise of the merged entity.   It could equally result in the rescuing of the weaker  
of the two firms and that, of course, may well, on Gold Fields’ own argument,  
promote the public interest. 22
65. But ‘systemic risk’ portends more than the mere bankruptcy of a single firm. It is  
a concept more traditionally associated with the banking sector where the failure  
or threatened failure of a large bank may cause a sudden loss of confidence on the  
part of depositors in the entire banking system.  Given the centrality of banking,  
the fear is that lack of confidence in the banking system may infect other spheres  
of economic and commercial life.  Extended to gold mining then we are asked to  
find that the collapse of the gold mining ‘system’ portends threat not only to the  
gold sector, but, given the continued  importance  of gold in the South African  
economy, to that latter terrain as well.    
66.   In   describing   systemic   risk   Dr.   du   Plessis   himself   argues   that   ‘the   relatively  
greater   corporate   risk   of   Harmony   could   undermine   the   sustainability   of   a  
Harmony/Gold Fields merger, with potentially grave implications for the South  
African   economy’. 23    This   is   then   the   high   watermark   of   Dr.   du   Plessis’  
allegations   and   accords   with   conventional   understanding   of   ‘systemic   risk’   as  
traditionally applied to the banking sector.
67. The evidence adduced by Dr.du Plessis’ in support of this dramatic conclusion is  
confined to his analysis of a range of conventional accounting ratios on which  
basis he concludes that:

basis he concludes that:
  ‘… there are various warning lights suggesting the possibility of financial distress at  
Harmony’.24 
68. He then concludes that because the Harmony management ‘ presided over the deteriorating risk  
profile of Harmony ’ that the same management would, if it presided over the merged entity, cause  
deterioration in its risk profile.
  
22  In his opening address, Mr. Gauntlet, Gold Fields’ senior counsel, characterised Harmony as an  
‘emaciated predator’ that is attempting to acquire Goldfields ‘in an endeavour to survive’.  We could  
reasonably conclude from this colourful invocation that should its prey elude it, then Harmony’s very  
survival is put at risk with all the negative consequences for the public interest arising from firm or even  
‘systemic’ failure that Gold Fields has identified – to state the obvious, this may well be an argument for  
approving the merger on public interest grounds!  
23  See page 4 of the Witness bundle.
24  See page 20 of the Witness bundle. Note that he warns, in footnote 13 on page 18, that a sharp change in  
the environment (such as depreciation of the Rand) could lead to a significant change in the assessment  
offered.
20

69. The   only   analysis   of   the   last   link   in   this   tendentious   chain   of   causation   is  
contained in the final sentence of his report where he baldly asserts: 
‘Given  the  size of these firms and the importance  of the gold mining industry  in the  
economy, the risk posed by the potential financial distress of a merged entity may rightly  
be called systemic ’.25  
70. And again in his evidence in chief:
 
‘I’m going to indicate that there are warning signs of financial distress at Harmony and  
given the size of the merger and the importance of the sector within which the merger  
occurs in this economy, from a macro­economic perspective, if a merger should proceed  
between these two companies, there is a concern that the management, which presided  
over the company now with the warning signs of financial distress would be in charge of  
a much bigger entity, which given the importance of this industry, would pose a systemic  
risk in this economy. ’26
71. Consider what it is that we are invited to decide:  Firstly, we are asked to determine, on the basis  
of   the   financial   stress   allegedly   revealed   by   Harmony’s   accounting   ratios,   that   Harmony   is  
threatened with imminent collapse.  And this despite du Plessis own admission that  ‘there is no  
consensus in the financial literature on the preferred measures with which to predict, for example,  
corporate financial distress’ .  Secondly, we are invited to decide that the financial stress allegedly  
suffered by Harmony will infect the merged entity ­ as we have already pointed out it is wholly  
conceivable that the merger would relieve rather than exacerbate Harmony’s alleged distress. The  
only argument advanced by Dr du Plessis in support of the more pessimistic prognosis that he  
favours is that the same management that presided over Harmony’s alleged decline would now  
secure the decline of the merged entity.  Accordingly, we are, thirdly, invited to judge the quality

of Harmony’s management and its contribution to Harmony’s travails relative to that of the Gold  
Fields’ management.
72. We are also expected to accept that while the merged entity and the South African  
economy is spiraling downward towards systemic collapse, nothing will be done  
to   reverse   it.     In   a   revealing   passage   of   his   evidence   in   chief   Dr.   du   Plessis  
concedes that: 
‘Financial stress is a slightly fuzzy term, which is meant to indicate that a corporation is  
in a position where,  if nothing changes, in other words, if the world stays the same and  
we   simply   turn   the   clock   onwards,   then   the   firm   is   not   sustainable   as   a   financial  
enterprise.’27 
73. This approach ignores market driven incentives. Would shareholders simply stand  
by idly as their company was driven into the ground?  Surely, they would seek to  
replace incumbent management. Or they may elect to withdraw their commitment  
to the ailing firm, thus depressing its share price and making it, in turn, vulnerable  
to take over.   These are the very mechanisms that we are mandated to protect  
25  See page 21 of the witness bundle.
26  See page 589 of the transcript dated 5 May 2005.
27  Transcript p588 (our emphasis)
21

rather than override which is what Gold Fields would have us do.
74.  In this vein, we should add that none of the other witnesses or any of the other  
affected parties who one may reasonably expect to be concerned at this potential  
economic Armageddon has alerted us to this threat. 28  Neither the unions, nor the  
community of gold analysts, nor Harmony’s auditors or shareholders, nor other  
gold   producers,   nor   the   World   Gold   Council,   nor   the   legions   of   experienced  
journalists   who   have   extensively   covered   this   merger,   nor,   indeed,   the  
government of South Africa, seem to have detected this pending cataclysm.  It has  
been exclusively revealed to a professor specialising in macro­economics who,  
through the examination of a number of standard accounting ratios, has identified  
this   threat   and   then   outlined   it   in  five   pages   of   his  report   to   the   Competition  
Tribunal.  Remarkably, or, perhaps, predictably, the person to whom this insight  
has been revealed  does not possess expertise  in the gold mining sector or the  
economics   of   finance   or   accountancy   and   therefore   does   not   even   pass   the  
threshold that must be applied in assigning weight to expert evidence, that being  
the possession of relevant expertise.
75. Having perused Dr. du Plessis’ report and having heard his evidence in chief, we  
can confidently reject the evidence and argument contained therein.   We repeat,  
the witness is manifestly not qualified to express these opinions, which have not  
been corroborated by any of other witness or commentator.   The quality of the  
evidence and the argument is indicative of his lack of expertise in these areas.
The effect of the merger on employment
76.   The employment consequences of the merger were examined at length on the  
record and in the hearings. In its recommendation the Commission concludes that  
the   “proposed transaction raises serious concerns with respect to job losses.”

the   “proposed transaction raises serious concerns with respect to job losses.”  
The Commission raised these concerns with Harmony who were willing to give  
an   undertaking   that   the   job   losses   be   limited   to   1500   so­called   ‘supervisory  
positions’. The Commission was of the view that if the undertaking was made a  
condition  of the approval  of the  merger this  would  obviate  the public  interest  
concerns. We have accepted this recommendation although we have imposed a  
lower limit – 1000 as opposed to 1500 ­ than that suggested by the Commission.  
The Commission’s recommendation that job loss be restricted to managerial and  
supervisory   categories   has   also   been   accepted   although   we   have   attempted   to  
28  The one possible exception is Dr. N. Segal, another expert witness called by Goldfields.  We say  
‘possible’ because Dr. Segal’s evidence and argument was at a high level of generality. It does nevertheless  
appear that he too apprehended that Harmony’s acquisition of Goldfields would threaten the sustainability  
of the latter and that this portended threat to the gold mining sector and, hence, to the South African  
economy.  The factors identified by Dr. Segal that portend systemic risk range through the fact that  
Harmony is smaller than Goldfields, that Goldfields’ mines are allegedly more complex operations than  
Harmony’s, that Harmony would inevitably undermine managerial capacity at Goldfields and that  
Harmony would withdraw Goldfields from the World Gold Council. 
22

identify   the   affected   categories   in   order   to   limit   any   ambiguity   in   the  
interpretation of our order.   The conditions that we have imposed also ensure that  
employees   whose   terms   and   conditions   are   governed   by   a   labour   contract   –  
referred to as ‘contract employees’ – be treated as employees of the merged entity  
for the purposes of this order.
77. Harmony’s evidence was that the merger was unlikely to have an adverse effect  
on employment and that it was unnecessary to impose a condition that related to  
employment effects. Nevertheless it indicated that from a pragmatic point of view  
that  it   was  not   opposed   to  having   the  condition  proposed  by  the   Commission  
made a condition for our approval as comfort to all concerned. Harmony argued  
that as it had not had an opportunity to perform a due diligence exercise it was  
unable to be more precise about the number of retrenchments that might arise out  
of the merger. However its past experience  in mining  mergers which includes  
mines previously owned by Gold Fields such as Evander, indicated that merger  
specific   retrenchments   were   not   extensive   because   it   employed   a   top   down  
approach. By this Harmony means that it retrenched from the highest levels first  
and   then   downwards.   This   approach,   because   high­level   employees   receive  
disproportionately better remuneration, meant that it could achieve high levels of  
savings with fewer job cuts. In addition it outlined that the likely retrenchments  
would   come   about   as   a   result   of   its   management   philosophy   the   so­called  
Harmony Way which favoured reducing the size of mining head offices and the  
layer of regional management that exists between the mine and the head office.  
The   figure   of   1500   it   suggests   is   a   conservative   estimate   of   what   these  
retrenchments might be in the current Gold Fields operation. The merger would

retrenchments might be in the current Gold Fields operation. The merger would  
not likely lead to retrenchments at lower levels for this reason.
78. Gold Fields, for its part, took the view – at some considerable length – that in  
order for Harmony to make the savings that it had ‘promised’ its shareholders, the  
scale of job loss would have to exceed that provided for in the Commission’s  
recommended ceiling by an order of some considerable magnitude.   There was  
some   dispute   –   resolved   in   Harmony’s   favour   –   as   to   whether   Harmony   had  
undertaken to achieve cost savings of R1 billion per annum or R1,6 billion per  
annum. If the lower figure was accepted on Gold Fields estimation there would be  
approximately 4000 job losses as a result of the merger.
79. This entire argument is, for the most part, of no relevance to our proceedings.  
Counsel for Gold Fields acknowledged that Harmony’s ‘undertakings’ regarding  
the savings that it expected to introduce into the merged entity are of no legal  
effect as they are undertakings made to the shareholders of the merging parties. It  
does   not   follow   that   even   if   these   cost   cutting   claims   are   extravagant   that  
Harmony will introduce extensive job cuts to implement the promised savings. If  
job cutting is irrational as Gold Fields witnesses suggest it is then Harmony is  
more likely to renege on its promises to shareholders than to cripple the company  
23

to   keep   its   promise.   Harmony   denies   it   needs   to   be   stressed   that   the   savings  
promised are as job cut related as suggested by Gold Fields.
80. Nor do we have reason for believing that the employment loss need exceed the  
number stipulated in the condition.  In our view employment loss arising from the  
merger   generally   refers   to   the   job   loss   occasioned   by   the   rationalisation   of  
production   and   support   facilities.     In   this   instance   there   will   be   limited  
rationalisation   –   certainly   none   of   the   production   facilities,   that   is,   the   mines  
themselves, are to be merged and so the need for rationalisation is limited and  
will, by and large, not affect mine level employees.  
81. Harmony as we noted makes much of its relatively flat employment structure, an  
approach that, it avers, enables it to dispense with several layers of management  
in the corporate head office and in the mining regions themselves. This goes to  
the heart of the distinction  between, on the one hand, Harmony’s approach to  
managing   a   gold   mining   group   and,   on   the   other,   Gold   Fields’   approach.  
Retrenchments   occurring   in   consequence   of   Harmony’s   leaner   management  
structure are then appropriately considered to result from the merger itself.  Our  
condition simply permits Harmony to exercise this management prerogative by  
providing for a possible level of retrenchment that includes both rationalisation  
and the flattening of the management structures.
82. Again Gold Fields’ witnesses opined at length on the violence  that this leaner  
approach   would   do   to   the   ability   of   Harmony   to   operate   the   merged   entity.  
Harmony disagrees.  We are unable to take a view on this debate and nor do we  
have to.  Again should Gold Fields’ prognostications be borne out and should it  
indeed prove impossible for Harmony to dispense with the layers of management

indeed prove impossible for Harmony to dispense with the layers of management  
it deems superfluous and counter­productive, this is not a matter with which we  
need to concern ourselves.  We are not obliging Harmony to dispense with these  
managerial   layers   –   we   are   simply   requiring   that,   whatever   it   chooses   to   do,  
retrenchments must be limited to the numbers and to the categories provided for  
in our conditions.
83.  We had requested the parties to submit a view on the impact that previous gold  
mining mergers had had on the affected communities.     Gold Fields submitted  
reports prepared by two experts, Dr. H. Bhorat and Mr. James Hodge. 29  Both of  
these experts testified at the hearings.  Dr. Bhorat’s report documented the impact  
that the reduction of employment in gold mining has had on poverty levels both in  
the directly affected mining areas as well as in the regions from which many of  
the   migrant   workers,   who   are   mostly   unskilled,   are   drawn.     There   can   be   no  
quarrelling with his conclusions.   We take some comfort from the fact that the  
merger­specific retrenchments to which our condition refers will implicate only  
29  See the G:enesis Report, page 48 of the Witness bundle.
24

those categories of employees best equipped to secure alternative employment. 30 
84. Mr.   Hodge   attempted   to   correlate   employment   loss   with   merger   events.     He  
concluded   that   these   were   robustly   correlated.     Counsel   for   Harmony   argued  
persuasively that it is not sufficient to demonstrate correlation.  If causation is to  
be   found,   what   is   required   is   a   theory   of   the   relationship   between   the   two  
observed   phenomena.   Mr.   Hodge   effectively   reasoned   that   retrenchments  
loosened the bond between employer and employee, or, as he would have it, it  
enabled the new management to break the implicit contracts that its predecessor  
management had developed over long periods.  
85. This is not a very persuasive theory.  In a relatively regulated labour market with  
national union representation, the regulatory regime and even the identity of those  
who operate within it, are unaffected by the fact of the merger – there are then  
‘bonds’ or ‘relationships’ with the national unions that will not be lightly broken  
because the impact will extend beyond the immediate area of contestation and  
there are ‘explicit’ contracts that are the provisions of the Labour Relations Act  
and, as is likely, industry and company wide collective bargaining agreements,  
that are unaffected by the merger. It is conceivable that the management of the  
merged entity may attempt to use the cover of persistent general job loss in order  
to   effect   merger   specific   retrenchments.     But   here   we   take   comfort   from   the  
monitoring mechanisms that are imposed as part of the conditions.   More than  
that, we are certain that the powerful mining unions – none of whom opposed the  
conditions   proposed   by   the   Commission   –   will   closely   monitor   future  
retrenchments in the merged entity and will ensure compliance with the terms of  
our order.

retrenchments in the merged entity and will ensure compliance with the terms of  
our order. 
86. Of   course,   as   pointed   out   by   Harmony’s   counsel,   the   observed   relationship  
between merger and job loss may well reflect the general and long­term decline of  
gold mining – the origins of both employment loss and merger activity may well  
be found in this general industry­wide decline. Hence it is common cause that  
Harmony is currently engaged in significant retrenchment activity on the mines  
that it owns.   This is likely to continue regardless of whether or not there is a  
merger.  On the other hand, it is suggested – not least of all by Gold Fields – that  
the merger represents an attempt by Harmony to place itself on a stronger footing,  
that   is, that   the  merger  (and  the  retrenchments)  represent  Harmony’s  response  
both to the long term decline of the sector and to the prevailing level of Rand­
equivalent price of gold.  Were the merger to go ahead under these circumstances,  
Mr. Hodge’s econometrics would once again demonstrate a correlation between  
the merger and job loss when in fact both of these phenomena are driven by long­
term   sectoral   decline   coupled   with   commodity   cycles   and   exchange   rate  
movements. 
30  Dr. Bhorat also conceded that he was not concerned about the categories of skilled labour referred to in  
the Competition Commission’s conditions. See page 251 of the transcript dated 4 May 2005.  
25

87. Certainly it seems clear to us, if not to Mr. Hodge, that the level of post­merger  
retrenchment activity will be critically influenced by the specific character of the  
assets   acquired.     Hence,   the   earlier   merger   waves   analysed   by   Mr.   Hodge   do  
appear to reflect Harmony’s acquisition of the marginal mines upon which it built  
its gold mining group.  This is manifestly not the case in the present transaction  
and there appears to be little reason to expect the merger to influence the rate of  
retrenchment in the mines presently controlled by Gold Fields.  There may, to be  
sure, still be job decline in these mines.   However, the pace of job loss will be  
dictated by the decline of the industry rather than by the merger event. 
88. It is conceivable that where, as in the earlier events, marginal mines are absorbed,  
the   pace   of   retrenchment   is   influenced   by   the   merger.     However,   under   these  
circumstances it is extremely difficult to construct the counterfactual.   Certainly  
there   was   much   unresolved   factual   dispute   over   Harmony’s   claimed   record   in  
extending   the   life   of   mines   –   despite   Gold   Fields’   denials   this   is   certainly  
Harmony’s widely held reputation.     It is then wholly conceivable that a post­
merger ‘rightsizing’ may well be predicated on a changed perception of the life of  
the mine  and the approach to the available  ore reserves that accompanies  this  
change.  The counterfactual – that is, where the seller retained control of the mine  
­ may well be a larger workforce employed for a relatively short period to harvest  
rapidly the ore reserve and then to close down the mine.  Indeed it may well be  
that the owners of the target mine, had calculated that maintaining the mine in  
production   required   a   level   of   investment   and   extraordinary   expenditure,  
including retrenchment, to which it was not prepared to commit.   Under these

including retrenchment, to which it was not prepared to commit.   Under these  
circumstances   the   merger   may   well   correlate   with   an   accelerated   pace   of  
retrenchment but it takes, nevertheless, a particularly distorted logic to ascribe the  
retrenchments to the merger.
89. We are nevertheless mindful of the fact that retrenchments of semi­skilled and  
unskilled workers in the present mining environment may well have an adverse  
effect on employment and hence the public interest, because they may lead to  
long term unemployment consequences, particularly if the evidence of Professor  
Bhorat   is   accepted.   Of   importance   here   is   that   Harmony   has   not   done   a   due­  
diligence   and   absent   the   condition   we   would   have   to   be   content   with   its  
estimation   that   from   past   merger   experience   there   is   no   likelihood   of   merger  
specific  retrenchments  of  this  class   of  employee.   We  find   that   taking   comfort  
from such a possibly uninformed estimation is insufficient in the circumstances of  
this case, and a condition is warranted to protect the public interest. We must also  
bear   in   mind   that   it   was   on   the   basis   of   the   condition   as   proposed   by   the  
Commission   that   certain   parties   such   as   the   unions   did   not   participate   in   the  
proceedings and they may well have done so to express concerns had it not been  
forthcoming.
26

90. Although Gold Fields had argued that the merger be prohibited on public interest  
grounds in the alternative it asked that the level of retrenchments be limited to  
head   office   employees   an   amount   of   112.   Neither   of   its   experts   on   the  
employment issues took issue with the proposed condition of the Commission,  
and hence it seems that Gold Fields needed to rely on the likelihood that Harmony  
would never be able to keep to this condition to achieve its promised savings to  
justify   why   the   merger   should   be   prohibited.   On   this   latter   point   there   is   no  
evidence to suggest that if a condition is imposed on Harmony that it will not  
honour it.
91.   We are then content to limit the merger­specific retrenchments to 1000 and to  
require  that  they all be drawn from the ranks of management  and supervisory  
staff. It seems to be common cause that these are employees that are able to find  
new   employment   in   the   event   of   retrenchment   because   they   have   marketable  
skills.   We   have   reduced   the   number   of   retrenchments   provided   for   in   the  
Commission’s recommendations  by 500. This we have done because the most  
detailed  accounting  by  Harmony   suggested   slightly   over  900  retrenchments. 31 
We are naturally aware that Harmony has been obliged to take a view without the  
benefit of a due diligence.   However, due diligence or no, the figure provided to  
us represents its best effort to detail the scale of merger related retrenchment and  
this is what we will accept as the maximum permissible level of retrenchment.
Conclusion
92. We   have   found   that   the   merger   will   not   substantially   lessen   or   prevent  
competition in any market. We have found that the merger may have an adverse  
effect on employment and hence the public interest but that the condition we have  
imposed will obviate any substantial public interest concern.
18 May 2005
D. Lewis Date

imposed will obviate any substantial public interest concern.
18 May 2005
D. Lewis Date
31  The Counsel for Harmony in its cross­examination of Mr McLuskie suggested that 921 retrenchments  
were envisaged. See page 167 of the transcript of 3 May 2005. Mr Swanepoel, in his evidence, anticipated  
1000 retrenchments. See page 983 of the transcript dated 8 May 2005.
27

Concurring: N. Manoim and Y. Carrim
28

Back to Law Reports