Association of Mineworkers and Construction Union and Another v Competition Tribunal of South Africa and Others (169/CAC/Dec18) [2019] ZACAC 1 (17 May 2019)

THE COMPETITION APPEAL COURT OF SOUTH AFRICA

HELD IN CAPE TOWN

CAC CASE NO: 169/CAC/Dec18

In the matter between

ASSOCIATION OF MINEWORKERS AND

CONSTRUCTION UNION First Appellant

GREATER LONMIN COMMUNITY Second Appellant

and

COMPETITION TRIBUNAL OF SOUTH AFRICA First Respondent

SIBANYE GOLD LIMITED t/a SIBANYE­STILLWATER Second Respondent

LONMIN PLC Third Respondent

MINING FORUM OF SOUTH AFRICA Fourth Respondent

SIKHALA SONKE Fifth Respondent

GREATER LONMIN COMMUNITY Sixth Respondent

COMPETITION COMMISSION OF SOUTH AFRICA Seventh Respondent

JUDGMENT: 17 May 2019

DAVIS JP

[1] Albert Einstein once said 'God does not play dice with the universe'. Whatever the controversy with regard to the meaning of this statement, there should be no difficulty in understanding that courts should not play dice with the welfare of workers, particularly in a country with notorious levels of unemployment as is the case in South Africa. This consideration must be uppermost in the mind of a court confronted with public interest concerns raised in respect of a proposed transaction between merging parties when these concerns relate to the employment of some 32 000 workers is potentially at risk.

The background

[2] On 13 March 2018 second respondent (Sibanye) and third respondent (Lonmin) notified the seventh respondent (the Commission) of a proposed large merger in terms of which Sibanye intended to acquire the sole control of Lonmin. Sibanye proposed to implement the merger by issuing 0.967 shares in Sibanye in exchange for each ordinary share in Lonmin. This would mean that, subsequent to the merger, Lonmin's shareholders would hold 11,3% of the enlarged Sibanye entity. In its competition filling Sibanye submitted that access to Lonmin's smelting and refining facilities would make it a fully integrated Platinum Group Metals (PGM) producer in South Africa. Sibanye contended that there would be a potential realisation of synergies between contiguous Sibanye and Lonmin assets and opportunities to further progress current developmental projects within the Lonmin business.

[3] In the merging parties filing, Lonmin set out its reasons for the merger: 'Lonmin has been suffering major challenges in recent years in respect of its debt structure, capital constraints and liquidity. As the headroom in the Lonmin group's tangible net worth had decreased, Lonmin's financial statements for the six months to 31 March 2017 disclosed the risk of a potential breach of Lonmin's debt covenant, which could reduce its liquidity. As at the end of Lonmin's 2017 financial year, the tangible net worth covenant was breached. Despite a series of restructuring initiatives, Lonmin has been unable to adequately restructure its debt so as to provide the liquidity required for the business to operate properly.'

[4] The Commission conducted an extensive investigation of the proposed transaction which included a market analysis of the proposed

transaction, a competitive assessment, and a public interest assessment. It concluded its competitive assessment by finding that

the proposed transaction presented both a horizontal and a vertical overlap. After evaluating the pre and post­ merger market

structure, it concluded that the proposed transaction was unlikely to substantially lessen competition in any of the separate PGM markets it had identified.

[5] The Commission found that the proposed transaction did not raise any unilateral effects. In its vertical analysis the Commission noted that the proposed transaction raised vertical issues because Sibanye did not have any smelting or refining operations in South Africa but sold its PGM concentrates to refiners and smelters in the downstream market where Lonmin was active. It, however, concluded that input or customer foreclosure was unlikely in the circumstances. In the light thereof, the Commission concluded that it was unlikely that the proposed merger would substantially lessen or prevent competition in any of the relevant markets.

The Tribunal's decision

[6] On 21 November 2018 the Tribunal approved the proposed merger, subject to certain conditions to which reference will be made

presently. The appellant (AMCU) together with certain other participants were granted the right to intervene in the proceedings

before the Tribunal. One of the parties which was granted intervenor status was second appellant (GLC), which belatedly applied to be heard in the appeal before this Court. I shall return to this application after analysing the Tribunal's decision.

[7] The Tribunal accepted the Commission's analysis that the proposed transaction was unlikely to substantially lessen competition in the relevant markets. It however accepted that the merger raised public interest issues, in particular the contemplated large scale retrenchments at Lonmin and post­ merger, the noncompliance by the merging parties with their respective Social Labour Plans (SLP), the effect of the merger on local suppliers and historically disadvantaged persons, the potential rolling out of an

Agri-lndustrial Development Program to create economic and social benefits for communities which were in the area where mining was conducted.

[8] Before the Tribunal, AMCU referred to the independent operational plans for the future of Lonmin which had been provided by Sibanye and Lonmin respectively. In a standalone plan, Lonmin had envisaged 12 459 retrenchments in order to cut costs and continue operations. Sibanye constructed a joint operational plan with Lonmin which envisaged 13 344 retrenchments. Sibanye considered that only 885 retrenchments were merger specific. By contrast the Commission considered that 3188 retrenchments were merger specific. AMCU considered that all the retrenchments, whether 12 459 or 13 344 retrenchments, should be considered to be merger specific.

[9] The Tribunal found, on the basis of the available evidence, that there was no justification for concluding that all retrenchments

proposed in Lonmin's plans were merger specific. In this connection it said 'The exact calculation of all merger-specific retrenchments

is difficult as it is in business decisions and plans based on imperfect assumption'.

[10] The Tribunal then examined the countervailing public interest arguments advanced by the merging parties, namely that, absent the merger, more workers at Lonmin stood to lose their jobs and Lonmin's assets were consequently sold on a 'fire-sale'. The potential loss of jobs was estimated to be as high as 32 000. While the Tribunal eventually found, in the light of the uncertainty regarding the exact number of retrenchments and when retrenchments were expected to take place, that it was unable to 'give the merging parties a free hand at the dismissal of whoever they wish without a thorough economic analysis and stakeholder engagement ... we have to balance the above commercial realities and cannot force unfeasible mines to stay open.'

[11] As a result, the Tribunal decided to adopt what it referred to as a 'balanced approach'. As a result, the order it made was, inter alia, that all retrenchments at Lonmin would be prohibited for a period of six months from the implementation of the proposed transaction, as well as imposing certain further conditions arising from an undertaking given by the merging parties to ensure that certain job saving measures were implemented. These further conditions were dependant on the realisation of certain PGM price and mining costs levels.

[12] The Tribunal then addressed the SLP commitments made by the merging firms to the Department of Mineral Resources (DMR) in terms of the Mineral and Petroleum Resources Development Act 28 of 2002 (MPRDA). These plans involved certain social responsibilities which were designed to uplift communities in the area in which the mining operations are located.

[13] Sibanye agreed to honour Lonmin's SLP obligations post-merger and to the imposition of a specific condition to that effect. In addition, the merging parties tendered to set up a community engagement forum for the purpose of providing information and soliciting the views of stakeholders surrounding the commitments in terms of the SLPs. The tender was included as one of the conditions ordered by the Tribunal.

[14] Sibanye agreed to honour four existing contracts owed to the Bapo Community Companies as well as an agreement to continue to pay the community and annual amount of R 5 million. In addressing the question of the effect upon local suppliers and historically

disadvantaged persons (HOP), the Tribunal added a condition that Sibanye would honour all existing contracts to HOP suppliers and

would endeavour to ensure that it would comply with current HOP procurement policies.

[15] The final issue concerned the Agri-lndustrial Development Program which was designed to promote the economic and social upliftment in the Rustenberg area which was most effected by the mining operations. Sibanye is to set up a memorandum of understanding with the 'West Rand Steering Commission' that seeks to develop agricultural and social benefits for the West Rand Communities affected by mining operations.' It was also accepted by the merging parties that, once the West Rand Development Program was completed, they would ensure that an independent body would conduct a feasibility to study to determine the suitability of such a project in the Rustenberg community. If the feasibility study found in favour of rolling out such a plan in Rustenberg, Sibanye would donate 500ha of land for use in this initiative. If the feasibility study did not find in favour of the roll out of the agriculture and industrial development program, Sibanye would investigate potential alternative programs and report the status of such to the Commission. In summary, the Tribunal approved the merger but subjected it to certain conditions all of which have been carefully articulated in its order.

[16] The substantive conditions which are relevant to this appeal are:

'Employment (the "Employment Condition”)

3.1 The target and acquiring firm will not retrench any employees for a period of 6 months from the Implementation Date.

3.2 For the sake of clarity, retrenchments do not include (i) voluntary separation arrangements; (ii) voluntary early retirement packages; (iii) unreasonable refusals to be redeployed in accordance with provisions of the LRA; (iv) resignations or retirements in the ordinary course of business; (v) terminations in the ordinary course of business, including but not limited to, dismissals as a result of misconduct or poor performance; (vii) any decision not to renew or extend a contract of a contract worker; and (viii) the initiation of proceedings in terms of s 189 of the LRA as long as such proceedings are not finalised before the expiry of the

period in 3.1 above.

3.3 The Acquiring Firm commits that it shall, provide that the variables and pre-requisites set out in item 1 of Annexure A1 are satisfied, save (through avoiding retrenchments and/or creating new jobs) 3714 jobs in the period 2018 to 2020. Annexure A1 herewith provides a breakdown of variables and pre-requisites such as the timeline and economic variable (including minimum price and reduction in cost base or operational costs) that would need to be satisfied, as well as technical and economic assessments required to be undertaken, per Short Term Project in order for such job savings to be realised.

….

The Agri-lndustrial Community Development Programme ('the Development Programme Condition”)

3.6 The Acquiring Firm shall ensure a feasibility study on an Agri-lndustrial Community Development Programme is conducted through the appropriate members of the West Rand Programme Steering Committee to understand the potential of rolling out a similar initiative as its Rustenberg platinum operations (including the Target Firm's operations) and the potential impact on job creation, within 1 (one) year from the finalisation of the Agri-lndustrial Community Development Programme project structure and roll-out plan. For the sake of clarity, the West Rand Programme Steering Committee shall be an independent committee.

3.7 Prior to the commencement of the feasibility study merging parties shall, enter into consultations with the Bapo Traditional Community, representatives of the Greater Lonmin Community (GLC), Sikhala Sonke, the Mining Forum of South Africa, Trade Union, and any other affected communities to discuss the envisaged Development Programme and feasibility study to be undertaken and to solicit their views.

3.8 In the event the feasibility study supports the rolling out of a similar initiative in Rustenberg, the Acquiring firm shall contribute land measuring approximately 500ha in extent as its Rustenburg operations to the initiative.

3.9 If the feasibility study contemplated in clause 3.6 above does not support the rolling out of a similar initiatives, the Acquiring Firm will, for a period of two years following receipt of the feasibility study, explore other options, in consultation with the stakeholders mentioned in paragraph 3.7 above, to achieve the objectives described in the Recordal above.

Target Firm SLP (the "Target Firm SLP Condition”)

3.14 The Acquiring Firm will honour existing commitments made by the Target Firm in terms of its SLP as at the Merger Announcement Date (colloquially known as "SLP2H), as well as any commitments being made by the Target Firm for 2019 in terms of the SLP that it is currently in the process of being finalised for submissions to the DMR ("SLP3"), which commitments shall at all times confirm with and be subject to then-current legislation and regulations.

3.15 The Acquiring Firm will establish a Community Engagement Forum ("Forum"), within a period of six months from the Implementation Date for the affected communities and stakeholders of the Target Firm, including but not limited to the Bapo Traditional Community, representatives of the Greater Lonmin Community (GLC), Sikhala Sonke, the Mining Forum of South African and Trade Unions. The purpose of the Forum will be to provide information and to solicit the views of the affected community and stakeholders of the Target Firm on the Acquiring Firm's commitments under SLP 2 and/or SLP 3, as applicable, and apprise the Forum of the Acquiring Firm's performance under the commitments.

6. VARIATION

The Merging Parties and Commission may at any time, on good cause shown, apply to the Tribunal for the Conditions to be waived, relaxed, modified and/or substituted. The Commission or merging parties will not be precluded from opposing such application.'

The appeal before this Court

[17] The appeal by AMCU against the decision of the Tribunal focussed ultimately on whether the Tribunal had taken sufficient account of the public interest concerns raised by it. AMCU contended that the Tribunal had failed to adequately assess the effect of the merger on employment. Hence, it asked this Court to refuse approval of the merger, alternatively impose further conditions or amend some of the conditions imposed by the Tribunal. It is to be noted though that the appeal specifically dealt with the application of s 12 A (3) of the Competition Act 89 of 1998 (the Act), as opposed to the determination as to whether the proposed merger raised cognisable competition concerns relating to the likelihood of the merger substantially preventing or lessening competition in the relevant market.

The admission of GLC

[18] Before dealing with the merits of the appeal, it is necessary to address the application by the GLC to be admitted as an appellant in these proceedings. On 5 March 2019 GLC essentially applied to be admitted as an appellant in the matter. It did so by simply lodging an appeal against the decision. This was some three and a half months after the Tribunal had delivered its decision to approve the merger and almost three months after the Tribunal had published the reasons for its decision. It was also a month and a half after this Court had set down the appeal. It also applied for condonation for the late 'lodging of the appeal'. Upon receipt of the application on 08 March 2019, this Court directed GLC to file written arguments in respect of its application for condonation. It was further directed to address the question as to whether it had locus standi in terms of the Act to appeal the decision of the Tribunal. IGLC's application to be admitted as an appellant was opposed by the merging parties.

[19] After hearing argument from both the merging parties and the GLC, this Court dismissed the application by GLC to be admitted as an appellant. It did so on the basis of s 17 (1) of the Act which provides that two categories of persons may appeal against a decision of the Tribunal in merger proceedings, being 'any party to the merger' and any person who was required to be given notice of the merger in terms of s 13 A(2), provided that such a person had been a participant in proceedings before the Tribunal.

[20] It is common cause that GLC did not fall into either of these two categories. It was not a party to the merger. It was not a registered union representing a substantial number of primary employees of the acquiring firm or of the primary target firm. There was however some suggestion that s 61 of the Act could justify GLC's case for admission. This section provides: 'a person affected by a decision of the Competition Tribunal may appeal against or apply to the Competition Appeal Court to review that decision in accordance with the rules of the Competition Appeal Court if, in terms of, section 37, the Court has jurisdiction to consider that appeal or review that matter'.

[21] This Court has dealt with this provision and scrutinised the issue of the limited classes who may appeal against a merger decision of the Tribunal in Competition Commission v Distillers Corporation (SA) Ltd and Stellenbosch Farms Winery Group Limited (31/CAC/Sep03) at para 38:

'...ss 61(1) and 37 [of the Competition Act] should not be read as altering or derogating from the provisions of s 17 in respect of appeal against Tribunal merger decisions. It follows that the categories of persons which may appeal against Tribunal merger decisions are those limited categories specifically set out in s 17(1) and not the class of 'affected' persons referred to in s 61 (1).'

[22] The further question was raised as to whether GLC had locus standi as a result of s 62(2)(b) of the Act, namely that this Court has jurisdiction over any constitutional matter arising in terms of this Act. GLC is concerned about the SLP and whether the conditions which were imposed by the Tribunal complied with the MPRDA. This concern does not raise a constitutional issue as contemplated ins 62(2)(b) of the Act. Hence, there is no basis upon which the GLC could be admitted as a party to these proceedings on appeal. Having reached this conclusion there is therefore no need to deal with the issues arising from its application for condonation. Accordingly, its application was dismissed with no order as to costs.

The merits of the appeal

[23] This appeal focussed entirely on what conditions were appropriate to justify the public interest concerns which arose in respect of the merger. In summary, AMCU raised two specific arguments. In the first place it argued that 13 344 employees would have to be retrenched as a result of the merger. In short, unlike the merging parties or the Commission, AMCU contended that all of the job losses which flowed, whether from Lonmin's precarious financial position or the effect of the merger were merger specific and had to be taken into account insofar as the imposition of conditions was concerned. Secondly, the two significant conditions imposed by the Tribunal, namely that no employees be retrenched for a period of six months from the implementation date of the merger and that Sibanye implement some short term projects in order to save jobs totalling 3714 employees over a three year period between 2018 and 2020 were fraught with caveats and were vague. AMCU contended that a six month moratorium on retrenchments was inadequate and that the sub (??) saving projects which were to be undertaken by Sibanye were conditional upon the increase in platinum prices to a certain level threshold level as well as to the costs of mining certain shafts being maintained at a particular level. In the event that these conditions do not materialise after a three month period of assessment, the merging entities would be relieved of them. A similar argument was raised with regard to the condition imposed by the Tribunal in respect of the Agri-lndustrial Community Development Program.

[24] Distilled to its essence, AMCU's argument was that the Tribunal had seriously misdirected itself by failing to properly consider the effects of the change of circumstances between the filing of the merger on 13 March 2018, the finalisation of the merger report of the Commissioner on 17 September 2018, the publication of Lonmin's Regulatory Releases on 26 October 2018, the metal exchange

agreement with Pangae Investment Management Limited (PIM) and the general improvement of PGM prices and positive market forecasts at the hearings on 12 November 2018. Before turning to evaluate these arguments, it is necessary to examine the existing jurisprudence

relating to s 12 A (3) of the Act.

The public interest grounds in a merger

[25] Section 12 A sets out the considerations which have to be taken into account in the evaluation of a proposed merger. Considerations of a merger by the Competition authorities. The Act envisages three separate but related inquiries:

1. Whether or not the merger is likely to substantially prevent or lessen competition;

2. If the result of this inquiry is in the affirmative, whether technological efficiency or other pro-competitive gains override the initial conclusion reached in stage 1 together with the further consideration based on substantial public interest grounds, which, in turn, could justify permitting or refusing the merger; and

3. Notwithstanding the outcome of the enquiries in 1 and 2, the determination of whether the merger can or cannot be justified on substantial public interest grounds.

The specific public interest grounds are set out ins 12 A (3):

'(3) When determining whether a merger can or cannot be justified on public interest grounds, the Competition Commission or the Competition Tribunal must consider the effect that their merger will have on –

a) a particular industrial sector or region;

b) employment;

c) the ability of small businesses, or firms controlled or owned by historically disadvantaged persons, to become competitive; and

d) the ability of national industries to compete in international markets.'

[26] As mentioned above, this case does not involve first two legs of the inquiry. The appeal therefore relates exclusively to assessing the merger through the prism of the public interest grounds.

[27] This court has considered the specified public interest grounds in Minister of Economic Development and others v Wal-Mart Stores Inc and others [2012] ZACAC (09 March 2012), where it did so in the context of an enquiry as to whether the merger should be disallowed on the basis of specific public interest concerns, rather than one based on competing public interest arguments.

[28] Subsequent to the decision in Wal-Mart, the Tribunal had occasion to further consider the test that should apply in assessing public interest considerations where it focussed on evaluating competing arguments regarding the public interest. In Metropolitan Holdings Limited and Momentum Group Limited [2010) ZACT 87 (9 December 2010} the Tribunal dealt directly with the question of the appropriate test to be applied in determining the issue of substantial public interest based on a loss of employment in a merger. The Tribunal found that there was an evidential burden upon the merging parties, once a prima facie case that the merger would result in a significant loss of employment had been established to show that the merger should nevertheless be sanctioned. To this end the Tribunal held that two criteria had to be satisfied. In its words:

'1) a rational process has been followed to arrive at the determination of the number of jobs to be lost, i.e. that the reason for the job reduction and the number of jobs proposed to be shed are rationally connected; and

2) the public interest in preventing employment loss is balanced by an equally weighty, but countervailing public interest, justifying the job loss and which is cognisable under the Act.' Metropolitan Holdings at para 70

[29] The second leg of the inquiry emphasises the public nature of the test; that is, if the merging parties are able to demonstrate that a loss of employment, for example, promoted efficiency and thereby could be justified by a gain to shareholders, this alone, cannot be considered to have satisfied the public interest ground for approving the merger. By contrast, a countervailing public interest ground would be one of the other grounds set out in s 12A(3), namely the ability of a national industry to compete in international markets, the ability of firms controlled or owned by historically disadvantaged persons to become competitive or the necessity of saving a failing firm or a firm in a precarious financial position where absent some loss of employment a far greater loss of employment could result if the merger was not permitted. See also BB Investment Company (Ply) Ltd and Adcock Ingram Holdings (Ply) Ltd (CT 018713) 28 August 2014 at paras 94-95.

[30] In the light of these decisions, the Tribunal held that two critical questions have to be answered: (i) whether a rational process has taken place with regard to the effect on employment pursuant to the merger and (ii) is the public interest in preventing any employment loss at least balanced by an equal weighty, but countervailing public interest justifying the job losses? Expressed differently, the first stage of the enquiry involves interrogating the reasons for job losses while the second involves a proportionality inquiry, which involves an examination of the competing public interest issues: that of job losses with any other counter-vailing public interest issue.

[31] Returning to evidence, in their merger filings, both Sibanye and Lonmin submitted operational plans for the future of Lonmin. Lonmin's plan (referred to as the Standalone plan) envisaged 12 459 retrenchments in order to reduce its costs and thus continue with its business operations. Sibanye constructed a plan together with Lonmin (the Sibanye plan) which envisaged 13 344 retrenchments. Sibanye submitted that 885 retrenchments were merger specific which figure was calculated as follows: Sibanye calculated that there were 1283 merger specific job losses comprising of 1132 jobs which would be lost because of the implementation of its operational plan, together with 151 jobs which would be lost because of duplication or consolidation of operations resulting from the merger. It then reduced the merger specific job losses by what it referred to as 398 'merger specific job savings'.

[32] The Commission examined both plans and concluded that the Standalone plan did not contemplate 12460 retrenchments but rather 10156 retrenchments. Therefore, the Commission concluded that 2304 retrenchments (i.e. the difference between 12460 and 10156) were merger specific ones. Adding this figure to the 885 proposed retrenchments specified in the Sibanye plan, the Commission maintained that the merger specific retrenchments totalled 3189. By contrast, AMCU contended that all the proposed retrenchments, (i.e. 13 344) were merger specific.

[33] In support of this contentions, Mr Puckrin, who appeared together with Mr Coetzee on behalf of AMCU, referred to a passage of the evidence of Mr Froneman of Sibanye. The passage emphasised by Mr Puckrin reads:

'We asked Lonmin to provide us with their base plan and help us develop our plan based of their base plan, because in the timeframe we had this company was about to hit the wall, we did not have the luxury of time to go and do that with our limited knowledge. That is exactly what - that is our plan, not their plan okay. We are asking for them give us your base plan and we will tell you how we want to develop our plan. That is not the Lonmin plan, that is the Sibanye plan. That is exactly, and that is what is being presented to this Tribunal and to the Competition Commission. That is the S1banye plan. Lonmin have their own plan which is not

this plan.'

[34] On the basis of this evidence, Mr Puckrin submitted that Sibanye had influenced Lonmin's plan to such an extent that all the proposed retrenchments identified there were a product of Sibanye's influence and was designed to make Lonmin an attractive target for Sibanye. Hence, al the proposed retrenchments (13444) were accordingly merger specific.

[35] Mr Puckrin also submitted that the evidence supported AMCU's contention that Lonmin's financial position was not so precarious as to justify all of its proposed retrenchments. In this regard he submitted that Lonmin had re-evaluated its operational planning and delayed placing some of its shafts on care and maintenance because of improved market conditions and an improved financial position. This argument was based on Lonmin's Regulatory Release First Quarter 2019 Production Report and Business Update', which evidence AMCU sought to have admitted at the appeal. I shall return to this application presently. For the moment it is necessary to examine the dispute regarding the merger specific job losses.

[36] In BB Investment Company (Pty) Ltd, supra at paras 55 - 61, the Tribunal sought to engage with the question as to what is meant by 'merger specific' losses. In its view, a merger specific job loss is one 'that can be shown, as a matter of probability, to have some nexus associated with the incentives of the new control...'. (para 56) Translated to an inquiry into the public interest effect on employment, the Tribunal sought to distinguish between merger specific employment loss and the 'operational employment loss', which would result regardless of whether the merger takes place or not. The latter job loss is a non-merger one.

[37] It is only the merger specific job loss that bears relevance when determining whether to approve the merger or not. This Court in Wal-Mart, at para 140, in dealing with the questions of retrenchments said 'a retrenchment, which takes place shortly before the merger is consummated may raise questions as to whether the decision forms part of the broad merger decision making process and would, accordingly, be sufficiently closely related to the merger in order to demand that the merging parties must justify their retrenchment decisions.'

[38] Applying this dictum to the present case, it is necessary to examine the reasons proffered by the merging parties for their retrenchment proposal. The merging parties provided a detailed set of spread sheets in which each job loss was categorised and the retrenchment decision justified. Four reasons were given for job losses: the close of shafts in which there was no more ore left to mine, volume reduction due to the end of a project, the implementation of the Sibanye-Stillwater operational model and duplication and consolidation as a result of the merger. In terms of Lonmin's analysis, 6680 job losses were attributable to the close of shafts and 4679 attributable to volume reduction due to the end of a project. These job losses were categorised correctly as operational job losses. They had nothing to do with the merger. They would have also occurred if a counter factual had been applied; that is that Lonmin was required to make a decision with regard to its existing labour force, absent a merger with Sibanye.

[39] In summary, while there might be a dispute between the figures of 885 merger specific retrenchments suggested by Sibanye and 3189 proposed by the Commission, there can be no doubt that, outside of these two sets of figures, none of the other job losses could be said to be merger specific.

The rationality enquiry

[40] Lonmin developed a number of business plans during 2017 in order to deal with its dire financial position. Its precarious business plan is illustrated in the Commission's report. The Commission found that Lonmin experienced losses during the period of 2014 to 2017 as follows:

'1. In 2014, the operational loss was $255m;

2. In 2015, the operational loss was $2 01Bm;

3. In 2016, the operational loss was $322m; and

4. In 2017, the operational loss was $1 079m.'

[41] Its retrenchment plans during 2017 were in direct response to this precarious financial position caused by huge financial losses. It completed a shaft by shaft analysis in order to determine the best manner in which it could reduce its high costs of production; in particular where the production of PGM was relatively more expensive due to, for example, mining methods used, depth of the mine, type and characteristics of ore or concentrate extracted. It explained that the retrenchments proposed in respect of each of the shafts was largely due to the reduced production profile as well as to shafts closing due to their mineral resources being exhausted or capital not being available in a short term to invest in further development of these shafts. Although there were a series of iterations of this plan, the upshot was that the proposed retrenchments, as indicated, were rationally based on Lonmin's precarious financial position and the need to engage in a significant restructuring to save the company from, at the very least, going into business rescue.

[42] To the argument that Sibanye influenced the entire proposal for retrenchment the Commission correctly found that, while Sibanye had indicated that whatever retrenchment numbers had been submitted to by Lonmin business plan, it was clear that these were inadequate, in its view, to save the company from being liquidated. This is made clear in the non binding proposal generated by Lonmin on 28 November 2017:

'Lonmin's standalone business plan envisaged a significant headcount reduction of approximately 1,800 employees over the next 24 months (with a further 2,800 in 2020), primarily as a result of the closure of the Generation One shafts, associated overhead savings and assumed increases in operational efficiencies. Sibanye-Stillwater understands that this process is essential for the sustained viability of Lonmin. We also noted that Lonmin had planned a savings of approximately ZAR500 million per annum in overhead costs from 2019 onwards, primarily associated with a reduced number of employees required to support the downscaled operations. Despite this significant restructuring on the basis of our review of the Lonmin operating plans, we do not believe they are sufficient to ensure the long term sustainability of the company. The primary reason for this is the cost reductions are not sufficient to

offset the very significant capital requirements to sustain mining volumes that ensure Lonmin's survival as a standalone entity.

The jointly developed sustainable business plan incorporates the expected synergies needed to ensure a profitable operation at current commodity prices that also does not put Sibanye-Stillwater at excessive risk. Sibanye-Stillwater continues to refine its thinking around the optimal headcount level based on the revised business plan and currently believes that a further 1800 employees, over and above the Lonmin standalone plan, may be impacted in the next 24 months to ensure the on-going profitability of the operations due to the current economic circumstances.'

[43] This evidence was not and could not be gainsaid. In my judgment it shows that a rational process was followed in order to determine the number of jobs that were to be lost, whether merger specific or not. It is not strictly necessary to decide whether the merging parties figure of 885 merger specific job losses or the Commission's number of 3189 merger specific job losses is correct. The important fact is that all the proposed job losses were rationally connected to the precarious financial position of Lonmin.

The justification enquiry: weighing the questions of loss of employment

[44] This enquiry requires the merging parties to show that the public interest in preventing employment loss is balanced by an equally weighty and countervailing public interest which justifies the job losses and which is cognisable under the Act. In turn, this enquiry requires an examination of the proper counter factual; that is the position absent the merger. Expressed differently, the initial question for determination turns on how many jobs would be lost if the merger does not take place. The Commission conducted an investigation which indicated that there was a risk that Lonmin would have be placed into business rescue absent the implementation of the merger. AMCU submitted however that this concern which had clearly influenced the decision of the Tribunal had not taken account of two important factors, namely the metal purchase agreement concluded between Lonmin and PIM in October 2018 and the improved market conditions.

[45] The agreement with PIM was that PIM advanced $200m loan to Lonmin. Hence AMCU argued that Lonmin's financial position had improved

significantly. But, this is only partially correct. This loan was firstly used to repay existing lenders ($150m), leaving a net amount of $34 m (after transaction costs and cash applied to collateralised guarantees issued by South African banks had been taken into account) to be utilised in its business. Lonmin remained indebted to PIM for the amount of $200m which debt it would have to service at an interest of 15% per annum while the capital would have to be repaid by way of delivery of platinum and palladium to PIM until the debt had been extinguished.

[46] As Mr Cockrell, who appeared with Mr Ngcongo and Mr Wild on behalf of the merging parties correctly noted, the net liquidity improvement $34m was marginal; it was less than one month salary for Lonmin's workforce and approximately 2.5% of Lonmin's annual operating costs. Mr Cockrell pointed out that Lonmin generated sales of approximately $1.1 to $1.3 billion per year and its cost base, which was largely fixed, had the same order of magnitude. A net liquidity of $ 34m was hardly likely to have any significant impact on Lonmin's financial viability or sustainability.

[47] AMCU also contended that Lonmin's production report for the quarter ending 30 September 2008 revealed that it was no longer in a dire financial position. There had been an improvement in the PGM prices which meant that average prices in rand/Oz exceeded unit costs in rand/Oz for the three months to 30 September 2018. But as Mr Cockrell submitted, it was impossible to predict whether the improved PGM prices would continue into the future. Furthermore, Mr Froneman gave credible evidence to the effect that quarterly reports were not a useful way to assess the financial health or position of a company. Examining Lonmin's position regarding its net cash position, despite the variation over the year, as at the end of the fourth quarter of 2017 was US $ 103m and at the end of the fourth quarter of 2018 it was US$ 14m. Nothing fundamental had changed. This is confirmed by the evidence of Mr van der Meiwe, on behalf of Lonmin, who testified further that for Lonmin to invest in new projects and therefore for a significant difference to be made to its long term financial viability, Lonmin would require about US $ 500m before new capital projects could be initiated.

[48] In summary, the correct counter factual was one in which, notwithstanding transient fluctuations in the price of PGM and currency (the value of the rand compared to the $, Lonmin's continued existence was in jeopardy. It had exhausted its capital. It only had

debt funding, absent the merger. It was unlikely to acquire new capital because given its precarious financial position, shareholders

are unlikely to take up a further rights issue and lenders are unlikely to advance any further loans. At best, Lonmin would, as Mr Cockrell described, continue to 'tread water'; that is, if it was not placed into business rescue, which, if it occurred, would hold significant risk for 32 000 jobs.

It is, thus, clear that the merger specific job losses (even taking the Commission's figure of 3189 retrenchments) were vastly outweighed by the potential job losses if the counter factual applied.

Application to admit new evidence

[49] On 20 February 2019 AMCU launched an application to produce new evidence on appeal. The evidence took the form of three publicly

available documents all of which were published after the Tribunal had delivered its decision on 21 November 2019. These documents were a summary of the Annual Report and Results extracted from the Annual Report and Accounts of Lonmin of 29 November 2018, the First Quarter 2019 Production Report and Business Update of 08 February 2019 and a newspaper report of 08 February 2019.

[50] Section 19 (a) of the Superior Courts Act 1O of 2013 does permit a court, which exercises appellate jurisdiction, to receive further evidence. However, as the Constitutional Court said in Rail Commuters Action Group v Transnet Limited tla Metrorail [2004] ZACC 20; 2005 (2) SA 359 (CC) at para 43: 'such evidence must be weighty, material and to be believed. In addition whether there is a reasonable explanation

for its late filing is an important factor. The existence of a substantial dispute of fact in relation to it will militate against

its being admitted.'

[51] In its justification for the admission of this new evidence, AMCU contended that the evidence it sought to adduce was relevant to the determination of the appeal and would be of assistance to the court. It submitted further that the interests of justice compelled its submission.

[52] The test, as set out in Rail Commuters Action Group, is clearly applicable to an appeal in respect of a merger case. Because markets do not remain static and because other economic and financial conditions fluctuate, great care should be taken to ensure that any evidence adduced at the late stage of an appeal is weighty and material, of such a kind which would probably cause the Tribunal to come to a different conclusion. {See: Simpson v SFMED Medical Scheme 1995 (3) SA 816 at 825 D-E.)

[53] The fact that evidence sought to be adduced might show a transient improvement in Lonmin's financial position does not detract from other non­ controversial facts such as Lonmin's lack of liquidity, the continued vulnerability to uncontrollable factors such as exchange rates, commodity prices, high fixed costs, the inability to raise funds for capital projects that could extend the life of some shafts, the depletion of mineral deposits at some shafts and the lack of geographic and metal diversification.

[54] In his answering affidavit in respect of this application, Mr van der Merwe placed this evidence into proper financial context:

'The slight improvement in some metrics (i.e. pricing and weakening in exchange rate) in Q1 2019 does nothing to alter Lonmin's position - in fact, Lonmin's position worsened in the first quarter of 2019 as evidenced by the unit costs increases of around 17% compared to Q1 2018 and 27% from Q4 2018; and a reduction in cash of over $30 million, despite the additional liquidity of $34 million contributed by the PIM refinancing transaction. As explained above, these factors are variable and unpredictable.

Over the course of the last decade, Lonmin consumed $1,6 billion of shareholder equality contributions. The aggregate operating losses of $3 674 billion (over R50 billion at current exchange rates) during financial years 2014 to 2017 is noteworthy in comparison to an operating profit of $101 million for the 2018 financial year and clearly indicates that this profit is immaterial compared to the capital losses suffered by the company in the previous decade.'

[55] Thus, the evidence which AMCU sought to be admitted is, at best, ambiguous in relation to the financial position of Lonmin, post the decision of the Tribunal. This is evident in the appendix which is extracted from the annual report to Lonmin accounts of 29 November 2018, contained in one of the documents sought to be admitted:

'In light of the challenges facing Lonmin and the PGM industry, the Company's strategic response in 2015 was to right-size the business, but costs, enhance working capital management and contain capital expenditure. These initiatives have proved effective resulting in the Company remaining net cash positive. Notwithstanding these improvements and the good performance achieved during 2018, notably form solid production, higher PGM basket prices and weaker USD/ZAR exchange rate, Lonmin still remains financially constrained. Further mitigating measures undertaken during 2018 led to refinancing the business in October 2018 by concluding the $200 million forward metal sale agreement; however this financial measure only provides relief during the short-term and regrettably does not provide an opportunity to avoid retrenchments and shaft closures. In spite of the effectiveness of the measures undertaken, the viability of Lonmin on a stand-alone basis is more vulnerable when compared to being part of a larger group. Consequently, failure to complete the Sibanye-Stillwater transaction will significantly impede Lonmin in funding the significant investment required in sustaining the business in the future.'

[56] This passage reveals the fundamental flaw in the case of AMCU, namely that it sought to cherry pick extracts from the documents which suit its case, whilst eliding over those parts which confirmed or supported the case of Lonmin. Read as a whole the evidence sought to be admitted is not weighty and material as to have caused the Tribunal to come to a different conclusion. There is, therefore, no basis to admit it.

[57] In summary, the merging parties have established that a the determination of the number of jobs to be lost as a result of the proposed merger was rational and that the public interest in preventing employment loss was balanced by an even more weighty public interest, namely the saving the jobs of the vast majority of Lonmin's workforce. Hence, there has been no misdirection on the part of the Tribunal in this regard.

The conditions

[58] Two significant arguments were raised by AMCU in relation to the conditions imposed by the Tribunal. Mr Puckrin criticised the condition which provided that, if certain variables and prerequisites were satisfied, some 3714 jobs could be saved during the period 2018-2020. His criticism was that the prerequisites for these jobs saving proposals depended on the price of PGM and the cost base. In his submission, the key variable should be the margin between price and cost rather than just the price and the cost themselves.

[59] The merging parties conceded to this argument by introducing the profit margin as a further variable. Mr Puckrin's argument has merit and accordingly the confidential condition must be altered in order to reflect the profit margin as the determinative condition in respect of the job saving proposals.

[60] Mr Puckrin also criticised the proposed Agri-lndustrial program condition. He submitted that this was little more than an undertaking

on the part of Sibanye to investigate the feasibility of a program, consult the community and other stakeholders, and should it fail, consider another alternative for two years. Consequently, he submitted, this condition was fraught with uncertainty and might

never come to fruition. It bears noting though that no details had been provided regarding the feasibility study undertaken by the West Rand Steering Committee and/or its potential viability.

[61] The problem with this criticism of this condition imposed by the Tribunal is that the Agri-lndustrial community development program depended on consultation with the West Rand Steering Committee, a party not before this Court and which also fell outside of the control of the merging parties. The entire program depended on cooperation with the Steering Committee. It is therefore difficult in these circumstances to alter the proposed conditions which directly involves a body which was neither a party to the merger, did not appear before this Court, nor did it fall within the control of the merging parties. There is no basis by which this Court can now interfere with the contents of this condition.

[62] Although AMCU succeeded in having one of the conditions altered, in my view, this does not constitute substantial success.

Conclusion

[63] For these reasons the appeal fails, save for the alteration of one condition in one minor respect.

Order

1. Subject to paragraph 2 the appeal is dismissed with costs including the costs of two counsel.

2. Paragraph 3.3 of the conditions outlined in the order of the Tribunal of 21 November 2018 is amended as follows:

'Annexure A1 herewith provides a breakdown of variables and pre­ requisites that would need to be satisfied, as well as technical and economic assessments required to be undertaken, per Short Term Project in order for such job savings to be realised.'

____________________

MGUNI and VALLY JJA agreed