Walmart Stores Inc v Massmart Holdings Ltd (73/LM/Dec10) [2011] ZACT 41; [2011] 1 CPLR 145 (CT) (29 June 2011)

COMPETITION TRIBUNAL OF SOUTH AFRICA

Case No: 73/LM/Dec10

In the matter between:

Walmart Stores Inc …..........................................................................Acquiring Firm

And

Massmart Holdings Limited …................................................................Target Firm

Panel : N Manoim (Presiding Member)

Y Carrim (Tribunal Member)

A Wessels (Tribunal Member)

Heard on : 09-16 May 2011

Order issued on : 31 May 2011

Reasons issued on : 29 June 2011

Reasons for Decision

Approval with Conditions

On 31 May 2011, the Competition Tribunal (“Tribunal”) conditionally approved the merger between Walmart Stores, Inc and Massmart Holdings Limited. The imposed conditions relate to the public interest effects of the proposed deal. The reasons for the conditional approval of the proposed transaction follow below.

The merging parties and their activities

The primary acquiring firm is Walmart Stores, Inc (“Walmart”), a company incorporated and listed on the New York Stock Exchange.1 No individual shareholder directly or indirectly controls it. The only shareholder with a shareholding in excess of 5% is Walton Enterprises, LLC. The following firms are controlled by Walmart: Walmart Stores East, LP; Walmart Property Company; Walmart Real Estate Business Trust; and ASDA Group Limited (“ASDA”). All these companies, except ASDA, which is based in the United Kingdom, are located in the United States of America.

Walmart, the largest retailer in the world, has three retail formats in the form of discount stores (stocked with a variety of general merchandise), supercenters (features products such as bakery goods; meat and dairy products; fresh produce; dry goods and staples; beverages; deli food; frozen food; canned and packaged goods; condiments and spices; household appliances; and apparel and general merchandise), as well as neighbourhood markets (which have a variety of products that supercenters also offer including health and beauty products; stationery and paper goods; drive-through pharmacies and one hour photo centres).

Walmart also has a chain of warehouse stores called Sam’s Club, which sells groceries and general merchandise, often in bulk. Customers buy an annual membership at Sam’s Club in order to be able to purchase merchandise from the club.

Internationally, Walmart currently operates in 15 countries, including Mexico, Puerto Rico, Canada, Argentina, Brazil, Costa Rica, El Salvador, China, Japan, Guatemala, Honduras, Nicaragua, Chile, the United Kingdom, and partnered with Bharti Enterprises in India.

In South Africa, Walmart through ASDA controls International Produce Limited (“IPL”). IPL does not directly or indirectly control any other firm. IPL purchases fresh fruit produce in South Africa for the export market and none of these products are sold back to the South African market. IPL is also responsible for giving practical advice to local suppliers relating to quality standards as well as communicating product information and shipping arrangements to ASDA.

The primary target firm is Massmart Holdings (“Massmart”), a company incorporated under the company laws of the Republic of South Africa and listed on the JSE. No individual shareholder directly or indirectly controls it.

Massmart has in excess of 10 subsidiaries nationwide and around the African continent. It is a wholesaler and retailer of grocery products, liquor and general merchandise. Massmart has four divisions namely: Massdiscounters, Masswarehouse, Massbuild and Masscash.

The Massdiscounters division trades under the names Game and Dion Wired. Game offers a wide range of general merchandise and non-perishable groceries to the value seeking end customers in the LSM 5 to 10 categories throughout South Africa and in Sub-Saharan Africa.2

Masswarehouse comprises the Makro chain of large wholesale outlets, which offers a broad range of food, liquor and general merchandise to commercially affiliated resellers and upper income end consumers predominately in the LSM 6 to 10+ group.

Massbuild comprises the Builders Warehouse, Builders Express and Builders Trade Depot chains, which sell hardware and home improvement/DIY products and building materials. These products are sold predominantly for the construction, augmentation, refurbishment or decoration of homes owned by consumers in the LSM 6 to 10+ group.

Massmart’s food and grocery offering to the low-end customers is predominantly at the wholesale level and through its Masscash division, it is also active in the retailing of grocery products. Masscash also comprises of retail/hybrid outlets, which sell grocery products, liquor and general merchandise directly to lower income customers in the LSM 2 to 7 socio-economic groups. The stores in the group include Buy-Rite, Sunshine, Mikeva, Cambridge, DF Astor Savemoor and Score (trading as Saverite).

The proposed transaction

On 27 September 2010 Massmart announced Walmart’s intention to acquire a controlling interest in Massmart.3

In terms of the proposed transaction Walmart intends to acquire 51% of the ordinary share capital of Massmart.

The rationale for the proposed transaction

Walmart wants to enter emerging markets, specifically South Africa and sub-Saharan Africa, accounting for approximately 20% of the consumer spending on the continent as a whole. Further, Walmart believes South Africa is sophisticated and has a stable economic, political and regulatory environment. South Africa therefore represents an attractive market on its own to Walmart.4

Massmart’s current strategy entails a comprehensive planned investment in expanding its operation in South Africa and further on the African continent. Walmart is renowned for its operating, retailing, marketing and merchandising skills and procurement and supply chain capabilities. Massmart is of the view that given Walmart’s collective skills and capabilities, they will enable the merged entity to implement its pre-merger expansion plans with more confidence and on an expedited basis, as the merged entity will be able to draw on skills, systems and processes already developed, tried and tested by Walmart.

Massmart also anticipates that Walmart, being a global leader in sourcing and retailing of fresh produce, will introduce new skills and technologies to assist Massmart in becoming a significant distributor of locally produced, perishable products, thereby complementing and supporting Massmart’s emphasis on expanding its fresh grocery operations.

The transaction will enable Massmart to gain access to Walmart’s procurement capabilities through a buying agency agreement and various other services (i.e. technology software and hardware, merchandise skills and other technical skills and services).

The intervening parties

Prior to the commencement of the hearing of the proposed merger, the South African Commercial Catering and Allied Workers Union (“SACCAWU”), the Congress of South African Trade Unions (“COSATU”), the Food and Allied Workers Union (‘FAWU”) and the National Union of Metal Workers in South Africa (“NUMSA”) (collectively referred to as ‘SACCAWU et al’)5, the South African Small Medium and Micro Enterprise Forum (“SMMEF”), the South African Clothing and Textile Workers’ Union (“SACTWU”), the Minister of the Economic Development Department (“EDD”), the Minister from the Department of Trade and Industry (“DTI”) and the Minister from the Department of Agriculture, Forestry and Fisheries (“DAFF”) (collectively “the Ministries”) filed notices of intention to intervene in the merger.

Hearing and witnesses

The hearing took place during the period from 9-13 May 2011 and argument was presented on 16 May 2011. The following witnesses gave evidence at the hearing:

For the merging parties:

As factual witness Mr Grant Pattison (“Pattison”), the Chief Executive Officer of Massmart Holdings Limited.

As factual witness Mr Andy Bond (“Bond”), the former Chairman of ASDA Stores Limited and an executive vice president of Walmart Stores Inc. 6

As factual witness Mr Enrique Ostale Cambiaso (“Ostale”), the Chief Executive Officer of Walmart Chile S.A.7

As economics expert, Mr. Simon Baker (“Baker”) from RBB Economics8.

For SACCAWU et al:

As expert witness Ms Sofia Scasserra (“Scasserra”), economic advisor to the Argentine Federation of Commerce and Service Workers (“FAECYS”).

As expert witness Mr Kenneth Jacobs (“Jacobs”), Chair of the University of California Berkeley Center for Labour Research and Education.

As factual witness Mr Noel Mduduzi Mbongwe (“Mbongwe”), the Deputy General Secretary of SACCAWU.

For SACTWU:

As expert witness Mr Etienne Doyle Vlok (“Vlok”), Director of the SA Labour Research Institute (“SALRI”), the research wing of SACTWU.

For the Ministries:

As expert witness Mr James Hodge (“Hodge”) from Genesis Analytics.9

Called by the Tribunal:

As factual witness Mr Gerhardus Ackerman (“Ackerman”), Head of Food Buying at Shoprite.

The following person’s statements formed part of the record but the individuals concerned were not called upon to give oral

testimony: Debra Layton10; Labour Relations Services (author not attributed) 11; Annette Bernhardt12; John Logan13; Stephanie Luce14; Barry Lynn15; Scott Nova16; Ashwini Sukthankar17; Claudio Alvarez18; Nelson Lichtenstein19; Alex Mahubetswane Mashilo20; Dannyboy Katishi Masemola21; Gaositwe Tebogo Khaas22 and Richard Michael Levin23.

While these statements have been considered they are given less weight than the evidence of witnesses who gave oral testimony and who were therefore subject to cross-examination.

Commission’s reasons

On 2 February 2011, the Commission finalised its investigation of the proposed merger between Walmart and Massmart. It found that the merger was not likely to lead to a substantial prevention or lessening of competition. In considering the public interest issues arising from the merger, the Commission looked at (i) pre-merger retrenchments at Massmart; (ii) the effect of the merger on suppliers; (iii) the effect of the merger on employment generally; (iv) the effect of the merger on the future terms of employment of Massmart employees and (v) the right to association and acceptance of unionized labour.24 Upon considering the transaction the Commission recommended to the Tribunal that the merger be approved unconditionally.

During its investigation the Commission engaged with labour unions SACCAWU, SACTWU, FAWU and the SMME Forum.

It was also brought to the Commission’s attention that the Economic Development Department had engaged with the merging parties and the trade unions to address the public interest issues and clarify certain commitments made by the merging parties.25 This process was however not concluded at the time that the Commission was required to submit its recommendation. There is some controversy over this process that emerged in a postponement application which we heard at the commencement of the original hearing dates. The merging parties and the government have given different accounts as to why this process did not result in any resolution. It is not relevant for our decision to go into this.

In its closing argument the Commission indicated that it had revised its position and decided to recommend a conditional approval of the merger.26 The evidence that the Tribunal considered differed in important respects from that considered by the Commission during its earlier investigative process. In our proceedings we have had the benefit of further discovery of documents at the instance of the government

departments’ team, and the testimony and examination of witnesses brought by these intervenors. This explains why the Commission changed its recommendation. We commend the Commission for not taking a static approach to the proceedings.27

The relevant market and the impact on competition

It is common cause that this merger raises no competition concerns. Walmart does not compete with Massmart in South Africa and its only presence in the country is a small procurement arm that sources local products for its stores globally.

In light of the above, we find that the transaction would not substantially prevent or lessen competition in any of the markets that Massmart presently operates in.

The public interest

One of the unusual features of the Competition Act, 1998 (Act No. 89 of 1998, as amended) (“the Act”) is that despite the fact that a merger may raise no competition concerns it may still be susceptible to prohibition, or approval subject to conditions, on public interest grounds.

In terms of section 12A of the Act even after the so-called ‘pure’ competition grounds have been evaluated and a conclusion reached that a merger does not lead to a substantial prevention or lessening of competition, the Tribunal must:

12A(1)(b) “…otherwise determine whether a merger can or cannot be justified on substantial public interest grounds by assessing the factors set out in subsection (3).” (Our emphasis)

Thus the public interest consideration is not open-ended. Subsection (3) limits this consideration to four factors, namely the effect the merger will have on:

a particular industrial sector or industry;

employment;

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

the ability of national industries to compete in international markets.

This merger concerns the effects referred to in subparagraphs (a), (b) and (c).

Subject matter and substantiality are not the only limitations in considering the public interest. A further consideration is that the public interest consideration must be merger specific. Expressed in less technical language, unless the merger is the cause of the public interest concerns, we have no remit to do anything about them. Our job in merger control is not to make the world a better place, only to prevent it becoming worse as a result of a specific transaction. This narrow construction of our jurisdiction has not always been appreciated by some of the intervenors who have sought remedies whose ambition lies beyond our purpose. It is not our task to determine whether those ambitions are legitimate public policy goals; only whether they lie within our powers.

The fact that a concern exists independently of a specific merger, however weighty that concern may be, does not bring it within our jurisdiction in performing merger adjudication.

A survey of the merger decisions since the Act came into being shows that in no case has an adverse public interest condemned an otherwise unproblematic merger, nor has a problematic merger from a competition perspective been rescued on public interest grounds. This does not mean that no public interest grounds have been found to exist in a merger context. Rather these have been thought adequately addressed by the imposition of conditions on a particular transaction.

The Tribunal has under certain circumstances been reluctant to venture too far in its public interest mandate. As an early decision showed, the Tribunal considered that it ought to show deference to other regulators when it tread upon territory outside its expertise.28 This however does not mean that the Tribunal has shied away from its responsibilities under the Act and in numerous cases conditions have been imposed to protect unjustified employment loss post merger. 29

The Tribunal has also viewed its public interest mandate as linked to its competition analysis.30 Although this does not go as far as amounting to a balancing exercise as required with an efficiency analysis, it also does not mean that the competition and public interest considerations are analysed without regard to each other. The choice of the language of justification as we decided in Harmony/Goldfields suggests this.31

It is not necessary in this decision to reconsider any of this jurisprudence. Since the merging parties offered their undertakings the issue for us to consider is whether these undertakings are sufficient. The merging parties made it clear that the undertakings were not offered out of any sense of legal compulsion but rather as a goodwill gesture. Whether this is the real reason or that in truth they felt compelled to react to the evidence led we do not know. But their motive for doing so is irrelevant. What matters is whether the undertakings were adequate to address the public interest concerns in this case.

The approach we have taken is to examine the undertakings and the evidence to which they are responsive to see whether they are adequate. Usefully, each undertaking matches, thematically, each of the material public interest concerns raised during the hearing.

Retrenchment moratorium

There is no evidence from the internal documents of the merging parties that retrenchments at Massmart are contemplated as a consequence of the merger. On the contrary, there is evidence that suggests, given the expansionist ambitions of Massmart, the group expects employment to grow between 2011 and 2013.32 The merger is expected to expedite this expansion suggesting that new jobs are likely to be created more quickly as a result. (Full-time employment positions are expected to increase by 2796 in 2011, 3147 in 2012 and 3147 employees in 2013).33 This was confirmed during the hearing by Pattison who stated that:

“...if nothing else changes, Massmart will create a significant amount of jobs over the next 3 to 10 years.” 34

All this might suggest that no retrenchment undertaking was necessary. This conclusion should be treated with caution for two reasons. First, expansion may take place outside of South Africa - internal documents and communications frequently do not make clear the distinction between South Africa and the rest of Africa, and we know that the merged entity post merger, intends to expand in other African countries – indeed this is a significant part of the rationale for the deal; second, Massmart’s current employment practice is to consider employment divisionally and not from the perspective of the group as a whole.35 It is thus possible that although some divisions may be creating employment, others may be contracting. Indeed the lesson from the 2010 retrenchments, which we discuss more fully later, is that retrenchments took place in a particular area at a time when group expansion was being contemplated in others.

The third concern arises from comments from Walmart executives in the recently discovered correspondence between the firms. Some Walmart remarks, albeit cryptic, made during the due diligence process tend to suggest that they consider that Massmart carries too many staff on its shop floor.36 Despite these remarks there is however nothing in the merging parties’ synergies document (which details savings that will come about as a result of the merger) that suggests retrenchments of the current work force are contemplated.37 On balance, retrenchments are, post merger, a possibility, but the more likely scenario is that either the workforce size will remain constant or will expand.

Despite this, the merging parties have still given an undertaking in respect of future retrenchments. This is set out in paragraph 1.1 of the Tribunal’s order. The undertaking is similar to the one imposed by the Tribunal in the Metropolitan merger.38 The one difference is the exception created for unreasonable refusals to be redeployed.

Since this is not a merger where redundancy is likely post merger - as was the case in Metropolitan - because in this merger there is no operational overlap in South Africa, the likelihood of merger specific retrenchments being disguised in the form of unreasonable redeployment is significantly less compelling than it was in Metropolitan.

The time period of the undertaking in regard to merger specific retrenchments accords with that in Metropolitan. Although SACAWU had sought three years we consider two years as being adequate. A longer period does not seem warranted given the probabilities of job creation being more likely than job loss going forward.

Reinstatement of retrenched employees

In June 2010 Massmart retrenched a number of employees who worked for Game in Nelspruit. A number of other employees working for regional distribution centres (RDC’s) were also retrenched. The total number of retrenched employees appears to have been 503. The union alleges that these retrenchments came about in anticipation of the merger. This is based not on any direct evidence, but on inferences about the timing of the retrenchments relative to the final phases of the negotiations in respect of the merger. We deal with this more fully below.

Game falls under Massmart’s Massdiscounters division. Recall that Massmart deals with its employees on a divisional and not on a group wide basis. Massmart had previously had two Game stores operating in Nelspruit. It decided to consolidate them into one large store in the newly established Ilanga Mall. It cited as the reasons for doing so, the difficulties of renegotiating a lease at one of the existing sites at the Riverside Mall, the fact that the two stores served the same catchment area and that they would have larger premises at Ilanga Mall.39 In short, Massmart advances operational reasons for the retrenchments independent of any merger specific considerations.

A second group of retrenchments at about this time occurred when Massmart was conducting a process of what it termed re-engineering

of its RDC’s. It appears that as part of this re-engineering, Massmart concluded that it needed fewer employees in these centres and so a number of them were retrenched. SACCAWU however disputes this and claims this was part of a process of casualisation and that now more casuals are employed through services provided by labour brokers than were the original number of full time employees.

The evidence in this respect is confusing, with neither side presenting a coherent picture of what happened and often talking past each other. However what seems not in dispute is the total number of workers affected by these retrenchments at Nelspruit and the distribution centres.40 The number asserted by the unions is 503 and the merging parties’ undertaking is in respect of the same number. It is not necessary for us to be certain as to whether these numbers all emanate from Nelspruit or from both Nelspruit and the distribution centres, since both the merging parties and the union agree on the total figure.

What is in dispute is the remedy. The union had as its primary demand that we impose a condition ordering reinstatement or re-employment

of all the affected employees.41 In the alternative however the union asks that the affected employees be the first to be hired as employment opportunities arise in the future in the Massmart group as a whole. It is this alternative undertaking that Massmart has in fact met, as contained in paragraph 1.2 of the Tribunal’s order. The Tribunal’s order furthermore requires Massmart to take into account the affected employees’ lengths of service with Massmart.

We note that the merging parties’ undertaking fully meets SACCAWU’s alternative request for a condition. Should the merging parties have given an undertaking to reinstate or re-employ the affected employees immediately?

Although in Metropolitan we held that the burden of justifying merger specific retrenchments fell to the merging parties in this case the burden has not yet shifted.42 This is because the retrenchments took place prior to the merger. The union would first need to show that retrenchments were merger specific. Only then would the burden of justification shift to the merging parties. The difficulty for the union is that they have not been able to cross this first hurdle. Massmart has given plausible reasons for the retrenchments that are not merger specific. The union would need to show on a balance of probabilities that this explanation is untrue and that but for the merger, the prior retrenchments would not have happened. It has not been able to prove this.

As we stated earlier, the union places primary reliance on the coincidence in timing. We know that Massmart has been courting an offer from Walmart for some time; which according to its chairman Mark Lamberti goes as far back as 1990.43 Whatever Massmart’s hopes were before then the relationship with Walmart only became a possibility in February 2009, when the two firms entered into discussions and signed a confidentiality undertaking. This type of undertaking is not unusual and contains the standard boiler-plate clauses that firms use when one is considering an offer and neither wants anyone else to know of it.

The Walmart evidence is that at the time Massmart was one of three possible South African targets under consideration and that

confidentiality agreements had been signed with the other two potential targets as well. Moreover, Walmart had not yet even concluded that it was willing to do a deal in South Africa as it was deciding between this country and another unnamed country. It would only choose to invest in one of the two countries so an investment in South Africa was not a foregone conclusion. The upshot is that on the Walmart version the date they decided to go with South Africa and Massmart, coincides with the 27 September 2010 public announcement of the deal. The deal was taken to the Massmart board in a rush over a weekend on 26 September 2010 and then announced to the market on the following Monday. Pattison for his part has also denied any linkage.44 This despite the fact that the record shows many meetings between himself and Walmart between 2009 and the date of the offer becoming public.45 He said that these discussions were a normal part of commercial negotiations between firms and did not commit either party to one another. As he expressed it pithily in his evidence – there may have been coincidence between the retrenchments and the deliberations with Walmart but no causality.46

The coincidence in timing of the deal’s imminence with the retrenchments is not strong enough to show its connection. Even if the operational justification for the retrenchments were exaggerated – we express no view on this – this might make, at best for the union, an unfair retrenchment scenario, but not a merger linked one. There is no evidence for instance that Walmart was requiring Massmart to engage in these particular retrenchments or that it knew of them at the time.

The remaining theory would then be that Massmart effected the retrenchments to entice Walmart’s bid i.e. even if no overt

agency between Walmart and Massmart can be discerned Massmart’s executives anticipated that Walmart would like to see some

downsizing of their labour force and that it would be expedient for them to make this demonstration in 2010 whilst the game of suitor pursuing bridal prospect was taking place. But it seems unlikely that given the total size of the Massmart labour force – about 26 500 – that this figure of 503 affected employees would prove material in persuading Walmart that Massmart was a sweeter prospect than its rivals.47 Whilst it is true that some of the due diligence reports done by Walmart, which we refer to earlier, might suggest that Walmart would prefer a leaner Massmart, there is nothing to suggest the former’s hand in the latter’s earlier retrenchments.

Indeed the merging parties have discovered all the correspondence between them during the period from when the confidentiality agreement was signed in February 2009 until the deal was publicly announced in September the following year.48 There is nothing in this documentation that suggests that Walmart was informed of the retrenchments or showed a specific interest in day to day employment issues at Massmart. The email that comes closest to this is a request from a Walmart executive to Massmart to inform them of the percentage of employees unionised and the duration of contracts with unions. The request however is of a general nature and cannot be linked to the retrenchment issue.

Prior to 2010, Walmart had a third party prepare a document for it on, inter alia, labour disputes. Labour conflict at Massmart in 2009 is mentioned, but the conclusion is that these issues would not affect Massmart’s ability to operate or its reputation. Thus this earlier report does not signal alarm bells about labour problems or the need to downsize.49 In yet another internal document prepared for Walmart in May 2010, and thus at the height of the retrenchment battle, labour issues at Massmart are also discussed; but there is no mention of the proposed retrenchments at all nor of the need to downsize the labour force. Rather the document emphasises the need for Walmart to deal with negative perceptions about its labour relations governance policy.50 Whilst not conclusive of anything taken on their own, the document trail is consistent with the merging parties’ version that Walmart had no involvement in Massmart’s retrenchment decisions in 2010. If they had it is likely that some mention would be made of this in Walmart’s internal documents.

On the whole, it is likely that the retrenchments were not merger specific albeit they may have been poorly handled as Massmart itself concedes. The union’s dissatisfaction with the management of the retrenchments was exacerbated when news reports, which had surfaced about the merger, were raised with management by the union in March 2010, but were met with the obdurate response that the firm would not respond to media speculation. It is not surprising that when the deal was announced in September that year, employees still bruised from the retrenchment battle became highly suspicious. To add insult to injury Massmart in its haste to announce the deal to the market implied that unions had been consulted about it. In fact Pattison conceded that SACCAWU leaders had only received an SMS from him prior to the press release further adding to their ire. In the circumstances whilst the retrenchments cannot be evidentially linked to the merger, the undertaking to give preferential employment opportunities to the 503 has been prudently made, but absent the showing of merger specificity cannot be expected to have been made an immediate offer of reinstatement.

Collective Bargaining

Walmart’s attitude to collective bargaining was a central issue in our proceedings. It is not necessary for us to go into all the detail with which these issues have been covered in the witness statements and the hearing as the undertakings made in this respect meet at least some of the core demands of the intervenors. Two undertakings were made; an undertaking to continue to honour existing labour agreements and an undertaking not to challenge the status of SACCAWU as the largest representative union within the merger entity for an appropriate period determined by the Tribunal. We have determined that this period should be three years. These undertakings are found in paragraph 1.3 of the conditions.

Walmart is the largest private employer in the world employing an estimated 2.1 million employees.51 Of this total almost two-thirds are employed in the United States. Not one of the workers in the U.S. belongs to a trade union. This is explained by the witness called by the unions, as due to Walmart’s origins in the southern states of the US with their traditional antipathy to organised labour and their philosophical preference for individualism.52 Also cited as concerns are the number of labour violations alleged by workers in the United States some in cases won against Walmart others in out of court settlements. Allegations range from discriminatory practices to what is termed wage theft.53 Taken on their own,these figures are daunting.

Walmart maintains two lines of defences. Firstly, in the US it claims that the number of labour related charges and complaints it has received are lower than the US national average. It says that notwithstanding that it employs 1% of the US workforce complaints against it represented only 0.06% of the charges filed with the National Labour Relations Board.54

The unions also produced an undated pamphlet, which Walmart apparently hands out, or once handed out, to store mangers entitled “how to deal with an approach from a union organiser”. The unions do not explain how they got the document. It was presented to Pattison during a meeting but presumably emanates from

one of the international solidarity groups. Confronted with this Walmart feigned ignorance; neither denying its authenticity nor seemingly able to account for it.

Although Walmart attempts to answer each of the charges levelled against its US labour relations record, we consider the unions have raised sufficient concerns about the firm’s attitude to collective bargaining. Thus the second leg of the defence is the more important one. This is the one advanced by Bond who was the only Walmart person to testify on this issue. Bond has not been US based, so he could do little to advance the cause of labour respect for his firm in its home country. He spoke largely from his United Kingdom experience. Bond is a former Chief Executive Officer and chairman of ASDA, Walmart’s U.K. subsidiary, which is the second largest grocery retailer in the UK. Here he said organised labour rights had always been respected by Walmart and that there was a healthy relationship with unions. Yet confronted with an incident at one of ASDA’s depots, where employees had been offered a bonus not to join the union, leading to an adverse finding by a UK Employment Tribunal he conceded that:

“...we very inappropriately handled the situation of the union representation in that site and were found to have balloted members

illegally and we were fined accordingly. I recognise it was a wrong thing we did and we recognised that at the time”55

This blemish notwithstanding, he testified Walmart respected the labour dispensation of the country in which it operated and the same would happen in South Africa. On this approach Walmart’s attitude to labour relations is chameleon like – it changes its colours to suit the immediate environment. The best evidence for this came not from Walmart, but from one of the unions’ witnesses. Scassera, an advisor to an Argentinean commerce and service workers union, testified that the labour environment in her country was demanding, requiring firms to centrally bargain with trade unions. Walmart had complied with this when it entered Argentina.56

The other evidence of Walmart’s attitude to unions is more specific to the South African situation. It concerns the correspondence post offer, between Walmart and Massmart. Whilst Walmart does not display what its strategic intent will be with regard to unions, the tone of the emails and the kind of questions asked whilst susceptible to different interpretation is at least, on one reading, consistent with the unions’ characterisation of it as a company not well disposed to collective bargaining.57 The undertakings made to respect present labour agreements and to continue recognition of SACCAWU for a period of three years are therefore appropriate.

Therefore the debate between the unions and the merging parties shifts as to whether further collective bargaining undertakings are justified. The unions sought a large number of demands under this rubric. Some were scarcely credible and the union legal team tactfully explained them away as acts of solidarity.58 Mbongwe when cross-examined abandoned reliance on many of these demands. The core demands the unions seem to make, are centralised bargaining and a closed shop. As noted earlier, Massmart, presumably for historical reasons owing to its growth by acquisition, bargains with labour per division. As a result wages for equivalent jobs vary per division.59 The union has been demanding that Massmart must bargain centrally and not per division. Other rivals such as Pick n Pay and Shoprite are said to bargain centrally. Thus far Massmart has not agreed to do the same. The same can be said of the closed shop demand.

It may well be that these demands of the unions are legitimate. However this has been Massmart’s position prior to contemplation of the merger. In short, Massmart may well lag behind its rivals in terms of attitudes to collective labour relations. But the question to be answered in this forum is whether the merger has brought about this attitude. All the evidence suggests that Massmart’s approach to centralised bargaining and a closed shop, is a policy pre-merger and there is no evidence that this policy has been formulated in conjunction with or in anticipation of the merger with Walmart.

There could be grave dangers if the Tribunal imposed itself on labour issues that must be thrashed out at the bargaining table. Whilst in this case protecting existing collective rights is a legitimate concern that our public interest mandate allows us to intervene on because we are protecting existing rights from the apprehension that they may be eroded post merger, we must be careful of how far down this path we go. Protecting existing rights is legitimate, creating new rights is beyond our competence. Recall our earlier jurisprudence about proper deference in matters in which we are not an expert. But the dangers of travelling further down the path of collective bargaining intervention is that we risk upsetting the balance of the quid pro quo that accompanies the winning of collective bargaining rights. Massmart might trail its rivals in recognising central bargaining, but we do not know if rivals have won concessions from unions that Massmart still seeks to extract before making such a concession.

Secondly, the unions need to appreciate that rights are symmetrical. If the Tribunal intervenes today to impose collective bargaining and a closed shop on an employer, we may create a precedent for intervening in collective bargaining more intrusively than is prudent given our limited mandate. Consider a hypothetical merger where the merging parties want to sanitise an anti-competitive merger by an undertaking to increase employment, but the proviso for doing so is a union concession that it will accept more flexible job hours. Not being able to achieve this at the bargaining process they seek to impose it through a condition on the merger, based on public interest grounds. The unions would no doubt consider that that is none of our business. We step cautiously into shop floor issues less we forget our purpose as a competition regulator.

We thus find that unlike the demand to protect existing labour rights which may well be merger specific, the creation of additional rights not presently enjoyed by unions is neither merger specific nor appropriately part of our limited public interest mandate in respect of effects on employment.

A concern was also raised that post merger, individual employee rights would also be degraded. Given Walmart’s history in the US, so it was alleged, wages and salaries would lag behind those of the industry and other employee rights would be diminished.60 The merging parties deny this will happen, arguing that if they do not offer competitive remuneration and working conditions they will not be able to attract competent employees which will undermine the competitiveness of the Massmart business. However if Massmart should, contrary to this protestation, attempt to lower levels of remuneration to below that of the rest of the industry, the strong protection given to union recognition in this undertaking would empower it to resist that tendency. Any remedy to extend a condition into setting levels of remuneration post merger would be disproportionate and an inappropriate interference with the process of collective bargaining.61

Procurement

Post merger domestic procurement of the merged entity was an issue for all the intervenors but was mostly explored by the government team and the team representing SACTWU.

Put at its simplest, the intervernors case is that pre-merger Massmart has some, but limited, capacity to import products - this depends on what the products are and Massmart’s current retail and wholesaling profile. Post merger, all this will materially change. Massmart will expand into other products – food and clothing being examples - and take market share from rivals. This acquisition of market share will be driven by lower prices. The source of the lower pricing will be Walmart’s superior buying power in sourcing goods from overseas. As a result rival firms will also have to re-engineer or form alliances to secure imported goods at prices that are competitive with those of a combined Walmart/Massmart. The result is a significant shift in purchasing away from South African manufacturers to foreign, particularly low cost, Asian producers. This portends a decline in demand for the products of these domestic firms who will have to close or downscale with a significant cost to local jobs.

Two solutions are suggested by the various intervenors. Either the merger is prohibited or is permitted subject to conditions to address this harm to the public interest in employment, industry sectors, BEE businesses and small businesses.

The condition favoured by some of the intervenors is the imposition on the merged entity of an import quota. They say the merged entity should, for a period, have its imports subject to some form of limitation. As we shall see later the devil is in the detail.

Although other witnesses contributed to it, the primary debate on the domestic procurement issue was between the two expert economists – Baker of RBB who testified for the merging parties and Hodge of Genesis, who testified for the Ministries.

Baker wrote the first report. In it he argued that we could predict what might happen to procurement in South Africa under a post merger scenario by examining what had happened in Chile when Walmart acquired a local firm Distribucion y Servicio D&S S.A. (“D&S”). Chile was chosen, he argued, because it had an economy of an equivalent size and level of development to South Africa, and Walmart’s entry was still recent enough to constitute a useful comparison.

The conclusion he came to was that there had not been a noticeable shift from procurement from domestic producers to imports and that domestic producers including small businesses were considerably better off.

Baker also made an attempt to show that Massmart at present is not heavily reliant on imports. This exercise was to prove unreliable, given that Massmart’s main defence against procurement conditions was to assert the impossibility of determining the extent of local manufacture in the products that they sell. If Massmart cannot perform this exercise credibly, neither can Baker.

Baker is not able to say much about a very important question in this merger. Can Walmart post merger source goods from overseas, in particular Asia, more cheaply than can Massmart and if so, will it? He cannot answer this because his clients have not told him and he has not performed this exercise. Hodge had wanted to, but when he asked for the data to do this exercise in a discovery application, the merging parties raised insuperable difficulties, contending it would lead to indeterminate collateral issues. We accepted this at the time and did not compel this information. It is highly probable that if Massmart was procuring at prices near to those of Walmart, this exercise -, entirely within the knowledge of the merging parties- would have been done. Is it likely that the two firms did not at some time, over their lengthy contact, not explore this possibility?

Baker then had to rely on repeating the central themes that Pattison and Bond had told him; savings contemplated in the synergy documents would come about not by substantial imports substitution but rather the end of disintermediation. To use more colloquial language, under Walmart, Massmart will cut out the middle man for already imported goods. With the middle man removed imports will remain at similar levels, but the margins made by the intermediary will be eradicated or reduced through Walmart’s superior procurement logistics.

There is something improbable in all this. First, it is counter-intuitive that a firm with Walmart’s superior buying power would not be sourcing any number of products more cheaply than Massmart can and that it would not take the opportunity to do so. Second, press clips put to Bond in cross-examination show Walmart crowing about how costs of procurement had been slashed in the UK post merger62. It is unlikely that such extensive cost reductions were solely attributable to disintermediation.63

From the merger record it is evident that Walmart has what it terms a ‘procurement supply toolkit’. It is an IT system with enough information programmed into it to enable it to plan where to purchase at any given time and from whom as a result of its worldwide operations. This toolkit it seemed might have been used to make some assumptions on cost savings in the Massmart business. At least so it appeared from some documents in the record. However the merging parties assert that this exercise was never done and we have to accept this assertion. The fact however, is that the toolkit exists and there is no reason it would not be made use of post merger.64

On this aspect the intervenors have the more likely version. Imports of goods will increase because Walmart’s superior buying power and logistics will allow for this. What is however not clear is its extent. Here it is where the intervenors have to make assumptions and run into difficulties.

Hodge filed a report in response to Baker’s report. He went to considerable lengths to show why the Chile model was not well founded. Not only had currency fluctuation factors not been taken into account of in Baker’s model, but also Chile’s

greater distance from Asian markets made it less likely to be effected by cheap Asian imports than would South Africa.65 Chile in short is not a good proxy. At the same time cross examination of Baker by governments’ counsel suggested that the choice of Chile was opportunistic, rather than illustrative, and that there was not a good reason why other countries where Walmart is present were not considered and factored into Baker’s analysis.

Having spent half his report refuting the comparative value of the Chilean experience, Hodge then used economic modelling to examine the employment impact of an assumed 1% switching in Massmart’s total procurement away from domestic procurement to imports. Hodge more specifically used employment multipliers to estimate how reduced domestic production (through greater imports) would impact on employment in the supply chain. For this he used the multipliers that were calculated using a 2000 Social Accounting Matrix (SAM)66 constructed by James Thurlow. Hodge’s modelling exercise ultimately estimates that there would be a potential loss of roughly 4000 local jobs for the assumed 1% shift in Massmart’s procurement to imports and away from domestic procurement.

The model makes a number of assumptions - a fact Hodge fairly concedes. The first is the above-mentioned 1% change in domestic procurement in favour of imports. This figure is not based on any evidential or probable finding. It is simply an assumption.67 Secondly, the calculation of the multipliers is based on the assumption that the employment output elasticity equals 1, i.e. a 1% increase in output leads to a 1% increase in employment. This is artificial, a fact that Hodge concedes as well. For instance this elasticity may vary per industry, e.g. it would be different for food and textiles and again is based on assumptions which have no real empirical basis.

The further criticism of Hodge’s calculation is that he fails to take into account so-called ‘second round’ or further multiplier effects, for example lower consumer prices resulting in possible job creation as opposed to job losses or increased exports from South Africa to other parts of Africa as Walmart expands into these areas and uses some competitively priced South African products to sell into these markets. Hodge conceded that the potential job creation associated with a hypothetical post merger consumer price reduction of about 5% by the merged entity could be as high as 20 000 jobs being created.68

In short Hodge’s economic modelling, whilst embodying a creative attempt to quantify the potential job losses, lacks sufficient

rigour to be relied upon as an accurate measure of the ultimate post merger employment effects.

A third witness to enter this debate was a witness called by the Tribunal, Ackerman, the head food buyer of Shoprite Checkers.

Shoprite is arguably the largest retailer in the country at the moment. It is likely to post merger face severe competition from the merged firm. Ackerman testified that the only way Walmart would be able to gain significant market share would be by reducing prices. They could do this as he put it as the “number one store in the world”. If that happened he testified Shoprite would have to respond. Part of that response would be to source products by way of imports. “If they were to import Shoprite would import”69

He gave as an example that Shoprite procures pasta from a local pasta manufacturer. If the product was sourced from Turkey this would lower procurement costs by 40%. If Shoprite was forced to price more competitively and use the imported product the local factory would have to close.

Ackerman’s view was that the response would be on a product for product basis rather than a generalised strategy to import.

Shoprite is a competitor of the merged firm and whilst we have no reason to be critical of Ackerman’s testimony, one must approach it with some caution. When asked by the Tribunal what Shoprite’s view would be on an import quota imposed on Massmart, Ackerman was perfectly frank. He stated that it would suit Shoprite as “whatever benefit they [Walmart] could get from the international sourcing point of view could be mitigated” but he added that he was not willing to express any view on the merger.70

But Ackerman’s testimony was more nuanced than making the cost from the supplier the sole consideration as to whether to import. It was not a simple matter to drop a local supplier, which had become uncompetitive because local supplier relationships were very important. If a firm was too quick to drop local suppliers and then had to go back to them later this might prove problematic. Thus security of supply is also an important factor from a retailer’s perspective.

Ackerman also distanced himself from a comment in the Shoprite letter to the Commission that suggested that importing through containers would be difficult for other firms as they may not be able to fill a container. He stated that the large competitors of Walmart in South Africa would not suffer from this disadvantage. Smaller firms that wanted to import might however have to use middle men adding to their costs and hence would be disadvantaged.

Ackerman also conceded that imports are less likely for perishables, a point made by the merging parties. Perishables are one area in which post merger the merging parties want to expand the Massmart offering.

What conclusions can we reach on this debate over domestic procurement? A change in the merged entity’s procurement patterns are likely, notwithstanding the merging parties protestations to the contrary. Its quantitative impact however is less clear. It is also likely that if Massmart post merger becomes more reliant on imports, then as a result other rivals will have to react and this may lead to greater substitution of imports over locally manufactured goods.

But we also should not assume that South African producers will not fight back to maintain market share. More likely they will do so. Given that in some sectors as Pattison testified supplier markets are characterised by oligopolies the threat by retailers that they can credibly switch to foreign sources of supply may mean that they can bargain for better prices. Greater retail competition post merger will ensure that the benefits of lower producer prices are passed on to consumers and not pocketed by retailers.

There is no doubt that the intervenors raise a valid concern. The problem is that the concern raised in relation to local procurement/imports is also associated with important benefits for consumers. A possible loss of jobs in manufacturing of an uncertain extent must be weighed up against a consumer interest in lower prices and job creation at Massmart. Since the evidence is that the likely consumers who will benefit most from the lower prices associated with the merger are low income consumers and those consumers without any means of support of their own , thus the poorest of South Africans, the public interest in lower prices is no less compelling.

This is the context in which we must consider the conditions proposed by the intervenors in relation to the domestic procurement concern. We have noted already that the extent of this concern is by no means clear and in its most articulate form in Hodge’s report, relies on a number of unproven assumptions and ignores the potential pro-employment effects post merger of lower prices in increasing consumer incomes and the export opportunities for South African manufacturers in Africa that the merger could create, at least on the merging parties’ version.

The procurement conditions proposed by the intervenors have varied during the course of the merger process but the final proposal made by the SACTWU legal team represents the most considered form in the sense that it has tried to meet some of the criticisms made by the merging parties.

The principle behind the suggested proposed procurement conditions is to hold Massmart to its existing volume of local procurement for a period of time. It is accepted - it seems by the intervenors - that post merger Massmart cannot be subjected to a more stringent regime of local procurement than it has currently, as the conditions must address harms that are merger specific and not pre-existing harms (on the assumption that the move to imports constitute this harm).

Thus one would need to determine what Massmart’s local procurement level is pre-merger and then hold it to this level for some period going forward. This all sounds fine at the level of principle, but as we shall see founders, when we get to the level of detail.

Given the nature of the Massmart group as a supplier of a wide variety of goods from basic groceries to consumer electronics and building supplies, drilling down to determine the extent of local manufacturing content per product, amongst 120 000 goods will in practice prove a daunting task.71

The first problem is to establish how much of the locally produced product supplied is in fact locally produced. The merging parties claim that such a task is impossible as sometimes suppliers themselves do not know this. A supplier may be a middle man supplying products of someone else whose local content level may not be known to it. Secondly, even if the manufacturer were known, the manufacturer from whom the good was procured and who is thus the final part of the production chain to produce the good may not have been the sole producer and may have imported some of the inputs into that good.

SACTWU’s solution to this was to deem any specific product one of local origin if more than 50% of its value was locally produced. The onus is put on the supplier to provide this information. This of course creates further problems. Deeming something 100% local, even when its domestic manufacturing content may be much lower, would start to dilute the very concern one was starting to mitigate through the proposed condition. Relying on suppliers to verify local content may also lead to problems, as suppliers, knowing of the condition, and wanting to supply Massmart, would have an incentive to falsify the information to ensure that they were deemed to supply a product of local origin.

The next problem is what base period to assume. This may be possible as Sactwu suggests by taking a measurement period of the immediate previous year that includes both the best and worst months of supply to iron out peaks and troughs in supply. That nevertheless starts to introduce complexity into the system and again an opportunity to incentivise inaccurate but opportunistic information being supplied.

But the problems do not end here. What period should the condition operate for? SACTWU suggests 5 years, the SMMEs’ suggested three years. If this were part of an industrial policy program the period of the restraint might be linked to some other measures designed to make the domestic firms more competitive whilst they benefit from the period of protection. No such other policy is contemplated here. This raises the question of the rationale for the time period chosen. This links to the further issue that the proposed procurement conditions are not linked to any industry in particular, but suppliers in general. Surely not all manufacturing requires protection and some because of other government policy may be more deserving than others. The only industry that got special attention in the hearing was textiles because of Sactwu’s presence. Vlok, who testified on their behalf, made an interesting case for why the textile industry is a typical employer of vulnerable workers.

However Massmart is not at present a large procurer of textile products from the domestic industry. Hence imposition of this local quota on it may be an attempt to stop the floods by putting a finger in the dike.

In short the system created would create massive complexity, opportunity for evasion or manipulation, balanced against dubious utility. A considerably diluted procurement remedy as ultimately suggested to render the proposal more practical, may in the end not be worth the candle and have only symbolic value, whilst being devoid of any real benefits to local industry.

These then are some, but by no means all, the problems of the detail associated with the proposed local supply condition. But there are both legal and economic objections as well.

The merging parties have strenuously argued that the procurement condition is impermissible as it would render the country in breach of its trade obligations under several trade instruments to which it is a party.72 They argue that the Tribunal is bound by these undertakings. We do not need to determine this point but the arguments raised may well be correct and add to the problems associated with this proposed remedy.

As an economic argument the merging parties contend that imposing conditions restricting procurement on one firm is asymmetric. If the problem is one of foreign goods displacing locally manufactured goods in retail outlets why is one firm subject to this restriction merely because of a merger, whilst rivals would not be? Hodge counters this by pointing out that all merger conditions are imposed on the merged firm only and hence, by definition, are asymmetric.

Hodge is correct in this respect. Perhaps the better objection to the concern is not asymmetry as such, but that normally when conditions are imposed on a merger it is to dilute the acquisition of possible market power. The asymmetry is intended to lead to greater symmetry in the market place, by regulating the apprehended market power. In this case the merged firm, post merger, will not have market power in any relevant market in South Africa. At best it can be described as being controlled by a firm with pre-existing international procurement power. But in the local retail markets the firm will be a number four or five, in most segments. In those where it is a leading firm its procurement is either already very high (consumer electronics and photography 81% to 100% respectively) or are in goods that are not economically susceptible to imports (building products). Its rivals will post merger still be procuring more than the merged firm in most of the segments in which Massmart competes for which import

substitution is a viable possibility. Yet these firms, will not be subject to import restrictions.

In this sense the proposed conditions do create an unjustified asymmetry and the merging parties are correct. Further the conditions will contradict the major objective of competition regulation – to secure lower prices – the procurement conditions would likely affect the merged entity’s ability to provide customers with the lowest possible prices. Competition authorities do not lightly impose conditions that contradict their primary mandate, unless there is overwhelming justification for doing so. If we are not for competition then who is? Here, as we have noted, the justification is premised on ambiguously established harms coupled to conditions of dubious utility.

If, as the intervenors suggest, the economy is harmed by retailers substituting local manufacturing for imports on a wide enough scale so as to harm domestic manufacturing and hence employment, the remedy for this concern is not merger regulation – at least on the facts of this case. Here it is common cause that the merged firm is not the largest procurer of merchandise; its rivals are considerably larger in most segments and there are also a number of other firms in these segments that have the capacity to import and do so presently. Thus understood, imposing a remedy on a single non-dominant procurer to remedy an industry wide concern, would not be a rational exercise of public power. Other industrial policy instruments, not limited to firm specific application, would seem more apposite for this than merger regulation.

For this reason we do not consider that the conditions proposed, even in the more nuanced form finally put forward by SACTWU, are appropriate, proportional rational or enforceable.

On the assumption that there will be some substitution of local procurement to imports post merger, the extent of which is unknown,

an investment remedy like the one proposed by the merging parties is more attractive than a quota. Indeed, Hodge himself suggested an investment remedy might be appropriate whilst being cross-examined.73

The merging parties investment remedy entails them expending R100 million over three years through the establishment of a program aimed at the development of local suppliers, including SMMEs. The program, although administered by the merged entity will be advised by a committee comprising representatives of trade unions, business and SMMEs. Government will also be invited to serve on this committee.

The investment undertaking is a more positive response to the domestic procurement concern. Instead of insulating local industry from international competition for a period, it seeks to make local industry more competitive to meet international competition. Whilst at a macroeconomic level the remedy is modest, at the level of a single firm commitment it is not. Expenditure of R 100 million over a three year period is significant. Further the remedy seeks to engage those very critics of Walmart in the decision making process over the disbursement of the funds, including representatives of SMMEs. It also obliges the merged party to account for the expenditure to the Commission annually on the anniversary of the effective date about its progress.

The other attractions of the investment undertaking are that it is appropriate, proportional and enforceable. It does not raise the concerns that the procurement remedy does as referred to above.

Nature of the undertakings

During the Commission process the merging parties made undertakings in respect of local procurement and labour conditions. Massmart and Walmart undertook: “(i) not to cancel any existing agreements with trade unions and honour pre-existing union agreements and abide by South African Labour Law; (ii) to afford all future employees the benefit of association with the trade union of their choice; (iii) to create jobs; (iv) to respect the rights of unions and workers and not engage in any activities which undermine its existence and activities of these unions; and (iv) to increase job security for all current and future employees.” With regard to local procurement “Massmart and Walmart undertake to support local suppliers and in particular small businesses and BBBEE suppliers and not significantly change the volume and value of purchases from these local suppliers in all product categories procured by Massmart in future”.74

However, the merging parties were unwilling to allow these to become conditions for the approval of the merger as they argued that no conditions were required by law. As such the undertakings would not have been legally binding and hence, if not fulfilled, would not have resulted in any legal consequences for the merged firm.

This attitude persisted in our hearings as the cross-examination of Pattison illustrates.75 He repeats the undertakings given to the Commission, but seems unwilling to agree to them becoming conditions to be imposed on the merger. The undertakings now given by the merging parties at the end of the proceeding are therefore a significant move from their previous position. The merging parties had continued to contend that they were not legally obliged to concede to public interest concerns, but offered the conditions to meet adverse perceptions that had arisen about the merger. Whether this is so or whether the concessions came about as a result of the force of the concerns raised, we do not know. Nor, as we noted earlier, is it relevant for us to speculate on what the actual motive was. The fact is that the merging parties have agreed to us making the undertakings a condition of the approval of the merger.76

Because of this the undertakings are enforceable. Non-adherence can lead to serious consequences for the merger, which is an illustration of the commitment to them and an indication that it is not in consequence a public relations gesture. Because this merger does not lead to an integration of businesses, as would the normal horizontal merger, it is easier to remedy in the event of non-compliance. Thus if there was a material violation of conditions that would justify the most extreme remedy – that of divestiture – that could be achieved without difficulty by requiring Walmart to sell its shares or part thereof. By contrast were the firms integrated post merger, divestiture would be much more difficult and time consuming.

For all these reasons we view the undertakings as an appropriate response to the public interest concerns raised and that the merged firm has sufficient incentive to respect the undertakings made.

Costs

When we heard the matter on 22 March 2011 the unions were not ready to proceed and sought a postponement, which was granted, but costs were reserved. Unusually, the party that benefitted from the postponement was the government team, although it had not sought it. Nevertheless we are satisfied, given that the intervenors co-operated in this hearing in dividing up the workload so that the time could be used optimally, the unions should be considered to have benefitted from the postponement as well. As the documents yielded important information and informed the preparation done after the postponement, particularly by Hodge, which was of benefit to the Tribunal process, we consider the postponement request justified and make no order as to costs, as is the default position in merger cases.

Conclusion

We find that the merger raises no competition concerns. It does raise certain public interest concerns, but these concerns are adequately remedied by the imposition of the conditions submitted as undertakings by the merging parties and which are annexed to this decision marked Annexure ‘A’.

____________________ DATE 29 June 2011

Presiding member

Norman Manoim

Concurring: Yasmin Carrim and Andreas Wessels

Tribunal Researchers : Rietsie Badenhorst and Kasturi Moodaliyar

For the merging parties: J Gauntlett SC and D Unterhalter SC, assisted by J Wilson, G. Marriot, F. Pelser instructed by Webber Wentzel Bowens and Edward Nathan Sonnenburgs

For the Commission: P Mtshaulana SC assisted by G Ngcangisa instructed by the State Attorney.

SACTWU: P McNally SC instructed by Werksmans Attorneys.

SACCAWU: P Kennedy SC assisted by M Le Roux, instructed by Cheadle Thompson & Haysom

The Ministries: R Bhana S.C. assisted by J Meiring, instructed by Deneys Reitz

SMMES: G. Englebrecht instructed by Deneys Reitz

1Walmart uses both the hyphenated and non-hyphenated versions of its name. It uses the hyphenated form to describe the acquiring firm and the non-hyphenated from to describe the business. We have attempted to follow the same convention. (See Bond witness statement record page 1 footnote 1).

2LSM or Living Standard Measurement is a tool used to measure the South African market according to their living standards. LSM 1 being the lowest and 10 being the highest. www.saarf.co.za

3On Friday 24 September 2010 executives from Massmart together with their advisors met with senior Walmart delegates and their advisors in London. They negotiated and agreed on a share price and Walmart issued a non-binding expression of interest. On Sunday 26 September 2010 a special meeting of the Board of Massmart was held to review the non-binding expression of interest received from Walmart. On Monday 27 September 2010 Massmart publically announced the expression of interest from Walmart. On 29 November 2010 Walmart confirmed the offer to acquire Massmart.

4A Walmart presentation suggests that consumer spending in Africa is expected to grow from $ 860 billion to $1.4 trillion by 2020. It also identifies sub-Saharan Africa as the third fastest growing region in the world.

5SACCAWU et al were all represented by the same legal team.

6ASDA is a subsidiary of Walmart and was acquired by Walmart in 1999.

7Walmart Chile was previously called Distribucion y Servicio D&S S.A. (“D&S”) until Walmart acquired a majority stake in D&S in January 2009.

8An economics consultancy.

9An economics consultancy.

10Chief Merchandising Officer of Walmart Chile and Vice President of Walmart Stores Inc

11Commissioned by SACCAWU.

12Policy Co-Director of the National Employment Law Project, USA.

13Professor and Director of Labor and Employment Studies at San Francisco State University and a Research Associate at the University of California-Berkeley Labor Center.

14Associate Professor at the Murphy Institute, City University of New York.

15Senior Fellow and Director of the Markets, Enterprise, and Resiliency Project at New America Foundation.

16Executive Director of the Workers Rights Consortium.

17Lawyer with expertise in international labour standards and transnational labour regulation.

18Partner at law firm Aravena, Pozo, Morales Abogados y Asociados, in Chile.

19MacArthur Foundation Chair and Professor of History at the University of California, Santa Barbara. Director at the Center for the Study of Work, Labor and Democracy.

20Head of the Organizing, Campaigns and Collective Bargaining Department at the National Union of Metal Workers in South Africa.

21General Secretary of the Food and Allied Workers Union.

22President and representative of the South African Small Medium and Micro Enterprises Forum.

23Director General in the Economic Development Department.

24Pg 4 of Commission’s report.

25Pg 5 of Commission’s report.

26The Commission’s recommendations were: that all 503 employees who were retrenched should be reinstated; and that the merged entity must honour the existing agreements with the trade unions for at least a period of three years.

27The Commission should however have asked for certain transaction specific documentation such as the due diligence reports and the transaction specific correspondence that was yielded as a result of the government discovery request.

28Shell South Africa (Pty)Ltd/ Tepco Petroleum (Pty) Ltd CT 66/LM/Oct 06.

29See: Wispeco (Pty) Ltd and the Business of AGI Solutions (Pty) Ltd (69/LM/Oct09) (employment conditions); Nedbank Limited and Imperial Bank Limited (70/LM/Oct09) (employment conditions); Harmony Gold Mining Company Limited and Pamodzi Gold Free State (Pty) Ltd (71/LM/Oct09) (employment conditions); Unilever Plc and Unilever N.V. and Sara Lee Corporation (14/LM/Mar10) (employment conditions); Metropolitan Holdings Limited and Momentum Group Limited (41/LM/Jul10) (employment conditions); AECI Limited, Acting Through its Division Plaaskem and Qwemico Distributors (Pty) Ltd (67/LM/Oct10) (employment conditions)

30: In Harmony at paragraph 76 we stated; “This prioritisation of the competition inquiry explains the use of the word justification in the public interest test. The public interest inquirymay 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 anti-competitive 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. 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." Harmony Gold Mining Company Limited and Gold Fields Limited CT Case No.: 93/LM/Nov04. We followed this reasoning in Metropolitan Holdings Ltd/ Momentum Group Ltd CT 41/LM/Jul10 at paragraph 76.

31See footnote 30 above.

32See record page 2474.

33See record page 2476.

34See transcript page 19.

35An example of this elision between South African interests and Africa can be found in the letter from Pattison to Richard Levin the Director General of the Department of Economic Development dated 26 October 2010. Pattison tells Levin that under its current strategy i.e pre-merger Massmart had intended to expand by 20%, but he is referring to both in South Africa and what he terms the region. (Record page 176.)

36Due Diligence Report entitled “Project Memphis-21st Century: HR Final Report” dated 29 October 2010, Walmart discovery item 84. For example the due diligence documents presented by Walmart relating to human resources refer to the “high levels of labour/associates in stores. The clarity of this position is clouded by the use of vendor colleagues, ‘brand associate advisors’ and other third-party employees (e.g. Decorland). There is an opportunity to reduce cost and drive productivity.” The solution or alternatives to mitigate risk in this report is to “review required structures, remove third-party labour where appropriate (seeking margin reduction where appropriate), establish a new model, amend contracts if necessary, introduce an automated scheduling system.”

37The synergies document came about as a result of the government discovery request. It was not before the Commission when it made its recommendation.

38In Metropolitan 41/LM/Jul10 para 68 we referred to Harmony/Goldfields CT 93/LM/Nov04 and held that the merging parties are not required to affirmatively justify a merger on public interest grounds. What we did not decide in that case is whether once a substantial public interest ground has been raised whether the merging parties face an evidential burden of justification. In this case we have decided they do. Once a prima facie ground has been alleged that a merger may not be justifiable on substantial public interest grounds, the evidential burden will shift to the merging parties to rebut it.

39In his evidence Pattison went further than in his witness statement to say that the landlord Old Mutual had cancelled the lease. Transcript page 210.

40At page 210 of the transcript Pattison says that approximately 60 staff members were retrenched because of a merger of Games stores in Nelspruit and the rest of the remaining staff, 434 staff, because of the regional distribution centres. Mbongwe gives different figures of 317 workers who were retrenched around 23 June 2010 (see transcript page 568). SACCAWU in its summary of key issues calls for the reinstatement of all retrenched workers but does not specify a number – it refers back to Mbongwe witness statement. See record page 296 paragraph 30 SACCAWU summary of key issues.

41See heads of argument paragraph 11.8.1.

42See Metropolitan 41/LM/Jul10 paragraph 68.

43Lamberti made this observation at the board meeting in September 2010 when the deal was first discussed. See Record page 2331.

44See Pattison witness statement record page 158. He says that despite having known of Walmart’s interest in making an acquisition in South Africa, the fact that Massmart was the preferred target was only revealed to Massmart on 26 September 2010.

45See Exhibit A which contains the correspondence between the parties at the time and refers to the various meetings and contacts during this period.

46Transcript pg 85.

47See page 3580 of record.

48See Exhibit A.

49Control Risks report for Walmart titled “Due diligence in South Africa” dated 3 November 2009.

50Walmart document entitled “International Mergers and Acquisitions Update” dated May 4 2010. Walmart Discovery File item 18.

51See Bond statement record page 32 paragraph 67.It was suggested in argument that only the Indian Railway service employs more people but this is probably apocryphal.

52See evidence of Jacobs transcript pages 485-490.

53Wage theft is defined as making workers work longer hours without compensation. Eg working through lunch breaks without commensurate

compensation. This is alleged to have been widespread in the Walmart stores in the US.

54See Bond witness statement record page 32 -3 paragraph 68. Jacobs the expert testifying for SACCAWU questioned the usefulness of NLRB statistics as evidence of a disproportionately low number of labour complaints. See transcript page 854.

55See transcript page 394. See aslo http://www.foodanddrinkeurope.com/Retail/Asda-Walmart-guilty-of-anti-trade-union-activity.html.

56See transcript page 438.

57We referred earlier to the email from Henry querying union size and termination of the union contract.

58By way of example the union sought that we impose a condition that Walmart cease its opposition to the US Employee Free Choice Act, and sign a Global Framework Agreement with UNI Global Union. (See Mbongwe witness statement record page 325.)

59See Hodge’s report pages 26-27.

60A debate over whether Walmart pays lower wages in the U.S than unionised rivals took place between the merging parties and Jacobs

for SACCAWU. Nothing conclusive could be reached on this issue with both sides alleging the others research or claims were flawed.

61The unions acknowledge this linkage as in their heads argument SACCAWU et al state, “Collective bargaining ensures that Massmart’s workers have countervailing power to the company which assists them in protecting and ensuring fair pay and working conditions.” See heads of argument paragraph 11.6.1.

62See transcript pages 341-343..

63See Exhibit B page 4. An ASDA executive is quoted in the press saying post the merger the firm is paying 50% less for its denim. He added that blanket fleece sourced previously by ASDA at $9 per yard is now sourced through Walmart at $3 per yard.

64“Sourcing Globally Toolkit” record pages 2483-2490.

65Hodge shows that during the period relied upon by Baker the peso had devalued making imports more expensive and hence the exercise in the country comparison less reliable. See Hodges report “Assessment of the Public Interest Impact of the Walmart- Massmart merger “ dated 19 April 2011 pages 15 and 16. (The Ministeries files record pages 1252-1253).

66The SAM is based on the 2000 Supply – Use table provided by Statistics South Africa. It represents flows of all economic

transactions that take place within the economy and thus allows for the calculation of the total effect on employment by incorporating

backward linkages in the economy.

67Baker states “But in any event, as Mr Hodge I think recognises in his report that there is no evidential basis for the 1%. It is a number used by him to illustrate the extent of employment effects were that to happen, so this is not evidence that it will be 1%. There is no evidential basis that I could see for a 1% assumption.” Transcript pages 676-677.

68Transcript pages 884-886.

69Transcript page 637.

70Transcript page 639.

71See Transcript page 198.

72General Agreement on Tariff and Trade (GATT) 1947 Article III (4) and (5) See Merging Parties written submissions dated 16 May 2011 pages 67-74.

73See transcript page 936.

74SACCAWU core bundle for cross examination file record page 2503.

75Transcript pages 187-193

76Perhaps the merging parties may have spared themselves of some of the length of these proceedings had these undertakings been made earlier in the process.

39