---

COMPETITION TRIBUNAL OF SOUTH AFRICA

Case No: 68/LM/Aug06

In the matter between:

Netcare Hospital Group (Pty) Ltd Acquiring Firm

And

Community Hospital Group (Pty) Ltd Target Firm

Panel : N Manoim (Presiding Member), U Bhoola (Tribunal Member)

and T Orleyn (Tribunal Member),

Heard on : 4-15 June 2007 and 18 July 2007

Order Issued : 2 August 2007

Reasons Issued: 5 November 2007

Non-Confidential Reasons for Decision

Introduction

1. This is a merger between two groups of hospitals. Netcare Hospital Group (Pty) Ltd (“Netcare”), the largest hospital group in the country is seeking approval to buy all the shares in the Community Hospital Group (Pty) Ltd (“CHG”), which owns five hospitals. Netcare presently owns 43.75% of CHG and seeks to buy the remaining shares. When Netcare acquired its 43.75% stake in CHG, in 2002, it acquired joint control over CHG, but did not notify this transaction as a merger, as the law requires it to do.

2. We are thus required to consider the effects of the merger not in relation to whether the change from joint to sole control by Netcare will result in a substantial lessening or prevention of competition, but whether the acquisition of control by Netcare ab inititio,will have this effect. The Competition Commission (“the Commission”) recommended that the merger be prohibited.

3. After hearings were concluded on 18 July 2007 we decided to approve the merger on 2 August 2007. We now give reasons for that decision.

Transaction details

4. Netcare which currently owns 43,75% of CHG will acquire the remaining 56,25 % of the shares from Community Health Care Holdings( Pty) Ltd ( 43,75%) and two management owned entities , Duelco Investments 65 ( Pty) Ltd (6,25%)and Private Preview Investments 27 ( Pty) Ltd (6,25%). The total purchase price is composed of a mixture of cash and shares in Netcare. As a result of the transaction, CHG becomes a wholly owned and controlled subsidiary of Netcare.

5. The material assets being acquired by Netcare in this merger are the five private hospitals owned by CHG. They are Montana Private Hospital, and Bougainville Private Hospital in Pretoria, Kuils River Private Hospital and UCT Private Academic Hospital in Cape Town, and East Rand N17 Private Hospital.

Background

6. This is an unusual merger case.¹ In the ordinary course, mergers are notified before they are implemented and our task is to evaluate, ex ante, whether they are likely to be in contravention of the Competition Act, 1998 as amended (‘the Act’). In this case it is common cause that:

• The merger has at least been partially implemented. According to the Commission this may have been as early as 2000.²

• That this implementation has taken place unlawfully. ³

7. Our merger regime is not designed for evaluating mergers after the fact. One only need examine the language of the section to see its futuristic inclination. For this reason we have a system of compulsory pre-merger notification. In some other systems which also require mergers to be notified before they are implemented this simply means the merging parties cannot implement for a certain period. Thereafter, unless the authorities indicate an intention to stop the merger the parties can consummate the transaction but still risk the possibility of post merger scrutiny by either the authorities or private parties.⁴ Our system requires an active decision by the competition authority once a merger is notified and for this reason a merger cannot be challenged again later; neither by the competition authorities nor private parties. In this sense, merger approval gives the merged firm immunity from future challenges and thus the comfort of business certainty going forward. In return for this benefit, firms are obliged to notify mergers before they are implemented and to delay implementation until they get regulatory approval. The entire edifice of merger control; notification, evaluation, and recommendation, is premised on this assumption of compliance. The leitmotif throughout this enquiry is “what will the world look like after the merger”. When a merger has already been implemented, and as in this case for some time, the premise of the entire system is undermined; a system designed to look into the future now has to gaze back in the past. Typically in the ex ante merger review, we, in the words of the Competition Appeal Court, “forecast a likely possibility” ⁵. We use the past to make an informed prediction about the future. When a merger has already been implemented and, as in this case for some time, the entire evidential premise of the system is compromised.

8. In the United States by contrast although a system of compulsory prior merger notification exists, mergers may be implemented following the observance of a statutory waiting period. Thus mergers can then be implemented by way of default and this is typically the way most mergers proceed. In contrast to our system however, the merger remains susceptible to later challenge, either by the government or at the instance of a private party. As a result, when adjudicating a merger a different approach is adopted depending on whether the challenged merger has been implemented or not. Thus, in a recent decision of the Federal Trade Commission (“FTC”) where a hospital merger had been litigated post consummation, the FTC compared prices being charged at the merged firm presently with those charged by the separate constituent hospital groups before the merger.⁶ In other words they looked backwards not forwards.⁷

9. In this merger, the evidence has not been prepared on this basis, but rather on the assumption of imagining what it would be like if the merger had not taken place. Thus, witnesses are asked what the future would look like, if CHG was a group independent of Netcare. But CHG has not been independent of Netcare for some years and no witness can recall what it was like when it was. Since part of CHG’s expansion took place at a time when it was already jointly controlled by Netcare, it has never in its present form existed as an entity outside of Netcare control. Thus any witness being asked to comment on what the world would be like with an independent CHG is being asked to imagine something that has never existed.

10. It is true, as Dr Stillman, the merging parties’ expert with whom we debated this issue, argues, that all merger control requires a measure of hypothesis. The witness in the pre-merger case, where there has not been implementation is being asked to speculate on what might happen in the market when the acquiring firm controls the target which, up till then, it had not. This requires an exercise of imagining what might happen in the market if the target became the object of the acquiring firm’s control.

11. But that is not the task witnesses in the present case have been required to perform. They have been asked to imagine what CHG might be like at present if it were independent of Netcare and then to ponder on what this hypothetical present case scenario might look like in the future if CHG was once again part of Netcare.

12. The response to this problem for the adjudicator is by no means clear. It may mean that in relation to witness evidence we adopt a cautionary rule about their ability to speculate. Even that is not helpful. Does it mean that we must be cautious about their optimism or their pessimism about the future of competitive inclinations of the merged firm? The other approach would be to presume against the interests of the merged firm. On the basis that no firm should benefit from its unlawful action, it may be appropriate to treat evidence of witnesses who testify that the merger will not give rise to anticompetitive effects with caution, whilst not adopting the same approach with witnesses who testify to concerns. The statute gives us no mandate to follow this approach and as an administrative tribunal we should tread cautiously in developing evidential rules or presumptions without the statutory remit to do so.

13. We are thus constrained to perform our functions under sub-optimal conditions. As we demonstrate later in this decision, when we evaluate the evidence of industry witnesses, many laboured under the difficulties we have identified of erecting a supposition about the future based on a hypothetical present. For some witnesses this exercise proved difficult, not because they were wanting in experience or intelligence, but because that was a conceptual demand they found daunting; others fared better, but even then the task expected of them was unfair and unreasonable.

14. The witness who did indeed have concerns would have to conjecture what Netcare and CHG might look like without the other; a situation last seen in 2002 when both groups were of a different size and so was the rest of the market. Some who testified were not in the industry at that time. The difficulty is further bedevilled by the fact that competitive bargaining between hospitals and funders is in its infancy. At the relevant pre-merger date, bargaining was still centralised making an assessment more difficult. Little wonder indeed that industry representatives either had difficulty in coming to any conclusion or concluded that the merger was not going to make much difference. It is also a question of degree. If the merger implicated half of Netcare’s hospitals or half in a particular region one can imagine an informed observer still being able to come to some meaningful conclusion. But located as it is with a small presence in the major metropolitan areas, CHG does not present an obvious case for easy post merger analysis.

15. Nevertheless we are required to make a decision on the case before us. We have decided to evaluate the merger following two approaches. First a post merger analysis, and secondly, a pre-merger analysis (the approach we adopt conventionally).

History of CHG

16. The hospitals that comprise the present CHG have had an interesting history. The rump of the present group originated in the stable of the now defunct Macmed Health Group Limited (“Macmed”). Macmed was involved in the supply of medical equipment and supplies to hospitals. In the mid 1990’s, Macmed decided to enter the private hospital market so as to create a customer base for its supplies. In pursuit of this strategy, Macmed acquired certain hospitals, including those presently in the CHG portfolio, Montana Park Clinic and Bougainville Hospital, in Pretoria, N17 Hospital in Springs and Kuils River Hospital in Cape Town. In addition, it acquired the licences to the Fourways Hospital, and Southgate Hospital.

17. Macmed’s ambitions to become a private hospital player soon displeased its customer base who did not welcome competition from a supplier. According to CHG’s chairperson Anna Mokgokong, Macmed tried to remedy the problem by “appointing a black economic empowerment partner and by transferring control of these hospitals to such partner.”⁸

18. Such partner was to turn out to be the Malesela Hospital Group (Malesela), a company controlled by Mokgokong and Joe Madungandaba. In 1998 control of the hospitals was transferred to Malesela. At the time of the transfer Fourways, Kuils River and N17 were then still under construction and the only operational hospitals were Montana and Bougainville. The latter two were managed by a company called Executive Health Management Services, a company owned and controlled by Dewald Dempers and Gerhard Ferreira, who were later to become minority shareholders in CHG.

19. Malesela’s stewardship of the hospitals was brief and unhappy. Cash flow from the two operating hospitals was insufficient to service its debts and in addition, the group was forced into an uneconomic and exclusive supply arrangement with Macmed. Macmed had to advance funds to it on an ongoing basis, to keep the group operating. By the last quarter of 1999, drowning in debt and overly dependant on Macmed for survival, Malesela began to negotiate with Afrox in the hope of finding a new and better partner. While these negotiations were taking place Macmed was liquidated. It was to prove one of the most notorious in recent corporate history. But in the aftermath, Malesela, which had also owned a stake in Macmed, suffered damage not only to its finances but also to its reputation. During the period November 1999 to February 2000 all the Malesela hospitals, which were owned by separate entities in the Malesela Group were eventually liquidated.

20. Mokgokong was determined to rescue her hospitals out of liquidation, but was unable to find any financier willing to support her. Instead, she looked to the three private hospital groups. She says that neither Afrox nor Medi-Clinic was interested in supporting her – both seemed intent on capturing some of the hospitals for themselves. Afrox, according to her, was keen on Fourways, N 17 and Montana. Medi-Clinic, through its empowerment partner Phodiso was intent on Kuils River. At this stage Netcare enters the picture as Mokgokong’s saviour. On her version, Netcare was looking for an empowerment partner at the time, as its previous partner had financial difficulties and had to dump its Netcare shares.

21. Netcare it seems had more than social objectives in mind when it came to the aid of Mokgokong. Anxious lest these hospitals fall into the hands of Afrox, Netcare engineered a situation which led to all the hospitals emerging from liquidation into the hands of CHG, which was to become the new creature of Mokogokong and her erstwhile partners Madungandaba, Dempers and Ferreira. All that is save for Fourways, which despite their best efforts fell into the hands of Afrox, now Life, who have developed this hospital and which is now operational. That Netcare was the lead partner in this rescue is not only acknowledged by Mokgokong in her statement, but is also acknowledged by Netcare in minutes, where in the Chairman’s report of March 2002, credit for this rescue is given to the way the process was “ smartly managed” by its attorneys. ⁹

22. What Mokgokong does not say in her statement and perhaps this is because she never knew about it, was that when Malesela was placed into liquidation the three private hospital groups had had an agreement amongst themselves not to bid for the Malesela hospitals. The reasons for this arrangement are not clear, we only have Netcare's comment on this that this was for “reasons which seemed cogent at the time.”¹⁰

23. Netcare then discovered that Afrox had “secretly breached this undertaking” and was making overtures to the liquidators to acquire some of the Malesela hospitals. Netcare then decided to assist Mokgokong to recover the hospitals. Netcare commenced involvement as a consultant to CHG, and from then on its involvement expanded. What seems clear is that because the hospitals subject to the recovery strategy required government consent for their licences to be transferred to the new owner, CHG had a competitive advantage over Afrox, as it was a black owned group.¹¹ As Sacks observes:

“..the strategy was to develop and promote a contest between Malesela [i.e. Mokgokong] and Afrox Healthcare, a contest which was politically far more easily manageable.”¹²

What emerges from this memo is why Afrox needed the hospitals more than Medi-Clinic or Netcare. N17 was a threat to its Parklands facility, whilst Montana was a threat to its Eugene Marais hospital. Sacks recognised this as well. He states in the same memo that once the ownership of N17 had been settled (i.e. it had emerged from out of liquidation in CHG hands) the “ turnover will increase through the recruitment of several medical specialists from Parklands Hospital an old and tired Afrox hospital in Springs.” relation to Montana, Sacks mentions that Afrox had made a ‘secret’ bid for this to the liquidators as this hospital “is close to Eugene Marais Hospital which would be seriously affected by an aggressive competitor in the area.”

Interestingly, the one hospital which was in the Malesela group and which was a potential threat to Netcare, was Fourways. According to Sacks;

“This hospital when and if completed could potentially pose a threat to Sunninghill Hospital and Olivedale Clinic and recognising the potential prejudice CHG acquired LTA’s secured claim of approximately ... [Confidential] to position itself in the project.”

Note that Sacks observes that CHG must act on a potential threat not to CHG, which has no hospitals in that area, but to Netcare which owns Sunninghill and Olivedale. Afrox ultimately secured the licence but only after considerable delay in a contest over the licence something that Netcare would later take credit for.¹³

24. It appears that the rescue was undertaken over a period, from early 2000 to mid 2001. Initially the hospitals were acquired by separate shelf companies, but the shares in these companies were then transferred to CHG, sometime in 2001. Mokgokong does not tell us who the initial shareholders of CHG were at the time, but they appear to have been held by an attorney as nominee.¹⁴ CHG only acquired its present shareholder structure in 2002, when Netcare formerly acquired joint control with Mokgokong and Madungandaba. Notwithstanding this, prior to the conclusion of the shareholders agreement in 2002, the hand of Netcare in the new group is writ large. Netcare provided the funding to acquire the hospitals, provided initial financial assistance and paid the legal fees involved in the recovery. EHMS managed the hospitals, except for Kuils River, which, when it commenced, was run by Netcare’s coastal division. Netcare also immediately implemented its IT systems across all the CHG hospitals and managed its pharmacies. Its IT systems are significant, because they inter alia contain the group’s billing, inventory and other hospital administration functions. In his witness statement Ryan Noach the chief operating officer of Netcare states as one of the reasons for the merger, the need to protect the intellectual property that it (Netcare) has extended to the CHG hospitals. He states:

“all the CHG hospitals are operated using Netcare’s accounting system, IT platform, nurse training programmes, management training programme and pharmacy infrastructure.”¹⁵

25. Thus, although the shareholders agreement in terms of which Netcare was to acquire joint control over the CHG group was only concluded in late 2002, it appears that from the outset of the so-called rescue operation, Netcare exercised significant influence over the new CHG. ¹⁶

26. In terms of the 2002 agreement, the present shareholding in CHG was structured as follows: Netcare (with a 43.75% shareholding), Community Hospital Holdings (Pty) Ltd (CHH) (with a 43.75% shareholding), Duelco Investments 65 (Pty) Ltd (“Duelco”) (with a 6.25% shareholding, and Private Preview Investments 27 (Pty) Ltd (“Private Preview”) (with a 6.25% shareholding). Duelco and Private Preview are the respective entities of Dempers and Ferreira.

27. The group did not remain idle once reconstituted. Kuils River, which recall was not yet complete in the Malesela days, commenced operating in July 2002. ¹⁷ In August 2002, CHG entered into a joint venture arrangement with the University of Cape Town in respect of the UCT Private Academic Hospital¹⁸ and a controlling interest was acquired in it in June 2003.¹⁹ When CHG was formed it had only 39% of UCT Private Academic, but the remaining 61% has since been purchased from the doctors of the hospital.²⁰

28. In summary then, currently, CHG controls five hospitals, with a total of 594 beds. Two of CHG’s hospitals are located in the Pretoria region and these are Montana Private Hospital (with 158 beds) and Bougainville Private Hospital (with 60 beds). The other two hospitals are located in Cape Town and these are Kuils River Private Hospital (with 120 beds) and UCT Private Academic Hospital (with 126 beds). The last CHG hospital is the East Rand N17 Private Hospital (with 130 beds).

Netcare

29. Netcare is the largest private hospital group in South Africa. It comprises 70 hospitals with over 8400 beds.²¹ In addition to the hospitals, Netcare owns Medicross, which comprises approximately 50 medical centres providing primary healthcare services. Medicross has acquired Prime Cure, a network of 45 Medical centres located countrywide.

30. Netcare further controls a range of medical related businesses such as Netcare 911, Ampath, and pharmacies, to name but a few. Netcare has recently extended its international business interest by acquiring a controlling interest in the leading private hospital group in the UK, GHG.

Market definition

31. Defining relevant markets in hospital mergers is a complicated task for several reasons. Firstly, this is not a conventional market where buyers and sellers are in a direct relationship. Ninety percent of private hospital revenue emanates from medical aid funds not individual patients. The presence of an intermediary has consequences for behaviour in the market. One writer Lawrence Wu explains it in this way:

“That is because substitution in hospital markets is not as straightforward as it is in other markets where there are no intermediaries between the seller and the consumer. In the case of hospital services, third party payers (e.g. health plans and employers) can take actions that could either facilitate or impede the ability of individual consumers to turn to alternative hospitals.”²²

Secondly, demand patterns are not those of a conventional market. The ultimate consumer of hospital services is of course the patient. Yet the patient’s preference for a particular hospital is rarely the determining factor in the choice of hospital. Frequently this choice is made by a general practitioner, who refers a patient to a particular specialist who practices at a hospital, or it may be made by the funder, an increasing likelihood with new products for low income schemes as we consider later in this decision.²³ Even if a patient is on an unrestricted medical scheme and is not channelled by a general practitioner the choice of hospital is still more likely to be a contingent one, dependant on a prior choice of specialist. I would like to see Dr X and I discover that he practices at hospital Y hence I will go there.

Thirdly, the modalities of private healthcare funding are changing constantly. What we call medical aid or the US would refer to as medical insurance is the subject of constant bargaining and strategic game playing between hospitals and funders. As Wu puts it:

“Arguments and evidence regarding the ability of consumers to turn to competing alternatives are complicated due in part to the growth and proliferation of new types of health insurance plans and innovative contractual arrangements among hospitals, physicians and third party payers. It is crucial to understand the commercial realities that influence these third party payors (sic) because they play an important role in determining the hospitals to which consumers can turn.”²⁴

32. As we examine later in this decision there is no consensus on the best funding models, and to the extent that funders seek to impose their own solutions and resist those sought to be imposed by hospital groups the degree of competition matters – and this has been the primary focus of the Competition Commission’s concerns in this merger. But it also impacts on how we define markets because if new funding models change the rules in which healthcare has traditionally operated the competition dynamics change as well.

33. Fourthly, hospitals provide differentiated services, because they typically provide a bundle of services varying in range and kind. This means that the closer the similarity in services the greater the likelihood that they compete with one another or put differently, they may vary in the degree to which they may be considered competitors. A consequence of this, conventional HHI analysis may throw up a skewed picture of a market as the extent of concentration it reveals may bear no relationship to the reality of competition.

34. Fifthly, hospitals do not just compete in providing services to patients. They also compete for the services of specialists, since specialists are a channel for patients. To this extent hospitals compete with one another to provide an attractive venue for specialists by investing in the capital equipment and facilities that will attract specialists.

35. Sixthly, the extent of geographic markets varies depending on what form of competition we are dealing with. Patients who are members of schemes that impose no restrictions on their choice of hospital would favour hospitals at a convenient distance from their home or work. By contrast patients on a restricted scheme that limits its members to a pre-determined list of hospitals, may find that they travel much further to find a hospital. Nor are the geographic choices of patients and doctors similar either. Specialists might be willing to consider larger geographic boundaries for hospitals competing for their services, especially if they wish to establish a second practice, than might a patient whose scheme gives her freedom of choice.

36. Seventhly, as we consider more fully later, new medical scheme products are being introduced into this market all the time, which work differently to the conventional reimbursement scheme with which we are most familiar. It is possible that each of these products in turn may form a sub-market subject to different competitive forces. Thus a hospital may be less constrained in raising prices to one class of consumers – say those on unrestricted schemes – than those on restricted schemes.

37. Finally, hospitals operate in partly regulated markets and to some extent regulation impacts on outcomes. Changes to the pricing regime for pharmaceuticals impacted on the price of unregulated services. The evidence from Netcare has been that when single exit pricing was introduced for pharmacies it compensated for losses in this area of its business by raising tariffs on non-regulated services.²⁵ Legislation also affects the design of medical schemes. In the course of this merger we heard that open schemes are required to accept members nationally even though it may be in their interests to focus on securing a membership base regionally and hence improve their bargaining power at a regional level. We are told that there are plans to change the legislation in this respect. ²⁶ This has implications for whether in a hospital merger one should be concerned more with regional concentrations than national ones.

38. In this merger we will eschew the adoption of any rigid market definition. This is not to say this is the approach we will always take to hospital mergers in the future. Rather, given the uniqueness of this case we will test theories of harm that have emerged during the proceedings and which posit different market definitions to assess whether the theory of harm is valid.

39. We will however accept that there is a consensus view that the product definition adopted in previous hospital mergers – the provision of private hospital services will suffice for the present.²⁷ As will appear more fully from this decision because the merger of hospitals implicates various forms of competition and that those forms of competition have different implications for the size of geographic markets, we lay down no hard and fast rule of what the geographic market is. Rather we identify areas where hospitals compete with one another and examine how the geographic market may differ.

40. We will examine the merger from two perspectives. First by adopting the post merger analysis approach and ask if there is evidence now that the merger has significantly lessened competition. We then ask if the merger will in the future significantly lessen competition. This second approach is the one we normally follow in our merger analysis.

Post merger evidence.

41. This analysis will be brief. Because the Commission has approached the matter on a prospective basis it has not led any evidence to compare the pre-merger market with the present one. Even if it had, this would not have been a simple task. The pre-merger market would have been an independent CHG before any Netcare investment. This takes one back to about 1999. Then the CHG group was smaller than it is presently²⁸ and Netcare, although roughly the same size in terms of numbers of hospitals, would have had fewer beds. But adding to this difficulty is that prices, which might be the basis for comparison, were at that time the subject of an industry wide tariff negotiation and hence would not provide a meaningful basis for comparison with rates of the merged entity of today. As we noted in the discussion of the Evanston case this comparison formed an important part of the conclusion of the FTC that the merger had had an anticompetitive effect.

42. Most of the witnesses who testified for the Commission indicated that they had not been in the market for long. Perhaps the most knowledgeable, David Strauss of Discovery, put it best when he said that he had been in the business for six years and during that time he had never known of CHG as independent from the control of Netcare.

43. We conclude that we have no evidence to be able to make this comparison and hence we are not able to conclude that on the basis of a post merger analysis, the merger has led to a significant lessening of competition.

Prospective evidence

44. The Commission identifies three prospective theories of harm-

1. the effect of the merger on competition for specialists;

2. The effect of the merger on competition for patients in a region; and

3. The effect of the merger on funders in respect of reimbursement rates and funding models

We will examine each one in turn.

1) Effect of the merger on competition for specialists.

45. This is the least developed part of Commission’s case. The Commission argues that hospital mergers can dampen competition for specialists, if hospitals competing for those services pre-merger, are inclined to make investment decisions independently. An independent CHG would make decisions to invest in infrastructure to attract specialists even though that investment may duplicate what has been made at a neighbouring Netcare hospital. Post merger, however, the combined group would have the incentive to avoid duplication in investment and from having its hospitals compete with one another.

46. At a theoretical level this is a valid concern. However the Commission produces little evidence to dress its theory in. The Commission tells us almost nothing of how this market operates in the present merger. Are specialists obliged to have exclusive relationships with hospitals? If they are it has more serious implications for a merger. If they are not, and the evidence in this case from Netcare suggests that they are not, it means that even if hospitals merge, because specialists can practice elsewhere they are still free to practice at hospitals not part of the merged firm. Dr Stillman has some figures for the number of CHG specialists who specialise elsewhere including, it seems, at non-Netcare facilities. Thus he tells us the 8 of the 19 specialists who regularly practice at CHG’s Montana in Pretoria also practice at Life’s Eugene Marais. On its own this a meaningless statistic; we do not know why the 8 practice at Eugene Marais and why the other eleven do not or if they practice elsewhere. A similar statistic is put up about CHGs’ East Rand facility N 17. Here we are told 12 of the 35 specialists also practice at Life’s Springs Parklands which is less than 1 kilometer away. The only documentary evidence on record emerges from a Netcare memorandum to its board where in March 2001, Sacks notes that CHG would be able to lure away specialists from Parklands, which he describes as an old and tired Afrox Hospital, to N 17.²⁹ The fact that Netcare saw N17 as a threat to Afrox’s Parklands and not its own facilities would support Stillman’s analysis on where competition for specialists was coming from in this area.

47. Presumably the reason that specialists will practice at more than one facility is to expand their practice to a patient base not served by their existing facility.³⁰ This means that specialists will be willing to travel further than patients. No specialist unless for some technical reason (access to theatre or equipment not available at their present facility) will practice in the area of a hospital serving an existing patient base – the duplication in rentals would make no commercial sense. ³¹

48. Thus a geographic market for specialists is not necessarily co-extensive with that for competition for patients. The geographic market for specialists would differ depending on whether a specialist wanted a single location for a practice or more. If a specialist wanted only one practice, the market for specialists might be one of hospitals in closer proximity. Nevertheless for the best opportunity a specialist may be willing to relocate or to travel a great distance to work, greater than might a patient similarly located. A specialist wanting to establish a second practice, will in order not to cannibalise his /her patient base, be willing to travel even further. Since most of the CHG hospitals are located in metropolitan areas there is much choice left in this wider notion of the geographical market for specialists to find alternatives. Of the CHG hospitals, two are unlikely to be implicated in competition for specialists. Bougainville is a modest facility, seeming unlikely to invest further under anyone’s control. UCT Academic was formed precisely as an outlet for academic doctors at the university and hence is not competing for anyone else. What remains behind is too modest to pose a concern about the market for competition for specialists. No specialists have given any information to the Commission indicating any concerns.

2) Effect of the merger on competition for patients.

49. As with the market for specialists, this is not a particularly well-developed part of the Commission’s case. The Commission’s case here is mostly reliant on the increase in concentration in the three regions where there is an overlap between CHG and Netcare hospitals. Based on the number of beds in the respective hospitals the Commission has calculated HHI’s which, when reworked by Dr Stillman, yield the following:

Table 1: Market Shares and HHI’s in the Pretoria region

Hospital Group	Hospital	Number of beds	Market Share (%)
CHG		218	6.1
Netcare		1230	34.7
Life Healthcare		1077	30.4
Medi-Clinic		685	19.3
Independent	Louis Pasteur Private Hospital	191	5.4
	Wisani Medical Centre	12	0.3%
	Zuid Afrikaans	135	3.8
Total		3548	100
Pre-merger HHI			2 577
Post-merger HHI			3 003
Delta			426

Source: Robert Stillman Expert Report pW165.

Table 2: Market Shares and HHI’s in the East Rand Area

Hospital Group	Hospital	Number of beds	Market Share (%)
CHG		170	7.3
Netcare		733	31.6
Life Healthcare		754	32.5
Independent	Arwyp Medical Centre	336	14.5
	Benmed Park Clinic	125	5.4
	Sunshine Hospital	200	8.6
Total		2318	100
Pre-merger HHI			2 425
Post-merger HHI			2 889
Delta			464

Source: Robert Stillman Expert Report pW166.

Table 3: Market Shares and HHI’s in the Cape Town Area

Hospital Group	Hospital	Number of beds	Market Share (%)
CHG		294	8.9
Netcare		476	14.4
Medi-Clinic		1 795	54.4
Life Healthcare		387	11.7
Melomed		265	8.0
Independent	Newlands Surgical Clinic	81	2.5
Total		3 298	100
Pre-merger HHI			3 458
Post-merger HHI			3 716
Delta			257

Source: Robert Stillman Expert Report pW167.

50. Looked at from a purely HHI perspective, with nothing else, these figures look alarming.³² But as Dr Stillman argues, and Dr Roberts does not dispute this, HHI’s are a starting point - a filter for agencies to determine whether a merger requires a more in depth look. Viewed in isolation they could offer a distorted picture of the state of competition. This is because hospitals are not homogenous providers of services. Modern private hospitals provide differentiated services. Whilst some offer a full range of services others choose to narrow their focus. To take an extreme example a maternity hospital may be next door to a hospital providing geriatric services - despite their proximity they are not competitors. Thus as Vistnes, in an article both the Commission and merging parties rely on for different propositions, states:

“HHI’s are likely to underestimate the competitive problem when the two hospitals are very similar (compared with other hospitals in the market) and to overestimate the problem if the merging parties are highly differentiated”³³

51. The merging parties in contrast went to great lengths to establish that no single CHG hospital has, as its nearest comparable competitor, a Netcare hospital. In his witness statement, Dr Stillman takes each of the five CHG hospitals and argues why its nearest competitor is a hospital of another group (Life or Medi-Clinic) or another independent (Melomed, Louis Pasteur) This evidence is largely established on the basis of interviews that Dr Stillman had with the respective hospital managers of the hospitals concerned. At only one hospital, Bougainville in Pretoria, did Dr Stillman conduct an independent test of these assertions by examining the postal codes of patients, but for the rest his research relied on the statements of managers.

52. Although the Commission in cross examination of Dr Stillman exposed his uncritical acceptance of his Netcare sources in certain instances, or inconsistency in his approach to issues such as distance, it has not given us any better picture of the extent to which CHG hospitals compete with their Netcare counterparts in a region. We are not in a position to dispute Stillman’s assumptions, serendipitous as the outcome always seems, that no Netcare hospital is ever the nearest competitor of its CHG counterpart and where it is (UCT Academic for instance is located close to Netcare’s Christian Barnard) there is an explanation for why they are not considered substitutes by patients.³⁴ However, given their location in the metropolitan areas even if Netcare hospitals may in some cases be strong potential competitors, the presence of other hospital choices makes CHG a less compelling competitive restraint to Netcare, than might be the case if its hospitals were located in under-served areas. Again, analysing this merger without a recent history of an independent CHG, makes it extremely difficult to hypothesise how it would operate as a competitive pressure on neighbouring hospitals, even if one took a less benign view of its metaphorical distance, competitively, from any Netcare hospital in the respective regions, than do the merging parties. The best evidence we have of how Netcare saw this issue at the time it began its relationship with CHG in 2001 was that it saw Fourways as the only immediate threat to itself of the erstwhile Malasela Group hospitals that CHG was attempting to recover and saw Montana and N17 as a threat to their respective neighbouring hospitals as we alluded to earlier.³⁵ As we know CHG failed to acquire Fourways which then was acquired by Life. Had CHG succeeded in obtaining Fourways this may have raised potential competition problems.

53. We cannot find that it has been shown that the merger will substantially lessen or prevent competition in the market for private patients in the regions identified by the Commission.

3) Market for funding

54. As we indicated earlier, more than ninety percent of private health care patients are funded by some form of medical insurance. Since funders do not operate according to the same incentives as self- paying patients, it is necessary to separately examine the effects of the merger on funders.

55. Competition between hospitals matters to funders in two areas. Firstly, around price or expressed differently, the rates at which funders are prepared to reimburse hospitals for services; secondly, over the form of product which funders can offer to their members. We examine each separately as the competition implications of the merger differ, depending which issue.

(i) Price

56. The hospital pricing regime has changed in recent years. Previously, until 2003, negotiations took place centrally between the hospital groups represented by their association, HASA, and the funders, through their association, the Board of Health Care Funders. Due to enforcement action taken by the Competition Commission, alleging that joint negotiations were collusive, both sides to this negotiation agreed in separate consent agreements with the Commission, to negotiate individually. Thus from 2003 onwards the hospital groups and funders have negotiated individually not collectively.

57. Thus, administrators negotiate separately with each of the three major groups and it appears collectively, via the NHN, with the independents.³⁶ Some are of the view that the change from centralised industry negotiations to a series of individual negotiations has suited hospitals more than funders. Among them are Alex Van den Heever, the witness from the Council for Medical Schemes, who stated that;

“The central weakness of the system lies in the mistake by the Competition authorities in outlawing centralised negotiations (i.e. market level) in respect of general tariffs. In reaching this decision it was presumed that forcing each scheme to negotiate their general fee for service tariff with every individual medical service provider (an impossibility) that competition would be enhanced. However, all that has occurred is a market power imbalance has been permitted that has enabled hospital costs increases to rise significantly where they were constrained before.”³⁷

58. We are not in a position to come to any judgment on this point, but it does appear that hospitals have huge informational advantages over all but the largest of funders and that this manifests itself in their ability to bring much greater sophistication to the bargaining table. Netcare and the Commission differed sharply during the hearing on whether funders have what competition law recognises as countervailing power. Netcare asserts that they have and that it (Netcare) does not get its own way in negotiations, because funders can refer members to other hospitals or threaten co-payment.

59. In their unsigned witness statements the Commission’s funder witnesses all stated that they lacked countervailing power. During their oral testimony and especially under cross-examination by Netcare’s counsel, some conceded that this was not always the case, whilst others introduced more nuance into what may have been a categorical denial ³⁸ and still others left us baffled with what they had to say.³⁹ Funders vary in size, resources and acumen, so it is unlikely that we can generalise about their ability to exercise some form of countervailing power. Certainly Discovery, the largest and most sophisticated funder, appears to be an equal in negotiations with any of its counterparts. If we judge it by the tone of its correspondence with Netcare, and on the evidence of its negotiations it can use its volumes relatively successfully. The same cannot be said of the others, albeit that they professed they had countervailing power. Presumably a funder which must compete with other funders is not about to make such a confession of weakness public. Yet there was little in their correspondence with Netcare or their descriptions of negotiations that suggested that there was any equality of arms at the negotiating table.⁴⁰ Netcare suggested that funders have a range of weapons open to them to force concessions. They can boycott a hospital or force members to make co-payments. Some have even published advertisements advising members not to go to a particular hospital. The funder witnesses whilst acknowledging that these tactics are open to them, suggested that they are more useful as threats than as weapons that have been successfully applied in practice. In practice members are not willing to boycott hospitals they wish to attend and co-payments may have as negative an effect on a funder as on the hospital.

60. Despite the end of central negotiations between hospitals and funders its culture still prevails. Negotiations occur once a year at the same time as they used to. Because hospitals have so many line items negotiations over tariffs appear to revolve more around the general than the specific. What happens in practice is that there is first a discussion on what medical inflation for that year is and once established, a negotiation of what increase will be on the previous year’s tariff for that group. Although the Council for Medical Schemes publishes recommended guidelines known as the National Health Care Reference Pricelist (“NHRPL”).⁴¹ Netcare negotiates off its own tariff for the previous year. It negotiates once a year with funders and agrees a rate that applies to all its hospitals. Until the end of 2006 the tariffs it negotiated applied to CHG as well. However, due to the Commission’s investigation of the non-notification of the present transaction they were advised by in-house counsel to negotiate separately.⁴² What happened then was that Netcare did all the negotiations first. Once concluded, CHG contacted the various funders and announced that it would be negotiating separately from Netcare. This appeared to confuse many funders who did not know why this break with past practice had occurred.⁴³

61. It seems that in general CHG settled on an increase slightly less than that of Netcare’s. Dr Stillman uses this as a basis for suggesting that when Netcare and CHG have negotiated separately the degree of difference was so slight so as to make a likely harm to competition as a result of the merger insufficient to raise concerns. However, this exercise in separate negotiations can hardly be a proxy for what would have happened had the groups been genuinely independent. CHG presumably knew where Netcare had settled, as this negotiation according to the witnesses had taken place first, knew that it was part of the larger grouping and hence the economic pressure on it was never akin to that of a small hospital group needing to fight the larger competitors it faces in the market place. The negotiation by their own admission was driven by legal not economic considerations. It would be highly artificial to regard this exercise as a proxy for what would happen in the market if CHG was independent of Netcare’s control.

62. However, despite this it appears that Netcare would still get what it does at the bargaining table over national tariffs, without CHG. Thus in cross examination by counsel for Netcare the following proposition is put to Mr Strauss of Discovery:

MR UNTERHALTER: And I assume that you will accept that whether the Community hospitals are in Netcare or wholly on their own or in the NHN network, ultimately in terms of their overall importance for Discovery’s tariffs and then ultimately premiums to members, these are just trivial in the big scheme of things.”

MR STRAUSS: That will be correct.⁴⁴

63. The Commission despite a diligent attempt to do so was unable to produce a funder witness who was either willing or able to state that the merger would make any difference on the outcome of national tariff negotiations on price. This leaves us with the Netcare evidence on this point which is that when negotiations take place, the big three have regard to the competition amongst themselves and no regard is had to the tariffs suggested by independents.⁴⁵

64. At best for the Commission was the evidence of Mr Allie of Melomed who claimed that his group is cheaper than its larger rivals. Part of the Commissions’ case is to suggest that Melomed is a proxy for an independent CHG and hence if it is cheaper than the three majors, so too would be an independent CHG. Yet internal correspondence from a funder suggests it considered Melomed to have made them a less competitive offer for a designated provider scheme than one of the major groups. ⁴⁶

65. One of the difficulties with this is that hospital pricing is notoriously difficult to compare. Strauss in his evidence talks of the four factors determining the cost of hospital care. In the first place one has the pricing of tariff items (ward, theatre and equipment fees etc) and then one has the prices for non–tariff items, such as pharmaceuticals supplied by the hospital. But the cost is not only dependent on the price of these services. How much usage is made of tariff and non-tariff items is just as determinative when costing services. When hospitals negotiate over tariffs it is a negotiation over only one of these three factors. Thus pricing the overall cost of treatment in a hospital is extremely difficult. Netcare and Discovery are engaged in an ongoing dispute over the former’s costs. Discovery claims in a study it conducted that Netcare is ... % [Confidential] more expensive than other hospitals. Netcare rejected this claiming that the methodology used was flawed and failed to account for the role specialists play in determining the extent of usage. As a result both firms agreed to commission a neutral party to conduct a further study. Using a different methodology, the firm concluded that Netcare was ...% [Confidential] more expensive.

66. Approaching the issue from a different perspective, Alex Van den Heever, an economist who presently serves as an advisor to the Registrar of Medical Schemes, the industry regulator, noted that there had been a 45,5% increase in the real per capita cost of hospital services to medical scheme members for the period 2001 to 2005. Van den Heever’s thesis is that there has been a steady rise in hospital expenditure since 1990. However, he contends that there has been a much higher increase since 1999 than before. According to Van den Heever:

“It is very probable that the key factor differentiating the period from 1999 onward was the fact that the hospital market in the main metropolitan areas became concentrated for the first time from 1999.”⁴⁷

67. Van den Heever believes that this merger should be prevented to at least retard this trend. Netcare disputes much of what Van den Heever has to say, in particular his contention that the increases we have seen in hospital expenditure must be a function of market power. Netcare has filed a report from its in-house analyst, Melanie Da Costa, who proposes an alternative theory of why costs have increased – a theory that predictably introduces several possible reasons for the increase in hospital expenditure, none of which is attributable to market power. Most importantly Da Costa alleges that Van den Heever confuses increased expenditure on hospital services with increased prices. Whilst Da Costa concedes that rising prices contribute to rising expenditure on hospital care, she points to the fact that there has also been increased utilisation of hospitals and this too contributes to increased expenditure, which she suggests, Van den Heever does not properly account for in his thesis.⁴⁸

68. Like the debate between Netcare and Discovery this debate is not one we are capable of resolving in the course of this merger. ⁴⁹ It may well be that an enquiry into hospital pricing is indicated, something on the lines of the present Commission enquiry into bank charges.

69. What is required for us to assess is whether CHG adds to Netcare’s pricing power or expressed differently, if it was lost to Netcare would it diminish its present pricing power. It does not seem it does. Netcare’s pricing power arises from its size as the largest hospital group in the country that can, negotiating nationally, bring that to the table. It seems unlikely that if it were to lose the five CHG hospitals that that power would be significantly diminished so as to meet a significant lessening of competition test. Nor is it likely, since tariff negotiations are national, that if CHG helped bulk up the independents that they would be able to constrain Netcare or either of the other majors from their present pricing strategies. None of the industry witnesses have expressed this view.

70. We conclude that the merger will have no significant effect on Netcare’s already existing bargaining power in national tariff negotiations.

(ii) Form of products

71. In the conventional model of reimbursement, a hospital charges a patient a fee and the funder pays it in accordance with an agreed upon tariff with the hospital or absent such, in terms of an agreed amount with the member, who pays the balance, referred to as a co-payment.

72. However, the medical rand is not infinite and funders and hospitals bargain to get their share of it. If costs continue to escalate as they appear to be doing and given that hospitals account for approximately 35% -50% of contributions, unless something in the system gives, fewer people will be able to afford medical aid. As it is the pool of beneficiaries has not increased beyond 7 million over a number of years. Fully alive to this both hospitals and funders are looking at ways to solve the problem.

73. In this hearing we have been told of two models that are in the process of being implemented. The first is referred to as alternative risk management (ARM), the second preferred provider networks or PPN’s. An ARM although subject to variation, broadly involves a hospital and a funder agreeing on what various procedures ought to cost and then agreeing that the hospital will be reimbursed at that standard rate for all procedures of that kind performed for that funds members, irrespective of what the actual cost may have been. The scheme can only work if it results in medical aid rates being reduced for a member, so that the pool of members on medical insurance is increased. This is good for hospitals as it increases their volume of patients and good for funders, as it increases their pool of contributors and hence risk profile.

74. But this type of scheme exposes both hospital and funder to risk. Both are bargaining over the cost of procedures and so the party with the best access to information will have an advantage. The risk for the funder is that it agrees to pay too much and that it does not recover enough in fees from members to meet the costs of treatment. For the hospital the risk is that it charges too little and that actual expenses exceed the agreed reimbursement rate.

75. But there is an added risk for both, a risk that applies equally to PPN’s. Since this type of scheme must charge cheaper rates than existing ones (if it is not it defeats the whole purpose of expanding the beneficiary pool) there is a danger that existing beneficiaries will, in the industry parlance “buy down”, that is if members see that they can get an equivalent benefit at a cheaper rate they will opt for the cheaper rate. For funders this is a serious threat and so they need to design a scheme that is capable of attracting new members, but repelling existing members on higher cost schemes. For hospitals the threat is much the same. Unless the low cost schemes bring in new patients to private hospitals, they have no interest in discounting fees on existing patients who are paying higher fees.

76. The key factor becomes freedom of choice. Funders and hospitals recognise that members used to freedom of choice in practitioners and hospitals, will be willing to pay for that privilege. Conversely, those unable to afford medical insurance presently will be willing to forego freedom of choice for the opportunity to access private health care. The art in constructing a viable low cost scheme is curtailing freedom of choice sufficiently to curtail buy-downs, whilst at the same time making the scheme sufficiently attractive so as to attract new members.

77. The object of the second type, the PPN, similarly involves an attempt to increase the beneficiary pool, but does not involve risk management. PPN’s involve schemes bargaining with a network of hospitals to get discounts in return for the promise of increased volumes in the form of new patients.

78. Strauss, whose company Discovery has developed KeyCare, which appears to be the only low cost PPN open scheme presently operating, explained the economics of such schemes.

79. Discovery’s market research indicated that members would require a ...% [confidential] discount to forgo a plan where they had freedom of choice. In order to achieve this because hospital costs account for ... % [confidential] of the costs of contributions, they would need to get a ...% [confidential] discount on hospital costs. ⁵⁰

80. Here too, there is great art in the construction of the scheme. A funder setting up a network needs to be able to provide enough hospitals to make the scheme attractive to its target membership, who may be located nationwide. If the scheme’s hospital base is too narrow, take up will be slight and the scheme will be unable to secure sufficient discount on volume. From the hospital group’s perspective it wants to drive volumes to underutilised facilities. The group would also want some degree of exclusivity. As Mr. Bishop of Netcare put it:

“All I would like is some level of geographic exclusivity, in other words don’t put me on and the Life Hospital next door. Try and give me something so I do gain in volume but feel free to choose any of the hospitals and add them to a network of [your] design …” ⁵¹

81. Another complicating factor is the nature of the scheme. Our legislation recognises two types of schemes, closed and open. Closed schemes are termed such because they confine their membership in some way, typically the employees of a firm. Closed schemes may be regional in nature and thus confined to the employees of a firm in one area if that is where the firm is located. Some closed schemes are national because the firm has branches nationally e.g. a large retailer. Open schemes are open to anyone willing to pay the fees charged. At present we are advised open schemes are required to accept members on a national basis. For this reason a good local deal from a hospital may not be that attractive if members located outside of the region are paying full rates elsewhere. What matters is the density of members in the discount region versus the density of members in non-discount regions. Some witnesses have suggested that regional membership will be allowed soon in terms of proposed amendments to legislation.⁵² Others have found means to make a scheme attractive for members in a region, but not others without contravening the right of anyone nationally to join the scheme.⁵³ We will assume for the Commission’s benefit that in the near future open schemes can be constructed regionally. Were CHG a hospital group with a substantial presence in any of the regions in which it is located, this might matter. As it happens it is not. Even in the Western Cape and Pretoria, the regions where it has two hospitals each, it is still not a major force.

82. To construct a national scheme requires more hospitals in more places. It is common cause that not even the big three have a sufficiently ubiquitous national footprint to construct a single scheme around, and while strong in some regions, may be under represented or not represented at all in others. Thus, a hospital group like Medi-Clinic has no presence on the East Rand, but has the only hospitals in some other smaller and growing regions such as Nelspruit. If a scheme is constructed around members who reside in Nelspruit and who already go to this hospital or if it has little room left for more patients, Medi-Clinic will have no incentive to offer a scheme a discount at Nelspruit – it is getting that business anyhow. If the scheme however has members elsewhere and they may be future patients at other Medi-Clinic hospitals not fully utilised or in markets where competition for patients is more intense, Medi-Clinic may be willing to offer a national discount (thus including its Nelspruit hospital) to get more volume at its other hospitals, but if the scheme was only regional it would have no incentive to do so.

83. Bargaining between schemes and hospitals takes place in the first place when the scheme is set up and secondly when hospitals are added or removed from the scheme.

84. A hospital merger may have consequences for this type of negotiation if it strengthened the bargaining power of a hospital group by either extending its national footprint or giving it dominance in a region where a scheme has members which it wishes to channel to a particular hospital. Thus if a scheme had a preponderance of members in Region X and pre-merger a choice between hospital A and B in that region to include on is network, A and B might be willing to bargain to get exclusivity in Region X. Post merger as the merged firm would be guaranteed that volume anyway it would have no incentive to bargain.

85. The focus of the Commission’s case in this merger has been to advance precisely that thesis. As expressed by the Commission’s expert, Dr Roberts:

“...market power at a regional level is clearly significant in negotiations by hospital groups to have their hospital included on preferred provider lists”.⁵⁴

He goes on to say:

“The presence of independents represents a significant source of competitive discipline in this regard, highlighted by the ability of hospital groups to exert influence in regions where there are no competitors.”⁵⁵

86. The problem for the Commission has been to translate this theoretical concern, which even Dr Stillman concedes is a valid theory of competitive harm, to the facts of this case. Stillman’s concession does not amount to an admission that it works in this case, because, as he puts it, none of the CHG hospitals are “must have” hospitals for any funder trying to develop a network.⁵⁶

87. The point Roberts makes is that this analysis is impeded by the fact that Netcare has been running CHG and so the evaluation exercise is about “assessing the likely future without the benefit of the immediate past.”⁵⁷ Note that this is the concern that we alluded to earlier in this decision.

88. What the Commission then seeks to do in the absence of a diagnostic history of an independent CHG is to use a proxy independent hospital group to draw conclusions of how an independent CHG may have behaved. The group chosen is Melomed, an independent group of a similar size to CHG. Melomed has 265 beds⁵⁸ and owns three private hospitals. Two of Melomed’s hospitals are in the Cape Flats and these are Gatesville Medical Centre and Mitchell’s Plain Medical Centre. Melomed’s third hospital is the Bellville Medical Centre (previously known as Jan S Marais hospital).⁵⁹

89. The problem for the Commission is that neither Melomed’s internal documents nor the views of those funders who testified, supports the fact that it offers more competitive pricing than the three majors or that it has been used as an alternative in the bargaining process. In short, nothing about Melomed, if it is to operate as a proxy for an independent CHG, supports the thesis that the Commission wishes to advance.

90. Roberts also argued that competition from independents mattered, not so much for the construction of a network, but for the replacement of hospitals on the network. Again while this may be a sound theoretical proposition the Commission did not develop a model for how this could be applied to the facts of this case. Bear in mind that hospital groups offer discounts based on volumes of new patients that funders introduce through a scheme. Removing a hospital that is part of a group offering discounts, in favour of an independent, could jeopardise the group’s willingness to offer a discount. The Commission needed to provide more analysis of how the merger would affect this bargaining in order to show that competition for replacement is likely and its implications.

91. Perhaps the most devastating blow to the Commission’s case comes from the testimony of Strauss. The Commission filed a witness statement from Strauss indicating the testimony it was anticipated he would give. This was drafted by the Commission on the basis of interviews with Strauss, but was not signed by Strauss, who had indicated that he was not willing to do so and required to be subpoenaed if the Commission wanted him to testify. Strauss in his oral testimony went through this written statement very carefully making a number of corrections. Crucially, he was to correct a paragraph in the statement that dealt with the type of discounts that could have been secured by Discovery from CHG for its KeyCare option. In the statement, paragraph 22 reads as follows:

“An independent CHG is not essential to the success of the Key Care Plan. Due to its size Discovery has been able to negotiate sufficient discounts with the major hospitals to ensure the viability of the plan. However, it is possible that if Discovery had known that the CHG hospitals were not part of the Netcare group, Discovery could have included certain of the CHG hospitals in the network sooner and at better rates.”⁶⁰

92. However, in going through this paragraph in his statement in his oral testimony, Strauss makes the following comment:

“ The point is that now that Netcare have come on to the network we have managed to secure from Netcare a better discount than I believe we would have from CHG had they been part of the NHN group. So it is actually the converse to what the statement says.” ⁶¹

93. What makes this comment crucial is the centrality of the KeyCare option to the Commission’s case. At present KeyCare is the only private sector low cost PPN. It is run by Discovery, the largest administrator of open schemes. If Discovery does not see an independent CHG as a significant price competitor then it is very difficult to see how the Commission can establish this without this evidence. Nor can Discovery be accused of being a sweetheart scheme for Netcare. Indeed a good deal of time spent on the dispute over pricing indicates the tensions between the two firms. Nor was Strauss for that matter a witness who was suggestible. His answer was given in chief, not during cross examination and he certainly indicated during the remainder of his testimony that he was not easily intimidated. ⁶²

94. Strauss also states that negotiations over the KeyCare option do not have a bearing on the national tariff negotiation.

95. The other witness who testified about establishing a low cost option was Mr Wambach from Old Mutual Healthcare. He testified about his company’s attempts to establish a low income scheme, but the plans have been put on ice. While he testified that independents had indicated that they could offer discounts, the scheme was never discussed with Netcare and so it is impossible to compare the discounts that were offered by the independents with those that Netcare may have offered.

96. Of course low cost PPN’s are not the only products being conceived of by funders and so we must consider the impact of the merger on other options. Mr Dawson, from an administrator known as Oxygen, testified about an attempt to establish a scheme based on managing doctor networks. The aim is to find low cost doctors who will, if they need to, refer patients to specialists, be required to channel patients to specialists who are also low cost.⁶³ Where a specialist is practicing at two hospitals the scheme will require channelling to the lower cost hospital. This scheme is still a work in progress. Dawson concedes that the aim of this scheme is not to influence hospital tariffs but doctor behaviour. Again he expresses no view on how the merger might impact on the efficacy of this scheme.

97. The one major PPN scheme beside Discovery, which has a low cost option, is the newly launched Government Employees Medical Scheme (GEMS). The Commission interviewed, but did not call any GEM’s representative as a witness. Netcare then approached GEMS to comment on the merger and elicited this comment from Eugene Watson of GEMS in an email to Richard Friedland of Netcare, dated 7 June 2007:

“I do not have a specific objection to the proposed acquisition.The number of hospitals represents a small percentage of the total number of private hospitals and our pricing strategy would not be significantly altered. If the transaction was not permitted and the hospitals remained independent, we would still need to negotiate reimbursement rates with all hospital groups, especially the larger ones who are responsible for a large proportion of the Schemes hospital spend.”⁶⁴

98. When asked in cross- examination why the Commission had not called anyone from GEMS as a witness, Roberts explained that his impression was that GEMS was still in the process of formulating its strategy and that it would be premature to get from them an understanding of what may happen in the future.⁶⁵

99. This may be a fair assessment, given that Watson’s views in the email do not seem to indicate any past experience in negotiating with hospitals, but rather an assessment of what is likely to happen. Yet even if this explains why GEMS were not called, it does not fill in the gaps in the evidence on this point. GEMS is a significant player in this segment. One would have expected to hear from it, if it thought the merger would have an adverse impact on its future strategies.

100. If the industry moves to an ARM model rather than PPN’s, the outcome is no different. These products, as we noted earlier, require an assessment of likely risk by the two players to the negotiation, the hospital and the funder. According to Strauss ARM’s are difficult for small players, be they funders or hospitals as they do not have the volume to offset volatility. ⁶⁶ He agreed with the proposition made in an internal CHG document that small groups will find it difficult to compete if the industry drifts towards ARM as the model requires advanced IT systems, integrated services, strong buyer power and a track record.⁶⁷ Bishop also makes the point that ARM’s require a strong balance sheet- something that independents typically do not have.⁶⁸

101. Of course this may beg the question as to whether ARM's are not products favoured by the larger hospital groups, precisely because they protect hospitals margins at the expense of funders and ultimately beneficiaries. Large hospital groups with their sophisticated IT systems can leverage the informational asymmetries in a risk based system and hence prevent erosion of their pricing power. Roberts in his testimony correctly refers to the fact that competition matters for the development and introduction of products. If a new product might erode the pricing power of hospitals they will have an incentive to resist its implementation. For this reason he holds that the presence of independents matters as it creates competition over the acceptance over new products. The difficulty in this merger is that we don’t know if the independence of CHG would have made any difference given the pre-existing market power enjoyed by the large groups and the relative insignificance of CHG.

102. It seems clear from the history of the only successful low cost PPN, KeyCare, that Netcare was resistant to it from the start, but failed in its attempts to establish an alternative, and eventually participated, offering a discount on its nationally negotiated tariff. This outcome suggests that notwithstanding the presence of CHG in its stable, Netcare was unable to impose its own product choice on funders and had to reluctantly accept one designed by a funder, albeit a very powerful one.

103. Let us briefly consider the history of how this came about. In about 2006⁶⁹ Netcare attempted to develop its own full capitation product called Netcare Plus with Discovery. The product failed. On Bishop’s version it failed because of so-called adverse selection. As the scheme had no savings plan the healthy avoided it while the unhealthy embraced it. As Bishop put it the last thing a scheme like this need is ‘bums in beds’. In other words, a risk management scheme can, if poorly designed from the hospital perspective, increase volume but not revenue. Medi-Clinic also attempted a scheme at its hospitals that was low cost and volume based but it too failed because of adverse selection. Consumers who were in areas where Medi-Clinic was the only provider ‘bought down’ as they were going to these hospitals in any event. This paved the way for the KeyCare option which suffered from none of these deficiencies, but required the funder to cherry pick which hospitals went on.

104. Initially the main backbone of the KeyCare plan was the Life Group, which had exclusivity in Gauteng and the Medi-Clinic group which had exclusivity in the Cape. Netcare did not participate and played hardball in negotiating access to its hospital in Uitenhage where no other private hospital existed and which Discovery was anxious to bring on to KeyCare because it was targeting employees of a large manufacturer in the area who were members of a rival scheme involving capitation.

105. That Netcare resisted this scheme initially, is best expressed in an internal document drafted by Bishop entitled “Summary of Netcare strategy regarding Preferred Provider contracting with Medical Scheme” In the document Bishop outlines the development in PPNs’ and concludes:

......

......

[Confidential]

106. He goes on to argue why Netcare’s strategy should be to partner schemes on a risk transfer basis. He explains that while Netcare could easily become the PPN for many schemes, if it was willing to discount tariffs, the discount would have to be “significant”in order for the scheme to offer members the kind of substantial discount that would make it worth their while to opt for a restricted option. Since, according to Bishop hospital costs are 35% of schemes costs, he describes the need for the discount to “exponential”. From the hospital’s point of view to earn back the discount, the increase in patient volume would need to be substantial.

107. Despite this reluctance Netcare came on board with KeyCare in 2007. It allowed Discovery to cherry pick hospitals, something it had resisted strenuously before. The reason for this change according to Bishop is that:

“it was a game we weren’t playing well enough in. We had made significant changes to our own strategy around cherry picking of hospitals etc, and we wanted in. It was on that basis that we asked, and we approached and we discuss and negotiated to be included”,

and

“I think this is the area where the real growth in our industry exists, the low income market and Key Care has been a success and it was a concern for us that we weren’t playing enough in this game” ⁷⁰

108. Thus Netcare is acknowledging that its strategy to impose a rival product has failed and that it needed to come to an accommodation with the PPN product or risk losing market share to rivals who had opted for the scheme. The probable reason for Netcare’s change in strategy was the rapid growth in numbers of people insured in 2007. After years of static figures this was the first increase in ten years driven by low cost options coming on to the market. The Commission suggests that this factor drove Netcare to realise that it had to change its policy against cherry picking.⁷¹ In short, Netcare, notwithstanding its control over CHG, could not defeat the launch of this new low cost product that requires hospitals to offer discounts off its tariff to members of the low cost scheme.

109. In brief the KeyCare concept requires hospitals to offer a discount from their nationally negotiated tariff in return for volume.⁷² Members of the scheme are attracted to it because its rates are lower than other available products. In return for the lower rates members’ choices of hospitals are restricted to those listed by KeyCare from time to time. This avoids the buy down problem referred to earlier as members are not referred to hospitals in areas with a wealthy demographic profile, but may be a less preferred choice. At present all the major groups have hospitals, selected by Discovery, on the plan, as well as some independents.⁷³

110. Melomed are also on the KeyCare list. But significantly, given that this group could be a proxy for an independent CHG they were not included on the list initially as Medi-Clinic was given Western Cape exclusivity. Only after Melomed laid a complaint with the Competition Commission does it seem that Discovery relented and included them on the network in January 2007.⁷⁴

111. We have no evidence about whether Netcare was able to defeat the implementation of any other proposed funding product. Some have failed because consumers did not like them, others are still works in progress, but those are the only ones we know of. There is therefore no evidence to suggest that control over CHG would give Netcare the market power to resist innovative new products that would inhibit its pricing power. Expressed differently whatever power Netcare may have presently to resist the implementation of new products, it has notwithstanding its control of CHG.

112. The merging parties also argue that PPN’s are a new product, have a very low market share out of present funding arrangements and are not the preferred choice of many players. ⁷⁵ We would be reluctant to dismiss concerns about the effect on this segment of the market simply because it is presently de minimus. If GEMS and Key Care are the growth plans, and they seem to be at the moment, their trajectory for growth may outstrip the rest of the market for private healthcare funding. In that case if a hospital merger was to lessen competition in relation to this segment of the market that might suffice to show a lessening of competition in the market, even if viewed as a market for private healthcare as a whole. But here the theory and the facts depart. Despite an elegantly crafted theory of harm the Commission has failed to find customers to support it.

Conclusion

113. We have considered carefully the various theories of harm advanced by the Commission and cannot find that on any, the merger with CHG will lead to a substantial lessening of competition. That is not to say that the Commission was misplaced in its concerns about this merger. Many of the witnesses, including some solicited by the merging parties, seem concerned about the private hospital sector.⁷⁶ What they have not been able to say is that the merger has contributed to this problem or whether it is a problem inherent in the present market structure, dominated as it is by the three major groups. Without such testimony it is difficult to impugn this merger. Whether such testimony might have been forthcoming if the merger had not been implemented already for some years, is one of those imponderables we cannot resolve. We can certainly say that many of the funder witnesses found it difficult to conceptualise an independent CHG – something they had never known - and then to hypothesise as to how it might have behaved differently outside of Netcare’s grasp. That this has redounded to the benefit of Netcare in defending the merger and to the detriment of the Commission in opposing it is also clear. It reinforces the reason why under our system prior implementation is unlawful.⁷⁷

114. In these circumstances the merger cannot be found to lead to a substantial lessening and prevention of competition and it is approved without conditions. There are no public interest considerations raised in this merger that would lead to a different conclusion.

115. A matter of some concern that arose during the course of this merger is the documentary evidence, which suggests that the three major hospital groups had at one stage entered into an agreement not to buy the erstwhile Malasela hospitals whilst in liquidation. Although the Commission had pertinently referred to this minute in its recommendation, no senior person from Netcare put up a witness statement, and thus no-one in a position to explain this comment, was tendered as a witness. Given that competition in the private health care sector is now so wholly dependant on competition between the three major groups any evidence that they may reach secret understandings over issues is disturbing. The Commission should not let this issue go without further enquiry.

________________ 5 November 2007

N Manoim DATE

Tribunal Member

U Bhoola and T Orleyn concur in the judgment of N Manoim

Tribunal Researcher : R Kariga

For the merging parties: D Unterhalter (SC) assisted by J Wilson, instructed by Webber Wentzel Bowens Attorneys

For the Commission : D Berger (SC), assisted by B Shabalala, instructed by

Eric Mabuza Attorneys.

---