image

Enforcing Competition Rules in South Africa

To Terry
(like I mean it)

Enforcing Competition Rules in South Africa

Thieves at the Dinner Table

David Lewis

Executive Director, Corruption Watch and Extraordinary
Professor, Gordon Institute of Business Science, South Africa

Edward Elgar

Cheltenham, UK • Northampton, MA, USA

image

Contents

Preface

vi

1. Beginnings

1

2. The new competition regime

24

3. Mergers

70

4. Abuse of dominance

130

5. Cartels

191

6. Competition enforcement on the world stage

228

7. Conclusion and a postscript

261

Index

283

Preface

To all the details that I undoubtedly got wrong in this book, I am most concerned not to add a failure to acknowledge and thank all of those who are responsible for much of what I got right. That having been said, I’ve undoubtedly omitted some important names from this list and to these people I should apologise.

Several people read extracts as the book unfolded in the fits and starts that seem to characterise the writing enterprise, especially for those who, like me, are cursed with a maddeningly short attention span. In particular I want to thank Yasmin Carrim, Ann Crotty, Eleanor Fox, Brenda Goldblatt, Norman Manoim, Wouter Meyer, Mike Morris, Shan Ramburuth and Simon Roberts. They helped me with both style and substance but are, as the traditional disclaimer goes, not to be held responsible for either.

Annalee van Reenen and Ntsako Mgwena at the library of the competition authorities and Rietsie Badenhorst, Lerato Motaung and Tebogo Mputle, dear friends in the Tribunal office, were constantly asked to recall illustrative cases and examples and track down references and the like, a key element of any research process and one that I am particularly ill-equipped to undertake. Lebo Moleko was my able and patient research assistant during an important period of the writing and research.

Nick Binedell, the dean of the Gordon Institute of Business Science (GIBS), and the GIBS staff, provided me with a stimulating and supportive home during much of the period that I spent working on this book.

Catherine Garson was my editor, the one person to whom the frequently-used tribute ‘this could not have been done without her’ truly applies. I would also like to thank Nicolas Wilson, my editor at Edward Elgar.

The International Development Research Centre (IDRC), the Canadian foundation that has played such a vital role in stimulating research in developing countries, provided me with a generous grant. I’m particularly grateful to Susan Joekes, my point person and friend at the IDRC, for her confidence in this project.

Terry – to whom this book is dedicated – and Jonah and Jessie, who have had to cope with the mess, the swearing, the irritability and, above all, the great sense of displacement and loss I experienced when my term of office at the Tribunal ended, are owed a debt of gratitude too great to express.

And finally I want to acknowledge the people with whom I worked. Some are recognised on this page and others in the text. But many are unacknowledged. If this book leaves its readers with no other lasting impression, I hope that they will be struck by how much I loved my work at the Tribunal. And that just has to come down to the people with whom I worked. They came from all over the world, and from many varied backgrounds. But mostly I think of the small group of people in the Tribunal office and on the panels with whom I had the privilege to work. On this score I really lucked out, and for that I am eternally grateful.

1. Beginnings

When I sat down in front of my computer, setting out to tell the story of South Africa’s competition institutions and confronted by a blank screen headed ‘Chapter 1’, the only opening sentence that came to mind – and this after chairing the Competition Tribunal for 10 years and writing many thousands of words on the subject of competition – was that which opens so many junior school essays: ‘It was a dark and stormy night …’ But, strange to tell, this may well be an appropriate opening for the story that I want to recount. At the time that this tale begins, darkness or, at least, secretiveness and a profound lack of accountability did indeed mark the workings of the South African business sector. Storminess did not characterise inter-corporate relations – indeed the marked absence of storminess or, conversely, the perceived cosiness that characterised these relations was precisely the problem that the new competition policy was expected to confront. However, a definite element of inclemency was provided by the nature of the interface between a new government and an old business establishment whose relationship, certainly at that time, may euphemistically be described as one of mutual suspicion, marked by fairly regular bouts of considerable turbulence.

In the eye of that particular storm was the stated intention of the new government and its allies in the trade union movement to introduce a robust competition policy, centred on a new antitrust statute.1 I vividly recall a conference on competition policy held sometime in 1994 at the University of Cape Town’s Graduate School of Business addressed by the newly-appointed Minister of Trade and Industry, Trevor Manuel. As Manuel has made the transition from community activist to international statesman, he has adopted a rather reflective, ruminative style of address. But he learnt his trade at mass meetings on the Cape Flats and he used this occasion to deliver a particularly stirring address, condemning, in the strongest terms, both the anticompetitive structure and conduct of business and promising a new competition statute equal to the task of reducing ownership concentration and correcting a century of anticompetitive conduct. It was just the sort of bugle call heralding the cavalry’s arrival that a wannabe competition enforcer wanted to hear.

I recall just as vividly the vote of thanks delivered by Michael Spicer, then the public affairs executive of the Anglo American Corporation, the largest of South Africa’s conglomerates and for long perceived by those now in government as a prime target of antitrust law. Spicer began his contribution with the observation that the only possible comfort he could draw from the minister’s speech was that ministers enjoyed ‘short shelf lives’. Ironically, although Manuel’s tenure as Minister of Trade and Industry was indeed to be brief, he was soon to assume the office of Minister of Finance and would ultimately become the longest-serving minister in that post in the world.

How does this story end? It is ‘common cause’, to use the term beloved of lawyers, whose quaint and pompous phraseology I have come to know so well, that the darkness has lifted somewhat and that this is, in very significant part, due to the Competition Act of 1998 and the manner of its enforcement. The investigations of the Competition Commission and the hearings of the Competition Tribunal have illuminated many of the dark corners of South African business. For the most part these revelations consisted of micro-economic data, as well as corporate strategies and cultures, that in many other countries were long part of the public domain. But the revelations have naturally also included many clandestine conspiracies to fix prices or rig bids. Many of South Africa’s leading business people – including, as it so happens, the then CEO of the Anglo American Corporation – have appeared before the Competition Tribunal obliged to give evidence, to submit themselves to the exacting cross-examinations that are exciting and fascinating to observe but clearly sheer hell to be subject to, and, on many occasions, to contritely admit guilt for one or another truly heinous conspiracy against South African consumers. With the assistance of a small cadre of curious and smart business journalists who soon learnt that the Tribunal hearings were extraordinary venues for getting behind the bland offerings of corporate public relations departments, sunlight began to filter into the hitherto hidden corners of South African business.

There is still, to be sure, plenty of storminess surrounding the enforcement of competition law. This is built into much of the activity of law enforcement. And it is a feature of the drama surrounding adversarial courtroom exchanges, especially where the adversaries are, on the one hand, pillars of the business establishment, people more used to giving instructions than answering questions and fielding public criticism, and, on the other, a public authority, the Competition Commission, and, frequently, competitors or customers of the firms under scrutiny. Needless to say, the sense of drama is inevitably heightened by the exceptionally able counsel who have been attracted to the practice of competition law. It is also a particular measure of how high the stakes are in the business of competition law enforcement.

But something about the quality of the turbulence has changed over the years. Most of the institutions and individuals involved have learnt a few things. Those working for the competition authorities have come to appreciate the acute complexity of many of the questions that face them and the far-reaching consequences of some of their actions. Our constitutional environment has taught the competition enforcers and adjudicators the merits of, or at least the requirement for, administrative fairness – even when invoked, as is more often than not the case, by those for whom the invocation of high-minded principles means little more than an opportunity to seize strategic advantage in litigation.

For their part, many business people have come to appreciate the deep contempt with which the public view anticompetitive conduct – conduct whose commercial legitimacy, if not legality, the perpetrators had long taken for granted. And they have come to appreciate that they will generally be treated fairly, if firmly, by the competition authorities. In short, if this still-robust relationship has not turned into one characterised by mutual affection, then it has, with predictable ups and downs, engendered a measure of mutual respect. That is about as good as it gets in this business; at any rate, that is about as good as it should get.

Indeed, the ‘big story’ about the South African Competition Act and the institutions that it spawned is that, on most measures, it has been a highly successful enterprise, and is widely acknowledged as such. It’s well-nigh impossible to settle on any single measure of success. But for the time being it suffices to observe that merger regulation is effective and efficient and that there have been many successful prosecutions of anticompetitive practices. The authorities have earned the respect of the business community, of government and of organised labour and they are highly regarded by their international peers. This alone makes it a story worth telling. Institutional success is all too rare a part of the ‘miraculous’ transition from apartheid to democracy. Yet here we succeeded in a field that until very recently had been deemed, partly because of its inordinate technical complexity, to represent activity in which only developed countries could build effective institutions.

There is, to be sure, a certain amount of criticism by the South African public of the competition authorities – at least, to the extent that those who write ‘letters to the editor’ typify the public. Although we will examine these criticisms in greater detail, my strong sense is that they reflect the impatience and somewhat grandiose expectations of a public that has come to understand the importance of the competition authorities’ work. In short, a competition culture has taken root in very infertile soil. This is something that many better-resourced competition authorities never achieve. It is truly remarkable that in a single decade the South African competition authorities have, through the quality of their work rather than any slick public relations campaign, firmly installed themselves as part of the business environment and, more than that, as part of national public discourse and debate.

Above all, though, I think that this success is reflected in the morale and calibre of those responsible for it, namely the staff of the Commission and Tribunal. Both institutions work in areas where the quality of their staff is highly visible to the private sector and to public opinion. This inevitably means that missteps and incompetence are immediately exposed, and so too are efficiency and competence. In the early years of the work of the competition authorities, those staff members who displayed the latter attributes – along with some who frankly did not – were very quickly picked off by a private sector, law firms in particular, offering significantly higher salaries and lucrative futures. Ten years on, the flow has, at the very least, become something of a two-way movement. Some of the best qualified and most competent staff members who had left the Commission and Tribunal have returned. Senior Commission staff members of the exceptional calibre of the chief economist, Simon Roberts, have drawn several of their outstanding students into the ranks of the Commission. And, most gratifying, young lawyers and economists have clearly begun to view the authorities as interesting learning and career-enhancing environments to work in. This probably represents the most important positive assessment of the performance of the competition authorities and the strongest basis for a successful future.

But I’m getting way ahead of myself. Let’s start at the beginning and, as with so much about South Africa, 1994 marks year one.

The year 1994, in which South Africa’s first democratically elected government took office, represented a radical disjuncture with its long apartheid and segregationist past. All such breaks – effectively a culmination or, at least, an important staging post of a ‘revolution’ by any other name – are represented by, indeed are driven by, narratives comprising at their most general level a range of stylised facts, circumstances and events that evoke dispossession, poverty, inequality, ‘unfairness’. In South Africa, the dominant narrative that underpinned the struggle for democracy was naturally the racially-assigned privilege that marked every sphere of life. And, to a significant extent, it is this narrative, expressed both in public policy and in the policies and behaviours of a range of non-governmental and private institutions, that continues to define and underpin the activities and discourse that predominantly seek their legitimacy in their efforts to confront the inequity of the past and the poverty and dispossession that remain its most powerful manifestation.

This book is about a public policy intervention – in the form of the Competition Act – that has deep roots in this dominant narrative. In particular, it is rooted in the concentration of ownership of private wealth in the hands of a small number of large corporations – for the most part, highly diversified conglomerates – that were in turn controlled by a select group of white families. The perceived counterpoint of this concentration of private wealth and economic power was precisely the dispossession, the poverty, and the unequal and unfair treatment of the majority black population.

And so the data, as likely to be read in respectable research and media publications as to be heard on the picket lines of the unions or in the underground cells of the liberation movements, revealed that the minerals-based but at that time highly diversified conglomerate, the Anglo American Corporation, controlled a significant proportion of the value of shares traded on the Johannesburg Stock Exchange. In addition, a large proportion of the remaining wealth represented on the stock exchange was controlled by a small number of other highly diversified mining, finance and industrial conglomerates – notably the Cape-based Rembrandt Group – and a trio of financial institutions, to wit the Old Mutual Group, Sanlam and Liberty Life.

Nor were these faceless corporations. Families, white families, loomed large in the control structures of many of the powerful business institutions. Some of these fabulously wealthy families – pre-eminent among them, the Oppenheimer family – traced their antecedents to the Rand-lords who controlled the gold, diamond and other mineral wealth of South Africa in the latter years of the nineteenth century. Others, most notably the Cape-based Rupert family, derived their initial wealth from lucrative consumer goods markets, namely cigarettes and liquor; although, like their northern counterparts, they too diversified their interests into many unrelated corners of the economy, into mining, manufacturing, and financial and other services. These were classic conglomerates, companies that held large, generally controlling interests in diverse companies across a range of sectors.

These aspects of South African economic and corporate life provided an important imperative for radical change, and the counterpoint for the sort of society and, in this particular instance, for the distribution of control of economic wealth, that was loosely envisaged by those determined to consign apartheid to history’s ash heap.

Of course, the factors underlying this corporate structure and the very data themselves are significantly more complex than what I have suggested above. For one thing, the pre-eminence of conglomerations did not always, or, at least did not only, express the predatory character of the owners or decision-makers of the corporations concerned. It derived from a diverse array of factors ranging from the nature of the mining-house form of organisation in which the South African corporate economy was rooted to the closed nature of the economy, specifically including the existence of exchange controls and, in later years, the advent of economic sanctions. Moreover, the two largest of the financial institutions that figured in the concentration narrative, Old Mutual and Sanlam, were at the time mutual societies, notionally owned by their policy-holders rather than by a dominant white family.

Nor of course do the conventionally-evoked data on concentration of ownership reveal the existence of many large listed and unlisted corporations that were not controlled by the dominant conglomerates or the existence of significant unlisted business interests that were similarly outside the dominant control structures. However, so wide was the reach of these behemoths that it would have been difficult for any player in the South African economy to avoid engaging, as customer, supplier or competitor, with interests that were controlled by the dominant conglomerates. The smallest independent spaza-shop2 bought cigarettes manufactured by a company in the Rembrandt Group and the largest steel manufacturer would inevitably count an Anglo American-controlled auto manufacturer or gold mine among its most important customers.

In any event, no amount of nuance served to alter the fundamental reality that an extremely large slice of South Africa’s productive asset base was controlled by a small number of very big corporations and that the most significant of these were controlled by conspicuously wealthy white families or, if not, by well-entrenched and richly-rewarded white managers.

And nor does it disguise the inescapable reality that this highly skewed ownership of South Africa’s wealth was acquired under apartheid and reflects the racial exclusion that defined that noxious system of social control and governance. Of course, surrounding even this bald fact – and that it is indeed factually correct is indisputable – there is also no end of debate, both scholarly and political. Scholars and political activists disagree deeply about the very fundaments of the relationship between capitalism and apartheid. Others dissect the diverse relationships of the various ‘fractions of capital’ to key features of apartheid – for example, whether mining, manufacturing and agriculture have an identical interest in the pass laws and influx control? More prosaically, leading business figures of the time held divergent political views – some were pillars and financial backers of the ruling National Party, others of the liberal opposition.

But again, none of this nuance, although of tactical and strategic importance to the unions and the liberation movements, served to alter the incontrovertible fact that the wealth was accumulated and concentrated under political conditions in which, for the overwhelming part, only whites were permitted to own and manage the institutions and assets in question.

I raise all of this at the outset because these are the narratives and the data that account for the orientation towards the corporate economy on the part of those who would, after many years of bitter struggle, come to displace the apartheid government. The concentration of ownership of private wealth was a graphic representation of the dispossession, poverty, inequality and ‘unfairness’ that characterised apartheid. The proposed solutions to this concentration problem that were ultimately pursued were many. Initially, nationalisation was the most loudly proclaimed solution proffered by the anti-apartheid activists, although a range of factors conspired to ensure that this never came to pass. Black economic empowerment, effectively a grab-bag of policies and mechanisms that enabled the small black elite to enter the corporate kingdom, was arguably the most prevalent of the options pursued by the post-apartheid government. But robust antitrust enforcement, perceived as an instrument to discipline, indeed to fragment, the ownership of corporate South Africa, was clearly conceived of as another such instrument for levelling the corporate and economic landscape.

There are two features of this early debate around concentrated ownership that bear immediate mention. First, although the venerable Freedom Charter, adopted in 1955 by the African National Congress (ANC), the party that ultimately came to power in 1994, clearly identifies ‘the monopolies’ as one of its pre-eminent targets – under the heading ‘The people shall share in the country’s wealth’, the charter proclaims that ‘the mineral wealth beneath the soil, the banks and monopoly industry shall be transferred to the ownership of the people as a whole’ – the reference to ‘the monopolies’ appearing to be a reference to the dominant conglomerates and to the concentrated ownership structures that they represented, rather than to the concentrated market structures that represent antitrust’s conventional understanding of, and concern with, monopoly. Moreover, the solution mooted by the Freedom Charter is, as already intimated, clearly nationalisation, a response in keeping with the political orientation of the ANC and its domestic and international allies at that time.

Second, in the discussion of ownership structures, then and in the significantly more detailed and advanced debates immediately preceding the demise of apartheid, there is little mention of state-owned enterprises and of competition policy more generally. In fact, again in keeping with conventional national development strategies of the post-World War II period, state ownership of basic industrial products and services – steel, energy, transport, telecommunications – was a prevalent and critical feature of the ownership structure of the South African economy. Indeed, far from being controversial or a subject of policy contention and debate, if the sparse pronouncements of the Freedom Charter are to be believed, the dominant view envisaged an extension of state ownership through nationalisation of ‘mineral wealth, the banks and monopoly industry’. By the same token, contemplation of a reduction in the regulatory presence of the state was also not much in evidence.

I want to suggest two early conclusions based on this extremely truncated representation of a slice of our history. The first is that those struggling to overcome apartheid and who would ultimately rule the country had few qualms about intervening in the existence and exercise of private property rights. The very nature of these rights represented precisely a critical pillar of all that apartheid signified. For this very reason it is not surprising that the predominant concern with property rights should be less preoccupied with the manner of their exercise (that is, with the exercise of market power or even collusion) than with their racially exclusive character and the exceptional degree of ownership concentration in a small number of white hands – ‘domination of the economy by a minority within the white minority’ is how the ANC’s 1992 Policy Guidelines for a Democratic South Africa expressed it.

The historically favoured mechanism for dealing with this concentrated ownership was, as is confirmed by the Freedom Charter, nationalisation, the most far-reaching intervention in private property rights. Indeed, in one of his earliest statements on being released from prison, Nelson Mandela himself reaffirmed – albeit on only one occasion that I can recall – the ANC’s historical support for nationalisation. I recall that this statement generated much controversy and anxiety in business circles. However, the truth is that by 1994 nationalisation was, for a variety of well-known reasons, no longer a viable option, notwithstanding its recent re-emergence in the schisms and shifting alliances that characterise the contemporary ANC. But the problem of concentrated ownership had not gone away and so support for an alternative robust intervention in the exercise of private property rights came to the fore, this being robust antitrust enforcement – largely because it was viewed as a mechanism for breaking up the conglomerates, for broadening ownership.

Never mind that these solutions represented polar-opposite public policy orientations: nationalisation, the substitution of state- for market-determined economic outcomes; antitrust, the defence of market-determined outcomes. What they shared was the view that in South Africa the manner in which private property rights had been acquired and exercised had, to put it mildly, not benefited the vast majority of the population and that a programme designed to curtail their exercise was not only considered legitimate, but a fundamental responsibility of the post-apartheid democratic government. This accounts for the easy, indeed enthusiastic, reception by the post-apartheid government and the general public of robust antitrust enforcement. In South Africa, antitrust was not part of a market liberalisation agenda; it was a central feature of the democratic project.

It also explains, just in case anyone was wondering, how a former trade unionist of long standing got so deeply involved in competition issues, a background that occasioned some bemusement from my newfound international competition colleagues. I have been fortunate to have had two jobs that engaged me heart and soul. First, the unions and then the Competition Tribunal. I know that a key common thread was that each involved a large agenda easily expressed in idealistic terms but with deep practical and institutional content – the first, building socialism from a capitalist society; the second, in the words of Raghuram Rajan’s memorable book title, saving capitalism from the capitalists. This role and importance of competition policy were, in my mind at least, bolstered by my increasing experience of the limitations – both managerial and political – that the state confronted in taking on the role as the instrument of economic growth and development. Expressed crudely, I began to have more confidence in the prospect of a JD Group disciplining an Ellerines than of a Minister of Furniture achieving similarly efficient and equitable outcomes.3

However – and this is my second conclusion – robust antitrust enforcement, the robust enforcement of rules governing the exercise of principally private property rights, did not extend, and never has extended, into a coherent public policy regarding the exercise of public property rights, regarding, in other words, the governance, the conduct or indeed the very purpose and role of state-owned enterprises. And this despite the historically poor and discriminatory service provided by key state-owned enterprises – think Transnet and Telkom – and despite the fact that the employment practices of many of these enterprises represented some of the very worst features of apartheid. It appears that, because the state- owned enterprises were part of the political kingdom inherited by the new government, there was no need to be concerned about the goodwill and conduct of their new owners and soon-to-be new managers. The new government did not have to concern itself unduly with imposing robust and independently enforced rules on their conduct. To the extent that the state-owned enterprises were subject to rules and requirements – and they were loosely viewed as major instruments for realising government’s key social objectives, vide the attempts by government to enhance telephone penetration – this could be achieved through the exercise of direct political control, largely through the ‘deployment’ of reliable cadres to their leadership, rather than independent regulation.

To be fair, the newly-installed democratic government did introduce, relatively expeditiously, a regulatory framework governing certain state-owned natural monopolies, notably in broadcasting, telecommunications and energy. But despite a formal commitment to independent regulation, the new government has never conceded genuine independence to the regulators or to the governing structures of the state-owned enterprises.

Indeed, this argument could be taken significantly further: despite the strong commitment to robust antitrust law, this has never been reflected in support for a strong competition policy. Proponents of a coherent competition policy are inherently sceptical of a leading role for the state, be this in the form of extensive state ownership, an interventionist industrial policy or a protectionist trade policy. However, competition policy proponents are not ‘free market’ fundamentalists. They recognise the existence of market failures.

Those to whom the proponents of competition policy assign responsibility for the enforcement of antitrust law are typically preoccupied with the market failures that arise from the existence and abuse of market power. They recognise, in other words, that competition – or, we may say, markets – must be subject to rules, antitrust law. This will serve to ensure the effective functioning of markets, so that market participants, whether in the shape of colluding competitors, aggressive acquisitors or exclusionary monopolists, do not, in their narrow self-interest, undermine the very structural and behavioural basis of a market economy – thus saving, as I’ve already noted, capitalism from the capitalists.

Market power is, of course, not the only market failure that competition policy must recognise: it recognises natural monopolies and the requirement to regulate them; it recognises the informational asymmetries and coordination failures that generally provide the raison d’être for industrial policy; and it recognises that government will have imperatives, usually expressed as a desire to influence distributional outcomes, which will on occasion trump those outcomes dictated by market-based interactions.

However, at base, competition policy’s default position is one that would envisage a limited role for the state. The state’s role, competition policy would hold, is essentially to provide an environment – the rule of law, the macroeconomic policy environment, the microeconomic regulatory environment and, critically, the public infrastructure – within which private investors flourish, producing goods and services and creating jobs.

And so there was undoubtedly an upside in the introduction of competition policy rooted in democratisation rather than liberalisation: this was to be found in the enthusiastic public support for antitrust enforcement even if the principal concern, namely ownership concentration, was strictly speaking somewhat outside antitrust’s central preoccupations and powers.

The downside of this historical lineage is represented by the absence of a coherent, complementary competition policy. Competition policy emanates from the market liberalisation process. So the same interest that enthusiastically embraced competition law because of its association with the democratisation project viewed competition policy as potentially compromising the consolidation and extension of that project, precisely because it seeks to limit the freedom of the state to assume powers and instruments that enable it to intervene in market outcomes, including state ownership. Hence the apparent paradox at the heart of South Africa’s competition policy – powerful enforcement of competition rules governing the conduct of private market participants, combined with a weak and incoherent competition policy specifying the role of the state in the economy.

But let’s return to the central concerns of antitrust law, namely market structure and the conduct of market participants, because while skewed and concentrated ownership structures played the major role in the enthusiastic response to antitrust, the competition law and its practice would ultimately have to take the ownership structure largely as given.

Potentially the greatest impact that antitrust would have on ownership structures would be in the greater ease of access to the markets that it defended. But greater ease of access served principally to discipline large firms, and its impact on the structure of ownership was indirect and incremental at best. Certainly merger regulation could, in selected instances, inhibit further concentration of ownership. Equally, the interlocking directorships characteristic of highly concentrated ownership structures play a role in identifying and prosecuting a cartel. The competition legislation provides for the possibility of structural remedies in both merger regulation and cases concerning anticompetitive conduct, and these too may impact on the concentration of ownership.

But, when all is said and done, effecting changes in the structure of ownership plays a secondary, indeed minor, role in antitrust enforcement. The central concern of antitrust with market structures is the extent to which structure underpins market power. And so antitrust law is preoccupied with keeping the structure of the market as open as possible through exercising vigilance over dominant large firms that are capable of engaging in exclusionary conduct (of actual and potential competitors) and through guarding against collusion and anticompetitive mergers, in other words with preventing regulation by means of private agreement of a competitively-structured market.

This is not to say that the new government was unaware of orthodox competition law’s central preoccupations with market structure and with the conduct of participants in those markets. These concerns are clearly reflected in the ANC’s 1992 policy document, and, more important, they are echoed in both the preamble to the statute that emerged from this policy process and in the core problems that this new law was responsible for addressing.

COMPETITION LAW BEFORE 1998

While the eventual passage in 1998 of the Competition Act represented a watershed in the role and prominence of antitrust law in South African corporate and public life, it was by no means the first such experiment.

It’s not my intention here to write a pre-1998 history of competition law in South Africa, but there is undoubted value in understanding something of the present competition regime’s antecedents. Several of the people tasked with drafting the 1998 Act, including myself, were appointed members of the predecessor of the present competition authorities, the Competition Board. I was appointed chairman of the Board in its dying months. The purpose of this was principally to effect the transition from one regime to the next. The experience undoubtedly did exercise some influence on what followed, particularly in what we learnt of the structure of the Competition Board and the manner in which it conducted itself.

Although on assuming membership and later the chair of the Competition Board, I encountered a small core of dedicated staff – a handful of whom remain, to this day, highly valued members of the new competition authorities – and although I learnt much from them and from the experience of actually starting to deal with competition issues, it was, on the whole, a singularly unedifying experience. Stuck away in a corner of the Department of Trade and Industry’s squalid and cockroach-infested building in downtown Pretoria, flanked by the mausoleum-like State Theatre and the South African Reserve Bank’s forbidding coal-black skyscraper, the Competition Board felt every bit like the Cinderella body that we, the representatives of the new order, firmly believed it to be.

The chair’s office was the size of a small hall. It was as dark as a grave with the most prominent piece of furniture being a massive vault-like safe, befitting the emphasis placed by those in charge of competition enforcement on ‘the utmost secrecy’. By the time I got to open the safe it was completely empty and I suspect had probably been that way for much of its existence.

Although most of the members of the small staff were welcoming and helpful, if understandably somewhat wary and uncertain, the former chair’s secretary was a fully paid-up member of the ancien regime, expressed in open personal hostility towards me and a numbing regard for indecipherable bureaucratic requirements. I recall, on my first day in the office, asking my newly-acquired secretary for a bulb to be fitted into the desk lamp – anything to relieve the all-pervasive gloom, both external and internal. This seemingly simple request occasioned a flurry of tut-tutting, the production of purchase requisitions to be completed in triplicate and a general dissatisfaction at the thought of any change – or was it merely light? – being introduced so early on in the new regime. Hearing all this going on, a young woman, the most junior member of the Board’s staff and clearly one of the few symbols of its recent de-racialisation, strode into the office, and, standing on a chair, removed the bulb from the corridor outside and installed it in my desk lamp. Needless to say, when I was appointed to chair the Competition Tribunal, Lerato Motaung became my first personal assistant and today is the exceptionally competent and tough-minded registrar of the Tribunal, as capable of dealing with the obstructive, entitled behaviour of Johannesburg’s most aggressive attorneys as she proved to be in dealing with the dyspeptic secretary of the Competition Board.

At least as depressing was the boardroom itself. This was a room next door to the chair’s office. But it was different in that it did not consist of the heavy teak that lent due gravitas to boardrooms (and that was much in evidence in the chair’s office), but was rather a peculiar ensemble of individual school desks arranged in a rectangle, at which the members of the Board and, on occasion, the petitioning lawyers and their clients sat. Without the dark furniture and heavy drapes of the chair’s office, it might have been a fairly light room, but for the fact that the windows were so grimy that precious little light filtered in.

The meetings of the Board were formal to the extent that they were minuted, but for the most part they were pretty informal, if somewhat eccentric, affairs. As I will elaborate below, the part-time members – that is, all but the chair – mostly appeared to be representatives of one or other government agency or special interest group. Their attendance was erratic at best – some members never appeared once during my tenure. This meant that it was quite conceivable – and it frequently occurred – that a matter spanning more than one meeting was discussed and decided on by two entirely differently-composed boards.

The meetings were not often attended by the lawyers or business people petitioning the Competition Board for a decision on one matter or another. They largely consisted of receiving reports from the Board’s investigatory officials and ratifying agreements that had been concluded between the full-time chair of the Board and his officials and the affected businesses. There were occasionally hearings: at times, a member of the Board, usually one of the post-1994 appointees, requested that parties seeking a decision appear before the Board, but these were exceptional. The frequency of hearings would rise as the new regime drew closer and its character became clearer, with the progress of the National Economic Development and Labour Council4 and related legislative processes.

The full-time chair of the Board was Dr Pierre Brooks, a quiet, modest, donnish man, who seemed to enjoy the intellectual challenges of competition law rather more than the thrill and excitement of enforcement. But he had a determined, stubborn streak and he played a constructive, if limited, role in the transition to the new competition regime.

When I and the others on the drafting team joined the Board, we also encountered several other members who had been appointed by the new government. I most clearly recall a rather odd pair of friends, Robin McGregor and Christine Qunta. Robin was the publisher of McGregor’s Who Owns Whom, the source of much of what was known about ownership structures in South Africa, and who possessed, as one might expect, an encyclopaedic knowledge of the South African business world, at least in so far as it concerned questions of ownership. He was passionate about the need for robust competition enforcement, which he saw principally as a mechanism for breaking up existing concentrations. That he wasn’t able to use his membership of the Board to achieve this didn’t mean that he had to accept any further concentration. The upshot was that I can’t recall a single merger that ever passed muster with Robin, who frequently was the sole dissenting voice on a number of merger decisions. I never agreed with Robin’s approach to competition enforcement and especially merger regulation, but his energy and enthusiasm and sheer belief in the power of competition law were a breath of fresh air, made more so by the mustiness of the surrounding environment.

Robin’s populist, anti-establishment approach to competition enforcement or, more likely, his hostility towards corporate executives, the vast majority of whom were of course white males, commended itself to Christine Qunta. Christine, who was ultimately appointed a part-time member of the first Competition Tribunal, was a strident bundle of unalloyed and unmediated aggression. She and I were both living in Cape Town at the time, so I was occasionally tasked with driving her from the airport to the Pretoria offices, during which time she regaled me with her views on whites, men, women, children, the spiralling AIDS debate, the media and every other touchstone test of tolerance, be it of one’s former (and current) oppressors or indeed of those oppressed by virtue of their age or gender or … whatever. It was like hitchhiking a ride with the proverbial Texan truck driver who had not a good word to say about anyone except, possibly, other Texan truck drivers. Christine later achieved considerable notoriety as a columnist on South Africa’s Business Day newspaper – she is a very accomplished writer and polemicist, assisted by a deeply-felt belief in the most outlandishly provocative viewpoints – and then, particularly, as the high-profile deputy chair of the board of the South African Broadcasting Corporation. As the SABC descended into ever greater chaos on her watch, I couldn’t help recalling how often she had told me, and anyone else in earshot, of her contempt for the media, a contempt expressed in her loudly and proudly proclaimed refusal to have a television set in her house!

So this was the atmosphere and character of the place – grimy, bureaucratic in a rather inefficient sort of way, people working alone in their offices, singularly lacking in urgency and, then, thanks largely to the newer members of the Board, these peculiarly eccentric board meetings and a degree of random, untargeted energy. It was of course an institution with a guillotine-blade suspended over its head. The staff members were concerned about the transition, since the leadership was on the way out: my recollection is that Pierre Brooks had already made it clear that he had no intention of serving in the new competition regime. I was aware of other parts of that same awful building in which new industrial and trade policies were being debated and which resonated with new ideas. In its earlier years, the Competition Board may have been a more dynamic institution. Certainly, the infinitely more aggressive, ambitious personality of Brooks’s predecessor, Stef Naudé, who moved from chairing the Competition Board to heading the Department of Trade and Industry, would have imparted more bustle and self-importance to the institution. However, I encountered an untransformed corner of a government bent on transformation and that was how it felt – a mild and ineffectual beast in its death throes.

But more than this, those of us who were imagining the new institution recognised that there were structures and practices built into the very institutional fabric of the Competition Board that accounted for its weaknesses: the centrality of the chair within the institution, coupled with his powerlessness relative to the minister; the orphan-like status of a body that, despite the centrality of the minister in the decision-making process, had been passed on from one minister to the next; the emphasis on secrecy and the role of confidential ex parte negotiation in the decision-making process; the lack of engagement with the public. These were the factors that really condemned the Competition Board to its dull and toothless existence. It is no coincidence that institutional structure and, particularly, the manner of conducting the business of investigation and decision-making were what ultimately came to represent the most decisive rupture between the old and new regimes.

I say this because a brief look at the history of competition law before democracy came to the country may well surprise those who imagine that the apartheid regime had little interest in competition matters. On the contrary, it had sufficient interest to pass two statutes and numerous amendments, to commission a major inquiry into the state of competition and to set up, through the second of these statutes, an institution dedicated to the enforcement of competition law and the promotion of competition policy.

Leaving aside the institutional arrangements that governed both of these apartheid-era statutes – notably the dominance of government ministers at every stage of the decision-making process – there are two features worthy of comment. The first competition statute – the Regulation of Monopolistic Conditions Act of 1955 – did not confer any merger jurisdiction on the Board of Trade, the agency responsible for the administration of the Act. Nor did the Act provide for any prohibitions of anticompetitive conduct. However, it did provide, in this era of executive domination, for the minister to proclaim, after informing Parliament, that specified conduct was prohibited on pain of criminal prosecution. Ultimately this power was only invoked in respect of the practice of resale price maintenance, the practice whereby producers or wholesalers prescribed the price at which their products could be on-sold.

Given the current controversy surrounding criminalisation of hardcore cartel conduct – of which much more anon – it’s interesting, and more than a little surprising, to note that following the prohibition of resale price maintenance, four companies were fined in 1975 and 1976 for persisting with this practice in the markets for photographic goods, commercial paper, denim jeans and television sets. In respect of one resale price maintenance transgression involving Philips South Africa (the local subsidiary of the Dutch multinational) ‘and others’, the maximum fine of R20 000 was imposed on the company, the managing director was fined R7000, six other employees were fined a total of R8500, and prison sentences suspended conditionally for 3 years were imposed on the managing director and two other employees.

In 1976 the Mouton Commission was charged with reviewing competition law. Its recommendations led to the repeal of the 1955 Act, which was replaced by a new statute, the Maintenance and Promotion of Competition Act of 1979.

Although the new Act did establish a body dedicated to its enforcement, namely the Competition Board, the Board’s independence was as severely compromised as that of its predecessor, with the minister retaining authority over all substantive decisions and actions of the Board. The Board thus remained a division of a government department, with advisory powers.

Nor did the new Act provide for any prohibitions of specified anticompetitive conduct. And so, while the prohibition (effectively, the criminalisation) of resale price maintenance in terms of the 1955 Act was carried over with certain amendments into the 1979 regime, no further prohibitions were specified until 1986, when hardcore cartel conduct was declared unlawful by government notice. Once an investigation conducted by the Board concluded that the prohibition of cartel conduct (or resale price maintenance) had been violated, it was obliged to refer the alleged contravention to the criminal justice authorities. Suffice it to say that this led to a single criminal conviction in which the members of a furniture removal cartel were fined the princely sum of R100! In its twilight years, the Board also referred Vodacom and MTN, the members of the cellphone duopoly, to the criminal authorities on the basis of the ‘London Agreement’, a price-fixing agreement between the two firms. As a portent of things to come, the two firms concocted legal stratagems designed to keep the issue out of court, where it remained until the 1979 legislation had run its course.

The 1979 Act did, however, confer merger review powers on the Board, although it did not introduce a mandatory pre-notification system of merger review. This, combined with its self-imposed requirement to conduct its merger review with ‘the utmost secrecy’ and to favour ex parte discussions with interested parties and negotiated outcomes as the principal mechanisms of merger review rather than formal investigations, makes it difficult to assess the actual impact on merger decisions.

This system has permitted some officials of the time to defend the apparently low level of activity of the Board in the area of mergers by pointing out that, while permission granted by the Board to merge was generally known, a decision to turn down a merger was not publicised unless it was the subject of a formal investigation. It was suggested to me that the privately expressed opposition by the Board to a proposed merger would effectively put an end to the transaction – that is, the merging parties would not opt for a formal investigation and then inevitably have to attempt to persuade the minister to overturn the Board, despite the latter’s decision-making power. It’s difficult to know how much weight to give to this assertion because there are examples – of which the liquor industry ‘restructuring’ is the best-known – of powerful private interests riding roughshod in the most public and humiliating manner over the Board and its attempts to support the most basic competition principles.

I soon came to appreciate that this secretive and informal mode of decision-making was ineluctably skewed in favour of those seeking a favourable merger decision from the Board. I learnt, at first hand, the asymmetry in power and information in favour of those approaching the Board in ‘secret’. This does not merely refer to the extraordinary persuasive power of a highly experienced deal-maker – for anyone working in competition law in the early 1990s, the celebrated corporate lawyer Michael Katz immediately comes into sharp focus – but it also derives from the sudden, sporadic ascription of insider status to those whose structural location was very much on the margins of corporate and public power. There is something indefinably seductive about one’s sudden inclusion at a very early morning private breakfast and being asked by powerful CEOs and their legendary lawyers to vet major corporate decisions.

I experienced this modus operandi in one of the most difficult decisions that the Competition Board had to make in its twilight months, namely the acquisition – which the Board approved – of SunAir by South African Airways. The Board – and I in particular – was eviscerated in the media for approving the transaction, which was immediately followed by the closure of SunAir. Ironically – or possibly predictably – the media charge was led by a talented Business Day journalist, Robyn Chalmers, who soon afterwards went on to become SAA’s aggressive spokesperson during even more controversial times for the airline. The merger was presented to the Board – and, given the workings of the Board, it was basically represented by me – in precisely the secretive, ex parte manner described. The fate of ticket-holding would-be SunAir passengers and airline employees resided in my hands, or so I was told by the persuasive lawyer and the hard-driving CEO at the very early-morning breakfast, and the clock was ticking. In these circumstances, on every possible measure, despite the formal locus of decision-making, power resides with those demanding an affirmative decision from the Board.

But the social relations aside, the greatest asymmetry is in the relative knowledge possessed by the parties across the breakfast table. It is not merely that the petitioning party possesses the superior technical knowledge and experience, or even that, despite the apparent speed with which a decision is required to be made, a phalanx of technical, financial and legal advisers has been preparing for days for precisely this meeting. It’s simply that in an ex parte discussion without the benefit of discovery, cross-examination or the mere presence of opponents of the desired decision, there is no way of knowing whether or not a large bluff is being perpetrated.

I’ve often thought back on that SunAir transaction and I have little doubt that more time would have been available for consultation had I insisted upon it. But, equally, I don’t believe that this would have changed the decision in any way. Indeed I have little doubt that, had that decision been taken by what would later have been a properly constituted Competition Tribunal acting as expeditiously as circumstances would have permitted, we would have arrived at the same conclusion. SunAir was a badly run, failing firm. Competition did not suffer as a consequence of its demise, and despite the sentimental attachment to small firms and the understandable aversion to big bullies, even the public interest considerations would have bolstered the decision sought by SAA. But I wouldn’t have said – as I did on a number of occasions – that I thought that the air travel market could sustain only a limited number of players. Not only has history proved me wrong, but I now know that those decisions and outcomes are best left to the market.

And so notwithstanding the insistence by some that the Board actively reviewed mergers, and notwithstanding the fact that the current system requires the authorities to undertake formal investigations of a great many mergers that clearly raise neither competition nor public interest (as presently defined) issues, I remain convinced that the pre-1999 merger review system was predisposed to accord merging parties a significant advantage in the merger review process. ‘Utmost secrecy’ should not be tolerated in the conduct of public life, except possibly in matters of national security, and then only with the most elaborate checks and balances in place.

But, this having been said, there were undoubtedly important merger decisions taken in the lifetime of the Board, and not all of them went the way of the merging parties. And so there is little doubt that the extension by the 1979 Act of merger review powers was an important moment in the history of competition enforcement in South Africa, although severely attenuated by the provisions of the Act and the practice of the authorities.

I can certainly recall instances of important merger decisions in the brief period that I was a member, and later chair, of the Board. Naturally, the modus operandi described above broadly prevailed, but because it was a period of transition it may not have been a truly representative period. And so, in that time, Pierre Brooks and his investigators would conduct discussions with prospective merging parties and they would place their recommendation before the Board. Invariably, the Board would be asked to approve the transaction and, as invariably, the Board, with the regular exception of Robin McGregor, would endorse the decision proposed. Occasionally, the Board would ask to hear representations from the parties.

This was the case in the proposed merger between South Africa’s two leading chemical companies, Sasol and AECI, and then, a short while later, in the attempt by a large South African pharmaceutical company, Adcock Ingram, to purchase Pharmacare, a significant wholesaler of pharmaceutical products. Both these mergers were turned down.

So, in summary, it appeared that the Board’s role in the merger review process was recognised by the business community; it became, in other words, part of the business environment. However, the absence of a pre-merger notification requirement and the emphasis on secrecy not only made it difficult to judge the quality of merger decisions, it effectively meant that the merger review process performed no part of the Board’s communication with the public, and it performed little function in inculcating a culture of respect for, and understanding of, competition law. The minister retained final decision-making powers in all aspects of merger review. Although we are told that the Board nevertheless managed to dominate the decision-making process in respect of mergers, we know of several important instances in which the minister exercised his powers in order to override the Board’s decisions.

I have already described the desultory environment encountered by those deployed to work with the Competition Board in its later years. As already acknowledged, I cannot vouch that this rather tired, grimy atmosphere always characterised the institution. What is reasonably clear, though, is that it perceived itself as a division of a government department, then the Department of Trade and Industry, although it had been relocated on several occasions in its not very long lifespan. And a very peripheral division at that. A possible interpretation of the marginalisation of the Competition Board, even in the early years of a government that was clearly prepared to pursue a significantly more robust competition law than its predecessor, was that it kept its distance from the Department of Trade and Industry precisely in order to carve out a degree of independence for itself. But this would be an unduly generous explanation. The fact is that its important decisions were subject to ministerial approval, and under those circumstances maintaining distance merely served to reduce the standing and influence of the Board vis-à-vis the minister, who occupied the decision-making function within the institution.

The more likely truth is that the new leadership of the Department of Trade and Industry was intensely engaged in attempting to develop and implement an industrial and trade policy in which competition law and policy played little, if any, role. In fact, if competition considerations played any role in the thinking of the department, then it was evident in some of the reasoning underlying the limited trade liberalisation reflected in South Africa’s offer in the Uruguay Round. Nor is this a feature introduced by the new government. Certainly, from 1990, when the unions and the ANC began to engage with the policy-makers of the apartheid regime, industrial policy and trade policy were the name of the game, with hardly any reference to competition law and policy.

Although the interface between the department and the Board (or between competition law and industrial policy) did not change during the interregnum period – that is, the period when the leadership and membership of the Board changed and when the new Competition Act was being piloted through the Nedlac and parliamentary processes – certainly the profile and proceedings of the Board changed considerably.

While I may be expected to attribute this to energetic new brooms – of which I was one – I don’t think that this was the most important factor by any means. The critical factors were that the legislation was changing and the direction of change, one that favoured the more robust application of competition rules through an independent, decision-making authority, was already clearly mapped out. This gave the Board the de facto authority to conduct itself as though the new Act, or at least the ethos of the new Act, was already in place. And the Minister of Trade and Industry, Alec Erwin, provided the necessary space for the Competition Board to conduct itself as though it were an independent body.

There were, in the final years of the Competition Board, a number of matters that presaged the significantly different era of competition law enforcement that was in the offing. A colourful precursor to the inclusion in the objectives of the current Act of the promotion of black economic empowerment and small and medium enterprises was the complaint submitted by Moosa Moosa, the owner of Avalon Cinemas, a small chain of movie theatres predominantly located in the Indian residential and business areas of racially segregated South Africa. Moosa, whose personality exuded showbiz, complained that his cinema chain was being denied access to the major shopping malls and that the cinema chain duopoly of Ster Kinekor and Nu Metro was being systematically preferred by the large financial institutions that owned the malls in question. In the first instance of the application of post-apartheid ‘public interest’, Pierre Brooks, who was uncompromisingly orthodox in his approach to competition law and decidedly not a showbiz type of personality, nevertheless negotiated an outcome favourable to Avalon Cinemas.

But complaints which from a competition perspective were significantly more far-reaching than that of Avalon Cinemas continued to receive short shrift. The clearest example of this was the alleged collusion between Vodacom and MTN, the owners and operators of South Africa’s only two mobile telephone networks. There were a number of other complaint proceedings initiated under what would effectively become the abuse of dominance proceedings of the current Competition Act. Most notable of these was a case involving pharmaceutical distribution, which I will deal with in the following chapter.

One of the most interesting restrictive practice cases that I recall from these pre-1999 days concerned the state-owned South African Broadcasting Corporation (SABC). The corporation, which enjoyed a huge statutory advantage over its competitors and hence a dominant position in television viewership and in the commissioning of filmed material for screening on television, introduced a requirement that any film commissioned by it had to be made using the studio and other facilities of the SABC. This naturally threatened the very existence of independent film-making facilities. After a lengthy wrangle with an arrogant and ignorant SABC – there are things that never change – we persuaded them of the error of their ways. But in the process they provided a foretaste of the difficulties the new competition authorities would experience in their dealings with state-owned enterprises, which saw dominance of their own and any other ancillary markets as a God-given right.

Merger regulation in these twilight years of the Board’s existence also presaged the more robust era on the horizon. I have already referred to the proposed merger between the chemical companies Sasol and AECI. Here was a case of two blue-chip firms with political connections, which in previous times would have ridden roughshod over the competition authorities, wanting to merge their considerable interests in the chemical sector. The merger was formally supported by the Department of Trade and Industry but was turned down by the Board.

I never really knew whether or not Alec Erwin or the leadership of his department ultimately agreed with the decision to prohibit the Sasol–AECI merger. I suspected not and I have little doubt that the companies sought to influence him, though again this was never communicated to me. But formally the Board’s decision was unhesitatingly and unquestioningly endorsed by Erwin. It was a decision of the Board and, in accordance with the new era of independence that was being ushered in by the pending new legislation, the minister made no attempt to intervene in the decision-making process, thus presaging the new regime.

Of course, there was a definite limit to the Board’s ability to transform itself. Aside from the obvious requirement to abide by the legislation in place, there were well-established practices that were difficult to summarily jettison in the absence of a clearly conceived, much less codified, alternative. Foremost among these were the ex parte communications – ‘negotiations’ as they were referred to – between the chair of the Board and the officials, on the one hand, and parties seeking some decision or another from the Board, on the other. Changing this would not only require a cultural shift but it would also require a separation of the functions of investigation and adjudication that was ushered in by the new legislation.

And so the Competition Board faded away as the legislation governing it was repealed and replaced by a wholly new statute, which established a trio of new institutions that would henceforth be responsible for the administration of competition law. The end was marked by a flurry of activity but no ceremony. All of those who may have – and possibly should have – paid their respects at its demise were too engaged with the complicated business of establishing the new competition order. And it is this to which I now turn.

A brief postscript. The new bodies, in particular the Competition Commission and the Competition Tribunal, immediately established a high profile and attracted intense media interest. However, for years after the establishment of the new institutions, indeed to this very day, they continue to be referred to collectively as ‘the Competition Board’, particularly in the business community. The brand was clearly well established, although I never really understood why.

NOTES

1. ‘Competition law’ and ‘antitrust law’ are used interchangeably throughout this book.

2. A ‘spaza shop’ is a small, usually informal, convenience store in the previously blacks-only townships of apartheid South Africa.

3. The JD Group and Ellerines are two large retail furniture chains that feature later in this story.

4. Known by its acronym Nedlac, this body is discussed in the following chapter.

2. The new competition regime

The process of drafting the Competition Bill, negotiating it in the National Economic Development and Labour Council (Nedlac), piloting it through the parliamentary process, and reaching its final outcome, represents one of the success stories of the early years of South Africa’s democracy. I make this bold claim because, despite the complex and controversial nature of competition law, a statute emerged relatively rapidly and relatively amicably through the law-making process, a process in which everybody interested in the outcome was heard. The broad acceptance achieved through this process and extensive media coverage meant that by the time the Competition Act was placed on the statute book, something of a competition culture was already implanted in South Africa, despite the aridity of the soil that history had bequeathed.

Stripped to its bare essentials, the process involved the preparation of a policy document that was approved by Cabinet and then released for public consumption and comment. The policy document was then tabled at Nedlac – the workings of which are described below – forming government’s input into the process of consultation and negotiation with organised labour, organised business and the ‘community’ over competition policy. Once the process of the Nedlac discussions had been concluded – effectively a negotiated outcome – a bill based largely on the principles agreed there was sent to Cabinet, which considered it and then sent it through the parliamentary process. This entailed submitting the bill to the trade and industry parliamentary portfolio committees,1 together with the Nedlac report that specifically indicated those elements of the bill on which the social partners could not agree. The bill that emerged out of the committees’ hearing process was submitted to a parliamentary vote and finally was signed into law and promulgated by the President.

The process was managed by the Department of Trade and Industry, the department responsible for the policy framework informing the bill, and specifically by Alistair Ruiters, then chief director of the section of the Department of Trade and Industry dealing with small business and cooperatives. He was smart, aggressive, unpretentious and, as many would later discover, ruthless. He was also looking for a route out of the Department of Trade and Industry or, at least, out of the rather marginalised position he held, and so had grabbed the opportunity to establish a new agency, which was to become the Competition Commission, and of which he would be the first commissioner. This lasted for a few months, whereupon Alec Erwin, the Minister of Trade and Industry, yanked him back into the Department of Trade and Industry, where he took up the post of Director-General. Although opinion is divided on the Department of Trade and Industry’s performance on his 5-year watch, he possessed an undoubted ability to get what he wanted. Without ever really engaging with the substance of the competition law that was being drafted, he was an unusually effective manager of the government and legislative processes, with a consummate ability to steer or bludgeon his way through tangled bureaucracies and byzantine politics. Aided by consultants and an efficient staff, he physically set up the competition authorities – from having the offices built to buying the furniture, from employing staff to ensuring that grievance and disciplinary procedures and pension and medical aid funds were in place.

Let’s briefly examine each stage of the process out of which the Competition Act ultimately emerged.

THE POLICY DOCUMENT

The policy document was drafted by a small team comprising Frederick Fourie, Willem Pretorius and me. My recollection is that the drafting team was appointed by Trevor Manuel in his brief stint as Minister of Trade and Industry. But it was largely Alec Erwin, who succeeded Manuel as minister, to whom the drafters reported. Frederick Fourie was a University of the Orange Free State economist, one of the few who had written extensively on competition law as practised in the previous regime. He would ultimately serve a term on the Competition Tribunal. Willem Pretorius was one of the few practising South African advocates with experience in competition law. He had studied competition law in Europe and had been a member of the Competition Board for several years. Willem is aptly described as a ‘rough diamond’, a ‘very rough diamond’. He ultimately litigated frequently before the Competition Tribunal. And I laid some claim to knowledge of industrial economics through my research interest in competition issues. I had also, since 1994, served in the public sector in one policy advisory capacity or another.

While the policy document suggests that it was the product of a number of inputs, including from the Competition Board and the Department of Trade and Industry, this is something of an exaggeration. Although I recall the drafting team engaging periodically with the then Director-General of the Department of Trade and Industry, the always thoughtful Zav Rustomjee, there was no systematic interaction with the department. While we revelled in and benefited from our independence from the department, our distance from what became an increasingly dysfunctional and leaderless institution ultimately came at some cost. However, the minister engaged regularly with the drafters. The drafting team was keenly cognisant of those aspects of the policy document that were particularly sensitive and that required a ministerial decision. The document was made available for public scrutiny and comment and was subject to a timeline that culminated in the passing of an Act of Parliament in October 1999. Remarkably, the time frames were largely adhered to.

It’s difficult to summarise the 37-page policy document. It manifests all of the verbose, ‘rhubarb, rhubarb’-type qualities of many of these kinds of documents. It is full of contradictory statements and characteristically extravagant undertakings – in this instance usually relating to parallel processes that were to be undertaken in relation to other laws and policy fields, important elements of which are yet to be completed. Both these features are the predictable products of an important policy document straining to be acceptable to as many divergent interest groups as possible in a still deeply divided society. But let me attempt to extract the kernel of the policy enunciated.

Above all, the document attempts to reconcile the application of competition policy with the national policy objectives of ‘competitiveness’ and ‘development’. The title of the document – Proposed Guidelines for Competition Policy: A Framework for Competition, Competitiveness and Development – says it all. And, partly deriving from this, it attempts to align competition law with trade and industrial policy and with the treatment of public enterprises. Apart from the intrinsic importance of each of these issues, the specific effort to reconcile them speaks to the complexity of addressing a range of constituencies, precisely those who brought apartheid to its knees and voted the government into power, whose most articulate representatives, to wit the unions, were intensely sceptical of government policies that foregrounded objectives termed ‘efficiency’, ‘productivity enhancement’ or, the term guaranteed to attract most suspicion from the unions, ‘international competitiveness’.

For the most part, the unions viewed these objectives as synonymous with labour repression. Certainly, they viewed the quest for ‘competitiveness’ as a euphemism for a business-friendly agenda. And they were partly justified. Businesses compete, not nations, and an agenda to enhance competitiveness is one that should seek to create an environment – critically including the market rules – and provide the public inputs necessary to generate and support competitive businesses. During the course of the Industrial Strategy Project – a large union-initiated industrial policy project that I co-directed – I recall trying to persuade the unions that the principal and direct objective of an industrial strategy was to raise productivity through industrial upgrading from which jobs, ‘decent’ jobs in current union and government parlance, would ultimately flow. However, they were intensely sceptical of the claim that well-performing companies would translate into more and better jobs – it was viewed as a variant of the dreaded ‘trickle-down’ economic policy.

In truth, then, government’s principal and direct difficulty lay not so much in selling the concept of competition, but with the much deeper problem of persuading key constituencies that economic development required a competitive private sector and that this should constitute a key objective of industrial policy. Because the gains from competition law and policy were even more closely identified with ‘efficiency’ or ‘competitiveness’ than was industrial policy, government was particularly hard put to demonstrate that its competition policy and the statute it was proposing contributed to ‘development’ – that, in other words, its gains were indeed broadly distributed beyond the private sector.2 The issue is highlighted in the second paragraph of the executive summary: ‘The overriding goal [of competition policy] is to achieve a more effective economy, which in turn requires us to better define what is meant by the “public interest” with respect to South Africa’s corporate structure and firm behaviour. The department believes that competitiveness and development are mutually supporting rather than contradictory objectives, if policies are properly aligned’.3

In essence, then, the policy document carried over the centrality of ‘public interest’ from the 1979 Act, but it was clearly concerned to imbue it with more precise content that combines competitiveness and development. This concern was ultimately reflected in the statute that emerged from this process.4

The summary goes on to say that this task of aligning competitiveness and development demands a ‘uniquely South African approach to competition policy’. It continues: ‘This set of policy guidelines fuses these different mandates, by assuring the public that on the one hand competitiveness and efficiency are pursued, and on the other that this process will ensure access to many more people previously denied an equal opportunity to participate in the economy’.5

Over time, as we’ll see when we discuss the Nationwide Poles case, the Competition Tribunal effectively reconciled these objectives by giving the objective of ensuring ‘an equal opportunity to participate in the economy’ both an orthodox competition law interpretation (lowering entry barriers and defending the competitive process) and, when appropriate, a public interest dimension (in the Nationwide Poles case, this being the promotion of small and medium-sized enterprises).

Successive paragraphs of the policy guidelines effectively identify – in suitably hedged and qualified form – the four concerns that the subsequent competition statute and complementary policies were required to address. Paragraph 2.4.11 foregrounds the core competition objectives: ‘The competition policy proposed here accepts the logic of free and active competition in markets, the importance of property rights, the need for greater economic efficiency, the objective of ensuring optimal allocation of resources, the principle of transparency, the need for greater international competitiveness, and the facilitation of entry into markets – all within a developmental context that consciously attempts to correct structural imbalance and past economic injustices’.

Paragraph 2.4.12 is focused on the broader social and industrial policy objectives, those objectives that were ultimately to be identified as the public interest content of the Act. ‘Competition policy seeks to incorporate the interests of consumers, workers, emerging entrepreneurs, and other corporate competitors, and to protect the ability of our large corporations to penetrate international markets, just as we allow foreign investors to do business in South Africa in the interests of enhancing overall efficiency and growth’.

Paragraph 2.4.13 addresses the procedural aspects: ‘Competition policy has to assume that the resolution of competition law cases be conducted in a procedurally fair, coherent, expeditious and decisive manner, and that new institutional arrangements for pursuing the policy will entail an appropriate division of labour within the relevant agency and independence’.

Finally, paragraph 2.4.14 addresses the requirement for compatibility between competition policy and other policy fields: ‘Finally competition policy seeks to be sufficiently flexible to incorporate existing policies and future modes of market regulation that extend in a coherent manner across the full spectrum of industrial policy, foreign exchange policy, the attraction of foreign direct investment, the restructuring of state assets, tax reform, labour market policy, financial market regulation, consumer protection, research and development incentives, small business and affirmative action programmes, corporate governance instruments and revised company law. Some of these – international trade policy, industrial strategy and public corporations – are considered in subsequent chapters’.

If these resemble the proverbial camel – a horse designed by a committee – that’s because they too were drafted by a committee of individuals from divergent backgrounds, not to mention writing styles, with several others, including the minister, injecting bits and pieces into the final document. And if every paragraph sounds as though it is attempting to be all things to all people – it even promises that competition policy will contribute to the transformation of gender relations!6 – then that too is because this is precisely what it attempts to do. However, when the inevitable knots are combed out of these four paragraphs, a reasonable summary is that competition policy would pursue orthodox efficiency objectives, while taking account of broader social and industrial policy concerns as well as the constitutional requirements for fairness, and would seek to align itself with trade and industrial policy as well as policy towards regulated sectors and sectors dominated by state-owned enterprises.

For the team assigned responsibility for piloting the document and then the bill through Nedlac and Parliament, the magic words are in paragraph 2.4.13: ‘that new institutional arrangements for pursuing the policy will entail an appropriate division of labour within the relevant agency and independence’. The ‘appropriate division of labour’ meant the separation of investigation from adjudication, and ‘independence’ meant freedom from political override. This is confirmed in a later passage of unusual clarity:

The government’s view is that monopolies law should be effected by a competent, professional agency with powers to investigate and to respond rapidly and robustly to anticompetitive conduct. The decisions of the tribunal envisaged will be subject to judicial review, but it is government’s intention to take enforcement of competition law out of the hands of the criminal courts and to avoid the prospect of lengthy, complex and costly litigation. The possibility of politically inspired intervention will also be removed by eliminating the exercise of ministerial discretion in the enforcement of competition law and by a more precise definition of both the mandate of the policy structure and its relationship to the minister and government policy. As already elaborated, our political choices will be exercised in the mandate extended to our industrial strategists and, from there, to our competition authorities.7

Were this policy position to be sustained through the law-making process, then the central weaknesses of the previous competition regimes would be eliminated: the competition authorities would not have to rely on the police and criminal prosecutors to go after competition law transgressors and political decisions would be removed from the equation. While there was yet to be much toing and froing over the extent of the independence implied by these passages in the policy document, there was never any question of the minister assuming any decision-making power regarding the core competition decisions. Indeed, Alec Erwin never hesitated in relinquishing the considerable powers that his predecessors retained over competition decisions. However, once a public interest test was inserted into the Act, the question of who would take those decisions became contested terrain. As we shall see, here too Erwin was ultimately satisfied to entrust even the public interest decision to the competition authorities. There are few ministers – then or now – who voluntarily relinquish any of their powers, few politicians who forswear the politicisation of decision-making. Indeed, the oft-made claim is that it is precisely the reluctance to empower the sectoral regulators by granting them independence that has so weakened, indeed corrupted, the overall regulatory process. Erwin’s willingness to give up his powers is a truly stand-out example of a consummate, professional politician who, in marked contrast to the vast majority of his colleagues, nevertheless understood the dangers, in particular circumstances, of ‘politicisation’.8

For the rest, the document is taken up with elaborating the requirement for policy compatibility. The most important areas of interface with competition policy that are identified, and hence, with which compatibility is required, are international trade policy and industrial strategy and ‘public corporations’. I’ll return to this critical issue of policy alignment in later chapters.

THE NEDLAC PROCESS

Nedlac distinguishes the South African law-making process from any other that I know and so would not be easily understood by non-South African readers. Regrettably, this may also by now be true of many South African readers.

Nedlac was established by one of the first major pieces of socioeconomic legislation to be enacted by the new government. It is a quadripartite body on which government, organised labour, organised business and ‘the community’ are represented. It emerged as a key by-product of the success of the negotiated political settlement but in fact has a substantially longer lineage in the role played by negotiation between labour and business. This frequently involved organised labour using business as a buffer and interlocutor between itself and the apartheid government. If the background to the ‘miraculous’ transition from apartheid to democracy is replete with personal heroism and spilt blood, then its ultimate form – negotiation between the parties to the political conflict that defined South Africa – was established by the unions’ keen sense of the essential interplay between, on the one hand, demonstrations of organised power and, on the other, a willingness to negotiate – albeit at a respectable distance – with those who were the object of the exercise in power. This characterised the unions’ approach even during those darkest days of the 1970s and 1980s, when it appeared that negotiation could not possibly bear fruit. In a strong sense, Nedlac represented the formalisation of a long-standing labour strategy, indeed vindicated the success of its strategy by giving the two old adversaries or, as they came to be termed, ‘social partners’ – organised labour and organised business – a formal role alongside government in the policy-making process.

Nedlac is structured into several chambers defined by the respective policy fields in which they operate – a labour market chamber, a public finance chamber, a trade and industry chamber, and so on. Competition policy was assigned to the trade and industry chamber. The process immediately struck gold in the form of the Nedlac official responsible for the coordination of that chamber. This was Shan Ramburuth, a deeply engaged social activist with a keen sense of strategy (and, equally important, as keen a sense of humour), who managed to keep a difficult process on track in the most unobtrusive, light-handed manner. When I knew that I was going to be appointed to chair the Tribunal, I immediately persuaded Shan to take up the position of CEO of the Tribunal, where he would effectively be responsible for establishing the institution and putting in place its management systems and structures. After several years in the Tribunal, he was appointed deputy commissioner of the Competition Commission and, soon thereafter, commissioner, a position that he still occupies today.

From the outset, the government delegation made an important decision with respect to the conduct of the Nedlac negotiations. Government was growing increasingly impatient with the constraints that it perceived the Nedlac processes imposed upon its requirement or, in the view of many ministers, its ‘prerogative’ to govern. This would culminate in the stubborn and short-sighted refusal by the National Treasury to debate the content of the controversial macroeconomic policy framework known as Gear.9 This is a decision that labour has neither forgotten nor forgiven and that immensely reduced the credibility, certainly in the ranks of organised labour, of the policy framework, the Treasury and the otherwise highly regarded former finance minister, Trevor Manuel.

The Department of Trade and Industry made no such attempt to sideline Nedlac in the process of developing the new Competition Act. However, the approach of the government delegation was to negotiate the principles of the policy document rather than the bill, the drafting of which would remain a government responsibility. Our drafting would be framed by the principles agreed at Nedlac. Several of the people involved in the competition policy negotiations – notably Jayendra Naidoo, the head of Nedlac and a deal-maker and negotiator extraordinaire, and I, in a previous capacity as an adviser to Tito Mboweni, then Minister of Labour – had first-hand experience of the horror of drafting legislation through the mechanism of Nedlac negotiations. These were the negotiations that ultimately resulted in the Labour Relations Act – admittedly a singularly more important piece of legislation from labour’s perspective than the Competition Act – where agreement was achieved through a process of trench warfare as labour, business and government argued through many long nights over every word of the Act.

Government was intent on avoiding a repeat of the Labour Relations Act fiasco, hence our insistence that the policy document, rather than the bill itself, form the basis of the negotiations, and that these discussions be kept separate from the drafting process. This enabled government to control the drafting process. I don’t think that this advantaged government in the negotiation process, but it certainly expedited the negotiations and alleviated a great deal of tension and irritability, not to mention sleep deprivation.

The government delegation comprised Alistair Ruiters, Alan Hirsch (another official of the Department of Trade and Industry), Menzi Simelane (a young junior advocate from the Johannesburg bar, who would ultimately become the Competition Commissioner and then, amidst no little controversy, rise to run first the Department of Justice and then head the National Prosecuting Authority), and me. In the background, with responsibility for drafting the bill, were Phil Knight and Norman Manoim. Phil, a Canadian specialist in ‘plain language’ legal drafting, had played an important role in the drafting of the Constitution and the Labour Relations Act. Norman was a human rights lawyer. He was then in private practice but would become one of the two full-time members of the first Tribunal – I was the other – and is now the chair of the Tribunal. Norman would come to have an incalculable influence over the development of competition law in South Africa.

Organised business was mainly represented by Michael Spicer, who was introduced in the opening paragraphs of this book. He was, as previously explained, the Anglo American Corporation’s public affairs executive and a formidable negotiator. Much of business’s technical load was carried by Stephan Malherbe, who would ultimately launch a successful and extremely competent private economic consultancy service that frequently provided expert testimony, usually on behalf of the private sector, to the Tribunal. Stef Naudé, a former chair of the Competition Board and former Director-General of the Department of Trade and Industry, was also a regular member of the business delegation. Business was challenged to reflect credibly the views of small business, which the guidelines identified as major beneficiaries of the new competition policy and law, but which participated only sporadically in the process. However, it was big business which, with clear justification, saw itself as the likely target of the pending new competition law and which was thus prepared to commit its resources to defending what it viewed as its essential interests.

Organised labour’s principal objective was to ensure that it was not short-changed relative to business on any stakeholder rights conferred by the new statute. Labour, which viewed the pending legislation as a mechanism for disciplining business, was intent on ensuring a statute whose tone and content were sufficiently robust, sufficiently prescriptive and ‘regulatory’ in its approach. This generally translated into a demand for extensive codification of prohibited practices. But above all, organised labour was anxious to ensure that the job security of its members was not threatened by elements of the business conduct subject to the new legislation that was being contemplated, in particular the job losses that frequently accompanied mergers.

Organised labour’s views are clearly outlined in a speech made during the course of the negotiations by its senior representative in the Nedlac process, Kenneth Creamer. In this speech Kenneth argued for ‘competition legislation most appropriate for the meeting of the economic imperatives of sustainable industrial expansion, including widened ownership patterns, as well as for promoting employment opportunities and avoiding job loss’.10 It’s noteworthy, though – particularly in relation to the position subsequently adopted by the Tribunal to public interest issues – that Kenneth recognised that while competition legislation and enforcement should be ‘employment sensitive’, it should play ‘a supportive role – not a primary role – in the programme for job retention and job creation’.

Labour’s demand for the inclusion of ‘public interest’ criteria did not end with the direct employment consequences of mergers. In the speech already cited, Creamer proposed that the public interest criteria should include ‘the likely impact of a merger on the objective of widening ownership patterns in the economy and on the objective of protecting key national or regional industry from the potential predatory behaviour of globalised conglomerates’.

As far as dominant firms were concerned, Creamer argued that ‘competition policy should be informed by the need to break up apartheid era monopolies and concentrations built up during a time of systematic exclusion and racial privilege’. To this end, he added that labour supported ‘the forced unbundling of inherited concentrations when this would be in the public interest and would advance the overall objectives of industrial and development policy’ and ‘for the enforced breaking up of conglomerates that abuse their dominant positions, for example through charging excessive prices or through profiteering by limiting output levels’.

The mechanism supported by labour for undertaking the public interest review was one in which ‘the Minister of Trade and Industry should be given the discretion to review merger decisions taken by the Competition Tribunal. On the basis of specified “public interest” criteria, including the impact on employment levels, the minister will be empowered to confirm, amend or overturn a decision of the Competition Tribunal’. In addition, labour argued for the inclusion in the bill of mechanisms to inform labour when a public interest review might be initiated and how it should participate in such a review.

Much of this was, of course, anathema to business. In fact – and this was confirmed by Stephan Malherbe, one of business’s key representatives in the Nedlac process – business’s preferred position was that there should be no antitrust legislation at all. As ever, while business was constrained to view competition as a ‘good thing’, it was preferably confined to someone else’s backyard. Nor is this surprising. Business had hitherto been subject to extremely mild antitrust scrutiny. Moreover, some of South Africa’s most venerable corporate icons – notably De Beers, the globally dominant diamond producer and organiser of the international diamond cartel and, one might say, the veritable jewel in the crown of the Oppenheimer family’s vast business empire – had already been at the receiving end of the United States Department of Justice. Directors of De Beers were under indictment in the US and, at some cost to the fortunes of the company, were not permitted to conduct business in or, indeed for fear of arrest, enter the country. The company had attained something of an iconic status in antitrust lore. Nicholas Oppenheimer, scion of the great Randlord family, had famously dubbed US Department of Justice antitrust investigations as ‘equivalent to the Spanish Inquisition’!

However, business was resigned to the introduction of new competition legislation and, given that reality, what it sought was what Stephan termed ‘normal’ antitrust, a conception that precluded easy resort, if any at all, to the power to break up firms. It also, in business’s conception, precluded the inclusion of non-competition or public interest criteria, which it viewed as politically loaded and arbitrary – although labour was quick to point out that these were at that time features of many developed country competition systems – and it was particularly opposed to the introduction of a political, that is a ministerial, role in the decisionmaking process.

But business’s position notwithstanding, key elements of labour’s demands found echo in the bill and ultimately in the Act itself. In particular, public interest criteria – including employment and black economic empowerment – were incorporated in the merger review process. However, although the first draft of the bill gave the minister the right to make the public interest decision and thus potentially override a decision of the Competition Tribunal and Competition Appeal Court, this was ultimately changed and decision-making responsibility, even over the public interest criteria, was given to the competition authorities.

Neither I nor anyone else to whom I have spoken (including Norman with his elephantine memory) is able to recall why it was decided to shift the public interest decision from the minister to the competition authorities, a decision of considerable significance. It’s a decision that would certainly have required the minister’s agreement. It possibly followed business lobbying to which we were not privy. I do know that the government team sent a memo to the minister, indicating, probably incorrectly, that there might be constitutional problems involved in the exercise of a ministerial veto over a decision of a quasi-judicial tribunal or when taken by the Competition Appeal Court (because it is a division of the High Court).

My best guess is that the logic of the substance of competition law and the negotiating process probably asserted itself in favour of business’s position. By ‘the logic of the substance of competition law’, I mean that the character of big business’s participation relative to that of labour confirmed what, to any outside observer, would have been obvious. This was a bill that implicated business’s vital interests. On the other hand, labour’s interest in the process was secondary. Unlike labour relations legislation, its vital interests were not directly affected by the details of the Competition Act.

As the process unfolded, the essence of what it meant to have a statute promoting competition emerged more clearly. To those well versed in competition law and policy, it would come as no surprise to learn that the overriding objective of competition law is to defend the market, principally against those market participants who choose to achieve market power through anticompetitive mergers, or collusion, or, in the case of dominant firms, through conduct aimed at excluding their competitors from the market. This may well involve taking robust action against those who choose to advance their interest in this way, but such action is clearly aimed at defending market or competitive processes rather than specific interest groups.

The horns of the rather crude dilemma that this posed for both business and labour are fairly obvious. Business could not credibly resist the introduction of market-friendly legislation. Indeed, whether the leadership of the new government recognised it or not, competition legislation had become one of the markers designating full membership of the new liberal global economic order.

So here was South African business, in this instance (as in so many others) under the leadership of an Anglo American director, talking about the introduction of dreaded antitrust legislation right into the front parlour, not to mention the backyard. While it was difficult for business to deny the salience and legitimacy of what one US judge famously dubbed ‘the Magna Carta of the free enterprise system’,11 South African business was determined to constrain it, to limit its consequences, particularly in the face of a new government replete with decidedly anti-big-business elements and sentiments and a perception of antitrust principally as a mechanism for the state to use to discipline business. Never forget that antitrust law restrains, albeit in the name of promoting efficiently-functioning markets, the use of property rights; and South African business was not about to give up without a fight its longstanding ‘prerogative’ to use its property as it saw fit. And, possibly of even greater concern, business feared the distinct possibility that both the unions and the government would, contrary to the received orthodoxy in antitrust, use the legislation as a means of imposing additional socioeconomic obligations and considerations, essentially non-competition considerations, on business decisions.

But, equally, the logic of the negotiating process, powerfully reinforced by South Africa’s particular history and contemporary economic circumstances, imposed itself on the outcome. Given the character of South Africa’s new political regime and the presence of labour in the process of formulating the new law, there was little prospect of excluding critical factors such as employment and the racially-skewed ownership structures (and hence black economic empowerment) from the ambit of a piece of legislation as important as the Competition Act. And, if these did not fit neatly into an orthodox competition statute, then they would enter by way of the introduction of public interest criteria. Nor were offences like excessive pricing and divestiture remedies going to be precluded. And this is precisely what happened.

However, again by dint of the logic of the negotiating process, business’s voice had to be heard. And so it was. Stephan Malherbe described how the business representatives experienced the Nedlac process as surprisingly ‘pleasant’ and ‘non-political’. And so while public interest criteria were, contrary to business’s strongly preferred position, incorporated into the Act, they were depoliticised.

In summary, business was ideologically committed to competition law and the rules necessary to defend it and promote it. But its leading representatives had already been bruised by their experience of competition law enforcement; it did not trust the South African government to enforce it, and it feared that its social partners, to wit organised labour, would use the legislation as a means of furthering objectives that had little to do with antitrust. And, of course, sections of business had their own special pleadings, their own reasons for limiting the reach of antitrust. These were generally expressed in one or other variant of the familiar small market argument that sought to justify domestic market dominance or export cartels by reference to the requirement to attain the scale necessary to compete on international markets and on liberalised domestic markets.

In the parliamentary hearings, these general industrial policy arguments for special treatment under the antitrust law were particularised. It was here that we learnt of the special character claimed for everything from diamonds to real estate, from intellectual property to wattle.

Organised labour, on the other hand, cared little for the market and its defence. It viewed the market as an institution of labour oppression. But deprived, at the finishing line, of the weapon or even the credible threat of nationalisation, it relished the thought of legislation that would enable big businesses to be fined and even broken up. Hence, labour’s conception of the legislation was essentially regulatory – for example, in its strong support for the prohibition of excessive pricing – and punitive. And, pragmatic and opportunistic (in the most positive sense of the word) as ever, labour viewed the legislation as an opportunity for blunting some of the sharper edges of the labour market. In particular it was, as I have already noted, concerned to limit the employment impact of mergers. So while business represented the competitiveness elements of the public interest as defined in the guidelines, labour effectively became the representative of the social elements of the public interest, ultimately expressed as the promotion of employment, black economic empowerment, small business and an equitable regional spread of development.

From the perspective of the government negotiating team, the differences between labour and business and the imperative to compromise provided a happy outcome. We always wanted an Act that would depoliticise the decision-making process in competition matters, that would give independent decision-making power to the competition authorities. The guidelines were silent on the locus of decision-making as regards the public interest because, while they clearly identified the ‘public interest’ as one of the objectives of competition policy and legislation, they did not contemplate that the public interest issues would form an explicit part of the merger decision. However, after they were carved out and incorporated not merely into the overall objectives of the bill but also into the merger assessment criteria, our expectation was that the minister would assume decision-making power over the public interest considerations. This would have placed us in a position similar to that of many other competition authorities – including such highly regarded institutions as the German and UK authorities – where, in merger decisions, the competition authority has sole authority over competition matters and the minister retains an override of the competition decision on public interest grounds. However, as I’ll elaborate later, we were delighted to be given decision-making authority over both competition matters and public interest. Certainly, the balancing of these two divergent sets of criteria would prove challenging, but at least we were not confronted with the distinctly unpalatable prospect of a ministerial override.

This was not the end of the issue of independence or public interest. As we’ll see, the issue was raised in Cabinet and resolved in the most bizarre circumstances. Ultimately, the unusual manner in which public interest issues were dealt with in our Act came to be regarded as a hallmark of our competition regime and something of a talking point in international antitrust circles.

The point that I want to underline here is that the really important victories were achieving ministerial support and then Nedlac support for the final authority of the competition agencies over competition matters. Just as I believe that the subordinated decision-making role of the Competition Board was the most important institutional weakness of the previous regime and one of the hallmarks of authoritarian governance, so I am convinced that the independent decision-making power of the competition authorities is the most important feature of the new competition regime. Regrettably, as the sector regulators know only too well, regulatory independence is not always a hallmark of South Africa’s democratic order.

The drafting process also brought those of us who would end up leading the new authorities into contact with the very supportive and active community of antitrust enforcers and scholars. We visited several agencies in countries ranging from the Netherlands, which (like us) was just starting out, through to the US with their vast experience and resources, most notably their extraordinary human resource capacity (I still recall my awe at being told that the antitrust division of the US Department of Justice employed more than 100 microeconomists with doctorates). And so we began the process of joining a global network of antitrust authorities, a network that would, in the course of our first decade, flourish and expand, both in size and quality, and in which we would play an active role and from which we would derive huge advantage. As important as the institutional relations initiated in this period, our preparatory work also brought us into contact with a number of people – most notably at this stage, with Eleanor Fox – many of whom were legends in the field of antitrust, and who would loom very large in the lives of the South African competition authorities.

THE PARLIAMENTARY PROCESS

The passage through Parliament itself was, appropriately, the most public part of the process.

Not only did organised business and labour have unfinished business, aspects of the bill with which they did not agree and which they were accordingly entitled to refer to the parliamentary process, but those who felt that their particular concerns had not been adequately dealt with in the Nedlac process also used the parliamentary hearings to air their concerns and make their proposals. Public comment was extensive. The parliamentary hearings were lengthy and vibrant. Both gave rise to significant revisions of the bill.

Certain of the submissions made in the hearings portended problems that we would grapple with for many years: the South African Telecommunications Regulatory Authority – which would later merge with the broadcasting regulator to form the Independent Communications Authority of South Africa (Icasa) – strongly argued for sole jurisdiction over all matters, including competition matters, pertaining in any way to telecommunications; food retailer Pick n Pay feared that the Competition Act would threaten fundamental features of the franchising model, as did the South African Liquor Store Association, while simultaneously lauding the positive impact that it expected the Act to have on the prospects of small enterprise; and De Beers was exercised about the use of a reverse onus in certain key sections of the Act and – surprise, surprise – argued for the introduction of wider grounds for exemption.

On re-reading those of the submissions that I have been able to lay my hands on, I am struck by the number of important amendments that were made to the bill following the public comment and parliamentary hearing processes. Most striking is that the structure of the authorities, outlined below, was drastically changed. And consistent with many of the submissions made to the parliamentary committee, the minister’s decision-making role in respect of public interest issues arising in merger reviews was removed.

My recollection – and in this I am supported by colleagues with decidedly superior memories to my own – is that it was decided, with a deep intake of breath, that because these changes, with the singular exception of the question of the powers of the minister, did not involve a change in policy, there was no need to take the amendments back to Cabinet. In this spirit, Alec Erwin submitted the Competition Bill as amended to Cabinet (although his report to Cabinet only highlighted the amendment that effectively took his office out of the decision-making process), with the amendment that depoliticised the process and that gave the competition authorities their unusual degree of independence. This clearly represented a controversial policy change.

I distinctly recall our nervousness, our expectation that this would be challenged. And so it was. The late Kader Asmal – then the Minister of Water Affairs and Forestry whom I had heard referred to in government circles as the ‘Minister of All Affairs’ for his proclivity to intervene in legislation that did not directly concern his portfolio – queried it. I think that there is every chance that the extent of the independence contemplated would have been watered down had it become the subject of a full Cabinet discussion. However, the discussion was never held because, just as Asmal raised the issue, Cabinet received news of a bomb that had exploded in a popular restaurant in the Cape Town Waterfront complex, derailing further discussion. When the Cabinet officials examined the minutes of the meeting, they could not clearly establish Cabinet’s decision. This was because there had been none. However, when approached by the Cabinet officials for clarification, the minister’s office advised that the Cabinet memorandum and the bill had been approved and the Cabinet minutes were signed off accordingly. Whoever planted that bomb – I don’t recall that ever being ascertained – may well be ultimately responsible for the independence of the competition authorities!

THE COMPETITION ACT

So when all the dust had settled, what did we have?

Thanks to Alistair Ruiters’s energy and his considerable talents as a property developer, we had a spanking new building in an office park in the east of Pretoria. For the most part it was a functional office block designed to accommodate the specific requirements of the Commission and the Tribunal. There were, however, some decidedly eccentric and dysfunctional aspects, none more so than the tables in the various boardrooms and committee rooms. Each room was named after some or other feature of the African continent. And so there was an ‘Africa’ room, a ‘Kilimanjaro’ room and a ‘Nile’ room. Wherever possible, the tables in these various rooms were designed to represent the name of the room. So the Africa room, the largest of the meeting rooms, had a table shaped like the African continent, with the Horn creating some intractable problems for seating arrangements – those sitting up the West African coast had their backs to those sitting along the south-eastern coastline. The Nile room had a long, narrow table with a sluice of water filled by some sort of pumping device representing the great African river running down the middle. After a very short while the pumping device broke down. I am told that it finally burst its banks during a meeting with one of Johannesburg’s leading corporate and competition attorneys, soaking his trousers in the process. From that time on, the Commission’s Nile ran dry. Thankfully, the furniture consultants did not attempt to design a table replicating Kilimanjaro.

But these small niggles notwithstanding, we had office premises that were, in every possible way, very distant from those that most public officials were obliged to suffer. There was one Finance Week journalist who was determined to prick our bubble and who chose to do so by exposing the opulence of our physical environment, which was, in truth, not extravagantly opulent, although it was admittedly a veritable Taj Mahal compared with what the previous Competition Board had had to put up with. The journalist chose the floors – which were apparently constructed from some or other costly cork-like material – to make her point. Ironically, the floors were the first to go – before long, yawning gaps opened up in parts of the building. Clearly, these floors were designed for more refined footwear than the jackboots of the competition officials!

But infinitely more important than the building and its eccentric furnishings was that we had our Act, and it was ready for implementation. The commissioner had been appointed, the Tribunal members had been appointed, and both institutions were actively recruiting staff. Only the Competition Appeal Court had not yet been appointed.

By now most readers will be familiar with the basic architecture of the Act, so let me outline its barest details. The Competition Act established three institutions.

The Competition Commission is the investigatory and prosecutorial body. It receives complaints of restrictive practices, investigates them – or, of course, it may self-initiate investigations – and, if it decides to prosecute any conduct, it does so before the Competition Tribunal. Where an investigation is initiated by a complaint and the Commission elects not to prosecute, it will issue a certificate to this effect to the complainant, who may then bring the matter to the Tribunal. Complainants are also entitled to approach the Tribunal directly for interim relief.

The Commission is also notified of all mergers above a specified threshold. Those that fall below a second, higher threshold are investigated by the Commission, which decides whether to approve a merger unconditionally or subject to conditions, or whether to prohibit the merger. Decisions of the Commission in respect of these mergers – referred to as intermediate mergers – may be appealed to the Tribunal. Mergers above the higher threshold – so-called ‘large mergers’ – are investigated by the Commission and referred for decision to the Tribunal. A later amendment introduced a third category of ‘small mergers’, which I’ll refer to later.

Applications for exemption are made to, and decided upon by, the Commission. These decisions may also be appealed to the Competition Tribunal.

Advocacy is also listed as one of the Commission’s functions. Regrettably, this function was not accompanied by the extension of any substantive powers. What would have been useful is if other branches of government had been obliged to submit their legislation or regulations or policy papers to the Commission for an opinion – or even a vetting – from the point of view of compliance with competition law and principles.

The Competition Tribunal is the adjudicative body established by the Act. As already noted, it decides all large mergers and decides appeals from decisions of the Commission in intermediate mergers. It decides all restrictive practice allegations. It decides appeals from decisions of the Commission regarding exemptions. The Competition Appeal Court has recently decided that the Tribunal has review jurisdiction over the Commission’s decisions in intermediate mergers, although general powers of judicial review do not rest with the Tribunal.12 The chair of the Tribunal appoints a panel of three Tribunal members to hear each matter.

The Competition Appeal Court is the body that hears appeals from decisions of the Competition Tribunal. It is a special division of the High Court, composed of sitting judges drawn from the various provincial divisions. Although the Act purported to give the Competition Appeal Court final appeal jurisdiction, the Supreme Court of Appeal decided that this was at odds with the Constitution and has asserted its constitutional right to be the final court of appeal in all but constitutional matters. The Supreme Court of Appeal did, however, hold that it would only hear appeals from the Competition Appeal Court on application for special leave to appeal and that higher thresholds than normal would apply in deciding to hear an appeal from the Competition Appeal Court.

In summary, the institutions established by the Competition Act have exclusive jurisdiction over all competition matters. If a matter surfaces before the High Court that falls within the jurisdiction of the Competition Act, the court is required to refer the matter to the Competition Tribunal and has done so on numerous occasions. There are, however, three qualifications to this exclusive jurisdiction. The first is outlined above: the Supreme Court of Appeal has decided that final appeal jurisdiction rests with it. The second, which I shall outline below, relates to banking mergers. The third is rather more complex and is related to the recent decision to amend the Act by criminalising hard-core cartel offences.

And so, on 1 September 1999 we were open for business. We were raring to go. While the process of getting to this point had been lengthy and arduous, it was, compared to much legislation and to the establishment of many new institutions, a veritable cake-walk, particularly given the importance and controversial nature of the legislation.

BUILDING AN INSTITUTIONAL CULTURE

I have stressed the significance of the independence of the competition authorities from the executive. At least as important was the independent relationship of the Commission and the Tribunal. We occupied the same premises although in physically separated wings of the building. We saw each other frequently – in our first premises we shared canteen facilities; when amendments to the Act or to the regulations were under consideration we worked together with the Department of Trade and Industry; and there were other occasions that brought us together.

But this has never interfered with the independent exercise of our substantive roles. Alistair Ruiters, the first commissioner, hardly served in that role; so, effectively, Menzi Simelane was the first commissioner during the time that I served on the Tribunal. Although Menzi and I enjoyed a perfectly cordial and amiable relationship, we were never personally close to each other and this may have assisted in maintaining the distance necessary to establish a culture of independence in those early years. In fact, Menzi was, in my experience, proper to a fault, so much so that I have always had some difficulty in associating what I know of him with the allegations of gross impropriety that have been consistently levelled at him in his post-Commission roles in the Department of Justice. However, where Shan Ramburuth, the third and current commissioner, was concerned, not only were we close personal friends, but he had headed up the management of the Tribunal for the first 5 years of its existence. The litmus test is whether the Commission and Tribunal arrive at different conclusions in important matters, and an examination will show that this has been the case since the first, and remains so until this day.

To a limited extent, we knew, or thought we knew, what to expect when we opened our doors. On many, though not all, scores we were in for a rude shock. Certainly those of us who, like me, were unfamiliar with the practice of the law were soon on a very steep learning curve. However much the substantive practice of competition law may represent the interface of law and economics, its procedures, certainly as provided for in our statute, follow fairly standard legal practice.

So the first task, at least for me – though, thankfully, not for Norman Manoim – was learning to run an essentially legal process, and to appreciate the importance of getting the procedures right from the first. Not for their own sake, but because the procedures of an adjudicative body are, in significant part, a tussle between the various counsel and the adjudicators over who controls the courtroom. It quickly became clear that it was imperative that we established in every rule, in every practice, in every gesture, our dominance of the hearing room.

I had never set foot in a courtroom before and so I had no idea – many episodes of Law and Order notwithstanding – of how to run a quasi-judicial proceeding. On one occasion, in the first weeks of our existence, I handed down a judgment on some or other procedural matter and, revealing my background in negotiated outcomes, asked the parties whether or not they agreed with it, only to have Norman hissing in my ear, ‘It doesn’t matter whether they agree with it or not. If they don’t like it they can appeal’.

I did obviously know that legal proceedings were highly formal affairs. However, the Competition Act specifically enjoined us to conduct ourselves in a less formal manner than that practised in the High Court. I recall when our hearing rooms were being built, Alistair Ruiters asked me whether we wanted an elevated bench, which I immediately rejected. I did, however, quickly come to recognise that to the extent that formality and rather elaborate decorum are ingredients of a very nourishing dish called ‘respect’, it was necessary to adopt certain of these practices. Particularly necessary, I should add, in the case of a new, untested decision-making body that had to establish its control over the proceedings before it. Hence the jackets and ties, the use of formal titles, the rising when the panel entered the room, may not have been my idea of the content of respect – in fact I often felt that those counsel that adhered most conscientiously to these outward manifestations of respect were often those who managed subtly to convey the greatest substantial disrespect. But they were the norms of the legal profession and, if that was the manner in which the lawyers demonstrated their respect in the High Court, then it became important that we be paid respect in similar, if not necessarily identical, coin.

There were other smaller, but no less important, mechanisms for establishing respect. For example, our hearings never started late. After years of waiting for trade union meetings or Nedlac negotiations to start, after waiting for hours in the ante-rooms of ministers and other high officials who, I am convinced, demonstrated their relative power, not to mention their sheer incompetence, by making their supplicants waste hours simply waiting for them, I knew something about the disrespect and deep contempt that conduct of this sort actually generated. And I was determined that we would not be that sort of body.13

This all seems to have worked. The critical, first-order test of an adjudicative body is, I think, in the distance it sets up between the adjudicators and those who appear before it, in establishing that the only access to the Tribunal or influence over its decisions is in the hearing room. This was established remarkably quickly, particularly given the practice of our predecessor, the Competition Board, which, as I have explained, was dominated by ex parte conversations and deals concluded in proverbial smoke-filled rooms. I can’t recall a single occasion on which any litigant or any legal representative attempted to ‘negotiate’ an outcome in the style of the erstwhile Competition Board. Lawyers periodically harassed our registrar and our case managers and still do, but these were largely around securing hearing dates and other administrative matters, and the Tribunal staff seemed able to give as good as they got.

Of course, in the Commission things were different. That is where the negotiations and the arm-twisting took place. And important decisions were taken on this basis. But these were legitimate engagements with a prosecutorial body. In circumstances where the Commission did not give way, this has led to persistent formal applications to the Tribunal, which still rear their heads, aimed at getting access to the Commission’s case notes and dockets. But these have generally been rejected by the Tribunal.14 And, of course, the Commission does make critical decisions – to prosecute or not, to recommend prohibition or acceptance of a merger. And occasionally, particularly in the earlier years, the Commission made some palpably incorrect decisions. But it was not the ultimate decision-making body; and where a decision of the Commission was particularly off the mark – as for example in the Afrox–Mediclinic merger, of which more anon – the wonders of the discovery process, a liberal interventions regime, and robust cross-examination by skilled counsel generally exposed the most egregious errors. I have to say, though, that even in the worst of these decisions, having been exposed in my brief stint as chair of the Competition Board to the charms and wiles and persuasive capacity of Sandton’s finest legal minds (not to mention the selective provision of information), I always understood, and empathised with, the capacity for making grievous errors in those circumstances.

But an open hearing before an independent decision-maker was far less susceptible to error than the chicanery of the smoke-filled room. This, of course, is not to say that Tribunal-type processes are immune from error – I presume that we made our fair share of mistakes. But discovery and cross-examination are extraordinarily effective instruments for getting at the truth and for getting to grips with contending arguments, which is why I think that most of the really serious errors were made in the decisions of the Commission and, at the other end of the hierarchy, the Competition Appeal Court. The former errors could be repaired by the Tribunal precisely through the truth-seeking nature of our procedures; the latter, we simply had to grin and bear … or despair.

The Competition Act, in what is something of a departure from normal judicial or quasi-judicial practice, granted the Tribunal ‘inquisitorial’ powers, which have never been clearly defined – and I doubt whether they could be. The manifest intention was to give the Tribunal panel a greater role in determining the nature and content of the proceedings than would normally be the case in the ‘adversarial’ High Court.15 These powers were liberally exercised in merger hearings, particularly in the earlier years when the Commission was still establishing its powers and earning respect. However, in later years, as the Commission has come into its own, it has become less important for the Tribunal to supplement the record or call witnesses who had not been called by either the Commission or the merging parties.

I don’t believe that the Tribunal has taken sufficient advantage of its inquisitorial powers in either merger or restrictive practices hearings. For the moment let’s just say that I think that we were excessively sensitive to the prospect of judicial review and so bent over backwards to remain within the boundaries of established adjudicative practice, and in the process sacrificed some of the advantages of our inquisitorial powers, our relative informality and our status as an administrative, quasi-judicial body. We took on the high standards of a court without possessing many of the critical powers of judges and courts. This was never mirrored in our substantive competition decisions, where we were confident in asserting our specialist expertise relative to that of any of the courts above us and, accordingly, we never made decisions with an eye to what might or might not be acceptable to these superior courts, except, of course, in those very rare instances where we were obliged to follow a clear competition rule laid down by a superior court. However, where constitutional and administrative matters were concerned, I believe that we were excessively conscious of their oversight and their practice.

Of course, gaining the respect of the legal profession is not only about forms of dress and address. It is also about the substance of our practice. And by confidently asserting our superior expertise in the area of competition law – even when that confidence was somewhat tenuously based – while simultaneously demonstrating greater circumspection and a more deferential approach in areas of administrative and constitutional law, we developed a reputation as a body that would set and demand high standards on ‘our’ terrain but would follow the highest standards of fairness and constitutional adherence when on ‘their’ terrain. I repeat, I think we were sometimes too generous in allowing these considerations to determine the nature of our proceedings.

Which brings me to the second rude shock. That, of course, was the amount of time and importance that would attach to these general law matters. Show a lawyer an area of the law with which he is unfamiliar and which is somewhat at odds with standard legal practice (I recall the palpable surprise of one eminent counsel when it dawned on him that we were entitled to void, on competition grounds, otherwise perfectly valid contracts), place him before a body that is able to display a degree of confidence in its grasp of the law in question, and an open invitation to the taking of general law points is effectively issued. And a new Act with far-reaching implications for the exercise of property rights, in a new constitutional order tested before a judiciary that had no experience of the substantive law or appreciation of its social significance or economic principles, was particularly vulnerable to challenge on constitutional and administrative grounds. And challenged it was.

THE EARLY ROUNDS

How then to characterise the first period of the competition authorities’ life? There was naturally no particular date, no particular time period or event that delineated the opening period. There were, rather, a number of important decisions that, in my view at least, characterised the birth of this new period of competition law, that set it on its path. First, there were a number of important procedural decisions that involved interpretations of our powers and procedures as distinct from substantive issues of competition law. Second, the early period was characterised by certain important abuse-of-dominance decisions. Third, there were several pioneering merger decisions.

Somehow my recollection of those early years is of adjudicating one procedural matter after the other. However, when I look at the list of early cases, the truth is that the work of the Commission and the Tribunal got going with a number of interesting and important abuse-of-dominance cases. The sector from which several of these emanated should not have been surprising, but it was.

South Africa’s agricultural sector had been heavily regulated. Central to the regulatory scheme in agriculture were statutory control boards managing single-channel marketing systems and cooperatives through which the output of their members, the farmers, was effectively channelled. The cooperatives essentially sold packing, marketing and other facilities to their members. In 1996, in an important instance of deregulation, the single-channel marketing system was abolished by the Marketing and Agricultural Products Act. Many of the key cooperatives that had constituted the institutional structure of single-channel marketing converted themselves into privately-owned companies and, in the process, converted their erstwhile members into shareholders. They purported to impose upon their shareholders – the farmers – the same obligations to use exclusively the services of the erstwhile cooperative, now public company. But new entrants attempted to enter these exclusive territories by offering some or even the full gamut of services provided by the former cooperatives. The raisin farmers were being offered better prices for their raisins by those who were effectively bidding for the opportunity to provide the downstream packing and marketing services.16 The citrus farmers were also being offered better deals on packaging and marketing.17

And so while we entered the competition terrain with our minds firmly focused on South Africa’s well-known monopolies and oligopolies – cement, paper, diamonds, banking, retail – what did we get? We got the humble raisin from Upington in the Northern Cape and aggrieved citrus farmers from the Gamtoos River Valley of the Eastern Cape. But in the process we got very interesting abuse-of-dominance cases of some significance to ordinary South Africans, not only those directly implicated in the matters at hand, but others who had been similarly locked into arrangements of this nature. While a lack of follow-up in order to assess the impact of competition enforcement has been a particular weakness of the South African competition regime, we later learnt, through the grapevine so to speak, that in the far northern town of Upington the complainant had, as a result of our intervention, succeeded in establishing a successful business.

These initial cases established a pattern that characterised the work of the competition authorities in the first 10 years of their existence and that would influence their strategy in the near future. Our experience demonstrates that abuse of dominance is heavily, if by no means exclusively, rooted in a history of regulation and state ownership. The early agricultural cases, cases that concerned quite unimaginable markets – I recall trying, in vain, to explain to my incredulous 10-year-old why my job should require me to care about the price of raisins, not to mention the impact the nasty little things had on my sleep – were all effectively centred on undoing the consequences of regulation, inasmuch as previously licensed monopolies were attempting to prolong their monopoly privileges by excluding new entrepreneurs from entering their long-captive markets. And so it will come as no surprise to learn that later abuse-of-dominance cases saw the big guns of the state, the likes of South African Airways, ArcelorMittal and Sasol in the dock, with Telkom taking determined advantage of jurisdictional uncertainties in order to stay out of harm’s way. All are or were previously state-owned enterprises. Some, like ArcelorMittal, the former Iscor, had been privatised many years before.

These were by no means the only abuse-of-dominance cases that we heard, but the fact remains that state ownership and regulation confer immeasurable and long-lasting advantage, and it requires determined and thoughtful action by the state to manage the transition from a market monopolised by exclusionary regulation and ownership into one governed by competition. But it is not beyond the bounds of imagination or possibility. Had the privatisation of Telkom been more carefully considered or had the state not permitted Iscor to own the Saldanha Bay steel mill – in both cases alternatives were clearly conceivable and were proposed – we would not today be confronting a notoriously uncompetitive telecommunications sector (or, at least, the inevitably pro-competitive consequences of innovation would have made themselves felt much earlier), nor would we be worrying about excessive pricing of steel products. Those are the consequences of a non-existent competition policy.

The early abuse-of-dominance cases also established both the value and the limitations of interim relief in the application of competition law. On the face of it, competition law – and particularly abuse of dominance – is tailor-made for interim relief. The image (and it’s not far-fetched) is of a large predator engaged in some or other conduct aimed at excluding its prey – competitors or would-be competitors – from the market. The dominant firm will simultaneously retain an army of lawyers instructed to engage in diversionary tactics sufficiently lengthy and costly to ensure that the patient is long dead by the time the cure, in the form of a final order pursuant to a full trial before the Tribunal, is administered. The absence of an effective provision for interim relief was thought to be a major weakness of the previous competition regime.18

An interim relief provision was provided for in the Act. It’s important to understand that interim relief is a remedy available to a complainant who, having filed a complaint with the Commission, is then entitled to approach the Tribunal directly for interim relief. The essential character of interim relief is that it is sought and decided well before the conclusion of a thorough investigation of the complaint. The matter is argued on the basis of affidavits. Oral evidence is rarely, if ever, entertained in interim relief applications. Moreover, because it is brought by the complainant and not the Commission, the losing party is vulnerable to an adverse costs award.

The approach to interim relief contains some peculiarly contradictory features. On the one hand, the requirements for proving interim relief are, by definition, less onerous than those required for a final decision following a full trial. For example, the adjudicator has only to establish prima facie evidence of a restrictive practice. On the other hand, the decision-maker must approach an application for interim relief with considerable caution. Because factual disputes cannot be resolved by a thorough examination of a full record, nor are there the critically important expedients of discovery and cross-examination, the accepted approach is effectively to accept only those facts that are not in dispute or, at least, when there is a factual dispute, to accept the respondent’s version. Finding the appropriate remedy is also important and difficult – in particular, care must be taken to ensure that the remedy imposed is not final in nature. So a claim for interim relief is at once easier and more difficult to sustain.

The citrus and raisin cases demonstrated the value of interim relief. Factual disputes were few and easily resolved and – of critical importance – clear, focused interim remedies were available. By contrast, the other early interim relief applications with which we dealt demonstrated the unsuitability of this procedure for dealing with factually dense and complex matters in which key facts were, inevitably, in dispute. Most abuse of dominance cases were of this latter kind, and this helps to explain why the provision for interim relief has been used less frequently than initially expected. This difficulty is most pithily illustrated in Natal Wholesale Chemists,19 the most straightforward in a long line of extremely complex cases dealing with the distribution of pharmaceutical products.

However, the problems posed for the granting of interim relief in Natal Wholesale Chemists paled alongside the larger cases involving pharmaceutical distribution. The largest of these was brought by nine wholesalers against five manufacturers and two logistics providers. The complaint in this case was initially filed in June 2000. This was followed by an application for interim relief, which, after much preliminary skirmishing, was heard in 2003 by the Tribunal, which granted relief. However, the Competition Appeal Court remitted this to the Tribunal for further consideration. The decision – in an interim relief matter, bear in mind – was finally handed down on 18 June 2003, a full 3 years after the complaint was filed!20

This taught another important lesson, at least to one unschooled in the ways of legal practice. And that is that the wheels of justice may indeed turn slowly, but in critical instances this isn’t the fault of an ineffective administration or incompetent adjudicators or even the prosaic, but very real, problem of finding suitable hearing dates. Equally frequently, it is the successful outcome of a deliberately conceived obstructive legal stratagem. Interim relief is, by definition, intended to provide expeditious relief. However, in this instance the delay was clearly orchestrated by the complainant, who did not want to risk a final outcome, and who preferred to leave the matter in limbo, thus increasing the uncertainty faced by the respondent. And so the complainant, while maintaining the fiction that it was seeking urgent relief, in fact contrived to drag out the interim relief proceedings for as long as possible.

As I write this, it is difficult not to hear the derisory laughter of seasoned lawyers who employ these tactics every day. It is difficult not to hear the standard arguments trotted out in patronising terms: every client is entitled to the most thorough-going defence even if that entails skirting as closely as possible to abuse of the adjudicative process; that the price of the powerful dominating the exercise of these rights is worth paying because it also establishes the standards that underpin the law’s avowed purpose of protecting the powerless against abuse by the powerful. However, in 10 years, I was never able to elevate strategies that purposefully set out to delay or obstruct the adjudicative process above, let us say, outright perjury. In the International Health Care Distributors case we were presented, replete with counsel’s hyperbolic courtroom dramatics of the most hackneyed variety – a substitute for a weak case and scant knowledge of competition law – with an argument by the complainants that interim relief was urgently required to save an entire market segment, the wholesalers, and, with it, competition from destruction while in fact everything possible was being done by the very complainants to prevent us from ruling on the matter. It’s little wonder that the law and the Constitution – and its accompanying high-minded principles of fairness and justice – are held in such low regard by the public when what is most frequently on display is the cynical distortion and abuse of these principles by well-heeled lawyers representing powerful clients in business and politics.

As already intimated, one regret that I have regarding the approach adopted by the Tribunal on my watch is that we did not adopt a more robust approach to these abuses, successfully deployed on many important occasions: the attempted acquisition of Goldfields by Harmony, the plethora of preliminary skirmishes in the alleged milk cartel and many other matters immediately come to mind. We didn’t take a more robust approach partly because we effectively decided to err on the side of fairness, and partly because we knew that the slightest appearance of a lack of fairness or due process was precisely what we were being goaded into displaying. This would then have enabled more delay as our alleged lack of fairness or disregard for sacred constitutional principles went to the High Court, to the Supreme Court of Appeal and to the Constitutional Court. Maybe the professedly aggrieved party would get lucky and win one of these courtroom battles. However, the real objective was simply delay and obstruction because these were, in themselves, often sufficient to secure the desired victory.

Just as we never wrote a substantive competition judgment while looking over our shoulder at what the higher courts would make of it, so too we should not have second-guessed what these courts would have made of our administrative law judgments. In any event, I have little doubt – and this is confirmed by recent statements made by former Chief Justice Pius Langa – that there are sufficient judges who share this frustration and that this, combined with our status as an administrative body, would have given us the latitude needed to be less tolerant of this type of conduct than the Tribunal has proved to be.

Finally, these early abuse-of-dominance cases – in particular the pharmaceutical cases, though all decided on applications for interim relief – began to reveal some important lessons about the sometimes difficult-to-draw line between, on the one hand, robust pro-competitive conduct (even when practised by an overwhelmingly dominant firm), and abuse of that dominance, on the other.

This may not seem like a particularly profound insight to those steeped in antitrust economics. But it was something of an epiphany for those who believed that the powerful – and who could fit this description more accurately than big pharma? – were bound to be wrong, and the victims of their decisions always right. So there is a certain irony in the consistent criticism levelled at the South African authorities that they took an excessively robust position on abuse of dominance: the first 10 years of our abuse-of-dominance jurisprudence began with a decision in favour of big pharma and ended with a decision in favour of big tobacco! However, as we shall see in our discussion devoted to abuse of dominance, this does not mean that we were unconcerned with abuse of dominance and that we did not support measures to ease the burden facing those responsible for prosecuting abuse-of-dominance cases. It simply impressed upon us the acute complexity of many of these cases and the vast quantum of resources required to mount them.

Leaving aside mergers, which were on our agenda from day one and which I’ll deal with in Chapter 3, we soon began to encounter our first cartel cases. Some low-hanging fruit was plucked, generally in the form of rate-setting by professional bodies.21 And the Commission referred the Ansac (the American Natural Soda Ash Corporation) matter to us. I will deal with an aspect of Ansac in the later discussion of cartels. Suffice it here to note that this was the ultimate lesson in obstructive legal stratagems. Ansac’s legal team managed to drag an incontrovertible cartel case through every court in the land for 10 years, whereupon Ansac conceded, hours before the conclusion of the final trial before the Tribunal, precisely what it had been accused of in the first place. I recall Ansac’s counsel once telling me that ‘a case well managed can last forever’. And so we spent 10 years watching that cynical adage being played out.

However, in this early period, the most significant event pertinent to the Commission’s ability to deal with cartels was a Supreme Court of Appeal decision declaring invalid the Commission’s first dawn raid. It probably set back the prosecution of cartels by years. There are a number of interesting facets to this case. For the most part, however, it suggests the importance of the enforcement agency selecting its early battles and its choice of weapons carefully. It demonstrates that power that rests on statutory authority alone will rarely suffice, if society is not yet ready to accept the exercise of those powers. And it demonstrates that while public displays of power – such as executing a search and seizure warrant or ‘dawn raid’ – are undoubtedly key weapons in the fight against clandestine cartels, the enforcement agency should, in its early years, before it has established any reputation, before the legitimacy of its mission has been established, in wielding this sword, err on the side of ‘strict regard for decency and order, and with regard for each person’s right to dignity, freedom, security and privacy’, as the Act itself puts it.

Those familiar with antitrust enforcement practice will not be surprised to learn that the target of the Commission’s first dawn raid was a prominent cement producer. Cement is a homogeneous product in which price is the overwhelming basis for competition, and so when competition breaks out it is likely to take the form of vicious price competition. It is also highly capital-intensive, involving large, lumpy investments in plant – investments to which investors are reluctant to commit unless they have reason to be confident that they will be able to utilise their costly capacity fully, and that the method of achieving this will not entail pricing at levels too low to realise what they deem to be the requisite returns on capital. For this reason, cement is a prime candidate for cartelisation – both price- fixing and market allocation. Indeed, in South Africa the cement industry had, for a significant period, and until relatively recently, been exempted from competition law; it had been permitted to operate a cartel. So here was an industry that not only had a powerful incentive to collude, but was well practised in this conduct. It is little wonder that the cement market featured high on the list of the Commission’s targets, as is the case in many young competition jurisdictions.

However, cement producers are not to be trifled with. Because of the sheer scale of the investments required, the producers – including several prominent multinationals – are usually very large, well-connected companies. And so while the Commission’s choice of target was well advised inasmuch as the likelihood of unearthing cartel conduct was strong, it should have reckoned on a very robust response when it raided Pretoria Portland Cement (PPC), a jewel in the crown of the Barlow Group, one of South Africa’s largest and most long-standing industrial groups, a veritable pillar of industry.

In brief, the Commission obtained the necessary warrant from a judge of the High Court. However, elements of its execution enraged the Supreme Court of Appeal, which ultimately set aside the warrant and ordered the fruits of the raid to be handed back to the company. In particular, the court was outraged by the Commission’s decision to have television crews accompany its officials on the raid. Indeed, the tenor of the judgment makes it clear that the court, citing the surreptitious manner in which the Commission secured entry to the plant by the television crews and a number of ill-advised remarks by the Commission officials, concluded that the principal purpose of the raid had been public humiliation via a public display of power. As Justice Schutz put it in his scathing judgment: ‘A perusal of the sections which I have quoted shows two things. The first is that the legislature has placed power in the hands of the Commission. That is as it should be, as monopoly is a canker that eats into a free enterprise economy. The second is that the legislature showed an awareness that power may be abused and so went to lengths to see that constitutional values were respected. In this connection see, among many other things, especially the references to decency, order, dignity, freedom, security and privacy in section 49’.22

By handing back the fruits of the search to the company, and by not merely censuring the Commission in the strongest terms or seeking an alternative remedy to demonstrate its disapproval of the Commission’s conduct, the court clearly placed ‘constitutional values’ above the ‘canker’ that is monopoly, and above the interests of consumers. In a much later case before the Constitutional Court, the fruits of an invalid warrant were ordered to be ‘preserved’ for possible production when the merits of the case came to be decided. This would have been an ideal process to follow in this case, but it was neither asked for, nor considered by the court. It would be some time before the Commission attempted another dawn raid. It would take even longer – about 10 years – before PPC applied for leniency for essentially the same offences that the Commission had sought to establish through the expedient of the dawn raid mounted a decade earlier. So although the Commission may have used a rocket launcher to attack a target that should have had a sharp stiletto applied to it, there can be no doubt that all those years back it did hit the right target.

However, in the first years of the competition authorities’ life, the really big hit concerned a matter that the competition authorities were prevented from deciding. This was the proposed merger between Nedcor and Standard Bank Investment Corporation (Stanbic), two of South Africa’s biggest banking groups. Or let me rephrase that: it concerned a successful attempt by Nedcor to persuade the High Court and ultimately the Supreme Court of Appeal that its merger did not fall within the jurisdiction of the Competition Act. It is undoubtedly for this reason that I associate the first year of our life with procedural matters. This matter not only eclipsed all others in its far-reaching implications, it damn near eclipsed the competition authorities altogether.

To cut a very long story short, Nedcor, the country’s fourth-largest bank, made an offer to acquire Standard Bank, the country’s largest bank. Because Nedcor was controlled by the insurer Old Mutual, while Standard controlled Liberty Life (one of Old Mutual’s most significant competitors), the long-term insurance market was also implicated in this transaction. The Standard Bank board opposed the attempted acquisition and prepared to argue its case before the competition authorities. Nedcor, however, contended that, by virtue of the operation of clause 3(1)(d) of the Act, jurisdiction over banking mergers was subject to the Banks Act and hence to the discretion of the Minister of Finance. Section 3(1)(d) provided that ‘(1) This Act applies to all economic activity within, or having an effect within the Republic, except … (d) acts subject to or authorised by public regulation’.

Nedcor argued that because the Banks Act required that banking mergers be approved by the Minister of Finance, these constituted ‘acts subject to or authorised by public regulation’ and, hence, were excluded from the jurisdiction of the competition authorities. This argument was upheld by the High Court. The raisin producers immediately proceeded to the High Court to press their own claim that the Marketing and Agricultural Products Act also made provision for ‘acts subject to or authorised by public regulation’ and so they too (and, it would appear, every other market within the agricultural sector) should be spared the scrutiny of the competition authorities. The High Court duly excluded the raisin producers from our jurisdiction.

Catastrophe loomed and this appeared to be confirmed when the Supreme Court of Appeal turned down an appeal by Standard Bank against the decision of the High Court.23 Although the implications of the reasoning underlying the Supreme Court of Appeal’s judgment ousting the jurisdiction of the competition authorities over banking mergers was somewhat more narrowly-based than the High Court’s judgment in the banking matter – and significantly narrower than the sweeping judgment of the High Court in the raisins matter – it was still pretty broad. It could easily be construed to exclude from the jurisdiction of the competition authorities all markets in which horizontal or vertical agreements, unilateral conduct and all mergers subject to any other public regulation are subject to other forms or instruments of public regulation. And so, despite Justice Schutz’s confident assertion that, notwithstanding his judgment, ‘the [Competition] Act is alive and well’,24 this was by no means certain. While a subsequent judgment of the Supreme Court of Appeal – heard by an entirely different bench – upheld the appeal from the raisins judgment, thereby restoring the competition authorities’ jurisdiction in that market, it appears that in this instance the Supreme Court of Appeal was prepared to concede the jurisdiction of the competition authorities because no regulations over the categories designated by the court in the banking matter – namely horizontal or vertical agreements, abuse of dominance, and mergers – had yet been put in place.25 Indeed, the one clear certainty emanating from the four judgments of the High Court and the Supreme Court of Appeal was that an enormous amount of room had been left for opportunistic litigation, portending endless disputes over jurisdiction.

The Supreme Court of Appeal judgment in the banking transaction was clearly going to prove particularly problematic from the perspective of future litigation. There was no doubt that the Minister of Finance’s permission was required for banking mergers. Nor was there any doubt that the Banks Act required the minister to consult with the Competition Commission. However, it was equally clear that the principal criterion that the Minister of Finance would bring to bear when exercising his discretion was, in the words of the Supreme Court of Appeal judgment itself, ‘the importance of maintaining the integrity and security of banks’. Although the Supreme Court of Appeal majority rejected the notion that the Competition Act – particularly in the light of section 3(1)(d) – in effect provided for dual regulation, this is asserted rather than reasoned. Acknowledging that there is a statute with jurisdiction to examine the competition implications over mergers in all areas of economic activity, which coexists with a statute that also has merger jurisdiction in a particular sphere of economic activity but that specifically establishes banking stability as its assessment criterion, appears precisely to infer the intention that the two statutes – and the respective regulatory regimes they established – should enjoy concurrent jurisdiction. It’s difficult to understand why the Supreme Court of Appeal decided against inferring concurrency.

The gulf between the drafters’ intention and the Supreme Court of Appeal’s interpretation of section 3(1)(d) was unbridgeable and fatal to the administration of the Act. In drafting section 3(1)(d), our mental picture was drawn from the telecommunications sector. Here the regulator was required to license new entrants. Even though this decision was of immense significance from a competition perspective, our jurisdiction was, we understood, ousted by the clear requirement that the sector regulator (actually, as it turned out, the minister) determined the number and identity of telecommunications operators. We could not, even in the name of promoting competition, presume to ease conditions or entry into licensed markets. Or price: the telecommunications regulator was empowered to determine the prices of certain telecommunications services. Again, despite the significance of this decision from a competition perspective, we understood section 3(1)(d) to confirm that the excessive pricing or discriminatory pricing provisions could not be invoked in order to challenge the decision of the regulator. So it came as something of a surprise, to say the least, when the Supreme Court of Appeal decided that any public regulation of the major categories with which competition legislation was concerned – agreements between firms, unilateral conduct and mergers – ousted the jurisdiction of the competition authorities. I can see how section 3(1)(d) permitted that interpretation. But I can’t see that it was the only interpretation that it permitted or that it was the most sensible interpretation.

The Supreme Court of Appeal’s decision potentially neutered the Competition Act, Justice Schutz’s views on the continued vitality of the Act notwithstanding. In the event, the Minister of Finance duly considered the banking transaction in question. As required, he sought the advice of the Commission, which counselled him to prohibit it. This he did, largely, it appears, on competition grounds. But the Act had to be amended. We had to assume that grounds would always be found in section 3(1)(d) to challenge our jurisdiction, at least grounds sufficient to provide space for mounting opportunistic appeals and reviews and introducing intolerable uncertainty into the administration of the Competition Act.

I can’t say that we weren’t warned. In a seminar on the bill, Michael Katz, whose firm represented Nedcor, is said to have referred to the clause as the ‘Pajero clause’, the clause that would generate the sort of fees that enabled lawyers to buy luxury 4×4 vehicles. He was right, as he so often is. I recall meeting David Unterhalter – whom I didn’t much know at that stage but who was to become South Africa’s leading competition counsel – at some or other gathering at the National Gallery in Cape Town after he’d participated in one of our sessions before the parliamentary committee, who warned me that we would come to rue section 3(1)(d). He was also right, as he too often is. But Phil Knight and Norman Manoim had thought about this and they were firmly of the view that the courts would understand it in the manner intended. It was the first of a range of important lessons – of the prospect of intelligent people holding, in good faith, diametrically opposed interpretations of the same few words and of the pre-eminence given by most judges to the black letter of the law. Good sense and context notwithstanding, the dictionary seems always to be a jurist’s most important textbook. We would encounter far more surprising – to put it at its most polite – interpretations of the Act and of our judgments. Fortunately, the implications of this judgment were so far-reaching that the only possible response was an immediate amendment of the Act. Even critical amendments can take forever. However, testimony to the importance of this particular amendment was the speed with which it was achieved. Section 3(1)(d) was repealed.

But amendments always come with a sting, in this case two stings. First, the banking regulators – and with them the Reserve Bank and the Treasury – which had not hitherto displayed any interest in the Act, suddenly became aware of an interloper on their turf. So, with the endorsement of the Minister of Finance, they persuaded Parliament to carve out an exceptional procedure for assessing banking mergers. The Minister of Finance and the banking regulator had in fact sided with Nedcor’s interpretation of 3(1)(d) in the various court proceedings. Accordingly, with the repeal of 3(1)(d), an amendment was inserted that required that banking mergers be notified to both the banking regulator and to the Competition Commission. Default jurisdiction over competition matters still resided with the Commission but the Minister of Finance was entitled to issue a certificate, assuming jurisdiction. All manner of red herrings were raised in the parliamentary hearings, the most common being that if the banking regulator wanted, in order to avoid the threat of systemic banking crises, a merger of a failing bank to go through, then it did not want the competition authorities preventing this on competition grounds. We naturally pointed out that the specific availability of a failing firm defence would prevent this from happening. But they would have none of it.

I suspect that financial market considerations were less important than turf. Evidence of this is that although the Act clearly provided that the minister was entitled to assume jurisdiction only when, after applying his mind to the question, he determined that the stability of the financial sector was at stake; in fact, whenever a banking merger was notified, a certificate was issued as a matter of course. Over a relatively short period, a number of smaller banks were swallowed up by South Africa’s big banks. I don’t know what a rigorous competition evaluation before the Tribunal would have concluded regarding these mergers. I do know that we have a highly concentrated banking sector, and a public that constantly complains about the level of bank charges. But to go up against the Minister of Finance and the Reserve Bank on a matter concerning banking supervision – or much else for that matter – is generally not a winnable proposition, as many other competition authorities have recently learnt.

Second, the issue that led to the amendment caused the parliamentary committee to apply itself to resolving likely further jurisdictional disputes between the competition authorities and the sector regulators. In brief, a clause was inserted into the amended Act that required the Competition Commission and the sector regulators to enter into memoranda of understanding that would lay down an agreed process for resolving jurisdictional disputes. This not only raised complex legal issues of its own, but it also provided a further platform for legal stratagems on the part of regulated entities that were intent on avoiding the jurisdiction of the competition authorities, notably Telkom, the state-owned fixed-line telecommunications provider. It’s something of an irony that a clause – 3(1)(d) – drafted with the telecommunications sector in mind had to be repealed in order to resolve unintended consequences in other markets, although I suspect that neither the original clause, nor its repeal, nor the parliamentary committee’s intercession, resolved jurisdictional difficulties in the telecommunications markets. As I’ll elaborate below, the nettle that had to be grasped in telecommunications was the existence, particularly in abuse-of-dominance matters, of genuinely grey areas that implicated both core competition considerations and licence conditions. However, where banking mergers are concerned, there are no grey areas; just turf and the innate conservatism, understandable if not always warranted, of financial regulators. I should underline that it is only over banking mergers that the minister is entitled to assume jurisdiction. The policing of restrictive practices in banking remains firmly within the jurisdiction of the competition authorities.

And so our early period, characterised by mixed but net positive outcomes, ended. We had our Act, warts and all. Important challenges had been mounted against it – necessitating an early amendment – and to the manner in which the Commission exercised certain of its most important powers. We had heard a number of important applications for interim relief concerning abuse of dominance allegations, and had in the process learnt much about the practice of the law, the utility of interim relief and the complexity of abuse of dominance. The Commission had stuck its toes into the shark-infested waters of cartel investigation only to have them snapped off by the Supreme Court of Appeal; and we had, still unbeknown to us, begun the longest matter that we would ever hear, in what was, from the outset, an apparently open-and-shut case of collusion. We had a regular roll of merger cases, and, within our first year, one important prohibition.

THE COMPETITION APPEAL COURT: A PRELIMINARY NOTE

The following three chapters deal with the three broad categories that comprise the substance of the work of most competition authorities, namely merger review, and the prosecution of abuse-of-dominance and cartels. However, I first want to clarify an aspect of what will follow in a discussion that inevitably involves a degree of assessment of the institutions established by the Act, and this particularly concerns the Competition Appeal Court.

I shall, in the course of discussing the competition jurisprudence that has developed in South Africa, lay bare some important differences between the Competition Tribunal and the Competition Appeal Court. These are matters of public record as reflected in the published decisions of the two bodies. As is to be expected, the two bodies arrived at different conclusions in some matters of considerable importance. While I am happy to concede that reasonable people may, in the utmost good faith, arrive at different conclusions in most matters of law and economics, I happen to think that the Tribunal was, in most important instances, correct and that the Competition Appeal Court erred on important occasions. I have spared little in laying bare these differences and in indicating where I think the Competition Appeal Court erred.

However, I am less concerned about these specific differences than I am about the general approach adopted by the Competition Appeal Court that, I believe, underpins many of these particular differences. This is an important general issue for antitrust enforcers everywhere because, to paraphrase Adam Smith’s oft-cited observation, antitrust enforcers from around the world seldom meet together, ‘even for merriment and diversion’ (of which there is a surprising amount), but the conversation ends in a widely-shared gripe about the approach of the courts and the judges who generally have the last word on a competition prosecution or merger review. We attempted to deal with this widely-expressed frustration by building into our Act the solution that many other competition enforcers imagine would solve their problems. We attempted to establish a specialist, expert court of appeal, a division of the High Court that would, or so we thought, enjoy the final word on competition matters. However, we have landed up with a court of appeal that is specialist, but is neither expert nor final.

South Africa has chosen an administrative system of competition law, a system in which decision-making is the responsibility of two administrative bodies: one, the Commission, the specialist, expert investigative and prosecutorial body; the other, the Tribunal, a specialist, expert administrative adjudicator. The expertise required does not refer to specialist training in competition law. It rather refers to the unusual combination of two disciplines: law and economics. Hence an expert competition body will be expert by virtue of a composition that includes both economists and lawyers.

Of course, the procedures of the Commission and the Tribunal are, as with all administrative bodies, appropriately subject to judicial review. As with all other administrative bodies, the Commission and the Tribunal are required to adhere to the relevant administrative and constitutional standards in the exercise of their powers. I am, somewhat more reluctantly, willing to concede that a right of appeal – as opposed to review – is necessary. But we are dealing with high-stakes matters and it seems fair – and it may, or may not, be constitutionally required – that affected parties are accorded a right of appeal on substantive administrative decisions. I do, though, observe again that, despite the quasi-judicial nature of the Tribunal, we are administrative bodies and the substantive decisions of other administrative bodies – such as, for example, the pricing decisions of the regulatory authorities – are not subject to appeal, although naturally their decision-making process is also subject to judicial review. However, we are – at least, when dealing with restrictive practices cases – concerned with alleged ex post contraventions of a statute and these seem to require a right of appeal that may not be required in respect of ex ante regulatory decisions. Of course, this suggests that the argument for a right of appeal is clearer in the case of a restrictive-practices decision than in the case of a merger decision where the Commission and the Tribunal are exercising an ex ante regulatory function.

However, while conceding that there should probably be a right of appeal from decisions of the Tribunal, there are pertinent questions regarding the appropriate body of appeal and the appropriate level of deference that the appellate body should extend to the Tribunal. These two issues – the forum and the level of deference – are closely linked.

When the Act was drafted, it was decided that a special division of the High Court, the Competition Appeal Court, would be the body enjoying exclusive jurisdiction over appeals from the Competition Tribunal. The intention was to put in place an expert court and also to facilitate expedition. However, what is clear is that the structure and composition of the Court differ significantly from those originally intended by the policy-makers. The original intention – and this is reflected in the initial drafts of the bill – was that experts in economics and commerce should be appointed to the Court and given a role akin to that of assessors. The obvious purpose underlying this intention was precisely to construct a dedicated, expert court of appeal.

However, the Judicial Service Commission, the body responsible for recommending the appointment of judges to the President, was not willing to consider appointing ‘laypersons’ to a court – that is, individuals who did not possess the necessary formal qualifications required for appointment to the office of a judge – even in the position of assessors who would participate in decisions of competition law and economics but not in the application of general law. It also sought an opinion from a leading advocate who advised them that this would not be permissible.

The upshot is that we have a specialist, expert investigatory body (the Commission) and a specialist, expert adjudicator of first instance (the Tribunal), but a court of appeal that is composed entirely of generalist judges and that enjoys (or used to enjoy) exclusive appellate jurisdiction in competition matters, but that is not an expert body for the purposes of engaging with competition law. As to the question of expedition, while access to the Competition Appeal Court is probably easier than access to the ordinary High Court, any expedition that a specialist court may have brought was severely compromised when the Supreme Court of Appeal decided that it, rather than the Competition Appeal Court, was the final court of appeal. So now there is a further stage of appeal in competition matters.

However, leaving aside the role of the Supreme Court of Appeal, the existence of a generalist dedicated court of appeal such as the Competition Appeal Court leaves us with the worst of all worlds. It leaves us with a specialist apex court that, although not expert, is, precisely because of its status as the specialist apex court, intent upon demonstrating and jealously guarding its seniority in matters of competition law. This has a profound impact on the question of deference.

I should emphasise that it was for good reason that we chose to have an administrative system. The decision to do so is founded in the role of economics in competition law and the requirement to have a decision-making body that specifically incorporates economic expertise into its ranks and its thinking. It follows that, in general, when a non-expert body – such as the Competition Appeal Court – is hearing appeals from an expert body, an exceptional level of deference should be exercised.

First, on questions of legal interpretation, competition law is, to be sure, concerned with the interpretation of a statute. However, the adjudicator is also frequently forced to grapple with economic laws and reasoning. There is no body of law in which law and economics – and I stress ‘economics’ rather than ‘commerce’ – interface as closely. If the appellate body has no expertise in economics, common sense should dictate that, even in the area of legal interpretation, it extends an exceptional level of deference to the body that does possess expertise in economics because a substantial part of that interpretative task, though legal in form, demands the application of economic reasoning and methodology.

Second, on the question of fact and evidence, my understanding is that appellate bodies are generally expected to exercise deference to the fact-finding body, the body that is required to read the entire record and that takes oral evidence and so is particularly well-placed to assess the reliability of the witnesses. In addition, an important body of the evidence presented in what are unusually fact-intensive inquiries is economic data, underpinned by economic reasoning and methodology, and so, again, an exceptional degree of deference to the expert body is indicated.

Third, there are, on occasion, important policy issues implicated in competition law decisions. While accepting that there is sometimes a blurred line between questions of law and of policy, I have no doubt that to the extent that policy decisions can be identified and separated from questions of law and statutory interpretation, a very high level of deference should be shown to the expert body.

I shall, in the following chapters, refer to several instances where, despite several statements by the Competition Appeal Court to the effect that it would exercise deference to the expert body and fact-finder, in reality it has not done so. I could postulate a variety of reasons for the Court’s reluctance to defer to the Competition Tribunal.

I think that the practice of deference runs strongly against the grain of the extremely hierarchical order that characterises the practice of law. In fact I know that it does, having been on the receiving end of a stern lecture from David Unterhalter on an occasion when he argued, quite correctly as it so happens, that I was guilty of showing insufficient deference to the Competition Commission. So here one has a strictly hierarchical order of courts as well as officers of those courts, where the higher courts and the most elevated officers are expected to show deference to those below them in the pecking order. This undoubtedly comes hard to those who are addressed and generally treated in an extremely deferential manner by those below them. And it undoubtedly comes even harder when judges are expected to show deference to mere ‘laypersons’.

And I don’t doubt that it comes even harder still when the laypersons constitute the expert body, while the judges are ‘mere’ generalists and, yet, are on a bench that is specifically dedicated to a single area of law in which they are the final decision-makers but, nonetheless, not the ‘experts’. The upshot is that the Court’s status in the area of competition law gives it an ‘interest’ greater than that of an ordinary court of appeal that would, when hearing an appeal, be on the alert for manifest factual and legal error – the stuff of appeals – but that would not be especially concerned to demonstrate its superiority in a single branch of the law, particularly one in which an expert body had been assigned jurisdiction.

I should say something more about this question of ‘expertise’. It’s not meant to suggest that the expert body requires the presence of Nobel prize-winning economists. It merely suggests that legal reasoning and form should be complemented by economic reasoning and substance, and this generally requires the presence of those who have both formal training in economics and practical experience in economic decision-making. There are few economists on the Tribunal, although it is generally the case that at least one is empanelled in matters in which economic issues are in the foreground. I have no doubt that although certain of the lawyers on the Tribunal are extremely knowledgeable about antitrust economics, the presence of those who are not steeped in legal form but who are inclined towards economic reasoning makes a material difference to outcomes on the Tribunal. In essence, it made the Tribunal an expert body, even if not all of the economists – I, for example – would consider themselves to be ground-breaking members of their profession.

By the same token, to deny that the Competition Appeal Court is an expert body is naturally not to question the competence of the judges as general law jurists. It is not even to question their competence or experience in competition law, although I am constrained to note the marked lack of experience of the Competition Appeal Court bench in the area of competition law. Only one of the judges who have sat on the bench of the Competition Appeal Court has had any experience in practising competition law at the bar and relatively slight experience at that. The upshot is that a single judge – the judge president – has occupied a permanent place on the bench for over 10 years and is thus the only one who could claim any wide-ranging experience in the area of competition law, which, of course, gives him a great deal of influence on the bench. And as judge president, and therefore the head of the apex court in competition matters, he is particularly susceptible to demonstrations of the Court’s seniority in competition law and thus is least likely to be deferential to the lower, expert body. But all that having been said, to say that the Court is not an ‘expert’ body is simply to say that the influence of the discipline of economics, which looms ever larger in competition law decisions, is absent from the Court. It’s interesting to note that even when the Court has invited amicus briefs, it has only ever asked lawyers to assist it.

The Competition Appeal Court’s reluctance to defer to the Tribunal is clearly discernible in a number of important decisions of the Court, to which I’ll refer in this and subsequent chapters. It is also graphically present in the gratuitously offensive nature of some of the Court’s comments on Tribunal judgments and the Commission’s conduct. It’s not difficult to gain the impression that this is a body determined to demonstrate its superiority, with considerations of deference a distant second at best. We are all adults and words don’t, after all, break our bones. But the Court’s manifest lack of regard for the Commission and the Tribunal undermines the entire competition law regime, including the Court itself. All it establishes is where formal power lies. The most striking, although by no means the only, example is in the decision in Netstar written by Judge Wallis, his first Competition Appeal Court decision as it so happens, where he strongly suggests, for no apparent reason – other than that he disagreed with the Tribunal’s decision – that the Tribunal has acted in bad faith and with bias, of reading and presenting evidence so as to support a predetermined outcome.26

There is no basis for these insinuations, neither in the case at hand – where, as it so happens, the Court’s decision is questionable, to put it at its mildest – nor in Judge Wallis’s previous, albeit limited, experience at the bar of the Tribunal. So if, indeed, I am being paranoid, this is not for want of evidence of the Court’s superior, rather than deferential, attitude to the expert bodies below it in the pecking order.

I should also add that the fact that the Supreme Court of Appeal has assumed appeal jurisdiction doesn’t promote deference on the part of the Competition Appeal Court. The Competition Appeal Court often gives the impression that it ‘writes for the Supreme Court of Appeal’ because, it appears, a reversal would be an indication that the lower court had ‘gotten it wrong’. The judge president of the Competition Appeal Court has publicly acknowledged the influence of what it perceives to be the likely view which the Supreme Court of Appeal has on the decisions of the Competition Appeal Court. In reference to a disagreement to which I later refer – Nationwide Poles – Judge Davis says as much: ‘I’m all in favour of helping the small guy, but you can help him only when you can sustainably interpret the Act to do so. If you twist the [Act] too much, it will only get overturned on appeal anyway, and then who are you really helping?’27

Note that Judge Davis is not referring here to an issue in which the Supreme Court of Appeal has established a binding precedent. And while I understand that there are established rules and precedents for interpreting statutes, we are in this instance, because of the unusual role of economics, dealing with very particular interpretative issues. Hence it seems that the point is not whether a decision is ‘overturned on appeal’ or upheld; the point is whether or not the decision-makers make a decision that they believe to be a correct interpretation of the competition statute. If the Court holds the view that in competition law the ‘small guy’ occupies a particular place – a complicated question but one that frequently has to be answered in interpreting competition issues – or if the decision-maker believes that in socially critical and evolving markets, as in the merger case of Prime Cure to which I’ll also refer, particular care and circumspection is demanded, then the decision is made and reasoned accordingly without second-guessing the possible outcome of a likely appeal. Certainly, the Competition Appeal Court risks being overturned, but if it is to exercise the function of an expert body, it should be willing to guide both those below it and those above it in the interpretation of an unusual statute. But the Court is not expert; it is an ordinary court of law and so it remains strictly within the bounds of standard interpretative rules.

It’s rumoured, although I’m uncertain how reliably, that an imminent constitutional amendment will soon remove the Supreme Court of Appeal from the competition equation. This may ‘free’ the Competition Appeal Court to take a more robust approach to the interpretation of the Act. However, I am not certain that this would resolve the issue of expertise, although it may speed up decision-making. My preference would be to re-examine the possibility of appointing economists to the Court in the form, at least, of assessors. If that is not possible, then my second-best option would be for an amendment to the Competition Act that eliminated the Competition Appeal Court and placed appeal jurisdiction in the hands of the ordinary High Court and the Supreme Court of Appeal. The optimal solution would be one that allowed direct appeals from the Tribunal to the Supreme Court of Appeal, but I’m not certain whether this is permissible. This may not give us an expert body of appeal, but at least it would remove the ‘special’ interest that a specialist court develops, with the negative implication that this has for showing deference to the expert body.

So if I appear to concentrate excessively on the errors that I think the courts have made, then I would like to be clear that I think this unfortunate decision-making structure is partly to be held responsible for those errors. It needs to be corrected. Either we should have a court of appeal that possesses the requisite specialist skills, and this involves appointing economists to the Court; or we should go through the normal High Court appeal route to courts that do not have such a stake in establishing their mastery and domination of competition law.

But enough with the health warnings, and on with a consideration of our merger review system and, then, of our enforcement regime.

NOTES

1. When public hearings are convened, the parliamentary process requires the relevant committees in the National Assembly and the National Council of Provinces to hold hearings. In this instance the hearings before the National Assembly committee were far more comprehensive and wide-ranging.

2. The policy document defines ‘development’ as ‘our willingness to address socioeconomic backlogs and capacity to correct, over time, existing racial and gender biases in ownership and control throughout the private sector’. Proposed Guidelines for Competition Policy: A Framework for Competition, Competitiveness and Development (henceforth ‘guidelines’) (Department of Trade and Industry, Pretoria, 2 November 1997, para 2.4.3). ‘Competitiveness’, which (along with ‘development’) is also identified as part of the public interest, is defined as ‘help[ing] to make South Africa more competitive by lowering costs along the entire value chain’ (ibid., para 2.4.2). For the unions, this point signalled potential wage repression and a potential compromise of hard-fought labour standards and rights.

3. Guidelines, executive summary, para 2.

4. As will be elaborated, the public interest criteria were applied in merger review and in the possible granting of exemptions. However, they were also included in the preamble to the new Act and in the clause defining the purposes of the Act, and so were clearly intended to influence the interpretation of the statute. This is more fully dealt with in the later chapter on abuse of dominance, and particularly in the discussion of the Nationwide Poles case.

5. Guidelines, para 4.

6. Guidelines, para 2.2.4.

7. Guidelines, para 8.3.1.

8. Contrast Erwin’s approach to the independence of the competition authorities with that of successive ministers responsible for telecommunications towards the telecommunications regulator. See Robert Horwitz and Willie Currie, ‘Another instance where privatisation trumped liberalisation: The politics of telecommunication reform in South Africa – A ten year retrospective’, Telecommunications Policy, 31 (2007), 445–462.

9. Growth, Employment and Redistribution.

10. ‘Proposed competition legislation: Outline of some key perspectives from labour’. Speech given by Kenneth Creamer at the competition law conference, Sandton, 10 July 1998.

11. Appellant v Topco Associates Inc. (405 US 596).

12. TWK Agricultural Limited v The Competition Commission and Another (67/CAC/ Jan07).

13. In Michela Wrong’s riveting book on corruption in Kenya, she recounts: ‘At a formal dinner in London, I found myself discussing with John and a British peer of the realm, in light-hearted vein, what were the little signs that betrayed the fact that once-reformist African governments had lost their way. “My measure is the time a person who’s agreed to an appointment keeps you waiting,” said the Lord. “If it’s half an hour or under, things are still on track; more than half an hour and the place is in trouble.”’ Michela Wrong, It’s Our Turn to Eat (Fourth Estate, London, 2009, p. 74).

14. Medicross Healthcare Group (Pty) Ltd / Prime Cure Holdings (Pty) Ltd (11/LM/Mar05).

15. See Judge Selikowitz in the Competiton Appeal Court: ‘Our courts have repeatedly stated that where proceedings are conducted informally or in an inquisitorial manner the decision-maker is placed in an active role to get at the truth and that the ordinary rules of evidence do not apply. Subject at all times to the requirements of fairness, the Tribunal is not precluded from having regard to hearsay evidence. Indeed it seems to me that expert evidence in disciplines such as economics and socially related sciences will inevitably be based upon hearsay evidence’. Patensie Sitrus Beherend Beperk v Competition Commission and Others (16/CAC/Apr02).

16. South African Raisins (Pty) Ltd and Johannes Petrus Slabber v SAD Holdings Ltd and SAD Vine Fruit (Pty) Ltd (04/IR/Oct99).

17. The Competition Commission v Patensie Sitrus Beherend Beperk (37/CR/Jun01).

18. These issues are traversed in some detail in South African Raisins (Pty) Ltd and Johannes Petrus Slabber v SAD Holdings Ltd and SAD Vine Fruit (Pty) Ltd (16/IR/ Dec99). This case concerned an attempt to suspend a grant of interim relief pending the outcome of an appeal.

19. Natal Wholesale Chemists (Pty) Ltd v Astra Pharmaceuticals (Pty) Ltd, Merck Pharmaceutical Manufacturing (Pty) Ltd and Pharmaceutical Healthcare Distributors (Pty) Ltd (98/IR/Dec00).

20. 68/IR/Jun00.

21. Competition Commission v Hospital Association of South Africa and Another (24/CR/Apr04).

22. Pretoria Portland Cement Company Ltd and Another v Competition Commission and Others (64/2001) [2002] ZASCA 63 (31 May 2002).

23. Standard Bank Investment Corporation v Competition Commission and Others, Liberty Life Association of Africa Ltd v Competition Commission and Others (44/2000, 50/2000) [2000] ZASCA 20; 2000 (2) SA 797 (SCA); [2000] 2 All SA 245 (A) (31 March 2000).

24. Ibid.

25. South African Raisins (Pty) Ltd and Another v SAD Holdings Ltd (176/2000) [2000] ZASCA 60; 2001 (2) SA 877 (SCA) (29 September 2000).

26. Netstar (Pty) Ltd, Matrix Vehicle Tracking (Pty) Ltd and Tracker Network (Pty) Ltd v The Competition Commission and Tracetec (Pty) Ltd (97/CAC/May10).

27. Financial Mail, 7 September 2007.

3. Mergers

In a large contested merger, the stakes are very high. In the usually packed room on the opening day of a significant merger inquiry, the full cast of characters is present. The advocates are primed, in most cases having become expert in a fact-intensive field to which they had never given 10 minutes’ thought prior to receiving their brief, a talent that is a hallmark of experienced advocates. The senior attorneys representing the merging parties shepherd their clients, who, no matter how powerful in their own worlds, invariably appear somewhat uncertain on alien turf, while the junior attorneys and articled clerks fuss around mountains of files. The transaction advisers, who have frequently initiated and driven the transaction with little thought given to the competition inquiry, exude indignation and anxiety at having to deal with this final hurdle over which they have little influence and for which they have zero sympathy. The Commission staff are on home ground and so relatively at ease. The media representatives, used to these little dramas, are also at ease, but expectant. For many in the room – including, I might add, the Tribunal members and its staff – many hours of intensive, head-cracking and often last-minute work have brought them to this point. The tension that pervades these occasions is palpable and invigorating, as is the frisson that I always experience on entering the hearing room.

There are two well-established misconceptions surrounding merger regulation. First, that it is, or should be, low on the list of priorities of a new competition authority. A second and related misconception is that, with the odd exception, it is boring, grunge work.

Only a very small proportion of mergers impact negatively on competition. Moreover, one can generally, although not unexceptionally, tell after a quick scan of the Commission’s recommendation and the competitiveness report filed by the parties whether a notified merger warrants deeper examination. Most do not and so the vast majority are approved quickly and unconditionally. Nor is this surprising. Mergers are not only perfectly legitimate business activities, but they are sometimes an important response to changing internal circumstances or shifts in the environment surrounding the firms in question. It’s thus widely held that merger regulation involves devoting a significant quantum of scarce resources to reviewing complex business decisions that only rarely impact negatively on competition. And because there is a rote element involved in looking at dozens of mergers that are of no apparent concern to competition authorities, they are thought to represent a tedious, rarely exercised gate-keeping function, and thus do not offer the thrill of going after an abusive dominant firm or a pernicious cartel. Both judgments are wrong.

Merger regulation is substantively important because of the long-term impact that these combinations, the most powerful and permanent form of inter-firm cooperation, may have on the structure of markets. They may result in single-firm dominance, thus strengthening the prospect of abusive unilateral conduct, or they may enhance the prospect of anticompetitive horizontal agreements. Moreover, mergers are likely to be particularly important – and to embody the most potentially serious implications for competition – in an economy going through significant structural change. It is precisely in these circumstances that firms will consider the benefits of specialisation, of ridding themselves of non-core assets in favour of a narrower specialisation. It is in these circumstances that firms will be most anxious to cement their relationships with customers and suppliers. And so it is competition authorities in economies undergoing significant structural change that must be most alert to the potential anticompetitive consequences of mergers. The burgeoning of new national competition authorities, including the South African authority, occurred precisely at a time of deep-seated structural change in the global economy and its component national parts.

In addition, merger analysis is a particularly effective platform for capacity-building in a new authority. Merger analysis doesn’t generally give rise to complex legal questions. Mergers also don’t encourage legal point-taking because, in a friendly transaction, the merging parties have no interest in obstructing the merger review process. For the same reason, merging parties tend not to withhold information. The upshot is that merger analysis generally provides case handlers and adjudicators with the opportunity, indeed the obligation, to get to grips with the core stuff of competition analysis: the examination and identification of markets and the analysis of competitive effects.

Moreover, provided that the competition authorities undertake their merger evaluation function conscientiously, it enables them easily to establish a reputation for thoroughness and professionalism. As I’ve already indicated, this is not necessarily true of restrictive practice cases. In the early years of competition enforcement, it is all too easy for the competition authorities to establish a reputation for tripping over constitutional and administrative law hurdles and for fighting costly battles in distant courts over issues that are difficult to reconcile with the mandate of the competition authorities and the public’s understanding of their role. By contrast, through merger review the authorities are easily able to get to the core of their mandate and, if they do it well, the reputational rewards will be considerable.

Merger review also enables the authorities to establish an immediate reputation for standing up to powerful business interests. Corporate leaders, not used to answering to anyone, are suddenly obliged, through the invasive medium of cross-examination, to put the strategic entrails of their businesses on public display. When a few smart and curious business journalists who have long had to try to peer underneath the bland offerings of evasive public relations professionals are thrown into this transparent mix, there are the ingredients, as well as the medium, for establishing a reputation for considerable power, arguably more than is actually possessed by the authorities at this early stage. I won’t easily forget the ‘what-the-hell-am-I-doing-here’ look of bemusement on the face of David Sussman, the powerful CEO of the JD Group, as he testified in favour of his firm’s attempt to acquire Ellerines, one of his leading rivals, or the dyspeptic irritation exuded by Eric Ellerine, the legendary entrepreneur and hard-driving boss of Sussman’s eponymous acquisition target, as he sat and watched our proceedings disrupt his well-laid plans for exiting the company that he had founded half a century earlier. Our purpose was not to demonstrate our power. It was to interrogate thoroughly a merger notified to us. But a reputation for possessing power came in handy. It meant that we were treated with some respect after a very short while, and that our elusive objective – the instilling of a competition culture – slowly began to filter through to at least some sections of society.

I make something of this because it conflicts with the conventional wisdom regarding how new, resource-strapped and inexperienced agencies should approach the range of tasks before them. The advice from developed country agencies is generally to concentrate first on advocacy and anti-cartel enforcement, then to move on to monopolisation or abuse-of-dominance cases, and then to mergers. The rationale is that advocacy is easy, doesn’t involve the expenditure of massive resources and will contribute to creating a competition culture; while cartels are universally accepted to be the most egregious antitrust offences. Mergers involve complex economic analysis and rarely fall foul of competition law, hence they should rank last on the list of a young agency’s priorities.

This advice is generally wrong on all counts. Effective advocacy is not only difficult, but also requires the prior establishment of a reputation. And particularly as competition advocacy frequently requires challenging powerful interest groups, an early emphasis on advocacy may, if anyone takes any notice of it, serve to arouse powerful opponents prematurely. Cartels, for their part, may not involve complex law and economics, but they do involve extremely difficult detective work – leniency programmes will not assist until potential participants in these programmes fear that they will be caught – and, as the PPC cement case shows, the use (and misuse) of powerful but necessary investigative powers may be viewed as draconian and overkill until society understands why they are necessary. Abuse-of-dominance cases are technically complex, and presuppose going into battle with well-resourced opponents who have no interest in cooperation.

Of course the advice to prioritise mergers is not cast in stone. In a recent peer review of the Armenian competition authority in which I participated, it seemed reasonably clear that, given the past history of central planning, the most urgent task facing the competition authority was persuading other key government agencies and the legislature of its legitimacy. This may apply to many new authorities in the transition economies. But legitimacy has to be earned, and for a competition authority this generally involves establishing a degree of authority in relation to the business community. How to go about this inevitably requires an intimate knowledge of the social, political and economic context. However, I’m confident that in our circumstances the early emphasis on mergers did the trick.

But aside from its strategic and tactical significance to the competition authority, merger review is both interesting and, as we’ll see, on occasion quite dramatic. The interest lies principally in the huge range of markets the competition authorities are required to examine, with the advantage of having people intimately familiar with those markets falling over themselves to provide information. Their analyses – and, I fear, all too often, the information they provide – are, of course, tailored to support the decision they want from the Tribunal. In a contested merger, however, discovery and cross-examination (bolstered if necessary by the Tribunal’s inquisitorial powers) work their wonders. And so we sat there receiving information and analyses that academic researchers and investigative journalists would have killed for.

The vast majority of mergers that we heard were uncontroversial and thus quickly approved at our equivalent of a motion court hearing. But many of even these uncontested mergers were fascinating, providing, as they did, a fairly detailed account of anything from platinum to poultry. And because it was usually, though not invariably, easy to identify an unproblematic merger, one could decide on the level of detail with which to engage with an uncontested merger. Occasionally, however, the Tribunal’s reading of an apparently unproblematic merger revealed real competition problems.

There is a peculiar aspect to learning in this manner. Possibly because it is undertaken with such a specific purpose and at such an extraordinary level of detail in so truncated a period, it seems to be relatively easily forgotten, almost erased as soon as the next transaction comes along. The mergers of chemical firms, in which we were required to examine a large array of diverse, difficult and unimaginable products, often proved particularly taxing to decide and the learning impossible to retain beyond the drafting of the judgment. We heard a number of cases in the poultry market, where much of the analysis involved understanding the relationship between different variants of breeding stock, tracing parent and grandparent birds, all of which, for some or other peculiar reason, ultimately emanate from Scotland. That’s about all I remember about the poultry industry.

But, though it was difficult to retain this knowledge, these strange excursions were always fascinating and provided an extraordinary, if massively underutilised, source of informed research and analysis for industrial economists.

THE SOUTH AFRICAN MERGER REVIEW REGIME

I have briefly described our merger review regime. Its salient feature is that, in common with an increasing number of other competition jurisdictions, it is a compulsory pre-merger notification regime. That is to say, all mergers above a specified threshold – defined by turnover and assets – have to be notified to the Commission. Those below a second, higher threshold (‘intermediate mergers’) are investigated and decided by the Commission with the possibility of an appeal to the Tribunal. Those falling above this second threshold (‘large mergers’) are investigated by the Commission, which then makes a recommendation to the Tribunal. The Tribunal convenes a public hearing and then hands down a decision. A merger may not be consummated until approval has been received from the relevant authority.

When, in 2009, the lower threshold was fairly significantly increased, the law was amended to enable the Commission to investigate mergers falling below the threshold (‘small mergers’). These do not, however, have to be notified but, partly in order to encourage voluntary notification in those exceedingly rare instances where mergers of this size may fall foul of competition law, the competition authorities are empowered to take action against small mergers for up to one year after their consummation.

When the Commission is notified of an intended merger, the parties are also obliged to serve notice of their intention on their employees or their recognised unions. The Commission is obliged to advise the Minister of Economic Development that it has received a merger notification. This is purely to enable the employees and the minister to exercise their right to make submissions to the Competition Commission or the Competition Tribunal. Other third parties who wish to intervene in Tribunal hearings have to apply for rights of intervention. However, the ministerial entitlement to make submissions to Tribunal hearings does not imply any further ministerial influence over the decision-making process and would be made, like any other, in an open public hearing. In fact, the Department of Trade and Industry, which became increasingly dysfunctional for most of the period I served on the Tribunal, has played hardly any role in the merger review process. Even in those instances when policy issues were potentially affected by merger decisions, the department has only ever intervened and made representation at the specific request of the competition authorities.

The time frames governing merger reviews are tight.1 In the case of an intermediate merger, the Commission has 20 working days in which to arrive at a decision. In the case of a large merger, the Commission has 40 working days. Given its notorious inability to meet the most important deadlines, the Department of Trade and Industry’s difficulties in participating in merger review were undoubtedly exacerbated by these time frames.

The Competition Act provides that the decision-maker in merger reviews has to decide whether or not the contemplated transaction is likely to lead to a substantial lessening of competition. Section 12A of the Act specifies a non-exhaustive list of criteria that are to be applied in arriving at this decision. If it is concluded that the transaction is likely to lead to a substantial lessening of competition, the decision-maker is then required to assess the efficiency consequences of the merger. If a positive merger-specific impact on efficiency is found, the negative effect on competition has to be assessed against the positive impact on efficiency, with the possibility that the latter may outweigh the former, resulting in the possible approval of an anticompetitive merger. This is an extraordinarily difficult balance – in 10 years, only one anticompetitive merger has been approved on pro-efficiency grounds. I’ll deal with this transaction – the Trident Steel transaction – more fully later in this chapter.

Finally, and most controversially, whatever the outcome of the competition analysis and, if necessary, the efficiency analysis, the decision-maker is required to examine the impact of the transaction on public interest grounds. Having taken all these steps, the decision-maker is then obliged either to prohibit the transaction or to approve the merger, with or without conditions.

The insertion of public interest into the merger decision-making process proved to be the most controversial element of our legislation, at least in the minds of antitrust professionals. However, in practice it rarely proved to be particularly challenging.

A merging party that is aggrieved by a decision of the Tribunal is entitled to appeal to the Competition Appeal Court. The Commission has no right of appeal.

THE MERGER REVIEW: FORM AND SUBSTANCE

Much of the rest of this chapter, as well as the following chapter, consists of a description and analysis of a selection of key decisions of the Tribunal and also the responses of the Competition Appeal Court.

Our compulsory pre-merger notification regime means that hundreds of mergers have been reviewed, with the ‘large mergers’ having to be decided after a public hearing before the Tribunal. There is no sound way of categorising them in order of their ‘significance’ or ‘importance’ and, even if we were to do so, it would still leave us with a list of merger decisions too large for the examination of each transaction.

The only way to approach the task of selection, in the limited space available, is to identify the key learnings derived from the review process, and then illustrate them by reference to a small sample of investigations and decisions of the competition authorities.

A Change in Control

I am not going to deal in any detail with what rapidly became the most – arguably the only – complex legal issue in merger review: deciding when a change of control had taken place. This is the prior jurisdictional fact that must exist in order for a transaction to constitute a notifiable merger. In the overwhelming majority of mergers there is no dispute regarding whether or not a change of control has taken place.

But in a handful of cases there are question marks and they have to be dealt with correctly because, if decided incorrectly, they will immediately impact on the notification of future mergers, indeed potentially on the manner in which future transactions are structured so as to avoid notification. So the question of control has given rise to several fascinating cases and decisions. Key aspects of the reasoning involved in most decisions were ‘competition-specific’: that is to say, we were deciding whether a change of control had taken place from the point of view of a body charged with examining the competition implications of that change of control. Most, despite the extent of legal reasoning and argument involved, also rested on an interpretation of the facts of the specific case. And, of course, each decision naturally hinged on an interpretation of the Competition Act. So while all this meant that there was plenty of room for a barefoot lawyer like myself to add his tuppence worth (in fact, the temptation to get involved in the nuances and complexities of the fascinating legal disputes was irresistible), there is no gainsaying that when dealing with control questions we were in hardcore legal territory. Enter Norman Manoim.

Norman was always willing – delighted I think – to take up the challenge. An extremely modest, almost self-effacing person, it was on this terrain, I always felt, that he demonstrated his huge talent as a lawyer. I clearly recall the number of occasions on which we sat deciding questions of control, invariably, given the nature of the issue, with some of the finest legal minds in the country arrayed before us, and thinking to myself, ‘This issue is very complex and you’re all very clever, but the cleverest lawyer in the room is sitting next to me and thank God for that’. The upshot is that there are few, if any, decisions on control that were not drafted by Norman. There are a surprising number of decisions on control – every occasion on which we decided a dispute over control I thought that the entire issue had been definitively settled, but given the particular bent of the legal mind and the incentives to which it responds, I don’t doubt that there are many yet to settle.

The problem, of course, is that it is not only those in front of us with whom we had to contend, but also those above us. A number of important decisions on control were overturned by the Competition Appeal Court, with far-reaching consequences for the matters in question and, potentially, for important elements of the merger review system itself. I refer particularly to two decisions of the Competition Appeal Court. The first was the decision upholding Goldfields’ appeal in the course of its protracted efforts to resist a hostile takeover by Harmony, a decision that obliged Harmony to notify a merger and thus subject itself to a merger review.2 The second is a more recent decision in a matter involving Primedia and its acquisition of shares in a radio station.3 Again, the consequence of the Competition Appeal Court’s decision was to oblige Primedia to notify a merger. On each occasion the Tribunal had decided that a change of control had not been effected by the alleged ‘acquiring party’ and accordingly that the requirement to notify had not been triggered.

Reasonable people may disagree on this complex issue of control. However, these Competition Appeal Court decisions are quite wrong on any reasonable interpretation. The first – the Harmony–Goldfields decision – effectively decided that Harmony had acquired control of Goldfields and was thus obliged to notify the transaction and subject itself to a competition review, when no discernible threshold of control had been crossed. It is fortunately not merely wrong, but it is also wholly unintelligible, and so its loud and unruly birth seems to have been followed by an unobtrusive death. It is thus unlikely to impact significantly on the merger review system, although at some stage it will doubtless be invoked in support of some or other interest. At that stage the Competition Appeal Court may have to admit that it erred in Harmony–Goldfields and set the law straight. Ultimately Harmony is interesting only because it illustrates the successful abuse of the regulatory process by a target firm in a hostile merger.

The Primedia decision is all too intelligible. It effectively requires that when Firm A assumes control over Firm B, thus triggering a merger notification and competition investigation, there also has to be a competition analysis done of the acquisition by A of B’s interest in firms C and D, even if B did not enjoy control of those firms. The implications of this, particularly for the workload of the Commission and for the parties preparing merger filings, are vast. The Competition Appeal Court, having decided that a change in control had been effected, remitted the Primedia matter to the Tribunal and instructed us to carry out a merger evaluation. This we did – finding no substantial lessening of competition – with Norman adding a postscript effectively outlining the errors in the Competition Appeal Court’s approach to control. Given Norman’s deep respect for the practices of his profession, including the hierarchies of courts and their various officers, his decision to engage in this unusual debate with the Competition Appeal Court is a measure of just how wrong he thought – and he is undoubtedly correct – the Court had got the control issue in Primedia.

So having said that I would not discuss control, I have been drawn into a discussion of an issue too intellectually intriguing to ignore. If I were to distil lessons from our experience with the vexed and difficult question of a change of control, I think it would be to take an expansive view of a change in control to limit the possibility of parties structuring a transaction so as to avoid notification when there has actually been a change in control. This, I believe, is the perspective that we have consistently adopted. Ironically however, the Competition Appeal Court, far from limiting our expansiveness, has, in this area if in no other, amplified it so as to divine changes in control where none has occurred.

The second and most important lesson is to have a lawyer of Norman’s calibre on your side.

THE MERGER REVIEW PROCESS

Our merger review process is unusually, if not uniquely, transparent. This is particularly so in the case of large mergers, which are adjudicated in an open Tribunal hearing. An opposed merger hearing is preceded by a preparatory meeting – a ‘pre-hearing’ – which is not open to the public. The merger may be opposed by the Commission or interested parties who have revealed their concerns in the course of the Commission’s investigation; these are then placed in the record filed by the Commission.

The merging parties and the Commission are required to attend the pre-hearing and an invitation to attend is generally extended to others who have expressed strong views on – usually objections to – the merger. The pre-hearing will decide on access to the record filed by the Commission (which inevitably contains documents over which confidentiality has been claimed); it will identify additional documents that are required and timetables for exchanging them; it will identify parties other than the Commission and the merging parties who wish to participate in the hearings (the interveners) and the scope of their intervention; it will arrange for the exchange of witness lists and, where necessary, witness statements; and witnesses who have to be subpoenaed may be identified. If any of these arrangements are contested – that is, confidentiality, discovery of documents and intervention – then a formal hearing will be convened for their determination.

Once all these pre-hearing matters have been cleared up, the chair of the Tribunal appoints a panel of three members to hear the matter. The period between the pre-hearing meeting and the start of the actual hearing may be relatively brief and painless, although arguments over intervention and discovery, the former potentially appealable, may take some time.

Because time is usually an important consideration for the merging parties, they are generally well-advised to cooperate with requests for discovery, access to confidential documents and intervention. Indeed, if their opposition to, say, an application for intervention is upheld by the Tribunal, this may well result in a review or appeal to the Competition Appeal Court and conceivably even the Supreme Court of Appeal, thus providing obstructive interveners or targets of hostile takeovers with endless opportunities for harassment or green mail. I’ll return to this below.

As I’ve already elaborated, merger hearings, though often tense and bitterly contested affairs, are inquiries rather than adversarial trials. That is to say, there is no onus on any participant to prove or disprove any claim or contention: it is for the Tribunal to decide, on the evidence and argument placed before it, whether or not the merger is likely to lead to a substantial lessening of competition. However, from a due process and fairness perspective, the key rules that characterise adversarial hearings are strictly adhered to. For example, in contrast to many other administrative systems of merger control, there are no ex parte communications between the decision-maker – the Tribunal panel – and any of the participating parties.

The hearings are conventionally-structured affairs. The Commission is given the opportunity to state its position and call its witnesses. It is followed by the intervening parties. And finally the merging parties are given the opportunity to state their case for the merger. An opposed merger hearing can last for anything between a morning and – in the largest and most exceptional cases – several weeks. The Sasol–Engen and Telkom–BCX hearings each occupied 19 hearing days. At the end of the hearing the participants are given the opportunity to present their arguments. Once this is concluded, the Tribunal is required to submit its decision, accompanied by the reasons for the decision, within 30 working days.

Merger hearings have been closely followed in the media. The courtroom style of the proceedings, replete with the drama of cross-examination, is compelling; the amount of information revealed about the participating companies is unusual; the ready access to corporate leaders is useful; and seeing them on the proverbial carpet is, shall we say, levelling. The story has human interest, a dramatic aspect, a deeply analytic aspect and some very accessible information and reportage material. It is not difficult for a Tribunal member to appreciate the basis for the intense interest displayed by those journalists who followed our work so closely. We were after the same things: reliable information, a good storyline and a coherent analysis. I would vouch that on a good day there was a small handful of journalists in that hearing room who knew as much about the intricacies of competition law as any of their peers anywhere in the world.

Openness and accessibility were the defining elements of our hearings. These were reflected most clearly in our impatience with excessive confidentiality claims and in our, possibly excessive, willingness to hear well-nigh anybody who wanted to be heard.

Confidentiality

There is a tiresome obsession in antitrust circles with the importance of confidentiality. While I am happy to acknowledge and respect legitimate business secrets that require protection, not least of all to limit cooperation and maintain competition, I am convinced that a large part of the claim to secrecy has to do with the puerile ‘laddishness’ of business advisers. Treating a client’s information as one would some deadly military secret is on a par with the dramatic code-names and battlefield analogies that characterise business strategy documents – lest it be thought otherwise, this language and the claims of secrecy strongly suggest that these are not merely besuited nerds, but rather latter-day Napoleons!

Many of the more outlandish confidentiality claims also clearly stem from overworked junior lawyers. As the clock approaches midnight – and I have never encountered a profession that works so close to deadline – the task of working through a mountain of files distinguishing actual business secrets from information that does not meet that standard, that on many occasions was actually found to be in the public domain, clearly proves too taxing. And so confidentiality is claimed by default and the buck for establishing genuine confidentiality is passed on to the formal process.

The problem of course is that, the importance of transparency in relation to the media and the general public aside, first principles dictate that it is simply not possible to conduct a fair decision-making process when one party relies on evidence not made available to its opponents. And this is not changed by the mere fact that one’s courtroom opponents are frequently, in competition matters, one’s competitors. The first task is then to sift out the information that is not confidential from genuine business secrets. The former frequently includes not only information in the public domain, but particularly covered information that, while not secret, serves to strengthen the opponent’s case – precisely the sort of information that needs to be exposed to sunlight and cross-examination.

When this task is carried out diligently, what is left is a much-reduced confidentiality claim, covering a small number of genuine business secrets. A modality has then to be agreed whereby even this information is made available, at least to the extent required by the exigencies of running a fair inquiry. We have duly worked out a broadly acceptable approach, although we have continued to be dogged – both at the pre-hearing and hearing stages – by excessive claims of confidentiality. We are frequently vindicated by witnesses, senior executives of the firm on whose behalf confidentiality has been claimed over one or other piece of information, simply brushing aside the claim from the witness box.

I am left, though, with the firm view that competition lawyers make far too much of confidentiality and are often abetted by national competition agencies – for example, the antitrust division of the US Department of Justice – that appear as anxious to protect confidentiality as the firms themselves. Of course, merger proceedings can be used as mechanisms for firms to unearth the competitive secrets of their rivals, and indeed as the basis for building cooperative relationships with their competitors, even if the merger does not, for one reason or another, actually take place. But this concern is significantly overstated.

Third-party Interventions

Third-party intervention in merger proceedings is a more complex issue. Again, a permissive approach to intervention is driven by principles of transparency and fairness, fairness to those whose interests are affected by a particular transaction. But it is also driven by the unique quality of information and analysis that knowledgeable insiders are able to bring to inquiries as fact- and context-specific as merger inquiries. This was particularly so in the early years of the Commission’s life, when the quality of its investigations could often not – and understandably so, given its lack of experience – meet the requirements of an important merger or restrictive-practices inquiry.

One need look no further – although there are countless instances – than the Afrox–Mediclinic saga.4 This was a particularly egregious attempt by the merging parties to deceive the authorities, so egregious that one might have expected a slightly more sceptical mind (even if backed by poor investigative experience) to have uncovered the true nature of this transaction. Although I think that a measure of healthy scepticism would have enabled the deceit to be exposed in the Tribunal hearings, there can be no doubt that its full extent was exposed as a result of the intervention of Netcare, the largest competitor of the merging parties.

This was the first occasion on which I had been exposed to the wonders of discovery, wonderful because of what it reveals, and wonderful because lawyers actually cooperate, and oblige their clients to cooperate, in revealing documentary material massively damaging to their clients. We have certainly encountered instances of less than full cooperation with discovery, but these have been the exception rather than the rule. The true wonder of discovery, though, is the extent to which it is maintained by self-regulation, by the reputational damage that would accrue to a legal team in consequence of dishonesty in discovery, and by the prospect that anything less than full disclosure may, at some later stage – possibly in another encounter – be visited upon those who fail to obey the rules of the game. I have often had to sit through the tedium of adjudicating excessive discovery claims, but the burden was considerably lightened by the knowledge that the process was not only vital to successful adjudication, but that it exemplified the benefits of cooperation with the rules of the game over the anarchy of lawlessness – over, one might say, thoroughly unregulated competition between contending forces engaged in a process of adjudication.

An adjudicator has to adopt an extremely cautious and sceptical approach to third-party intervention, particularly in merger cases where the most energetic, invasive and informative intervener is generally a competitor of the merging parties. In the Afrox–Mediclinic matter, the intervener, Netcare, was not merely an intervener defending its own commercial interest, but an extremely angry, vengeful competitor who had effectively been denied the equally anticompetitive prize that it was now seeking to prevent its competitor from claiming. We have come to know Netcare well. I have little doubt that, given half a chance, its combination of notoriously aggressive attorneys, advisers and executives would not have hesitated to employ precisely the same tactics as those employed by Mediclinic. But this was not about moral high ground or low ground. For Netcare it was about red-blooded competition and for us it was about access to critical information and insights. On this occasion Netcare and the competition authorities – indeed competition itself – were the undisputed victors. I can’t say that this happy outcome was always achieved in our dealings with Netcare.

I could recount many more occasions on which intervention has accounted for our ability to take the correct decision in merger cases. We could not have successfully adjudicated the Sasol–Engen matter without the intervention of the other oil companies.5 In this instance it was not so much deceitfulness on the part of the merging parties that was exposed by the interveners – although there was some of that too – but rather the inordinate technical complexity of the markets we were examining. It’s instructive that in both of these instances the Commission recommended approval of the transactions – in Afrox–Mediclinic the Commission actually stuck to its patently flawed recommendation, while in Sasol–Engen, the greater strategic and tactical nous and self-confidence of its acting commissioner, Shan Ramburuth, and its head of legal services, Thembinkosi Bonakele, enabled the Commission to withdraw its approval at the argument stage.

My point is not to belittle the Commission, nor to criticise its decisions (although its recommendation in Afrox–Mediclinic ranks with the worst decisions it has ever taken). I rather want to convey the superior quality of the knowledge acquired through discovery and cross-examination in a public inquiry, over that acquired through private encounters with individual parties, even given the investigatory powers of the Commission. These powers are the Commission’s investigatory instruments and, in the intervening years since these two matters, they have honed their techniques and so, as I shall elaborate, a permissive approach to intervention becomes less essential, even counter-productive. However, in many cases, the Tribunal processes will reveal information and insights that are not exposed through the Commission’s investigatory processes and, where that happens, the Commission must, as in the Sasol–Engen matter, be willing to shift its stance without concerning itself with loss of face or allegations of indecisiveness.

As the Commission’s investigatory prowess has improved, the utility of permissive intervention has decreased and its dangers have increased concomitantly. The danger is not so much that interveners, particularly those who are competitors, will provide self-interested information and analyses, but rather that they will use intervention as a mechanism for delaying and obstructing transactions in which time is often extremely costly. Recent years have been marked by interventions that have not contributed an iota of useful insight to the adjudicators, but have simply served to harass their competitors.

Consider Caxton’s intervention in the M-Net–SuperSport transaction.6 I won’t describe this case in any detail. Suffice it to say that Caxton applied to intervene in a transaction involving its arch nemesis, Naspers. It’s a well-known rivalry that, on Caxton’s part certainly, borders on the obsessive and manifestly resides in Caxton’s exclusion, many years ago, from participation in South Africa’s first pay-television licence. This background should have immediately alerted the Tribunal to Caxton’s true purpose in intervening. But our permissive, liberal attitude prevailed. We granted Caxton rights of intervention but limited the scope of this intervention. It then predictably appealed the limitation that the Tribunal had imposed on its scope. Its appeal was rejected as was its petition to appeal further to the Supreme Court of Appeal. Eventually, it ran out of options and the hearing went ahead with Caxton in tow, having successfully delayed the transaction and raised the costs for the merging parties. Its intervention brought nothing of value to our decision-making process. This was pure harassment – Naspers was, as far as I could see, under no time-pressure to conclude the transaction and so was not about to walk away from it. For the sheer pleasure of irritating Naspers, Caxton wasted its shareholders’ money and obliged Naspers to do the same. We vigorously debated imposing costs on Caxton but regrettably decided against even this mild sanction.

This is by no means the only instance of vexatious intervention – other particularly egregious instances include Altech’s intervention in the MTN–Verizon transaction,7 and AME’s intervention in the Primedia–Khaya FM merger.8 But it’s illustrative of a well-recognised ploy in which the processes of an adjudicative body, and particularly that body’s respect for the right of corporate and other citizens to participate in administrative decisions that appear to affect their interests, are grossly abused. In these circumstances, not only should the firms have had an adverse-costs order imposed upon them, but on the next occasion on which they intervene they should have a higher bar placed on their intervention than would normally be the case. They should be required to satisfy the Tribunal of the merits of the intervention application by being compelled to provide a detailed account of the evidence they intend placing before the Tribunal. Failure to do so should result in a rejection of their application to intervene. Intervention should be viewed not as a right of the would-be intervener, but as a means of assisting the Tribunal. This is precisely why the Competition Appeal Court has held that a decision to allow intervention is at the discretion of the Tribunal. Those who wish to make a statement before the Tribunal can always request permission to do so and they are very unlikely to be turned down. The Tribunal for its part can subpoena witnesses whom it wants to hear. But there is, particularly in reasonably straightforward mergers, generally little reason to grant formal rights of intervention, with all that that implies.

Hostile Mergers

The strategic use of litigation is both more obstructive and more difficult to prevent in the case of hostile mergers. In these instances, where timing is all-important, the party with an interest in employing the regulator as a mechanism for delay is the reluctant target of the merger. It is significantly more difficult to limit the participation of a party to a merger than the participation of a third-party intervener.

But it’s not impossible. A merger hearing is an inquiry into whether or not the prospective transaction is likely to lead to a substantial lessening of competition. The Tribunal is under no obligation to hear evidence from anybody, including a party to the merger, that does not assist it in answering that question. There can be no doubt that, particularly in these circumstances, a refusal to hear irrelevant evidence will give rise to appeals and reviews which may well serve the dilatory objectives of the party that wishes to submit the irrelevant evidence. But that is not the Tribunal’s concern. Its concern is to prevent abuse of its own processes. If the processes of one or all of the High Court, the Competition Appeal Court, the Supreme Court of Appeal or the Constitutional Court are abused, then that is for them to address. It is not for the Tribunal to permit an abuse of our processes because of a fear that an adverse decision will initiate a conveniently time-consuming round of appeals. While reviews may initially result from the Tribunal adopting a more uncompromising stance in relation to dilatory and vexatious litigation, I’m convinced that were the Tribunal to take a stronger stance the ultimate outcome would be to limit this sort of conduct.

We have heard one hostile merger – the attempt by Harmony to merge with Goldfields.9 We spent several days hearing evidence largely submitted by Goldfields. It was clear, even before the start of the hearing, that there was no prospect of Goldfields, in its unaccustomed role as competition champion, establishing that a merger between it and Harmony had any implications for competition, let alone that it portended a likely and substantial lessening of competition. Had the merger been a friendly transaction, it would have occupied all of 15 minutes on a Wednesday morning. But simply because it was hostile, we sat and listened to a range of empty arguments that everybody in the room knew were only adduced in order to present a veneer, the appearance, of a merger inquiry mandated by the Competition Act. In fact, so nonsensical was it, that I lost it at having to listen to ‘evidence’ presented by an expert witness called by Goldfields, a respected economist from Stellenbosch University, who cheerily explained how the merger portended the likely collapse of the South African economy, and in the process I came as close as I ever did to having to hear an application for my recusal on the grounds of bias.

There was but one pleasurable moment in this whole unedifying affair and, though I digress, it bears recounting. Much of the opposition to Harmony’s audacious bid for Goldfields was, for want of a better word, ‘cultural’. Goldfields was a pillar of old, Anglophone money in the South African gold mining industry, a 120-year-old remnant of the Rhodes empire. Harmony, on the other hand, was led by Bernard Swanepoel, as far as I could tell the only executive in the South African gold mining industry who may have been described as an entrepreneur. But he was an Afrikaans arriviste from Virginia, a small town in the heart of the Free State goldfields. And, in taking on the establishment, he appeared to delight in camping up his ‘otherness’, which included an extremely unrefined Afrikaans accent.

Jeremy Gauntlett was Goldfields’ senior counsel. He is certainly one of South Africa’s most able, if not necessarily best loved, advocates. One of the great pleasures of Tribunal membership was seeing, at extremely close quarters and over a long period, some of South Africa’s best advocates in action, with Gauntlett undeniably in the top rank of those whom I regularly encountered. I can think of many complimentary terms with which to describe Jeremy: ‘brilliant’, ‘great sense of humour’ and ‘charming’ would all describe him. ‘Nice’ and ‘humble’ would not (although in fairness to Gauntlett, the only eminent silk that I ever encountered who managed to combine outstanding advocacy with humility was Arnold Subel).

In my experience, Gauntlett’s particular skill is in the art of cross-examination, where he closely tailors his style to the personality of the individual witness. I have observed him using a sharp knife to remove thin slices from the liver of his victims. It’s not a pretty sight, but it’s memorable and highly effective. I have also seen him be affable and charming in cross-examination, although frequently as a prelude to bringing out the knife.

On this occasion, a Sunday morning, with Swanepoel having flown in from Europe in time for the hearing, I watched Gauntlett observing Swanepoel as he was being led by his counsel, David Unterhalter, another scintillatingly brilliant advocate. Gauntlett’s mode of observation made a powerful impression on me. It was like watching a great portrait photographer line up a shot; through slightly narrowed eyes, he took Swanepoel’s psychological portrait.

And then the cross-examination began. I write from memory but I am confident of my recall of what turned out to be a riveting bit of theatre. Gauntlett was just getting into his stride – ‘affable with menace’ is how I would describe his opening gambit – when he was interrupted by Swanepoel turning to me and, in his deepest Afrikaans accent, asking, ‘Sir, may I know the name of the person who is questioning me?’ I replied, ‘This is Jeremy Gauntlett’. Swanepoel then removed a pen from his jacket pocket, although he somehow made it look as if he was retrieving a pencil stub from behind his ear, and, after asking me to spell Gauntlett’s name, laboriously wrote it down on what looked like his airplane boarding pass. Swanepoel then asked, ‘And what shall I call him?’ I replied, ‘Mr Gauntlett’, while Jeremy trilled, ‘Oh anything vaguely polite will do’. Swanepoel leant back in his seat and said, ‘Ja, I like to know the names of people who ask me questions’.

I have little recollection of the rest of the cross-examination. There was no point – the battle was over, with Swanepoel the undisputed victor. In fact, by then I think that everyone knew that the war for Goldfields was over, with Goldfields the victor. But the war had been won by and for the Goldfields management, employing the most deplorable, if utterly predictable, litigation strategy. This particular battle, on the other hand, had been won with great panache, in a perfect rendition of why there was little place for someone of Swanepoel’s character in the gold mining industry – at the end of the nineteenth century maybe, but no longer at the beginning of the twenty-first century.

There are, in the Harmony–Goldfields matter and in this cameo, substantive lessons for the whole business of merger review. We worry about the impact of mergers on competition, and so we should. But for the most part, mergers are driven, not by considerations of competition or efficiency, but by the large egos of entrepreneurs and corporate hacks wishing to extend or retain their empires, and by the fees of investment bankers. Which is another good reason for an early focus on merger review, for using it as a platform for capacity-building: an occasional early mistake, an occasional prohibition of a transaction that more experienced merger regulators would have approved, impacts on a few individual egos and bank balances but has little influence on the economic welfare of the firms involved, much less the country and its citizens.

THE SUBSTANTIVE REVIEW

As outlined, when reviewing mergers, the competition authorities are required to consider three sets of criteria. First, we ask whether the merger is likely to lead to a substantial lessening of competition (SLC). Second, in the event that we do find a substantial lessening of competition, we are mandated to investigate whether or not efficiency gains resulting from the merger outweigh the negative impact on competition. Third, we are required to examine the merger’s impact on a defined set of public interest criteria.

Unlike the efficiency test, the public interest test is undertaken regardless of the finding of the merger’s impact on competition. That is, should we find an SLC, it is conceivable that this may be countervailed by a positive impact on the public interest. On the other hand, it is possible that finding there is no SLC may be countervailed by a negative impact on the public interest.

Once these determinations have been made and the balancing between the various sets of criteria undertaken, we must either prohibit the merger, or approve it, with or without conditions, these being designed to mitigate a negative impact on competition or the public interest.

While in the antitrust community it was the introduction of the public interest criteria in our merger review that attracted the most attention, South African business appeared to experience real difficulty in grasping the nettle of merger review in any shape or form. Certainly it was viewed as an unwelcome political intervention in an important area of business decision-making.

Particularly aggrieved were the investment analysts, that most deeply self-interested, most parochial part of the business community. They simply couldn’t come to terms with the notion that a deal which they had put together – and investment banks appear to be the initiators of a great many of the large transactions that came before us – and on which, if successful, they stood to earn significant fees, could be subject to regulatory oversight, no matter that it had long been a critical factor in business decision-making in the US and Europe. As far as they were concerned, any decision taken by the competition authorities that was not to their liking was ‘politically’ inspired. I particularly recall them railing against our decision to prohibit the proposed transaction between two furniture retailers, our first prohibition, which they insisted had been taken on employment grounds, even though a glance at the reasons for our decisions would have quickly confirmed that employment factors played no role in our decision whatsoever.

Nor were they the only ones who had difficulty accepting our merger review powers. I recall Cedric Savage, the CEO of Tongaat-Hulett, a sort of corporate variety of a Natal sugar baron, publicly threatening that his company would reconsider future investment in South Africa in the event that we rejected Tongaat-Hulett’s proposed acquisition of TSB Sugar, one of its only two rivals. We blocked the transaction and, needless to say, the sky didn’t fall in.

A Substantial Lessening of Competition

This is going to have to be a very selective tour through the huge bulk of merger review investigations and decisions by the Commission and the Tribunal. The task of selection is eased by the fact that we have followed a pretty standard approach to evaluating the competition implications of a merger. The major distinction in our approach lies in our role with respect to public interest. I’ll deal with that later.

For now, I’ll focus on the competition evaluation. I’ll draw for the most part on decisions in which I was directly involved. My direct involvement would have been as chair of the panels of which I was a member and as the drafter of many of those panels’ decisions in opposed mergers. This selection is not as egocentric as it sounds. It implies nothing about the relative importance of the decisions taken by the various panels. Indeed, because our merger decisions have been pretty consistent, it really doesn’t make a lot of difference which ones are chosen to illustrate the key points I want to highlight. Also, the conduct of the hearings and the drafting of the decisions were pretty collegial affairs, particularly between the full-time members, Norman Manoim, Yasmin Carrim, my esteemed colleague and friend, and me. This is not to say that the part-timers did not play an important, sometimes critical, role. But in the nature of things they had less time to be involved in the preparation of hearings and the drafting of the decisions, hence I stress the collegiality between the full-time members.

While speaking of collegiality, it is interesting that in 10 years, there has been only one dissenting decision, and that by Christine Qunta in a relatively trivial point in an early complaint referral.10 I’ve often wondered about this, wondered whether there is any inference – positive or negative – to be drawn from this. Does it reflect too great a uniformity, too great a willingness to compromise, too much importance attached to maintaining a ‘united front’? Or does it reflect the sort of intense debate that produced agreement between people who deeply respected each other’s views? I wonder whether the lawyers have believed that the composition of a panel, and particularly the identity of the full-time members, materially impacted on the outcome of a case.

On balance, I think not, because what I mean by collegiality is that although one panellist was assigned responsibility for drafting, the others were very closely consulted. Predictably, consultation between the full-timers serving on the panel – and for most of the large, time-consuming matters I generally had to empanel at least two full-timers – was particularly close.

Collegiality does not mean that there were not, on many occasions, stinging disagreements. However, these only rarely concerned substantive matters: Should we approve a merger? Should we find a contravention of the Act? How should we decide a control dispute? These questions were settled relatively easily and amicably among ourselves. Our disagreements – particularly between Norman and me – rather concerned issues relating to how to deal with dilatory conduct and with blatant point-taking, how to deal with intervention applications, costs orders and the like. As is I think well known, I was more easily enraged than were my colleagues by the opportunism that most lawyers displayed on one occasion or another. I loathed witnesses who blatantly lied to us, and particularly lawyers who, if not exactly complicit in those lies, were nonetheless prepared to construct a case on the basis of what their palpably dishonest clients and witnesses were telling us.

This is not to say that Norman condoned this conduct, but he, with Yasmin – ever tough-minded and ever able to find a middle path – somewhere in between, were more prepared, albeit reluctantly, to view the sort of conduct that enraged me as part of the accepted ways of the legal profession and the practice of law. (I recall my ever-wise deputy chair, Marumo Moerane, once telling me that if every lying witness was charged with perjury, the courts would do nothing other than hear charges of perjury.) On balance, though, I am not sorry that the first chair of the Tribunal was not a lawyer. I think that my relative lack of prior experience with, and therefore intense intolerance of, this conduct did have a little influence – and for the better – on the working style of the Tribunal.

But back to the matter at hand – the evaluation of the competition implications of mergers.

At the heart of our merger review regime is a determination of the likelihood of the merger resulting in a substantial lessening of competition. This is determined against a non-exhaustive list of factors specified in section 12A(2) of the Act, including the level (and potential level) of imports, barriers to entry, the history of collusion in the market in question, the ‘dynamic characteristics’ of the market including innovation and product differentiation, and the prospect of failure of either of the firms party to the transaction. And so, as in merger reviews almost anywhere, we began our analysis with a definition of the relevant market. And then, having done that – often the determinant element of a merger analysis – we proceeded to predict the likely outcome of the merger against those of the standard factors listed above that were pertinent to the case at hand.

I think that the major issues – and our approach to dealing with them – emerged quite early on in our merger cases. Predictably, this was the period in which we relied most heavily on the great wealth of international jurisprudence and scholarship and on the experience and quite remarkable generosity of the large, but very close and supportive, community of antitrust enforcers and scholars.

We were fortunate to get some interesting early merger cases that we were able to examine thoroughly, and to draft sound, comprehensive decisions. We were extremely fortunate that the most important of these decisions were not appealed and so they have set the standard and approach to which we have adhered ever since.

The key lessons that I drew from merger review – apart, that is, from the strategic lessons detailed earlier in this chapter – were largely related in one way or another to issues of evidence. In the oft-quoted words of Judge Richard Posner, merger review calls for ‘a predictive judgment, necessarily probabilistic and judgmental rather than demonstrable’.11 But this does not license an exercise in guesswork and crystal-ball gazing. On the contrary, merger analysis is extremely fact-intensive and is precisely concerned with predicting likely responses to changes and likely changes in the matrix of factual evidence, both exogenous changes, for example in tastes and technology, as well as those changes that occur in consequence of the merger itself. So solid evidence is critical, first in identifying the relevant markets and then in predicting the impact the merger will have on the conduct of participants in the restructured market. In order to make the latter decision – predicting the impact of the merger on the conduct of the participants – the decision-maker must rely, in part at least, on evidence of the current behaviour of participants in the market in order to predict how they will respond to the change in the structure and changes in the incentives that will be wrought by the merger.

However, massive analytical and evidentiary problems are soon encountered when evidence is sought to prove the standard tests applied in market definition. The clearest example of this is seen in the application of the universally used ‘SSNIP’ test, which purports to measure the consumer response to a ‘small but significant and non-transitory increase in price’, and so the boundaries of the relevant market. But, its simple elegance notwithstanding, in real life this is not an easy empirical test from which will emerge a definitive and dispositive number. The quality of available evidence simply does not allow for that. And even if it did, the time and resources required to gather the evidence would put it beyond the reach of merger review. The much-cited US Staples case arguably represents the most celebrated reported instance of the ‘hard’ empirics of market analysis.12 But Staples is the exception that proves the rule. Here, a massive quantity of reliable data was available in the form of checkout slips and an extremely resource-intensive econometric analysis was applied to those data. I don’t believe that this has ever been repeated on this scale in the US and it is certainly way beyond the resources of a South African competition authority.

So the first lesson learnt in merger analysis is that the standard empirical tests – such as the SSNIP test – have to be approached as thought experiments, as methodologies to which the available evidence, some quantitative, other qualitative and anecdotal, is creatively applied. I don’t believe that our Competition Appeal Court has ever appreciated this data limitation or, accordingly, the manner in which the test must be applied. The Court is still waiting for that magic number to pop out of a mechanical application of the test and to provide an answer that I fear can be supplied only by intelligent application of qualitative evidence and by looking into the eyes of the witnesses, which, of course, is why deference on the part of the appeal body to the fact-finding body is so obviously important. It is also a very basic indication of why the Court is required to understand the methodologies and thinking of economics, a discipline in which theories and models frequently do not easily translate into definitive answers.

Second, consumer behaviour is not only sensitive to changes in technology – notably in the form of new product development – and taste, but also to particular historical circumstances, national cultural factors and income distribution.

If I were to distil a single lesson from merger evaluation, it would centre on the rather unusual combination of the value to be derived from the general approaches so thoroughly traversed in international jurisprudence, together with the simultaneous importance of local historical, social and economic conditions.

Take, for example, the first merger that we ever prohibited, the proposed transaction between two of South Africa’s largest furniture retailers, the JD Group and Ellerines.13 For one thing, these two firms enjoyed an iconic status in South Africa that is unlikely to be accorded to mere furniture retailers elsewhere. The reason for this is that these were, until very recently, the principal source of credit for low-income black consumers. Indeed, furniture retailers – and Ellerines in particular – were the first firms to extend credit to black consumers, who, bear in mind, did not for the overwhelming part have access to banking services. This immediately accorded a degree of significance – and a particular set of facts to be considered – to this merger that would not have been relevant elsewhere.

When defining the relevant markets, the paucity of hard empirical evidence quickly led us to embrace the notion of ‘practical indicia’ accepted by the US Supreme Court in the Brown Shoe case as the evidentiary basis for identifying product markets.14 However, the indicia that we did use were highly particular to South African circumstances. Hence, in attempting to resolve the boundaries of the relevant market, we had to decide whether the particular nature of the customer base of each of the various branded national chain stores placed the chains directed at low-income consumers in a market separate from those chains directed at higher-income consumers, despite the functional substitutability of the products involved. Internal documents and clearly differentiated advertising strategies indicated that the merging parties – as well as the other chains – targeted particular chains at particular income groups. The distinctively targeted chains did not compete with one another; they rather competed with similarly targeted chains. However, one unusual but particularly revealing bit of evidence that helped establish the deep social and economic divisions among the targeted groups proved to be a promotional campaign mounted by the JD Group on behalf of two differently targeted chains: the low-income chain offered a free sheep for purchases above a specified amount; the high-income chain offered a coffee-table book on 101 ways to cook lamb for purchases above a specified amount!

In a similar vein, when determining the geographic market, we had to decide whether the large number of owner-managed ‘independent’ stores were in the same market as the large national chains. This could have had a significant impact on the outcome of the evaluation because it determined whether the geographic market was national, thus excluding ‘independents’. Alternatively, the geographic market might, as the Commission contended, have comprised a large array of local markets, each incorporating the ‘independents’. Clear documentary and oral evidence demonstrated that the national chains determined their key competitive strategies – including pricing, advertising and purchasing – on a national basis, which allowed local store managers limited latitude and hence a limited ability to respond to competitive conditions in local markets, even if this meant occasionally conceding sales to local stores offering promotional deals and the like.

While this clearly pointed in the direction of a national market comprising national chain stores, we had to consider the seemingly prosaic reality that the bulkiness of furniture and large home appliances prevented customers from turning to sources outside their local market. The Commission was sufficiently swayed by this reality to conclude that there were indeed a myriad of local markets. However, the evidence demonstrated that in South Africa, where historical circumstances dictate that a large number of low-income breadwinners still work and live considerable distances from their family homes, this was not necessarily true. The purchase would be effected at a store near to the breadwinner’s place of work where it was easy to make monthly payments, while the national character of the chains enabled the family to take delivery of the product from a branch store conveniently located near their residential area.

When examining the competition implications of a merger between these two large firms in the relevant market that we had identified, the fact that over 90 per cent of the customers purchased on credit loomed extremely large. This significantly raised barriers to entry. The credit issue took on even greater significance when the evidence showed that the other competing national retail chains were tightening up their credit-granting facilities, thus enhancing the competitive significance of the merging parties, which, it was generally acknowledged, were distinguished by their extensive, but tightly managed, debtors’ book. It also meant that customers, most of whom had no bank accounts and, frequently, whose only credit relationship was with a furniture chain, would tend to remain with the firm with which they had an established credit record. Thus, whereas the conventional wisdom held that in most mergers the merging parties would expect to lose a certain amount of the combined business to competing firms, this was very unlikely to occur in the South African low-income retail furniture market, where holders of that precious commodity, credit, were likely to remain with the retailer with whom they had an existing credit relationship. Hence the acquiring party was not merely purchasing considerable capacity, it was purchasing a largely loyal, even captive, customer base. This interplay between international jurisprudence and local conditions constantly reappeared.15

For obvious reasons, among the greatest challenges confronting antitrust analysis is the dynamism of the environment surrounding the market. This generates particular challenges in the case of inherently predictive merger analysis. There are few markets that are not subject to a degree of dynamism. So even a market as seemingly mature as furniture retail is subject to relatively rapid change. These changes may range from demand shocks – for example, sudden changes in macro-economic variables, or massive government spending on infrastructure – through to improvements in credit assessment mechanisms that may simultaneously enable the furniture retailers to expand their credit granting but will also bring new competitors into the credit-granting market.

However, the impact of dynamic market forces is particularly well documented in the case of those product markets subject to rapid technological change. We encountered this in our review of the proposed acquisition by Telkom, the state-controlled fixed-line telecommunications giant, of BCX, one of South Africa’s largest suppliers of information and communications technology. In this instance – as in a raft of mergers as well as enforcement actions in technologically dynamic sectors – the merger was prohibited not because of an apprehension that Telkom would dominate the new technology, but rather because the panel concluded that it would enable Telkom to retard the introduction of new technologies that threatened the dominance of fixed-line telephony, its key competitive asset.16

Unquestionably, the most difficult mergers to call are those in relatively new markets and those that are subject, for one reason or another, to an elaborate regulatory regime. Although one may well be entitled to imagine that the uncertainties of technological change render merger analysis in technologically dynamic product markets particularly complex, it is the vagaries of regulation that are most difficult to predict. Hence, in prohibiting a merger in the highly regulated sugar market, we heard evidence from a Department of Trade and Industry official of government’s intention to deregulate the market, something that some 10 years later is yet to occur.17 Similarly, in the Sasol–Engen merger we were told of government’s intention to deregulate the retail price of petrol. This too has not occurred, and nor does it appear likely to happen anytime soon.

I don’t think that we allowed promised deregulation – or, for that matter, its current existence – to determine the outcome of mergers in regulated markets, although in the sugar merger the parties argued that because regulation had, in their estimation (with which we largely concurred), effectively eliminated competition from the market, by definition we could therefore not find a substantial lessening of competition in a market where none existed. However, the regulatory environment and likely future changes in that environment certainly influenced our thinking. In particular we were careful to ensure that mergers in regulated markets did not determine possible post-regulation levels of competition. Hence, we accepted the Commission’s recommendation to prohibit the merger, a decision influenced by the assurance given by the Department of Trade and Industry (which opposed the merger) that the sugar market was to be deregulated.

Of course, the important point remains that the government’s failure to move on the deregulation of the sugar and the petrol markets exemplifies the disconnect between competition law and competition policy. So, to this day, despite our efforts to defend a competitive or potentially competitive outcome in the petroleum and sugar markets, competition remains stunted by government’s failure to carry out promised deregulation.

One of the most difficult and important mergers that we decided concerned a then small, embryonic market in the health sector, embryonic in the sense that the product was newly in the process of establishing itself in the broader and highly regulated health-care market. It carried the promise of extending health insurance to a lower-income rung and so was of enormous social significance, or expressed otherwise – and the pertinence of this alternative phrasing will become clear – ‘public interest’. This was the merger between Medicross and Prime Cure and concerned the provision of low-cost health insurance options.18

It’s difficult, in the space of a few paragraphs, to do justice to this merger. But it’s worth spending some time on it, as it illustrates some of the difficulties involved in dealing with dynamic markets. It’s also illustrative of some substantive differences in approach between the Competition Tribunal and the Competition Appeal Court as well as the latter’s treatment of its supervisory role with respect to the Tribunal.

South Africa has a deeply segmented health-care system comprising, on the one hand, a relatively well-resourced, comprehensive private health-care system with excess capacity and, on the other hand, an under-resourced and over-used public health-care system. This problem, although by no means uniquely South African, is rooted in income distribution and so is naturally exacerbated by South Africa’s unusually skewed distribution. This greatly limits the number of people who can afford the insurance premiums that effectively fund access to private health care. One possible solution, then, is for the sophisticated private health-care insurance system to develop insurance products that a significantly greater proportion of the population can afford, thus somewhat relaxing the burden imposed on public health care and increasing capacity utilisation in the private system.

However, affordable private health insurance is a near contradiction in terms, partly in consequence of the incentive structure that it supports. Stated most simply – and this admittedly glosses over many important issues – the purchaser of a costly insurance product is incentivised to ‘over-utilise’ insured services, and doctors and hospitals are incentivised to ‘over-treat’. This has spawned, largely at the initiative of the funders, a major new market in managed health care, the various products of which are designed to ‘manage’ or reduce costs, by exercising supervision over clinical decisions, by reducing the incentive on the part of health-care providers to ‘over-treat’, and by reducing the incentive on the part of the purchasers of insurance products to ‘over-utilise’.

Managed care as described above may well have succeeded in lowering the costs of health care, although I make no judgment on which part of the health value chain has captured the benefits of this cost reduction, if any. Assuming that managed care has actually reduced cost, in countries with flatter income distributions this may well have enabled a greater proportion of the population to afford private health-care insurance. However, in South African circumstances, the distance between those who can afford any form of private health insurance and those who cannot is so great that the best that managed health-care as a mechanism of cost reduction can achieve is to lower costs pertaining to that part of the population already insured. Its impact could not be sufficiently great as to swell the ranks of those for whom private health-care insurance is an affordable option.

In order to offer private insurance to income groups lower than those currently insured, at least two things had to happen. First, risk had to be capped (through, in the lexicon of the industry, the development of ‘capitated’ insurance products) and shared. Although there is a spectrum of possibilities, the most complete form this would take would be the payment of a fee to selected primary care providers – that is, to a network of general practitioners, who would then, in exchange for this ‘retainer’, agree to provide an array of specified medical services to the members of the scheme, thus shifting the risk of insuring primary care from the insurer to the health-care provider. For this to get off the ground, the doctors who signed up to the scheme had to be assured of a large fee income. In other words, the network of doctors signing up for the scheme had to be induced by sufficient ‘upside’ prospects to accept the ‘downside’ risk they were assuming.

The second prerequisite, therefore, is that a large employer, employing a large number of low-paid workers, had to sign on to capitated options, thus providing the scale necessary for these new low-income insurance products to get off the ground. The state was the obvious employer in question. While many of the mid- and higher-level state employees were members of private health-insurance schemes, the great bulk of low-paid state employees were not. Accordingly, the state was in the early stages of introducing a health-insurance scheme for state employees, one that would first seek to recruit membership by means of an inducement to those of its employees who were already members of a private scheme and ultimately would extend it to its lower-paid employees in the form of a capitated option.

We were then confronted with a proposed merger between two of the three firms offering a capitated health-insurance option. This by no means represented the universe of firms offering a variety of managed health-care products and services. But, as our judgment was at pains to point out, while general managed health-care of the sort already described may have contributed to lowering the cost of health-care, it did not, on its own, offer the possibility of an insurance product that low-income consumers could afford. Only capitation – or risk-sharing – offered that possibility. Accordingly, we defined the market as that for ‘capitated primary managed health-care products’. Certainly, capitation was a form of managed health-care and it presupposed elements commonly found in other managed health-care options, for example supervision over the clinical decisions of the health-care provider in order to suppress costs. But while all capitated options may have involved cost management, by no means did all managed health-care include capitation, and so managed health-care products did not, on their own, offer an insurance solution for low-income earners.

This merger posed particularly complex problems for ‘predictive’ merger analysis. The market was still very small. The low-income products were in the early stages of development, and although all presupposed capitation or risk-sharing, there was a variety of ways of achieving this. We had to be satisfied that the post-merger market structure encouraged innovation on the part of the participants. These features played a significant part in the Tribunal’s decision to prohibit the merger. ‘There is undoubtedly significant room and an urgent requirement for experimentation and innovation. We have little doubt that a significant merger in this embryonic market will slow the pace of innovation. It will reduce the number of alternative modes of provision on offer, and it will likely slow the pace at which new forms and concepts of low-income health care insurance are introduced’.19

The Tribunal took a strong view that this fluidity in the regulatory and market environment, and in the development of the product itself, dictated that we view any private conduct – to wit, the merger – that might stunt the development of innovative new products with particular caution and circumspection. This imperative was, in our estimation, strongly bolstered by the particular nature of the product. We were dealing with a product of huge potential social significance, of huge ‘public interest’, and this supported the view that we take a particularly cautious view of a private intervention that might inhibit the emergence of a dynamic, innovative market, indeed the creation of a new product.

Our view of the merger’s likely impact on competition was powerfully influenced by our conclusions regarding barriers to entry. Our interpretation of the evidence concluded that barriers to entry were unusual and high. A successful capitated option required a tightly organised network of doctors and a large number of ‘lives’ per doctor, so that in exchange for accepting the risk of capitation they were guaranteed significant fee income. In other words, a provider’s network capable of supporting successful capitation involved more than doctors simply signing on to membership of one of the many loose networks already in operation. It meant limiting the network to a relatively small number of doctors who would abide by the rules of the scheme – which would include intervention in their clinical decisions in order to hold down costs – in exchange for guaranteed fee income. The scale economies issue in turn presupposed that the scheme management enjoyed the confidence of those individuals and institutions – notably, the trade unions – in a position to recommend these products to large numbers of potential purchasers. In other words, entry into this market required the accumulation of considerable ‘social capital’.

For these and other reasons laid out in our decision we accepted the Commission’s recommendation and prohibited the merger. However, the Tribunal’s decision was reversed on appeal, so the merger was ultimately approved unconditionally.

I recall the despair that I felt when I first experienced a reversal by the Competition Appeal Court, despair that frequently turned to intense frustration when I saw the errors that were regularly made by the Court and its reluctance, several of its own dicta to the contrary notwithstanding, to defer to the Tribunal’s findings of fact. However, I soon came to view these as an occupational hazard of being a lower court and so I learnt to accept, with some degree of stoicism, even the most frustrating decisions of the Competition Appeal Court.

There were several grounds on which we were reversed and they are indicative of the rather mechanistic, unimaginative approach that the Competition Appeal Court has adopted in its task of building a new body of competition jurisprudence. It is also indicative of the lack of deference that the Competition Appeal Court accorded to the Tribunal as the fact-finder, despite assertions to the contrary.

First, the Court held that we had misdirected ourselves by adopting a ‘cautious and circumspect’ approach towards this merger as a result of conflating the public interest test we were mandated to undertake with our ‘primary task’ – that of deciding whether the merger was likely to substantially lessen competition – and, moreover, that the public interest ground we invoked did not fall within the ambit of that specified by the Act. The Court’s conclusion arose from our reference – in a preamble to our competition evaluation outlining conditions in the health-care sector – to the need for circumspection in dealing with a market in which ‘public interest’ is as deeply implicated as the one in which the product is concerned with access to health care.

It’s perfectly clear that this reference to the ‘public interest’ has no reference whatsoever to the public interest test mandated by the Competition Act’s merger review evaluation. Indeed, the final paragraph of the judgment explicitly states that public interest grounds as defined in the Act were neither raised nor considered. All that our earlier reference to ‘public interest’ is manifestly intended to convey is that, were we dealing with the market for silk scarves, we might have been less concerned with factors like the uncertain regulatory environment and the dynamic nature of the fashion industry. However, given that we were dealing with the question of promoting the access of the poor to health care – something of great public interest and concern – we were inclined to err on the side of circumspection and caution when evaluating private actions that might have the effect of limiting the development of a market for products that could enhance that access. I can only imagine that David Unterhalter and Alfred Cockrell – the exceedingly smart lawyers who appeared for the appellant – must have felt vindicated by their shared tendency to advance, when all else failed, oblique and inventive arguments. Although I don’t think that the Tribunal often fell for many of these ‘too smart by half’ arguments, on this occasion they hit the target. Someone – in fact, no fewer than three appeal court judges – had actually been blindsided by one of their more outlandish arguments!

This does, however, point to a deeper issue and that is the Competition Appeal Court’s refusal to take account of the social content of competition law. Not only is this mandated by the preamble to the Competition Act itself, but I would have thought that any reasonable reading of the character of competition law and policy would have dictated that decisions occasionally implicated broad policy questions. And what could be less controversial from a policy point of view than an argument which insists that, in dealing with access by the poor to health care in a country that has achieved a degree of notoriety in its failings in this area, particular ‘caution and circumspection’ is mandated, if not by the personal sensibilities of the decision-maker, then at least by the expressed concerns of the public? Instead the Court cited approvingly the appellants’ contention that ‘The mechanisms by which government intends to achieve its policy objectives and the place and role of the private sector in these objectives are issues which fall within the purview of Parliament and the executive; they are not issues that fall within the remit of the competition authorities’.20

The Court thus implicitly accepted the argument that the competition authority should have no regard to the character of the market under examination – a resource misallocation is a resource misallocation, whether or not it occurs in the market for silk scarves or in health care. And what consideration could be more pertinent to, more at the centre of, competition policy than ‘the place and role of the private sector’ in the attainment of socio-economic objectives?

Second, the Competition Appeal Court rejected our finding on the relevant market. The respective judgments of the Competition Tribunal and the Competition Appeal Court reveal some of the evidentiary issues that, I believe, are so central to antitrust analysis generally and merger analysis in particular. As noted, we found capitation or risk-sharing to be an essential element of an insurance product for low-income consumers. The merging parties argued that all ‘managed health care products for low-cost medical scheme options’ should form part of the relevant market. The Court accepted this despite the fact that there were no low-cost medical scheme options in existence that did not include capitation. There were three low-cost medical options of any significance in existence – those offered by the merging parties and one other – and each of these included capitation.

The Tribunal was criticised by the Competition Appeal Court because of ‘an unfortunate absence of a rigorous exercise to determine the scope and nature of the market. The Tribunal did not perform any of the traditional exercises used for determining the dimensions of product market [sic]. There is no analysis in its determination which [sic] sector to compare prices of competing products or the functionality of those products from a consumer perspective. No customer substitution test was performed: that is an analysis of price substitutability or functional substitutability’.21

I remarked earlier on the apparent unwillingness of the Court to view these tests as methodologies, as thought experiments, that are rarely capable of producing definitive empirical outcomes. This particular case represents a clear example of how difficult it is to apply these tests in the ‘rigorous’ manner suggested by elementary economic textbooks and, apparently, by the Competition Appeal Court. We were dealing with a new product for which price data and evidence of consumer substitution were not available. We had to be inventive and careful in our search for the evidence necessary to establish the relevant market. It was not going to come from the application of a simple arithmetic formula. And search we did, although, as it happened, in this instance we did not have to search very far for the pertinent evidence. It was, rather, persuading the Competition Appeal Court to accept the evidence that proved elusive.

We were significantly aided in our investigation of this new market by the fact that when the merging parties were asked in the standard filing to identify their competitors, the only firms they identified were those that provided a capitated managed-care option, omitting to mention a range of firms that provided general managed-care services but without the crucial element of risk-sharing or capitation. The competitors named were the two merging parties and one other firm, and then a handful of fringe players. And so, predictably, when, in their initial filings, the parties, aided by their experienced competition lawyers, identified the relevant market, they stated that ‘the merging parties are, in the broadest sense, competitors for the administration of capitated managed care options’ (my emphasis).

However, when this was subjected to the scrutiny of a Tribunal hearing, it proved to be an inconveniently narrow definition for the parties. Indeed, the transaction began to look precisely like the 3–2 merger that it so clearly was. And so, when they saw the prospect of a looming prohibition, the merging parties sought to broaden the definition by removing the adjective ‘capitated’ from their definition, thus bringing into the market all manner of firms that provided elements of ‘managed care’.

As the Tribunal judgment confirms, we considered the evidence and we did indeed decide, on an examination of all the evidence, that capitation was a critical element in the make-up of a low-cost insurance option. We did make it clear that we placed considerable weight on the parties’ initial view, which agreed with our assessment, particularly on the parties’ submission that confined the competitors they identified to those who provided capitated options.

The Court rejected this. It took the view that by assigning the weight that we did to the parties’ initial submissions, we effectively did not take account of the evidence amassed in the course of further testimony, testimony that, in the view of the Court, broadened the market from ‘capitated managed care’ to ‘managed care’.

A reading of our decision will confirm that, while we did give considerable weight to the parties’ initial submissions, we also considered evidence that ranged considerably beyond the initial findings but in our estimation confirmed the parties’ initial submissions. However, more than this, we outlined clearly why we deemed it appropriate to accord significant weight to the initial submissions. We would not necessarily assign a high weighting to a business person’s identification of a ‘relevant market’. This is a term of legal and economic art that we would not necessarily expect a business person to appreciate. But we would expect a business person to be able to identify his competitors. And where, as in this instance, those competitors are listed to include only those who provide capitated options, strong judicial notice should be taken. Where this is bolstered by their lawyers declaring that the relevant market is that for capitated managed-care options, the burden on them then conveniently broadening the market – and perforce the range and identity of competitors – is considerable.

There is much else that could be said about this decision and other decisions of the Competition Appeal Court. But I’ve illustrated the most important points that I want to make. The first is the Court’s unwillingness to pay heed to the social content of the competition law. This resulted in the Court’s decisions that were based on a very narrow interpretation of the letter of the law. Second, the Court showed very little deference to the Tribunal’s fact-finding role. Its decision was strongly influenced by an unrealistic notion of the character and quality of the economic evidence upon which our decisions had to be made and a concomitantly ill-informed understanding of the nature of economic reasoning. I’ve already outlined why I think that the Court is predisposed to errors of this sort.

Efficiencies

Where the competition evaluation concludes that a merger is likely to give rise to a substantial lessening of competition, section 12A(1)(a)(i) of the Competition Act obliges the decision-maker to determine ‘whether or not the merger is likely to result in any technological, efficiency or other pro-competitive gain which will be greater than, and offset, the effects of any prevention or lessening of competition, that may result or is likely to result from the merger, and would not likely be obtained if the merger is prevented’.

While the public interest test has attracted considerable interest and criticism, the efficiency defence is, if anything, more difficult to apply. This is despite the fact that countervailing efficiencies are commonly examined in merger evaluations. In some jurisdictions – the US for example – they are incorporated into the competition finding, while in other competition regimes they are treated as a defence against a finding of substantial lessening of competition. We have adopted the latter approach. The relevant section of the Act, which is cited above, is lifted directly from the Canadian legislation, where it has also proved controversial and difficult to apply.

An apparent premise underlying the consideration of efficiencies is, presumably, an abiding belief in the notion that a rational decision-maker in a firm – the responsible manager, or board of directors, or shareholders – would attempt or accept an acquisition only if it generated cognisable gains for the merged entity that were not attainable by the separate firms. These gains may be realised in a variety of ways – cost-cutting through a rationalisation of overlapping functions or through the ability to realise that economies of scale are the most common source of efficiencies claimed. Others, less frequently claimed, may include the opportunity to combine know-how manifest in the introduction of new or improved products. As we shall see, the source of the claimed efficiencies matters.

The very purpose of merger review suggests that these claimed efficiencies be considered when deciding whether to approve a transaction. After all, through merger review the competition authority defends a competitively-structured market, because it believes that this will promote efficient outcomes. Prohibition of an anticompetitive merger will limit the rise of dominant firms with their pricing power and ability to exclude rivals from the market; or it will reduce the likelihood of collusion; or it will reduce the incentive and ability to foreclose a rival’s access to inputs or customers. It would thus appear to defeat the key purpose of merger review if cognisable, countervailing efficiency gains generated by the merger were not put into the mix that determined the ultimate fate of the transaction.

That having been said, it is widely accepted that the extravagant efficiencies characteristically and confidently predicted are seldom realised. Certainly those that claim the most elaborate departures from existing products or processes – for example, ‘the marriage of content and delivery platforms’ to which mergers in the entertainment and communications sectors often laid claim – generally turn out to be hazy chimeras. And even those more sober managers who rely on gains most easily measured in advance – say, gains from the rationalisation of duplicated facilities – tend to underestimate the cost of combining diverse institutions, even those that produce the same product using similar technologies.

We also know that mergers are frequently not driven, and are certainly not initiated, by the prospect of efficiency gains. Many of the largest mergers are driven by considerations no more elevated than the obscene fees that the investment bankers who characteristically initiate and promote the mergers are pursuing. The structure and character of remuneration packages often explain the support by the managers of target firms for a merger, while the sheer ego of the managers of the acquiring firm, the desire to manage ever-larger empires that pay ever-larger salaries and bonuses, frequently explains the decision to embark on an acquisition. These realities are not altered by a rationalising framework that talks up the ‘synergies’ and prospects for cost reductions.

Nor is the balance easily made between the claimed efficiency gains – if any are indeed proved – and the efficiency losses occasioned by a likely lessening of competition. It might be assumed that, because we are expected to place efficiency gains and losses on both sides of the scale, the balancing act is straightforward. But this is very far from true. There is tremendous uncertainty on both sides of the scale – the efficiency losses occasioned by a likely substantial lessening of competition cannot be measured, and the claims of efficiency gains are unreliable at best. So, the scientific veneer notwithstanding, the efficiency balance is, in truth, no less a matter of pure judgment than the public interest balance. The key factor in weighing up both the countervailing efficiency gains (as with the public interest considerations) is a judgment call regarding the likelihood and the substantiality of a lessening of competition and, in my view certainly, the prospect that any plausibly claimed efficiency gains will find their way to the consumers of the product in question.

So while there may be intuitive good sense in weighing the efficiency gains to which the combination of the merging firms gives rise against the efficiency losses that result from a substantial lessening of competition, the efficiency gains requirement may well be impossible to administer. Judge Richard Posner notes: ‘Not only is the measurement of efficiency … an intractable subject for litigation; but an estimate of a challenged merger’s cost savings could not be utilised in determining the total economic effect of the merger unless an estimate was also made of the monopoly costs of the merger – and we simply do not know enough about the effect of marginal increases in the concentration ratio … to predict the price effects’.22

In other words, the efficiency defence in mergers suffers from the same shortcoming that bedevils several other complex economic concepts that appear to make sense in, and to belong to, antitrust legislation and enforcement, including, as I’ll presently show, the vexed issue of excessive pricing: these concepts are rarely capable of being effectively administered through an adjudicative process. Certainly, if they are to be administered – and if they are in the statute they can’t simply be ignored – it is difficult to operate within the black letter of the law, which, with no hint of the immense difficulties of translating economic theory into statutory language and courtroom practice, glibly requires that a ‘Williamson trade-off’23 be undertaken in merger review or that an excessive price be prohibited.

So it’s as well to treat claims of efficiency gains with a healthy degree of scepticism, and this is precisely what we have done. And even then, if the claims are proved, the other side of the balance is likely to prove at least as difficult to measure. While I cannot recall the process by which the efficiency defence was incorporated into our legislation, I have no doubt that, just as the prohibition of excessive pricing was a product of the unions’ participation in the negotiating process, so the efficiency defence was inserted at business’s behest. Both insisted on the incorporation of their narrow interest into the Act; neither gave much thought to how they would or could be effectively determined.

Certainly, business was loudly concerned at the prospect that robust merger review in a small market would condemn local firms to inefficiency through their inability to achieve scale economies, and so arguably end up with two bites at the efficiency cherry, first in the efficiency defence itself, and then in the inclusion of ‘international competitiveness’ in the limited list of public interest considerations. Indeed, I recollect the review of the proposed merger between AECI and Sasol undertaken shortly before the demise of the Competition Board, with the new pending competition regime looming large. The merging parties argued that efficient production of explosives – the most important market implicated in the merger – required the realisation of scale economies if South African producers were, in their limited domestic market, to achieve international competitiveness. A cursory examination revealed that explosives were, for obvious reasons, extremely difficult to trade internationally and that South Africa was, at that time, the largest national explosives market in the world, thus simultaneously manifesting both business hysteria regarding this issue and the hyperbolic nature of so many of the efficiency arguments.

In the early years of the current competition regime, we were frequently obliged to sit through minutely detailed efficiency defences. However, as it became clear that we were going to set a high bar on a successful efficiency defence, so the amount of attention paid by merging parties to this element of the legislation dwindled. I recall only two occasions on which the Tribunal accepted an efficiency defence. The most important of these decisions involved the acquisition by Trident Steel of the three processing plants of a competitor, Baldwin Steel.24

The Trident Steel matter involved a merger to monopoly, which was approved on efficiency grounds. The real value of the decision resides in the elaboration of a clear and, ironically, stringent approach to the treatment of efficiencies. I say ‘ironically’ because, on re-reading the decision, I am not confident that on its facts this merger should have passed the test laid down. Indeed, I don’t believe that a substantial lessening of competition that arises from a 3–2 merger, let alone a merger to monopoly, should ever be permitted on efficiency grounds. Low entry-barriers or a failing firm defence may well permit such a merger on substantive competition grounds, thus removing the requirement for an efficiency evaluation. However, having found a substantial lessening of competition, I don’t think that the efficiency defence should succeed under the structural conditions of the Trident merger. While, to use Posner’s words, we may ‘not know enough about the effect of marginal increases in the concentration ratio … to predict the price effects’ of a merger, I think that we can infer the overwhelming likelihood that these will be unfavourable in a merger to duopoly or monopoly.

I say this with no small measure of circumspection and humility. The panel arrived at this conclusion by demonstrating persuasively the significance of the dynamic efficiencies that the merger would realise for both the monopoly producer of steel and the post-merger monopoly consumer of these steel products. More than that, the decision was careful to identify the limits to the pricing power of the merged monopoly, namely the import parity price, which, as the panel noted, was not significantly greater than the price then charged. And, possibly more important than the decision itself, the reasoning generated (in a manner that manifests Norman Manoim’s persistent and admirable effort to define administrable rules of law capable of dealing with the most intractably complex economic formulations) a set of rules guiding the practical application of no less than the Williamson trade-off.

However, my inclination is to support the application of a less flexible rule, one that would exclude the use of the efficiency defence in the case of a merger to monopoly. There are cogent arguments for doubting the proposition that imports will constrain the pricing behaviour of the merged firm. The Trident Steel judgment supports this by demonstrating that ‘imports are not competitively priced on present exchange rates’, thus justifying a finding that ‘the market for processed flat steel products (outer blanks) is a national one, subject to some import competition’. It also concludes that the countervailing power (of the automobile manufacturers) ‘in a post-merger market place where there will only be one domestic supplier of ISF outer blanks to the automotive industry, is unpersuasive’. I have little doubt that, had the efficiencies been considered as part of the competition evaluation, the merger would have been prohibited.

The comeback, of course, is that these arguments are the basis for finding a substantial lessening of competition, a finding that, as required by the Competition Act, is made prior to a consideration of the efficiency gains. The really difficult question – at least in this merger – is then contained in the requirement to weigh up that finding against the efficiency gains, which the panel found to be considerable. The $64 000 question is then about pass-through, whether to use consumer welfare or total welfare as the appropriate standard. The panel proposed the following test: ‘Where efficiencies constitute “real” efficiencies and there is evidence to verify them of a quantitative or qualitative nature, evidence that the efficiencies will benefit consumers is less compelling. On the other hand, where efficiencies demonstrate less compelling economies, evidence of a pass through to consumers should be demonstrated and although no threshold for this is suggested, they need to be more than trivial, but neither is it necessary that they are wholly passed on’.25

A merger to monopoly effectively entrenches the import parity price and ensures that none of the efficiency gains will be passed on. In fact, the incentive to sustain the efficiency gains is eroded by the post-merger structural conditions and so may well prove to be short-lived.

In the scale of things, these are minor quibbles. Certainly the standard set for the efficiency test ensures that it will be difficult to pass and the importance of having developed an administrable standard is not to be underestimated. I repeat, though, that I am not sure that the standard set should ever permit a merger to monopoly.

The Public Interest Test

The public interest grounds are clearly specified. They incorporate the impact on employment, on economic regions, on black economic empowerment (BEE) and on international competitiveness. In practice there are two sets of public interest grounds that have surfaced before the competition authorities when conducting merger reviews. These are, first, the employment impact where the form that the argument would characteristically take is a plea for turning down, or imposing conditions on, a merger which was acceptable on competition grounds but which was likely to lead to job loss.

The second is the promotion of BEE, where the argument would generally be for approving a merger that might not pass muster on competition grounds, but that nevertheless carried sufficient promise of black economic empowerment to justify approving the merger. The international competitiveness arguments that, like the BEE arguments, would have been invoked to justify approval of an anticompetitive transaction tended to be subsumed in the efficiency arguments for a transaction that, absent the efficiencies, might have been prohibited or have had conditions imposed on their approval.

The inclusion of public interest grounds in merger review aroused huge controversy. I’ve already referred to the role they played in the negotiating process. However, in truth, no matter how controversial, and no matter how much the business community in particular would have preferred their omission, everyone involved in the negotiating process recognised that no major piece of socio-economic legislation would have passed muster without incorporating job creation and BEE into the overall objectives of the policy and the statute. This, of course, did not mean that these had to be included in the criteria for evaluating mergers, but once the unions took this up, there was, in the prevailing political and economic climate, no way they were going to be denied. And so the only real question concerned the identity of the public interest decision-maker, and while the drafting team was fully prepared for this role to be assumed by the minister, to our delight he ceded this power to the competition authorities.

It was in international circles that the inclusion of public interest criteria into merger review generated the greatest controversy. During the first years of our life, whenever any of us were invited to contribute to an international conference, it was inevitably this issue that we were asked to address. However, South Africa is by no means the only country in which public interest criteria are applied in merger decisions. In a country with as strong a tradition of independent antitrust enforcement as Germany, the minister retains an explicit veto, on public interest grounds, of merger decisions of the venerated Bundeskartellamt. German competition officials will point out that it is infrequently invoked, but it’s there, and in at least one important recent instance it was used to trump a merger prohibition taken on competition grounds.

However, it appears that nowadays – at least in the mature competition jurisdictions – a public interest test appears to be applied only in the case of cross-border acquisitions, that is acquisitions of domestic firms by foreign-owned firms. Hence in the US there is the capacity – through the Exon-Florio Amendment to the Defence Production Act of 1950 – for blocking mergers on national security grounds. Similar issues arise from foreign acquisitions in other countries, where similar foreign investment reviews have been invoked, athough it’s often difficult to understand the national security concerns that led the French to establish such a review when PepsiCo attempted unsuccessfully to purchase Danone, the French yoghurt manufacturer, or what national security concerns led Italy to introduce a similar review after a French company attempted to acquire Parmalat, the Italian dairy products firm. The truth of course is that ‘national security’ or, for that matter, ‘public interest’, if too loosely defined, is open to be invoked opportunistically, usually by one or other group of stakeholders – employees, regional interests, shareholders and boards holding out for a higher price – when they organise their public interest appeal around the sentiment attaching to one or other iconic national brand.

Canada is an interesting case in point, as exemplified by two attempted mergers. In 2009 the Canadian government blocked an attempted merger in the potash market that would have resulted in the acquisition of Canada’s largest potash producer by a large Australian mining company. There were no competition issues involved, only unspecified ‘strategic’ issues that seemed to derive from a view that the deal did not offer a ‘net benefit’ to Canada. Indeed, the only competition issues involved predisposed in favour of permitting the merger: the Canadian target firm was the leading member of an officially-sanctioned export cartel from which the prospective acquiring firm undertook to exit. This action not only represents an overt departure from the competition principles that govern merger regulation, but also probably prolongs the cartelisation of a product of vital economic and social interest – because it is a basic input in the production of fertiliser – to the developing world in particular.

Nor is this an isolated action. In 2011, it appeared that the Canadian government would apply the same criteria to a proposed merger between the London and Toronto stock exchanges. A columnist in the Globe & Mail wrote that ‘Thanks to a mixture of history and happenstance, Canada treats a clutch of industries as national treasures. Think of banks, broadcasters, cable companies, newspapers, airlines and liquor stores’, and he goes on to ask, ‘do we protect liquor stores to grab the profits or to control distribution of a potentially dangerous substance? Do we keep foreigners from owning our banks in order to protect consumers and borrowers, or for blatantly protectionist reasons?’26

And so it seems that the really unusual element of South African practice was not that public interest – or, it might equally be said, ‘non-competition’ – issues played a part in deciding whether to approve a merger, but that the public interest criteria were incorporated into the Competition Act, that they applied to all mergers (as distinct from only cross-border mergers) and that the competition decision-makers were assigned responsibility for balancing the competition and public interest considerations.

So what to make of all of this?

It’s perfectly common practice for mergers to be subject to multiple criteria and decision-makers. Hence, a takeover involving a listed entity is subject to stock exchange rules and requires the authority of the securities regulators. The merging of two licensed telecommunications or broadcasting operators would require the permission of the relevant licensing authorities. In both instances, the mergers in question would also be subject to competition jurisdiction. Decision-making would be assumed by each regulator in its sphere of competence and a prohibition by one regulator would always trump an approval by the other. This is perfectly uncomplicated and uncontroversial. The South African oddity lies in having a single decision-maker, the competition authority, responsible for deciding a merger on two, potentially conflicting, sets of criteria.

While the law placed the public interest inquiry on the same plane as the competition and efficiency assessments, the fact is that the sequencing of the decision is such that the competition and, if necessary, efficiency investigations are concluded before the public interest considerations are evaluated and balanced against the competition and efficiency conclusions. An authority principally and specifically charged with promoting and defending competition would have been hard pressed to prohibit or approve a merger on grounds other than the impact on competition. As I’ve already elaborated, even when efficiency arguments, which could be said to be core to the concerns of a competition authority, were invoked to justify approval of an anticompetitive merger, the Tribunal set a very high bar.

And so with public interest arguments: the fact is that no decision to approve or prohibit a large merger has thus far been on public interest grounds alone. In contrast with the more standard approach where a minister took the public interest decision, with at best a passing knowledge of the competition considerations and a powerful lobby across the table, in our case the public interest ground was examined in a public hearing by the competition decision-maker through the prism of the competition evaluation. It was not so much a situation that downplayed the public interest argument, as one that ensured that the outcome of the competition evaluation received the weighting it warranted. Accordingly, the public interest argument had to be extremely powerful if it was to result in the prohibition of a pro-competitive merger or the approval of an anticompetitive merger.

I have no qualms about this approach. After all, the promotion of competition itself is in the public interest (though, given our Prime Cure experience, I should hasten to add that I do not mean ‘public interest’ as defined in the Act. I simply mean that it’s a public good). However, the supporters of competition – largely consumers and small businesses – are atomised and poorly organised, in marked contrast to the powerful bodies that back employee interests or BEE or international competitiveness. Our approach simply levels the playing field on which, on occasion, the better-organised and more powerful are ranged against the weak and disorganised.

The Public Interest: Black Economic Empowerment

The evidence of BEE gain from a merger generally involved the acquiring company in finding a black partner, the more politically prominent the better, who was then presented in the hearing as a beneficiary of the transaction and representative of a class – black investors – whose interest the Competition Act, among other more important and direct instruments, is committed to advancing.

Although BEE was invoked on a number of occasions, four are most illustrative. The first was in the Sasol–Engen transaction, a huge and complex transaction that required us to assess the likely impact on competition in a market – liquid fuels – critical to the country’s economic interest and to the immediate well-being of all its citizens.27 The BEE argument centred on the commitment by Sasol to sell a share of the merged entity to a BEE group led by none other than Penuell Maduna, a former minerals and energy affairs minister. In fact Maduna’s grouping had for some time been Sasol’s empowerment partner. For its part, Engen was not only able to empower black shareholders without the expedient of the merger, it was in fact required to do so in terms of a charter operative in the industry in question. In other words, the possibility of empowering a black shareholder in Sasol or Engen or a merged entity owed nothing to the merger.

Indeed, generally, BEE was not merger-specific – that is to say, the firms promising to empower black investors as a condition for the approval of a deal could each have empowered different groups of black investors without the deal, and were in fact required to do so by a variety of other instruments and points of commercial and social pressure. The Sasol–Engen merger, which we prohibited, is memorable not for the salience of its BEE argument, but for, inter alia, the manner in which it raised the pernicious, unhealthy relationship that was rapidly developing between business and government. I’ll return to this theme later.

The second illustrative occasion of BEE’s place in merger review represented, in many ways, one of the lowest points in the Competition Act’s young life. This involved the attempted acquisition of one of South Africa’s three large private hospital groups, Afrox Healthcare, by a consortium blandly referred to as ‘Bidco’. In this case, the participation of a BEE investor was not only invoked to support the public interest content of the transaction, but it was used to disguise the thoroughly anticompetitive nature of the merger.

The transaction arose from the decision of African Oxygen – a subsidiary of British Oxygen – to sell Afrox Healthcare (‘Ahealth’), one of three large private hospital groups. Both of the other large private hospital groups, Netcare and Mediclinic, wanted to get their hands on Ahealth. However, Netcare was rebuffed by the transaction advisers on the grounds that its participation would not pass competition muster. Although identical competition considerations applied to Mediclinic’s interest, it devised a ploy, as crude as it was cynical, to enable it to participate in a bid. Essentially it sought a BEE partner to counterbalance the problems that would inevitably arise at the competition evaluation stage. It settled on two firms, Brimstone and Mvelaphanda, two of the country’s most successful and highly-regarded BEE groupings. The two BEE companies held 75 per cent of Bidco, with the remaining 25 per cent held by Mediclinic. In addition to Mediclinic securing the funding that permitted the BEE companies to participate in what was being touted as the largest BEE deal ever, Mediclinic had also entered into an agreement with its Bidco partners in terms of which Mediclinic would acquire from Bidco several hospitals accounting for 2500 beds, approximately 33 per cent of Afrox’s capacity, once the Bidco deal had been concluded. In other words, not only were the BEE companies being used as a front to enable Mediclinic to acquire a large interest in a key competitor, the competitor’s position – in which the BEE companies would hold a majority stake – was going to be massively weakened by the subsequent transfer of key assets to Mediclinic itself.

In a remarkably poor exercise of judgment, the Commission recommended to the Tribunal that the transaction be approved. Moreover, its approval was not based on the conclusion that an anticompetitive merger was acceptable on the grounds that it introduced BEE investors to the health-care sector – the Commission seems to have believed that the BEE partners were actually the controlling shareholder, with Mediclinic a passive, minority investor, a ploy described by the Tribunal as ‘fronting’ on a grand scale, with Mediclinic, though holding a minority equity stake, thoroughly dominating its BEE partners. However, on the Commission’s extraordinarily naive reading, no competition problems arose and so the merger was approved on competition grounds. Needless to say, all hell broke loose, led by a vengeful Netcare intervening in the Tribunal’s process and initiating a revealing discovery process. In the words of the Tribunal:

This sordid story does have a happy ending. The certainty that the Tribunal would have prohibited this transaction together with the escalating damage to Mediclinic’s reputation as a result of the public nature of the Tribunal’s hearings, led to the withdrawal of the transaction and its replacement by a new transaction in which Mediclinic’s place in Bidco was taken by a number of financial institutions and the agreement to sell the 2,500 hospital beds to Mediclinic was cancelled. And so, thanks to a transparent adjudicative process that allowed liberal intervention and discovery, a competition disaster was averted and Mvelaphanda and Brimstone got their private hospital group, thus promoting a defined public interest.28

However, neither story shows BEE in a good light. Both demonstrate that in two transactions involving important South African companies, the BEE investors were minor players in a larger game whose manifest intention was to thwart competition. This does not necessarily dispute the value of black economic empowerment. But it does question the wisdom of trading off two critical programmes – the promotion of competition and the expansion of black ownership – that cannot be measured on the same scale. Indeed, the third BEE story demonstrates the strength of clearly focused mechanisms to promote black economic empowerment and the prospect of these being undermined by inappropriate interventions by the competition authorities.

Thebe Investment Corporation, a black-owned company, proposed to sell a small subsidiary, Tepco Petroleum, to Shell South Africa, the South African marketing arm of the giant Royal Dutch Shell group of companies.29 Tepco, which was in dire straits, consisted of some 14 branded retail petrol stations and a small presence in the commercial fuels industry. However, it represented a fairly significant investment for, and a consequent threat to, its parent, Thebe. Shell, undoubtedly mindful of the BEE charter for the liquid fuels industry, which required the oil companies ultimately to ensure that 25 per cent of their equity was owned by empowerment investors, offered Thebe a stake of some 25 per cent in Shell South Africa, plus one of four board seats in exchange for its 100 per cent stake in Tepco.

Thebe was pleased to accept this offer. Indeed, it appears to have been a very generous offer and an indication of Shell’s anxiety to comply with the charter. However, although no competition issues arose – Tepco’s market share and its accretion to Shell’s share were minuscule – the Commission, concerned at the public interest implication of the likely disappearance of a black-owned petrol brand, recommended that the Tribunal impose conditions on the approval of the transaction. The principal condition required that the Tepco brand be maintained under the joint ownership of Shell and Thebe. In the unlikely event that Shell accepted such a condition, it would undoubtedly have been priced in the form of a lower offer being made to Thebe. The Commission also sought to impose conditions related to employment equity and skills development.

In response to this recommendation, the Tribunal and Thebe executives who appeared before the Tribunal pointed out that this effectively discriminated against a black-owned company (Thebe). Whereas a white-owned firm would not have been constrained to hold on to an asset that no longer served its purpose, that indeed threatened its owners’ commercial interests, the Commission proposed to constrain Thebe in this way because it was black-owned! Or expressed positively, whereas a white-owned firm was able to dispose freely of its assets to raise capital, Thebe, a black-owned firm, was only able to sell its assets to a firm that would accept the condition recommended by the Commission or, at the very least, would be obliged to accept the discount that the condition imposed on the price of the asset. In the course of the hearings, the Commission responded to this observation by pointing out that it was charged with protecting the public interest and not with protecting Thebe. Effectively, it argued that the public interest was served by maintaining a black-owned petroleum brand, with the commercial interests of the black owner a secondary concern.

The Tribunal’s response to the Commission’s contention has defined its approach to the application of the public interest: ‘The role played by the competition authorities in defending even those aspects of the public interest listed in the act is, at most, secondary to other statutory and regulatory instruments – in this case the Employment Equity Act, the Skills Development Act and the charter itself immediately spring to mind. The competition authorities, however well intentioned, are well advised not to pursue their public interest mandate in an over-zealous manner lest they damage precisely those interests that they ostensibly seek to protect’.30

A transaction in which the outcome of the competition analysis interfaced successfully with remedies that enhanced both competition and black economic empowerment is the Coleus–Rheem transaction.31 Again it is worth examining this merger in some detail. Not only does it have a public interest dimension, but it is a vertical merger and it raises some interesting issues regarding the selection of appropriate remedies. Fortunately the facts are relatively straightforward and uncontested.

Coleus Packaging made an offer to acquire the Rheem Crown Plant belonging to Highveld Steel. Rheem Crown is a producer of bottle tops. Coleus Packaging is a wholly-owned subsidiary of South African Breweries (SAB), which, as the near-monopoly supplier of beer in South Africa and the owner of Amalgamated Beverage Industries (ABI), the largest Coca Cola bottling operation in South Africa, is, by a country mile, the largest purchaser of bottle tops. Rheem enjoyed a sole supply agreement with SAB.

Rheem had a single domestic competitor, MCG Ltd, a relatively new entrant to this market, which was owned by BEE investors. Indeed, it appears that MCG had been encouraged to enter this market by Distell, the country’s largest supplier of wines and spirits, which in the wake of the earlier exit of Rheem’s then only competitor, Crown Cork, feared a dependence on a single supplier of bottle tops.32 Hence, while MCG enjoyed a significant share of the non-SAB purchases of bottle tops (41 per cent to Rheem’s 59 per cent), once the SAB–ABI purchases were factored in, MCG’s share of the total market dropped to 10 per cent.

The bases of opposition to this merger are easily identified. MCG feared that SAB’s custom would be placed permanently out of its reach. It appears that MCG could not attain a minimum efficient scale of production without obtaining a share of SAB’s requirements.

For their part, other users of bottle tops – who competed with SAB in the broad alcoholic beverages market and particularly in the growing niche market for flavoured alcoholic beverages – feared input foreclosure, that is, that their access to the country’s largest manufacturer of bottle tops would be foreclosed or that they would be discriminated against in terms of price and availability of supply. They also feared that information-sharing between Rheem and its parent would give SAB advance warning of the sales volumes and promotional activities of its competitors. Moreover, they recognised that SAB’s support for Rheem would enable Rheem to achieve economies of scale and undertake investment that ensured that it remained South Africa’s most competitive supplier of bottle tops. MCG, on the other hand, was, in the absence of any custom from SAB, left with significant excess capacity and little prospect of undertaking the investment necessary to match Rheem. At a minimum, then, MCG was dependent on retaining all of the custom of the non-SAB purchasers of bottle tops. The latter, in turn, feared that if they did not use Rheem, they would be reliant on a single supplier. Unlike SAB, the volumes of the other beverage suppliers were not sufficiently great to enable them to import economically.

The Commission recommended that the transaction be prohibited. The Commission’s recommendation indicated that MCG, Rheem’s competitor, would oppose the transaction, as would a number of SAB’s competitors who were customers of both Rheem and MCG.

However, prior to the hearing, SAB assuaged MCG’s concerns by entering into a supply agreement with it. It also entered into discussions aimed at addressing the concerns of its competitors. SAB also announced an undertaking to sell 40 per cent of Rheem’s equity to an empowerment partner within 2 years. These undertakings, both those addressing the various competition-related concerns and the public interest undertaking, were captured in a draft order, which was submitted to the Tribunal. The Commission was persuaded that this combination of undertakings allowed it to support a conditional approval rather than a prohibition.

Public Interest: Employment

The employment test does not embody the conceptual ambiguities or the potential for opportunism contained in the application of the BEE criterion. This does not mean that it is easy to weigh competition and employment considerations on the same scale. It is nevertheless wholly possible to conceive of a number of rules of thumb in applying the employment test. Hence a merger that is supported entirely by the rationalisation of duplicate facilities which it enables and which generates large employment loss may well be prohibited on employment grounds. We have also tended to give a higher weighting to employment loss suffered by relatively immobile unskilled workers than to the prospect of job loss among more skilled employees who could reasonably expect to find alternative employment.

Strange to tell, I don’t think that the competition authorities had, until very recently, been seriously confronted by the employment test. Had the Supreme Court of Appeal granted us jurisdiction in the proposed Nedbank–Standard Bank transaction and in the unlikely event that the merger would have overcome the competition hurdle, we would have had to confront the fact that many thousands of relatively well-paid jobs would be threatened by the rationalisation of overlapping branches and this may well have resulted in a prohibition on employment grounds. But in our first 10 years we were never confronted with cataclysmic employment consequences arising from a merger. This may change with the economic downturn – employment loss following mergers might be greater and sensitivity to employment loss will undoubtedly be heightened. A recent criticism suggests that the Commission is causing lengthy delays in evaluating mergers principally because of the time it is taking to evaluate employment loss. While I am not persuaded that this is indeed so, nor would I be surprised to learn that the Commission and the unions, which, you will recall, are notified of mergers, are paying enhanced attention to employment outcomes in the current environment.

Our approach to employment consequences is similar to that cited in Shell–Tepco, namely that there is a comprehensive set of laws and institutions set up to deal with employment, specifically including retrenchment, that negotiated outcomes between labour and employers are at the core of our labour relations framework, and that in the first instance our approach should encourage respect for this system and its outcomes. While unions or employers may welcome one or other decision of the competition authorities regarding employment, I doubt either would welcome a competition authority that is overly interventionist in employment issues. In summary, then, the Tribunal’s approach is to encourage the merging parties and the employee representatives to use the provisions of the Labour Relations Act and to agree on the employment implications among themselves, and then, if requested, make approval of the transaction conditional upon that agreement. This is more than a mere rubber stamp because, of course, once embodied as a merger condition in a Tribunal order, contravention may result in labour law and contractual contraventions as well as a competition law challenge to a consummated merger.

In the absence of an agreement or credible employee representation, we are inclined to make the merging parties’ estimate of employment loss a condition for approval. At the very least this should discourage the merging parties from understating the employment consequences in order to secure an easy approval.

The Tribunal’s approach is clearly illustrated in an early merger decision that involved the sale and leaseback by Telkom of a large number of its properties and the conclusion of an agreement with a facilities management company to take over from Telkom responsibility for a range of management functions in relation to the properties in question.33 This involved the transfer of a significant number of Telkom employees to the facilities management company.

There had been consultation among the parties to the merger and the employees’ representative, the Communications Workers Union (CWU). Following these discussions, Telkom and the facilities management company concluded an agreement that imposed an obligation on the new employer of the erstwhile Telkom employees to refrain from retrenching any of the transferred employees for a period of 20 months. The union appeared to accept the time period. For its part, the Tribunal considered it a particularly adequate undertaking, given the dynamic circumstances in the market concerned, but acceded to the CWU’s request that the agreement between the merging parties be made enforceable by the employees concerned, rather than simply by the merging parties who were also the parties to the employment agreement in question. The Tribunal also considered a concern raised by the union that, should Telkom find that it had not transferred sufficient workers to the facilities management company, retrenchments would take place within Telkom itself.

The Tribunal then made approval of the merger conditional on the facilities management company not retrenching any of the transferred employees for a 20-month period, on the agreement being enforceable by the workers concerned, and on Telkom not retrenching any of its employees as a result of the merger for a period of 20 months. The union also requested that we order the merging parties to recognise the CWU as the collective bargaining representative of the transferred employees. While our decision reflects that we did not respond to this request, it exemplifies what the Tribunal would consider to be a matter for the collective bargaining system to resolve and some way beyond the remit of the employment public interest provision in the Competition Act. This is a fairly elaborate version of our standard employment condition, which would simply specify a maximum number of merger-related retrenchments.

However, on occasion, we have imposed more elaborate employment conditions. Ashton Canning is unusual on a number of measures.34 It was one of those rare instances where a merger was found to lead to a likely substantial lessening of competition but was nevertheless permitted on the grounds that the efficiency gains would outweigh the negative impact on competition. However, for present purposes, it’s the employment condition that we are concerned with.

Management of the acquiring firm estimated that the merger would lead to a loss of 45 permanent jobs and 1000 seasonal jobs, and this in an extremely deprived rural area. Although the seasonal workers were, as the name suggests, employed only for a portion of the year, they were regularly employed, some for up to 15 successive years, by the firms involved in the merger. The Tribunal approved the merger subject to a 3-year moratorium on any further merger-related retrenchments. The conditional approval also required the merged entity to endow a training fund to the tune of R2 million, to which retrenched employees affected would be entitled to apply, in order to support accredited training.

Employment issues will inevitably be weighted most heavily in situations of extreme economic deprivation, whether a consequence of regional factors, as in Ashton Canning, or general economic downturn, as in the recent Momentum decision.35 The Tribunal has also consistently focused employment conditions on protecting unskilled workers whose prospects of securing alternative employment are, relative to those of their more highly-skilled counterparts, slim.

The Tribunal’s treatment of employment loss is most comprehensively outlined in the Momentum decision. In Momentum the merging parties initially estimated that the merger would result in the loss of approximately 1500 jobs. While the Tribunal found no basis for any competition concerns, it also found that the merger could not be justified on public interest grounds. While the parties did not concede that this amounted to a ‘substantial public interest’ concern, they were prepared to accept limited conditions. The union representing a large number of those targeted for retrenchment argued for the prohibition of the merger on the grounds that the parties had failed to justify the job loss. The Tribunal, however, approved the merger but subject to a 2-year moratorium on any merger-related retrenchments. The condition did not apply to the retrenchment of ‘senior management’.

The decision is tightly reasoned and difficult to summarise. It certainly gives the lie to the notion that the public interest evaluation, as it relates to the employment impact of a merger, is made on the basis of a purely subjective assessment of the scale of job loss. Clearly, the reasoning suggests that there may well be circumstances in which a larger number of jobs are lost as a result of a merger, but which may not give rise to a prohibition or the imposition of an employment-related condition in order to protect the public interest. Conversely, there may well be cases in which a smaller number of job losses may give rise to the imposition of a condition.

The principle laid down in Momentum is that, once it has been established that there will be substantial job loss arising from the merger – and this was not seriously in contention – then, in order to have the merger approved unconditionally, the evidentiary burden imposed on the parties requires that they prove that (1) a rational process has been followed to arrive at the determination of the number of jobs to be lost, that is, that the reason for the job reduction and the number of jobs proposed to be shed are rationally connected; and (2) the public interest in preventing employment loss is balanced by an equally weighty but countervailing public interest, which justifies the job loss and which is cognisable under the Act.36

The upshot is that even if the merging parties successfully demonstrate that the employment loss will lead to substantial efficiency gains (that is, that the ‘employment loss is rationally connected to an efficiency claim’), it will not be sufficient to show that the gains accruing are purely private in nature, particularly when there is also a public interest loss arising from the transaction.

At the start of the hearings, the merging parties made an undertaking to limit the number of job losses to 1000 and provide a re-training allowance to a class of qualifying retrenchees. However, on the basis of the evidence placed before it, the Tribunal concluded that the quantum of job loss was principally dictated by the requirement to achieve a level of savings, which the management teams of the merging parties had assured their shareholders the merger would deliver. That is, the merging parties failed to establish a ‘rational connection between the efficiencies sought from the merger and the job losses claimed to be necessary on their worst-case scenario. Rather we find that this figure has been arrived at in an arbitrary manner on the basis of sweeping assumptions made in a broad brush fashion’.37

And the parties also failed on the second test. For example, they were not able to establish that the job loss was necessary to rescue one of the merging firms: both were prosperous entities. In other words, it was not established that it was necessary to accept some job loss in order to forestall the prospect of greater job loss. Nor were they able to show that the savings accrued from the retrenchments would be passed on to consumers in the form of lower prices or new products. The savings arising from the job loss would go straight to the bottom line, to the shareholders, and so the public interest loss arising from the job loss was not offset by any discernible offsetting public gain to consumers.

In line with earlier approaches to the employment-related public interest test, the Tribunal’s decision to impose a condition appears to have been bolstered by evidence that the merging parties had failed to engage in meaningful consultation with their employees.

So, difficult though the balance between competition and public interest is, there is a developing jurisprudence that attempts to take the judgment out of the realm of pure subjectivity and that, in so doing, provides a degree of certainty in the application of the public interest test.

Generally, however, the competition authorities have resisted using their employment public interest mandate as an instrument of industrial policy, particularly when a pro-producer industrial policy intervention is countervailed by a palpable loss imposed on consumers.38

There is little doubt that in times of economic downturn the public interest considerations – and particularly the employment considerations – will loom larger. I think that Momentum demonstrates that the Commission and the Tribunal are willing to respond to changed economic circumstances by adjusting their approach to their public interest mandate.

However, the danger lies not with the flexibility required to deal with changed objective circumstances, but rather with how to deal with excessively interventionist economic ministries. The competition authorities have gone through three ministerial regimes. The first was Alec Erwin, who was responsible for ‘depoliticising’ the public interest decision and who, as one might have expected, adopted a hands-off approach to the decision-making powers of the competition authorities and who was, in any event, not a micro-manager. The second was the Mandisi Mpahlwa ministry, where it seemed the poor capacity of the department, and notably the division responsible for competition policy, ensured a complete absence of intervention. On occasion the Tribunal – and, I don’t doubt, the Commission – requested a submission from the department. I recall on a number of occasions in this period bemoaning the inability of the Department of Trade and Industry to put forward a coherent government view, although there were exceptions, as in the Mittal excessive pricing case. The Rob Davies ministry was, in relation to the Department of Trade and Industry’s responsibility for competition policy, relatively short-lived; he was saddled with the same incompetent departmental division and so there was not much change from the inactivity of his predecessor.

In 2009, ministerial responsibility for the competition authorities was transferred from the Department of Trade and Industry to the newly created Economic Development Department. The minister, Ebrahim Patel, is extremely interventionist and given to micro-management. His preferred modus operandi is to secure bargained outcomes between adversaries.

Knowing all of this, I had at the time of the handover from one department to the other wondered – not without a gnawing concern – why Patel wanted responsibility for the competition authorities. Competition law is, at best, an indirect instrument and so does not lend itself to bargained outcomes. Nor is the process of competition decision-making – the quasi-judicial nature of the Tribunal and the relative independence of the investigative process – compatible with a political bargaining process. This means that a policy department has only a limited ability to intervene in the competition investigative and adjudicative processes by trading off a decision entrusted to the competition authorities in exchange for a particular public interest or industrial policy-related outcome. And the policy department runs the considerable risk of severely undermining the independence and hence the credibility of the competition authorities. Nowhere is this better illustrated than in the merger involving the proposed acquisition of a large local retailer, Massmart, by the iconic and notorious US-based retailer Walmart.

It is difficult to capture the whole sorry Walmart saga in the space of a few paragraphs. However, to cut an extremely long – and still continuing – story as short as possible, it is common cause that there are no competition issues at stake in the merger. Walmart is a new entrant and, as already noted, the South African grocery and general retail market is, for the most part, intensely competitive. But Walmart carries some unpalatable baggage, not least a notorious reputation for generally poor employment practices and a particularly aggressive anti-union stance. It was thus to be expected that the South African unions, abetted by a US union-based anti-Walmart coalition, would take a jaundiced view of Walmart’s entry into the country. And so they did. They attempted to argue that Walmart’s entry would result in job losses in Massmart, a public interest criterion that the competition authorities were obliged to consider. However, the merging parties effectively dispelled this. Although the unions have not conceded this, the evidence clearly suggests that the direct employment consequences of the merger are, if anything, likely to be positive. The unions also argued that Walmart would engage in practices contrary to South African labour relations legislation. This consideration is not part of the mandated public interest criteria but, given Walmart’s reputation, it was bound to be raised.

However, it is the Minister of Economic Development’s intervention in this matter that is, for present purposes, of most interest. His stated concern was not merely with the direct employment consequences of the merger, but rather with the prospect that Walmart would substitute imported – read ‘Chinese’ – goods for South African products in its newly-acquired stores. And so the minister wanted a commitment that the merged entity would maintain local procurement at the same level as Massmart, the South African target entity. In short, what was demanded was a wholly protectionist condition, effectively the imposition of a local procurement quota. Walmart made it clear that it would not willingly submit to this condition.

Moreover, not only did the minister intervene extremely late in the investigatory process, but he chose to intervene not by making submissions to the Commission’s investigators – as he is entitled to do – but rather by entering into private negotiations with the merging firms, clearly holding out the promise that, were the firms to reach agreement with him, he would ensure that the agreement would be rubberstamped by the competition authorities. In other words, it would be supported by the Commission in making its recommendation to the Tribunal and by the Tribunal in accepting the recommendation.

While these negotiations were under way, the Commission’s time frame for conducting a merger investigation expired. It recommended to the Tribunal that it approve the merger unconditionally but advised that the merging parties and the Minister of Economic Development were in negotiations regarding the public interest implications of the transaction. The Commission effectively reserved its position on the public interest, pending the conclusion of the discussion between the minister and the merging parties.

However, because of the procedural requirements of the adjudicative process, this meant that the Minister of Economic Development, with the ministers of trade and industry and agriculture in tow, was then in effect compelled to emerge from his smoke-filled room and have his officials appear in front of the Tribunal to argue his case. He could no longer hide behind the skirts of the Competition Commission. The first of the minister’s three contributions was to petition for a delay in the entire proceedings. When the hearing finally got under way, the minister’s legal counsel argued for the procurement-related condition or, failing that, outright prohibition. The unions asked for the merger to be prohibited on employment grounds. In the alternative, they proposed certain employment-related conditions.

The Tribunal ultimately approved the transaction but imposed a number of conditions on the merged entity, all of which related to employment and industrial relations issues, and all of which were uncontested by the merging parties. However, the Tribunal refused to impose a procurement-related condition.39

There was a variety of reasons given by the Tribunal for its refusal to accept such a condition. It was found to be unworkable, unenforceable and asymmetric. Although the Tribunal did not finally have to decide this point, it also appeared to accept that the imposition of a procurement-related condition of the kind sought would be contrary to South Africa’s international trade obligations.

But, these reasons aside, the Tribunal rested its rejection on the consumer impact – and, in this case, particularly low-income consumers – of a protectionist condition. ‘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?’40

Remarkably, it was left to the Tribunal to point out to senior government representatives, supposedly wedded to a contemporary version of industrial policy, the availability of competition-friendly conditions that would simultaneously meet industrial policy objectives. It unearthed a pro-competitive ‘investment remedy’ earlier proposed by the merging parties to the ministers, as distinct from the protectionist ‘procurement remedy’ favoured by the same government representatives. This investment remedy obliged the merged entity to set up a R100 million fund to assist in developing the competitiveness of its domestic suppliers. In deciding on disbursements of this fund,

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 R100 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.41 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.42

The unions have filed an appeal against the decision of the Tribunal to permit the merger to proceed. Unlike the unions, the ministers do not have a right of appeal against the substantive decision of the Tribunal. However, as with any litigant, they do have the right to review the proceedings. Remarkably, they have chosen to do so on the extremely flimsy and transparently contrived grounds that the Tribunal had limited the extent of discovery they had asked for and that the Tribunal had imposed limits on the number of witnesses it would hear. They have accordingly petitioned the Competition Appeal Court to remit the matter to the Tribunal for another hearing that will be untainted by the procedural defects they allege. The Tribunal had set these rules for the hearing largely to expedite an already delayed process. The review is clearly designed to achieve the opposite. It is designed to prolong the decision-making process, the better to extort a larger industrial development fund from the merged entity – it appears that the ministers would settle for a bounty of R500 million rather than the R100 million imposed by the Tribunal.

The message that the Minister of Economic Development is sending is clear. He is effectively saying that merger decisions – and this need not refer only to cross-border mergers – will be used to exercise leverage to attain industrial policy goals. This will be done by politicising the decision-making process, by effectively compromising the ability of the competition authority to make independent, evidence-based decisions.

It’s sad to note that the response of the CEO of the merged entity indicates an understanding of competition policy and industrial policy, not to mention the rule of law, that is significantly more advanced than that of the trio of government ministers. A report in the local media notes:

Massmart CEO Grant Pattison said yesterday he was prepared to meet the government to talk about the Walmart takeover, but was not prepared to negotiate terms.

‘It would not have set a good precedent for other foreign investors into South Africa if every transaction that ever happened required a business to have a negotiation with government first,’ he said.

The appropriate bodies to assess the merger were the Competition Tribunal and Competition Appeal Court, he said.43

The ministers have also made it clear that neither competition nor consumer interests will be considered when making these decisions. The senior official of the Department of Trade and Industry has explicitly stated that it is government’s intention to expand – it appears by their ad hoc, opportunistic interventions rather than by a legislative amendment – the definition of public interest. In short, government has decided to elevate the place of public interest over competition in the decision-making process and take upon itself a primary role in determining public interest remedies. These ‘remedies’ will be imposed even where there is little evidence that the public interest, as defined in the Competition Act, has been compromised. Government will simply use the existence of a public interest test as leverage, either by the simple expedient of imposing costly delays on merger approvals or by the threat of an adversarial future relationship with government, a prospect few businesses entertain lightly.

Nor are the ministers who engaged in the Walmart saga constrained by the consumer interests defended by the Tribunal. Quite the contrary. Interviewed by one journalist, the Minister of Trade and Industry ‘acknowledged that Walmart’s arrival might help people as consumers but said “we have taken the decision [to intervene] in favour of the production sector at the expense of other sectors such as the consumer sector”’.44

Public Interest: Conclusions

This discussion of public interest is disproportionately larger than the time and energy it actually consumed in merger evaluations. However, I think that it’s important to demonstrate that, textbook orthodoxy notwithstanding, it is possible to incorporate public interest issues credibly into core competition evaluations, without doing violence to the principal mandate of an agency charged with defending and promoting competition.

This has been achieved through a combination of two factors.

First, the public interest is clearly defined in the Act and we have resisted extravagant expansions of its ambit. This has massively reduced the uncertainty associated with catch-all terms like ‘public interest’, which, in truth, almost invariably reflects the efforts of a single interest group to equate its interests with those of a conveniently defined public. In applying the public interest, a competition authority has constantly to bear in mind that in representing consumers it is not only representing a public interest, but arguably the only interest shared by all of the public. When making the pure judgement call that is involved by attempting to place competition and ‘public interest’ on the same scale, I’ve always found it useful to think of this as balancing two elements of public interest. It is not surprising then that the scales generally tip in favour of the larger public.

Second, and in direct contradiction of the views of the Act’s critics, we have been immensely well-served by the incorporation of the public interest criteria into the Competition Act and by the decision that the competition authorities be given responsibility for the balance between the competition and the public interest criteria. I can easily imagine that, had the balance between a well-reasoned competition decision and the public interest been struck behind closed doors by a minister and a powerful political lobby, the outcome would have favoured the well-organised insiders over the broader public interest represented by less well-organised consumer interests. This must not only be an extremely demoralising experience for the competition authority; there is also the fear in the real danger that it might tailor its competition decisions to limit the prospect of ministerial override. I have heard a previous head of the Bundeskartellamt torturing competition law to explain why protecting small retailers against their larger and more efficient competitors is consistent with competition orthodoxy. Far preferable would be to balance the pro-competitive orthodoxy represented by the large retailer against the many public interest arguments that may be made in favour of retaining a class of small traders.

However, the Walmart fiasco has been a sharp wake-up call. It bears out the caution advocated by those who warned us of the potential abuse of the public interest provisions when employed by an executive power that has little respect for regulatory independence or competition, and that is determined to use the public interest test as a lever to attain ill-considered industrial policy objectives, even when the attainment of those objectives is directly in conflict with consumer interests.

However, in the wake of the current crisis of liberalised markets, the defenders of the market are going to have to acknowledge the salience of public interest to a far greater extent than the orthodoxy has permitted. The alternative prospect is the increasing marginalisation of the orthodoxy. In this new world, our treatment of public interest may well become an element of a new orthodoxy.

NOTES

1. Lee Mendelsohn and H.B. Senekal, ‘Assessment of mergers face[s] new deadlines’, Business Law and Tax Review, Supplement to Business Day, 10 May 2010.

2. Harmony Gold Mining Company Ltd / Goldfields Limited (43/CAC/Nov04).

3. African Media Entertainment Limited v David Lewis, Norman Manoim, Yasmin Carrim, Primedia Ltd, Capricon Capital Partners (Pty) Ltd, New Africa Investments Ltd, the Competition Commission (68/CAC/Mar07).

4. Business Ventures 790 (Pty) Ltd / Afrox Healthcare Limited (105/LM/Dec04).

5. Sasol Ltd and Others / Engen Ltd and Others (101/LM/Dec04).

6. Caxton and Another v Naspers Ltd and Others (16/FN/Mar04).

7. Mobile Telephone Networks Holdings (Pty) Ltd / Verizon South Africa (Pty) Ltd and Another (81/LM/Jul08).

8. Primedia Ltd and Others / The Competition Commission (39/LM/May06).

9. Harmony Gold Mining Company Ltd / Gold Fields Ltd (93/LM/Nov04).

10. There has never been a dissent in a judgment of the Competition Appeal Court.

11. Hospital Corporation of America v Federal Trade Commission 807 F.2D 1381 (1986).

12. Federal Trade Commission v Staples Inc 970 F.Supp. 1066 (D.D.C. 1997).

13. JD Group Ltd / Ellerine Holdings Ltd (78/LM/Jul00).

14. Brown Shoe Co. v United States 8Ct. 1502, 370 U.S., 8 L.Ed. 510.

15. For a particularly clear example of the influence of national characteristics in market definition, see Distillers Corporation (SA) Limited / Stellenbosch Farmers Winery Group Ltd (08/LM/Feb02).

16. Telkom SA Ltd / Business Connexion Group Ltd (51/LM/Jun06).

17. The Tongaat-Hulett Group Limited / Transvaal Suiker Beperk (83/LM/Jul00).

18. Medicross Healthcare Group (Pty) Ltd / Prime Cure Holdings (Pty) Ltd (11/LM/Mar05).

19. Ibid., para 194.

20. Medicross Healthcare Group (Pty) Ltd and Prime Cure Holdings (Pty) Ltd v The Competition Commission (55/CAC/Sept05) para 16. And this despite the Court citing a few paragraphs of its own dictum enunciated in an earlier case that ‘In its decision as to whether to set aside, amend or confirm the decision of the Tribunal, this Court must be cautious before imposing its own conception of the policy considerations upon the decision adopted by the Tribunal’.

21. Ibid., para 39.

22. Cited in Trident Steel (Pty) Ltd / Dorbyl Ltd (89/LM/Oct00).

23. This is a trade-off whereby the gains resulting from cost savings are set off against the losses resulting from the effect of market power on prices.

24. Trident Steel (Pty) Ltd / Dorbyl Ltd (89/LM/Oct00).

25. Ibid.

26. Globe & Mail, 14 February 2011.

27. Sasol Ltd and Others / Engen Ltd and Others (101/LM/Dec04).

28. Business Venture Investments 790 (Pty) Ltd / Afrox Healthcare Limited (105/LM/Dec04).

29. Shell South Africa (Pty) Ltd / Tepco Petroleum (Pty) Ltd (66/LM/Oct01).

30. Shell South Africa (Pty) Ltd / Tepco Petroleum (Pty) Ltd (66/LM/Oct01).

31. Coleus Packaging (Pty) Ltd / Rheem Crown Plant, a division of Highveld Steel and Vanadium Corporation Ltd (75/LM/Oct02).

32. Crown Cork had been the dominant supplier of bottle tops. However, in 1999 it failed to reach agreement on a supply contract with SAB, which then switched its allegiance to Rheem and, in so doing, effectively caused Crown Cork to exit the market.

33. Telkom SA Ltd and TPI Investments and Praysa Trade 1062 (Pty) Ltd (81/LM/Aug00).

34. Tiger Brands Ltd, Ashton Canning Company (Pty) Ltd, Newco and Langeberg Foods International, Ashton Canning Company (Pty) Ltd (46/LM/May05).

35. Metropolitan Holdings Limited / Momentum Group Limited (41/LM/Jul10).

36. Ibid.

37. Ibid.

38. The saga surrounding Walmart’s acquisition of Massmart is the landmark case here and is detailed below. For an earlier foretaste of what was to come in Walmart, see Edgars Consolidated Stores / Rapid Dawn 123 (Pty) Ltd (21/LM/Mar05).

39. Walmart Stores Inc. / Massmart Holdings Ltd (73/LM/Nov10).

40. Ibid.

41. Small, micro and medium-sized enterprises.

42. Ibid.

43. Business Day, 26 August 2011.

44. Business Report, 10 August 2011.

4. Abuse of dominance

How to characterise practices variously referred to as ‘abuse of dominance’ or ‘monopolisation’, or, in the anodyne and purposely neutral language of international debate, simply as ‘unilateral conduct’? ‘Complex’ and ‘controversial’ are the first descriptors that come to mind, with ‘intellectually challenging’ and – a consequence of the obstructive legal stratagems inevitably employed by a well-resourced, determined respondent – ‘downright tedious’ as close seconds. All that is certain is that, in all but the most straightforward transgressions, there will be as many views as there are economists and lawyers contesting and commenting upon the case at hand, and that those views will change, sometimes diametrically, over time and between cases.

But equally certain is that, as with all law enforcement, those who enter the world of competition enforcement are, by and large, driven by a notion of justice that has a likely set of miscreants firmly in sight. In competition law enforcement, the ‘bad guy’ is inevitably represented by a large powerful firm, acting either on its own or in combination with other large firms to compromise the interests of those – consumers, or actual or potential new entrants – less powerful than themselves. Certainly, the lived experience of antitrust enforcement will refine that approach; it will prove many of those a priori suspicions unwarranted, and will, indeed, engender respect for those large firms that reproduce their success by renewed investment, by innovation, by penetrating new markets. And experience will also expose that much abusive conduct is rooted in public action, past and present.

There are, of course, all too many instances of senior competition officials whose engagement with the activities of their agencies is dominated by a view that stresses the infrequency of bad mergers, the instability of cartels, the potentially chilling effect of abuse-of-dominance prosecution, and the likelihood of monopoly (and especially its abuse) attracting new entry. I recall my surprise at hearing one respected antitrust official telling an OECD committee that he encourages his colleagues to welcome monopolies because they inevitably attract new entry. However, while a check on rampant populism is necessary, constant ‘nay-saying’ has a demoralising, depressing effect on law enforcement and regulation. If you don’t come to work driven by a determination to get the bad guys and are not animated by the view that there are plenty of them out there, then you’d better stay at home or go and work for the other side.

And so those entering the world of competition enforcement will, for the most part, do so sceptical of the icons of the corporate landscape. For the new recruit into the professional ranks of the South African competition authorities, names like Sasol, South African Airways, Arcelor-Mittal, Telkom and SABMiller will be emblazoned on their minds. These represent the firms that have the means and, in popular perception, the will to exclude their competitors from the market or exploit their customers. Experience may establish that many of these giants, including those that are proven offenders of competition law, are also innovative competitors. But good sense will nevertheless dictate that these are the firms that will command the attention of an agency for whom a monopolistic market structure represents the ultimate obstruction to pro-competitive outcomes.

This is neither inappropriate nor surprising. While the dangers inherent in law enforcement agencies developing a priori profiles of likely law-breakers are widely recognised, many decades of law enforcement experience and scholarship have helped identify the underlying conditions conducive to collusion, just as it has taught us that only a select number of firms are able to engage in unilateral conduct capable of excluding competitors or exploiting consumers. This does not imply that every dominant firm will engage in anticompetitive conduct, any more than it implies that commodities markets are always rigged. Nor, on the other hand, does it preclude the possibility of anticompetitive conduct emanating in unlikely circumstances. Our own early enforcement experience was full of surprises. But knowing the profile of a potential law-breaker does help focus the activities of a resource-strapped law enforcement agency, and in antitrust law that profile is well known and widely accepted.

Chapter 2 of the Competition Act prohibits restrictive horizontal practices, restrictive vertical practices and abuse of dominance. While I don’t intend summarising the contents of the chapter, the pertinent provisions will naturally be referred to when I discuss a sample of illustrative cases. The language of the entire Act – and this includes chapter 2 – draws on an eclectic mix of foreign statutes and case law, with restrictive practices codified at a high level of detail, certainly relative to the sparse treatment in US law and in the EU treaty.

While all complaints regarding restrictive practices must be submitted to the Competition Commission, which will then investigate the complaint and decide whether to prosecute, there are limited rights of private action, essentially two routes through which a private party may come directly before the Competition Tribunal.

First, while the Commission is in the process of investigating a complaint, the complainant may apply for interim relief. While this is clearly intended for dealing with abuse-of-dominance allegations, I’ve already elaborated why the provision for interim relief, although it remains an important element of the Act, has not realised the promise held out for it by the Act’s drafters.

Second, if the Commission decides not to prosecute a matter and provides the complainant with a certificate to this effect, the complainant is entitled to refer the complaint to the Tribunal. In addition, if the Commission has neither referred a complaint to the Tribunal nor issued a certificate within a year of its submission, nor has it secured an extension on its period of investigation – which would require the complainant’s agreement – the complainant may refer the complaint directly to the Tribunal. Some of the most important complaints adjudicated by the Tribunal have been referred to it by the complainant. Of course, in the case of self-referrals (including applications for interim relief) the complainant not only bears its own considerable litigation costs but is also, if unsuccessful, vulnerable to an adverse costs order.

As in the discussion on mergers, I’ll confine my review to a small sample of illustrative cases. I’ll look at Federal Mogul, a complaint brought under section 5 in terms of which restrictive vertical practices are prohibited.1 This case is interesting, not least because it was the first instance in which the Tribunal imposed an administrative penalty. In this instance it was for minimum resale price maintenance (RPM), which is, or, at least used to be, a ubiquitous practice despite the fact that it has been a per se contravention of competition law since the 1970s.

Aside from this per se prohibited offence – where, of course, it is not necessary to prove harm to competition in order to secure a conviction – section 5 also provides for the prohibition of all those vertical agreements that are proved to harm competition.

I’ll naturally examine several of our most important abuse-of-dominance cases, in particular the loyalty rebate case brought by the Commission against South African Airways;2 the Commission’s referral of the complaint of exclusionary conduct against British American Tobacco South Africa,3 the dominant cigarette manufacturer; the excessive pricing complaint brought by Harmony Gold Mining Company against the ArcelorMittal steel monopoly;4 and the price discrimination case brought by Nationwide Poles against Sasol Oil.5

In the following chapter, I’ll turn to examine the cartel cases that have defined enforcement of the Competition Act in recent years and have wrought far-reaching changes in the practices and strategic priorities of the Commission, in the understanding of business, government and the broader public of the role and place of competition enforcement, and, ultimately in the provisions of the Act itself.

But what to say about enforcement generally? It is, as I have said, the heart of the antitrust matter, an assertion verified by the public response to the authorities’ handling of its law enforcement functions. There can be no doubt that the transition from an authority that was in its early years principally focused on merger review to one focused on enforcement has dramatically strengthened the culture of competition in South Africa.

We have heard, as I trust this chapter will bear out, some fascinating and important abuse-of-dominance cases. However, the overwhelming recollection is of dilatory legal stratagems and case records that must have accounted for veritable forests of paper. Because of the adversarial character of these proceedings and because the respondents’ lawyers are keenly sensitised to any administrative conduct that may allow for an appeal and, hence, delay, it is difficult to truncate hearings by, for example, disallowing tangential evidence and excessive cross-examination.

In a recent judgment of the Competition Appeal Court, Judge Wallis, a newly-appointed acting member of the Competition Appeal Court bench, took issue with the inordinate length of heads of argument and the size of the record filed in the Netstar appeal.6 He accordingly prescribed strict limitations on the scale and character of filings for a hearing before the Court. It’s obviously easier to impose these limitations in an appeal hearing. However, the Tribunal should take comfort from the Court’s willingness to prescribe standards in order to limit the duration of hearings. Following the remarks made in that judgment, I think that the Tribunal is entitled to assume that the Court would look favourably on the Tribunal’s imposing reasonable limitations in respect of its own hearings. As I’ll elaborate, it’s certainly the only aspect of the Court’s Netstar judgment that could give any comfort to those interested in promoting competition.

But let’s take a look at some of our defining cases involving vertical agreements and abuse of dominance.

SECTION 5: VERTICAL RESTRICTIVE PRACTICES

Resale Price Maintenance

Section 5(1) of the Act provides for a general prohibition of agreements between parties in a vertical relationship (that is, agreements between input suppliers and manufacturers or between manufacturers and wholesalers or retailers) that have the effect of substantially lessening competition. This general prohibition does, however, allow the parties to prove countervailing efficiency gains that arise from the agreement and offset the lessening of competition.

I’m not going to spend any time on section 5(1). Although it has been invoked on occasion, it’s extremely difficult to prove a substantial lessening of competition unless at least one of the firms party to the vertical agreement is dominant. Hence, most vertical restraints are brought under the abuse-of-dominance provisions, with section 5(1) offered as an alternative basis for a finding when the complainant is not confident about the boundaries of the relevant market or about securing a finding of dominance. However, where dominance is in doubt, it will inevitably mean great difficulty in proving that the effects of the conduct – which must be proved to have derived from an agreement between vertically-related parties – will harm competition.

Section 5(2) prohibits the practice of minimum resale price maintenance – typically an arrangement between a manufacturer or brand owner and a wholesaler or retailer whereby the resale price of a product or service is specified – but it does not provide for an efficiency defence. In other words, minimum RPM constitutes per se prohibited conduct.

Antitrust law has long treated minimum RPM as a per se prohibition. It not only eliminates intra-brand price competition – price competition among sellers of the same brand – but it may also constitute a mechanism for collusion among either the producers of different brands or the wholesalers or the retailers. However, the recent 5–4 decision of the US Supreme Court in Leegin7 reversed a long-standing earlier decision in Dr Miles Medical Co.8 that had, in 1911, imposed a per se prohibition on minimum RPM in that country. The Leegin decision reversed Dr Miles by ruling that RPM should henceforth be subject to a rule-of-reason test. In other words, in the US it is now necessary to prove that RPM has lessened competition.

The majority in Leegin held that while particular instances of RPM may indeed harm competition, it’s also conceivable that, in other instances, the same practice may promote competition. Hence, lest pro-competitive conduct be prohibited, the court deemed it necessary that RPM be examined on a case-by-case basis rather than be prohibited outright. While we shall examine the danger of ‘chilling competition’ when we turn to our experience of abuse-of-dominance cases, it’s as well to introduce the subject now in the context of one of the few vertical restraints that are prohibited per se. Clearly for those concerned at the chilling effects of a possible prohibition of pro-competitive conduct, the prospect is enhanced in the case of a per se offence where all that has to be established are the elements of the impeached conduct, where, in other words, there is no requirement on the part of the plaintiff to establish anticompetitive effects and where the defendant is barred from invoking a pro-competitive defence.

Federal Mogul

Resale price maintenance, which appears to be a fairly commonplace practice, has proved to be one of the few low-hanging fruits in the enforcement arena. And so in the early years, when it was very difficult to prosecute anticompetitive conduct, RPM became extremely useful in establishing a degree of credibility for the competition authorities in law enforcement. It’s also illustrative of some important general issues surrounding enforcement.

A number of RPM agreements were unearthed, ranging from automobile assemblers and auto components to luxury sunglasses. In each of the auto manufacturers and sunglasses matters, consumers found, after shopping around, that they were being offered an identical discount on a popular auto brand or an identical price on a brand of luxury sunglasses, and they reported the apparent contraventions to the Commission. In the auto industry this opened a veritable Pandora’s box: most of the large global manufacturers filed through the Tribunal hearing room to have consent orders approved, including having to pay administrative fines.

The Tribunal’s judgment in Federal Mogul is our most important and illustrative RPM matter.9 The proof required for the finding of a contravention is not onerous. This is examined in some detail in Federal Mogul, which concluded that all that has to be established is knowledge on the part of the retailers of a minimum resale price determined by the producer or its agent or wholesaler backed up by a sanction for non-compliance. These elements were clearly present in Federal Mogul. Moreover, while the per se nature of the offence meant it was not necessary to prove the anticompetitive effects of the conduct, the Federal Mogul case does, I believe, support the view that it should indeed be prohibited per se.

Without attempting a comprehensive summary of the matter, the essential facts are that AZ Friction Products, the leading retailer of the highly regarded Ferodo brand of braking components in South Africa, began to lose market share after it had been acquired by Midas, a large general auto components retailer, itself part of a large South African-owned engineering conglomerate. An employee of Midas – a Mr Koos Erasmus – a long-standing and highly successful salesperson in the erstwhile AZ Friction Products, disaffected at the new corporate environment in which he found himself, left Midas and managed to acquire franchise rights for Ferodo from Federal Mogul, the US multinational that acted as Ferodo’s wholesale distributor in South Africa.

Although Federal Mogul clearly imposed a minimum resale selling price on its Ferodo products, Erasmus, the complainant in this matter and recently enfranchised Ferodo retailer, began to take market share from his erstwhile employer, Midas, Federal Mogul’s largest South African customer for its considerable portfolio of auto component brands, now including Ferodo braking products. Erasmus achieved this without contravening the RPM requirement. That is, he did not compete on price. The elements of his successful competitive strategy are not altogether clear – he may have offered a better service; he may have been a more energetic and aggressive salesperson; he clearly exploited the networks that he’d established after many years of employment in a specialist braking products retailer; he may simply have been well placed to take advantage of the instability that so often follows a merger, in this instance the acquisition of AZ Friction Products by Midas. Recall that after the acquisition he had worked at Midas for some time and so he knew where its weak points were, including the identities of disaffected customers.

Midas sought to regain its market share by cutting the prices of Ferodo products, thus contravening the RPM arrangement. After Erasmus and the other brake product wholesalers had appealed, unsuccessfully, to Federal Mogul to enforce the agreed resale price on Midas (in the process establishing, I think, that RPM was, in part at least, the platform upon which a retail cartel was organised), Erasmus – who had been principally responsible for Midas’ decline in market share – responded by cutting his Ferodo prices. Intra-brand competition had broken out with a vengeance and, it appears, on all fronts simultaneously.

However, Federal Mogul was not prepared to countenance this. The explanation proffered was that because Midas, as a general auto components dealer rather than a specialist braking products retailer, was a larger and more valued customer of Federal Mogul than Erasmus or the other specialist braking products resellers, Federal Mogul elected to reimpose its preferred resale price by disciplining Erasmus for engaging in a price war even though he had not initiated the price competition. It did this by cutting the rebate that Erasmus received – which was a standard volume-based offering applied throughout the trade and which determined the retailers’ margins – thus significantly reducing Erasmus’s margins. The evidence showed that while Federal Mogul was prepared to countenance a temporary transgression of its prescribed retail price on the part of its favoured customer in order to enable Midas to regain its lost market share, it was not prepared to countenance ongoing price competition. It was, in short, not prepared to allow Erasmus, Midas’s most successful competitor, to compete on price in order to hold on to his newly-won market share. Federal Mogul’s clear expectation was that once Midas had, through its price cut, restored its pre-competition market share, it would revert to the prescribed resale price.

Because RPM is subject to a per se prohibition, Federal Mogul was not entitled to put up a pro-competitive defence. It did, however, suggest that the basis for its support for RPM was to ensure that price competition between Ferodo retailers did not result in retail margins too low to sustain an effective retail distribution system. However, while I have no difficulty in accepting that a supplier may be legitimately concerned with the financial viability of those franchised to retail its products, this does not seem to necessitate lessening competition through the imposition of a minimum resale price on its franchisees. There are many alternative mechanisms for monitoring financial viability and for withholding franchise rights from a franchisee who does not meet objective standards of financial sustainability.

In short, it seems that Federal Mogul didn’t want any competition in the retail market at all. It actively discouraged non-price competition by allowing the loser in non-price competition, namely Midas, to recoup its position through granting it a temporary and exclusive right to contravene the RPM rules and engage in price competition. It’s interesting to note that once Midas started reducing the prices of Ferodo products, Erasmus’s non-price-competitive strategies were not able to measure up. In Leegin, there is a perverse suggestion by the majority to the effect that permitting RPM (that is, prohibiting intra-brand price competition) will encourage other forms of intra-brand competition. This may well be borne out by Erasmus’s ability to gain market share from Midas despite not engaging in price competition. However, it’s clear that once price competition broke out, South African consumers of braking products clearly indicated their preference for lower prices by returning to Midas, and this is why Erasmus was ultimately obliged to resort to price competition himself, even at the risk of incurring the wrath of his supplier. Effectively, then, Federal Mogul discouraged any ‘disorder’ in the marketing of Ferodo products, and this it achieved by eliminating all forms of intra-brand competition.

Because the Competition Commission was not required to prove competitive harm, RPM being a per se offence, we learnt little about the strength of competition, inter-brand or intra-brand. We did know that Ferodo was the largest brand on the market. However, in considering, for the purpose of deciding the size of the administrative penalty, the loss or damage caused by the contravention, the Tribunal noted that the extent of harm depended in part on the impact that Federal Mogul’s conduct had on the rest of the brake products market – that is, on inter-brand competition. We had been told that Ferodo’s market share was approximately 30 per cent. Given this significant market share, we contemplated the possibility that Ferodo’s anxiety to prevent intra-brand competition between its distributors may well have reflected a fear that a discount on Ferodo products might lead to a price war with other braking brands. In other words, the RPM agreement in Ferodo was, as it often is, also the platform for a cartel in braking products. Were this to be the case, the harm to competition arising from RPM would have been very serious indeed.

While the evidence necessary to establish the existence of a cartel would probably not have emerged through a rule-of-reason RPM investigation, by prohibiting RPM in respect of the most important braking products brand we in all likelihood put a stop to a cartel or, at least, to price leadership and price ‘stability’ in this product market. So the Leegin majority’s speculation about the possibility of enhanced inter-brand competition attributable to RPM may just as easily be counterposed by speculation of much weakened inter-brand competition attributable to the same conduct. Of course, the answer to this is that a requirement to prove the effects would end the speculation. However, cartels are difficult to prove, especially for a fledgling authority. RPM, on the other hand, is easy to prove. Moreover, its negative impact on intra-brand competition is clear and its role in creating and maintaining cartels is well documented. So it’s legitimate to ask whether competition enforcement is well served by having to prove the anticompetitive effect of RPM.

An interesting dissenting judgment in Leegin – authored by Justice Breyer on behalf of four of the nine members of the Leegin bench – argues for the retention of per se illegality for RPM. One of the arguments underpinning the dissenting judgment is ‘administrability’, essentially an argument for clear, easily understood and applied rules, even if they end up curing no great mischief (because of the relative and continued strength of inter-brand competition), and indeed even if they eliminate a rare instance of significant pro-competitive RPM. Breyer argues that eliminating per se rules in favour of rule of reason standards significantly increases the difficulties facing prosecutors, who would be obliged to apply complex economic evidence and reasoning to easily identified conduct, much of which is indeed anticompetitive.

I’ll return to this general issue when I look at abuse of dominance. However, it’s important to note that RPM is, after all, a vertical price restraint and so it eliminates what is the cornerstone of most competitive strategies. The same considerations may not apply to non-price restraints. Hence, if Ferodo required that its branded products could be sold only by dealers capable of providing a particular level of after-sales service – and this capacity may well include evidence of financial stability – this too may be easier to sustain than a requirement that limits the retailer’s ability to determine, within the need to provide a prescribed level of after-sales service, his competitive strategy, in particular his pricing strategy.

So, with all due respect to the US Supreme Court, I would not be in a hurry to convert our per se prohibition of RPM into a rule-of-reason standard. The gains would be small and infrequent. The losses may be counted in various forms – certainly less intra-brand competition and possibly compromised inter-brand competition. Certainly, the consequences would be manifest in significantly more complex legal proceedings. As it is, the Federal Mogul matter was submitted in late 1999. The appeal was handed down in December 2003, and this after several interlocutory hearings before both the Tribunal and the Competition Appeal Court, the latter arising from a constitutional point taken by Federal Mogul against the Tribunal’s power to impose an administrative penalty. Delay was also caused by the inevitable late filing of papers by the respondent, Federal Mogul, and also by dilatory conduct on the part of a resource-strapped and inexperienced Commission. Had we been compelled to hear complex economic evidence and argument in order to make a finding on the competitive effects of the conduct in question, the matter would have taken longer, it would doubtless have involved expert witnesses, and everybody’s costs would have multiplied several times. And the prospect of a different outcome would have been slender, to put it mildly.

Reading the majority judgment in Leegin, the strongest take-home message I receive is one that says that if firms are free to conduct themselves in a thoroughly unrestrained manner – including conduct that, as with RPM, clearly eliminated an important element of price competition – they would then be inspired to find other means of competing. But where does this end? As I’ll show in the later discussion of abuse of dominance, it ends in praise of dominant firms, because their monopoly rents will inspire new entrants. I can’t make up my mind whether we have in Leegin a particularly red-blooded notion of the competitive process or a strangely perverse notion. I think it’s the latter. The majority in Leegin effectively say: allow traditional and well-established forms of competition to be suppressed because that will bring forth ever more robust and creative alternative forms of competition, presumably because to compete vigorously is part of the natural order of things that will tolerate no absolute suppression. If competition, whether through new entry or product innovation, is so much part of the natural order of things, then there’s no need ever to act against dominant firms. And that of course is the essential view of the majority of the US Supreme Court and, many believe, of its enforcement agencies, in particular the Department of Justice during the Bush era.

The other significant feature of Federal Mogul is that this was the first occasion on which an administrative penalty was imposed in a contested matter. Although fines are relative to turnover, the media and the public tend to react to the absolute level of the penalty and, at that time, the R3 million fine imposed on Federal Mogul was considered to be very significant. Indeed, I recall a lengthy debate among the panel members about the likely reaction to such a large penalty.

Inevitably our imposition of an administrative penalty in this matter attracted a constitutional challenge – which sought to characterise our penalties as civil in name only, while being criminal in kind – which Norman Manoim easily disposed of in a judgment upheld by the Competition Appeal Court.10 I am not sure that the taking of this point was related to perceptions concerning the size of the penalty or simply because it was an opportunity for delay and obstruction. Certainly, today a fine of this size would not merit any attention whatsoever. In fact fines hundreds of millions of rands and greater are nowadays considered – by the public at least, if not necessarily the courts – to be derisory slaps on the wrist, a clear indication that the public has come to appreciate the price it pays for Competition Act contraventions.

Federal Mogul was really the first full-blown restrictive practices case we had heard. It was a good case on which to cut our teeth. The legal and economic complexity was limited by the per se nature of the offence and the salient facts were easily discerned and understood. But, as I later came to learn, in common with most adversarial proceedings we spent much time dealing with factual disputes that ended up having little or no bearing on the matter at hand.

Relatively straightforward though this case was, on re-reading the judgment I am nevertheless struck by the importance of the factual narrative, by the importance of identifying and telling a clear and plausible story. All antitrust inquiries are intensely fact-based. However, in mergers, where the sheer quantity of factual material is legion, the predictive character of the inquiry always lends an ineluctably speculative dimension to the interpretation of the facts. But restrictive practice inquiries, although legally and economically complex, are backward-looking and so the understanding of the case is reflected in the ability to assemble a coherent and plausible storyline from the facts presented and, of course, to decide when a witness is indeed presenting fact or fiction. This is why deference, on the part of appeal bodies, to the fact-finding tribunal is so important. It is this body that is responsible for assembling the facts drawn from the documentary and oral record into a coherent narrative, and that is uniquely entrusted with assessing the honesty of the witnesses. Of course this judgment is not only based on observing the demeanour of the witness – although that is important – but also on testing the oral evidence against other evidence and against rationality, which leaves the appeal body with some limited room for disagreeing with the fact-finding tribunal.

We’ve tried in our restrictive practices judgments to preserve, amidst the legal and economic complexities, the narrative, the storyline. To the extent that we succeeded in this we owe much to the experience of some of the counsel appearing before us. Not all of them: sometimes those with the finest understanding of the law and the economics were the least capable of, or possibly attached the least importance to, assembling the storyline. But those who were attentive to this essential requirement in the making and drafting of judgments appeared before us with an immediate advantage because they spoke the language in which we ultimately had to render our judgment.

The person who immediately springs to mind is Owen Rogers, a senior counsel from Cape Town. His ability to marshal a vast and complex array of facts into a coherent story, usually with the aid of a particularly ancient pocket calculator, is truly awesome. I recall my palpable relief on seeing him in the huge legal teams involved in the Sasol–Engen merger, the mother (and father) of all fact-intensive matters that we handled. From previous experience of his participation in Distell, I knew at once that, among all the brilliant minds before us, this was the person who would most assist us in finding the essential threads that had to be drawn out of the hundreds of thousands of pages of documentary evidence and the hundreds of hours of oral evidence we were going to have to consider. And I was right. Of the array of senior counsel representing the many parties involved in that matter, he was the least showy, the least dramatic, the least inclined to smart-ass argument. But from the opening bell he commanded our most rapt attention, because he assisted us in identifying the essential storyline, without which a judgment, however theoretically sound, simply ‘won’t write’.

SECTION 8: ABUSE OF DOMINANCE

This is the most intellectually engaging and controversial area of competition law enforcement. Abuse of dominance is conduct perpetrated by a single firm rather than by a combination of firms. It is not the product of an agreement, but rather the conduct of a single firm capable of acting without regard for its competitors, suppliers or customers, hence ‘unilateral conduct’. I’ll illustrate this through an examination of some of our most important decisions, notably those in South African Airways, British American Tobacco South Africa, ArcelorMittal and Nationwide Poles.

But first, a brief summary of the relevant provisions of our statute. Section 7 of the Competition Act provides that a firm with a market share of 45 per cent or more is presumptively dominant. A firm with a market share of 35 per cent or more but lower than 45 per cent is considered dominant unless it can prove that it does not have market power. And a firm with a market share lower than 35 per cent may be dominant if it is established that it possesses market power. Market power is defined as the ‘power of a firm to control prices or to exclude competition or to behave to an appreciable extent independently of its competitors, customers or suppliers’.

Section 8 lists the conduct that is prohibited. These acts can be categorised in a variety of ways, one being the distinction between acts subject to a per se prohibition and those subject to a rule-of-reason test. Another line of demarcation would distinguish between ‘exploitative’ abuses and ‘exclusionary’ abuses. The only exploitative abuse named in section 8 is that of ‘excessive pricing’.

There are only two ‘full’ per se offences listed, namely the charging of an excessive price (the only ‘exploitative’ abuse, the rest being ‘exclusionary’) and refusal of access to an essential facility. In regard to these offences, respectively section 8(a) and 8(b), there is no pro-competitive defence and so no competitive harm that has to be established. An administrative penalty is permissible in respect of a first contravention of these provisions.

The remaining section 8 offences all provide for the possibility of proving (or disproving) countervailing efficiency or pro-competitive gains arising from the conduct in question. In the case of a contravention of section 8(c), which prohibits a general ‘exclusionary act’, the complainant or plaintiff bears the onus of proving that the anticompetitive effects of the conduct in question outweigh any ‘technological, efficiency or other pro-competitive gains’ arising from it. An ‘exclusionary act’ is defined as ‘an act that impedes or prevents a firm entering into, or expanding within, a market’. An administrative penalty may not be imposed in respect of a first-time contravention of section 8(c).

Section 8(d), on the other hand, describes a number of exclusionary acts that include well-known conduct such as predatory pricing, a refusal to supply and tying. In these instances the onus shifts to the perpetrator to prove the countervailing efficiency gains. While I don’t think that the description in South African Airways of the section 8(d) offences as ‘limited per se’ offences is accurate, it is clearly the case that the complainant only needs to prove the existence of the elements of the conduct in question – for example, a refusal to supply a scarce good to a competitor when supplying the good is economically feasible – in order to prove that the conduct is exclusionary. Unlike the case of section 8(c), there is no requirement to prove that the conduct ‘impedes or prevents a firm entering into, or expanding within, a market’ in order to establish that it’s exclusionary.

However, these sections of the Act specifically provide that the respondent may put up a countervailing pro-competitive defence and seek to prove that the pro-competitive consequences of the conduct outweigh the anticompetitive effect. South African Airways confirmed that because ‘exclusion’ can as easily be the consequence of pro- as anticompetitive conduct, it was necessary to establish anticompetitive effects in order to secure a conviction under 8(c) or (d). So although an irrebuttable presumption enters into the equation when defining dominance and although legal rules may predominate when deciding that, for the purposes of section 8(d), particular conduct is exclusionary, the requirement to nevertheless prove anticompetitive effect makes this a full rule-of-reason inquiry, bolstered by the explicit provision for a pro-competitive defence. An administrative penalty may be imposed in the event of a first contravention of section 8(d).

So, in a nutshell, apart from the single instance of an ‘exploitative abuse’ – namely the proscription of excess pricing – abuse of dominance is concerned with limiting the ability of a dominant firm to maintain its position through exclusionary conduct rather than through continued innovation and investment (or competition ‘on the merits’). And then, cognisant that competition ‘on the merits’ may also result in the exclusion of a competitor, the reasoning must distinguish between robust competition on the merits and anticompetitive exclusionary conduct by proving harm to competition. In respect of general exclusionary conduct prohibited by section 8(c), the Commission or complainant must prove that the harm to competition outweighs any countervailing pro-competitive consequences, while in respect of the exclusionary practices identified in section 8(d)i–v, the onus is on the defendant to prove that the pro-competitive consequences countervail the harm to competition that the Commission or complainant must prove in order to successfully prosecute an alleged abuse of dominance.

Section 9 prohibits price discrimination by a dominant firm. It is, in a variety of aspects, accorded a treatment distinct from the section 8 abuse-of-dominance provisions and is examined below.

EXCLUSIONARY ABUSES

Chilling Competition?

The source of the controversy that attaches to potentially abusive conduct is apparent from the rule-of-reason approach to most of these practices, an approach which implicitly accepts that the effect of dominant firm conduct may be to undermine or else promote competition. As such, over-zealous prosecution of unilateral conduct may well land up proscribing pro-competitive conduct, not only on the part of the firm prosecuted, but on the part of all those dominant firms who desist from practising this conduct for fear that they too will be prosecuted. This is popularly described as ‘chilling competition’.

On the other hand, an excessively permissive approach to unilateral conduct may permit a firm to engage in conduct that undermines competition and, as the converse of the outcome of over-zealous prosecution, will signal to other firms that they too will get away with similar anticompetitive conduct. The welfare losses consequent upon either error – respectively referred to as ‘false positives’ or ‘type I’ errors arising from ‘over-prosecution’, or ‘false negatives’ or ‘type II’ errors arising from ‘under-prosecution’ – are thus potentially significant, conceivably extending way beyond the firm whose specific conduct was the subject of the erroneous prosecution and conviction or acquittal.

The orthodoxy is overwhelmingly focused on the threat of false positives rather than false negatives. Essentially, the premise that false positives represent a greater danger than false negatives derives from the application by courts or tribunals of statutory or court-made rules to economic phenomena. Given these fears, the use of a per se prohibition in circumstances where there is any conceivable doubt about the impact on competition – as in the case of RPM – is clearly anathema. Where a per se rule is applied, the interpretation and application of legal rules are all-important.

However, the fear of false positives also extends to the application of the rule of reason. Here too, of course, courts and tribunals are applying statutory or court-made rules to economic phenomena. This fear is exacerbated by the application of precedent, a key, almost sacrosanct, aspect of court practice, which may result in the application of legal rules made under wholly different economic circumstances: technologies, tastes, economic thinking and many other factors may have changed since the precedent-making judgment but the courts will still be reluctant to overturn precedent. Essentially, the anxieties of those who live in dread of false positives can be ameliorated only by minimum resort to legal presumptions, whether in the form of statutory rules or court judgments.

Aspects of South African legislation are wide open to attack from those who fear an inherent tendency towards over-enforcement and, hence, false positives. In particular, it is thought that our presumption of dominance on the basis of market shares will predispose us in the direction of over-enforcement of abuse of dominance. This, the argument continues, is less likely to occur when a finding of dominance is based on the reasoned presence of market power.

Just as our detractors suspect and fear, we incorporated a market-share test of dominance in our legislation precisely in order to limit the prospect of litigation over as slippery a concept as market power and because it enhances certainty. Given that our process of defining markets is rigorous, given that we are unlikely to make a finding of dominance when the market share is at the boundary and subject to sudden shifts, and given that a finding of dominance does not suggest the conduct in question is exclusionary, let alone an abuse of that dominance, I can’t see that our market share presumptions regarding dominance predispose us to false positive errors in abuse matters. Moreover, I have no doubt that unnecessary litigation is thereby avoided. And it’s most fruitfully avoided precisely in those cases where market power is easily established – which will inevitably coincide with very large market shares – but will, in the tradition of dilatory, cost-raising adversarial litigation, rarely if ever be conceded.

Why did we want to ease the litigation burden in pursuing abuse-of-dominance cases? Simply because we believe that in small economies with a considerable history of state participation in the economy, we are likely to encounter significant instances of dominance and abusive conduct. I think that we’ve been vindicated, particularly in relation to the source of abusive conduct. The companies that have come before us accused – and, on occasion, found guilty – of abusive conduct read like a ‘who’s who’ of South Africa’s state-owned and former state-owned or regulated enterprises: Sasol, ArcelorMittal, South African Airways, Senwes and, had jurisdictional issues not intervened, doubtless Telkom too. Each of these commands market shares considerably in excess of that required to establish dominance, as do those few firms, British American Tobacco South Africa for example, which have never been state-owned and which have been accused of abuse of dominance.

Our reluctance to subject the dominance finding to extensive litigation has also been vindicated. The lowest market share on which we have found an abuse of dominance is 56 per cent. The abuse-of-dominance cases that have actually come before the Tribunal for final adjudication have involved market shares of 56 per cent, 57 per cent, 66 per cent and 81 per cent. In fact I can’t recall many abuse-of-dominance allegations that have been filed when dominance is at or even near the threshold of 45 per cent.

I am, however, happy to acknowledge that market share and market power don’t coincide perfectly, and so, occasionally, a firm whose possession of market power is open to question will, on our market-share test, nevertheless be found to be dominant. However, market power will come back into play when the competitive effects of the alleged conduct are assessed. So the firm that, though dominant, doesn’t possess market power is suitably protected from a finding of abuse based on legal presumptions. And, given that we will, in countless other cases, have been spared costly and time-consuming litigation where it is perfectly reasonable to presume the possession of market power on the basis of market shares, I remain convinced that our approach is sound.

However, our critics will point out that it is precisely the greater certainty the market-share presumptions bring that will cause the anticompetitive chill. A firm with a market share in excess of 45 per cent will, in our competition regime, be well aware that it is vulnerable to a claim of abuse of dominance. While the uncertainties surrounding the boundaries of the market always inject an element of the unknown into the existence or otherwise of dominance, it is certainly true that a market-share test more easily permits a firm to calculate whether it is likely to be adjudged dominant than a market-power test. This may cause it to tread more carefully in devising and employing elements of its competitive strategies than may otherwise be the case, for fear that these may be found to be abusive, thus precisely engendering the undesirable chilling effect.

The same can be said about the approach taken to the potentially abusive practices listed in section 8(d). As already explained, a series of common and potentially harmful practices are identified as ‘exclusionary’ as long as the elements of the conduct are proven. That is to say, unlike the general prohibition of exclusionary acts provided for in section 8(c), the complainant in an 8(d) matter is not required to prove that the conduct ‘impedes or prevents a firm entering into, or expanding within, a market’ in order to establish that the conduct is indeed exclusionary. All that is required is to prove the existence of the elements of the practice in question.

Hence, if the evidence establishes the essential elements of a tying or bundling arrangement, then that conduct will be sufficient to establish the existence of an exclusionary act. The aim was to forewarn those engaging in these practices that they were potentially placing themselves in the firing line because they were committing an exclusionary act. So the spectre of chilling competition rears its head again – first, the dominant firm is expected to tread carefully; then, the dominant firm that engages in practices which section 8(d) identifies as exclusionary receives, as it were, a second cautionary. This is clearly outlined in South African Airways.11

However, in South African Airways the Tribunal also decided that, in order to sustain a finding of abuse, it is necessary also to prove competitive harm precisely because it recognised that exclusion is consistent with both pro- and anticompetitive conduct. Both 8(c) and 8(d) permit a pro-competitive defence – a showing that ‘the anticompetitive effects’ of the conduct are countervailed by ‘technological, efficiency or other pro-competitive gain’.12

There are three answers to those who are anxious that this would lead to false positives and hence a possible chilling of competition. First, we are dealing with law enforcement, and central to all regimes of law enforcement is deterrence. I’m comfortable with the idea that dominant firms should more self-consciously question their conduct than nondominant firms. They need, however, to ask themselves the right question. To pose the question as a lawyer might – ‘Is my planned conduct likely to contravene the law?’ – is more likely to chill pro-competitive conduct (particularly if the lawyer actually becomes the arbiter of competitive strategy) than ‘Is my competitive strategy likely to benefit my customers, or merely exclude my competitors?’ The antitrust enforcer or adjudicator will, when assessing the competitive effects of the conduct, have to pose the latter question, the business person’s question. It is the requirement to establish anticompetitive effects, rather than the presence or absence of legal rules or presumptions in assessing dominance or finding exclusionary conduct, that provides the most effective safeguard against type I (‘false positive’) error.

Second, the worst of all possible worlds is one in which proxies influence decisions but are left unspecified. In these circumstances, decision-making is inevitably characterised by uncertainty – indeed, the business person doesn’t even know when to subject his or her strategy to the test posed above. It is absolutely clear that whatever the doctrine dictates, those charged with administering competition law – both in Europe and the US – make extensive use of market-share proxies. However, a bewildering array of measures is used that, at once, indicates the ubiquity of market-share presumption in establishing dominance and the danger of leaving it to every judge or enforcement official to decide what proxy they will actually use.13 Moreover, economic theory and research are sometimes treated with scepticism by antitrust enforcers, and adjudicators in particular, because of its apparent faddishness, because its most recent insights have often received scant verification and are usually subject to stringent assumptions and ultimately qualifications. All these factors make economic reasoning difficult to administer in a judicial framework. Hence Breyer’s appeal, on the grounds of administrability, to use easier-to-apply legal rules, even if, on the rare occasion, they lead to erroneous conclusions.

The US Supreme Court judgment that has, for fear of chilling competition, most explicitly elevated the primacy of economic reasoning over the application of legal rules is Trinko. Ironically this, the judgment most hostile to legal presumption, is effectively predicated on a presumption that monopoly be viewed as the positive outcome of competition on the merits, and, moreover, that the pricing power and commercial returns that accompany monopoly will inevitably attract new entrants. In a widely-cited passage, Justice Scalia famously notes: ‘The mere possession of monopoly power, and the concomitant charging of monopoly prices, is not only not unlawful; it is an important element of the free-market system. The opportunity to charge monopoly prices – at least for a short period – is what attracts “business acumen” in the first place; it induces risk taking that produces innovation and economic growth. To safeguard the incentive to innovate, the possession of monopoly power will not be found unlawful unless it is accompanied by an element of anticompetitive conduct’.14

Certainly, South African practice would support Justice Scalia to the extent that the ‘mere possession of monopoly power [is] not unlawful’. But to ascribe ‘risk taking, innovation and economic growth’ to the possession of monopoly power seems little distant from asserting that the freedom of a dominant firm to conduct itself unfettered is, in Justice Scalia’s view at any rate, a proxy for consumer welfare!15 And this despite the powerful evidence accumulated over decades of antitrust enforcement that dominance lends itself not merely to the charging of monopoly prices, but enables conduct whose effect is precisely to exclude those wishing to enter a market whose structure has enabled the dominant incumbent to extract monopoly rents.16 It is strikingly similar to the majority view in Leegin, which I’ve characterised as supporting the elimination of intra-brand competition because it will induce new, robust forms of other competition. Both the Leegin and Trinko decisions are predicated on a decidedly perverse, contorted logic.

Third – and this is where legal advice may prove useful – a successful abuse-of-dominance prosecution confronts so many serious obstacles that all but the most risk-averse business people would, unless the conduct in question was unusually brazen and the anticompetitive effects uncharacteristically obvious, probably take their chances on getting away even with conduct that did not pass the test outlined above. Notwithstanding the anxiety surrounding our dominance presumption, it plays no role in making a finding that the dominance has been abused. The hurdles to be crossed in arriving at such a finding are considerable. Not only do the elements of the exclusionary conduct in question have to be established, but so too do the anticompetitive effects. And with the exception of excessive pricing and denial of access to an essential facility, which are both per se offences but whose essential elements are extremely difficult to prove, the Competition Act explicitly requires that the other instances of abuse are subject to an efficiency defence.

So the data show that in a period of nearly 11 years, nine abuse cases have been referred to the Tribunal. Five of these have been referred by the Commission and four have been referred by private complainants after the Commission declined to prosecute their complaints. The Tribunal has found an abuse in six of these cases. Three of the Tribunal’s abuse decisions have been appealed. The appeals were upheld on one occasion.17 In one case – ArcelorMittal – the Competition Appeal Court decided that the Tribunal had applied the incorrect test in determining that the steel monopoly had contravened the prohibition on excessive pricing, and remitted it to the Tribunal so that the test deemed by the Competition Appeal Court to be correct might be applied. The parties settled the matter before the Tribunal could re-hear it. One abuse-of-dominance referral by the Commission – a referral against Sasol – was settled and confirmed in a consent order approved by the Tribunal. It would, on the basis of these data, be difficult to conclude that either the Commission or private parties were over-prosecuting abuse of dominance and, therefore, that competitive strategy would be unduly chilled for fear of prosecution in terms of section 8.

Nor is this surprising, if for no other reason than that abuse-of-dominance prosecutions are notoriously resource-intensive. Take Arcelor-Mittal. The matter was first referred to the Tribunal in January 2004 but the judgment was delivered only at the end of March 2007. Eight advocates were used, two by the first and second applicants, four by the first respondent and two by the second respondent. British American Tobacco South Africa is another case in point. Japan Tobacco International’s complaint was referred to the Tribunal on 17 June 2005. After 33 hearing days the matter was eventually decided on 25 June 2009. Seven advocates were used: one for the Commission, and three each for the intervener and the respondent.

It’s also worth noting that an administrative penalty cannot be imposed after a finding of abuse arising from the general category of exclusionary conduct, that is from a contravention of section 8(c). Civil damages claims – which rely on the Tribunal finding an abuse (a finding that will generally be appealed and so a civil damages claim will have to wait for the outcome of the appeal) – are difficult to mount, given the difficulties entailed in filing a class-action suit. Since 1999 there has been only one claim for civil damages. Rights of private action are limited, although two of the most important abuse cases heard by the Tribunal were self-referred after the Commission declined to refer. In the case of self-referral the complainant will have to fund its own substantial legal costs and, if the respondent prevails, usually the latter’s costs as well.

Thus the risk–return calculation that influences a decision to prosecute an abuse-of-dominance case is markedly different in the US, given the relative ease with which private suits may be pursued there and with the prospect that a guilty finding may give rise to an award of treble damages. Accordingly, the prospect of vexatious, competitor-driven abuse-of-dominance suits is far more likely in the US than in most other regimes, and clearly plays some part in the high level of anxiety at the prospect of error in US scholarship and courts. However, this is a peculiarly US problem and should not influence approaches to abuse cases in the vast majority of jurisdictions that do not incentivise litigation in this way.

The long and the short of it is that the prospect of chilling competition through excessive prosecution of abuse-of-dominance cases is significantly blunted by a number of factors: the resources that the Commission or a private complainant has to expend in prosecuting these cases; the fact that the respondent will, by definition, have extremely deep pockets and will in all likelihood be called upon to defend an important element of its competitive strategy and will have no incentive to cooperate in expediting the matter or lowering the costs of litigation; and, not least, the uncertainty surrounding the outcome, particularly in view of the requirement to establish competitive harm.

South African Airways

The case that has had the greatest influence in establishing the reasoning employed by the Tribunal when adjudicating abuse-of-dominance claims is South African Airways. As already mentioned, this case clearly established the necessity to prove anticompetitive harm and it laid down the factors that must be present in order to identify anticompetitive harm. It also attempted to construct guidelines for the calculation of administrative penalties.

South African Airways was concerned with a loyalty rebate scheme, whereby the dominant airline rewarded travel agents by paying a commission based on the share of SAA tickets in the total sales of the individual agents. Note that this was not a discount based on volumes of SAA tickets sold, but rather on the proportion of SAA tickets sold relative to sales of tickets for competing airlines. The commission was increased after a specified share was reached with the increase calculated not only on incremental sales but back to the ‘first rand’ – that is, total sales of SAA tickets. The effective commission paid on the marginal ticket, on the sale of tickets after reaching the target, was thus enormous.

Although several of SAA’s competitors used similar schemes, because their market share was dwarfed by SAA, the ‘back to rand one’ commission paid by the smaller competitors was bound to be proportionally overshadowed by that of the dominant state-owned airline. Hence the loyalty rebate scheme was effectively available only to a market participant with a dominant share. The upshot, alleged the Competition Commission, was that travel agents were incentivised to maximise their sales of SAA tickets even if this involved failing to disclose to their clients the availability of competitive offerings.

SAA was found to have contravened the Act and was fined R45 million, at the time the largest penalty ever imposed. Apart from the reasoning regarding harm and the calculation of administrative penalties, the hearings served to confirm the rather unsavoury and incompetent manner in which SAA conducted business at that time and it was the first in a series of adverse findings against SAA, both on the rebate scheme and for cartel conduct.

From the perspective of our jurisprudence, much the most significant aspect of South African Airways was our finding on anticompetitive harm. In brief, we found that in order to secure an abuse-of-dominance conviction in terms of sections 8(c) and (d), it was necessary to prove anticompetitive harm. However, the panel decided that evidence of significant foreclosure constituted sufficient evidence of competitive harm. In other words, it was not necessary to prove direct harm to consumers.

This is the principal answer to those concerned about our approach to abuse of dominance giving rise to a chilling of competition. There can be no conviction without proof of competitive harm, but the standard used is probably lower than that required in the US. South African Airways was preceded by two similar cases, one in the US, the other in Europe. The former resulted in an acquittal, the latter in a conviction.

This did not mean that we had set a low bar on proving harm to competition, or that we had established an impossibly high standard on proving pro-competitive effects. Indeed, cognisant of the dangers of chilling competition as a result of erroneous convictions, we have, if anything, set a somewhat lower bar on proving pro-competitive efficiencies in abuse cases than in merger inquiries.

SAA did not appeal our decision, which was a major relief because it meant our finding on the question of competitive harm was not tampered with. When I see the extremely narrow and increasingly conservative jurisprudence emanating from the Competition Appeal Court, I find myself wondering what it would have made of the characterisation of harm that we established in South African Airways. However, it did not get to intervene in this judgment, on the basis of which Nationwide Airlines, the complainant, sued SAA in the High Court for damages. This, the first such damages claim, was settled although the terms of the settlement were not made public.

British American Tobacco South Africa

The clearest illustration of the significance attached to proving competitive harm is to be found in BATSA. This matter was referred to the Tribunal by the Commission, which alleged that BATSA, the South African subsidiary of BAT, the UK-based multinational, had abused its dominant position in the South African market. The referral arose from a complaint filed by Japan Tobacco International, the global tobacco giant, and owner of the Camel brand. JTI, by alleging a contravention of section 5, also contrived to attain the status of a second complainant in this matter and so had full rights of participation. This case consumed 33 hearing days, oral testimony by a small army of witnesses who represented a small proportion of the witness statements actually filed, and tens of thousands of pages of documentary evidence.

The case also involved three counsel representing each of BATSA and JTI. JTI’s team was led by David Unterhalter, South Africa’s leading competition lawyer, whose unusual, indeed unique, ability to appreciate the interplay between law and economics, combined with his truly awesome articulacy, will undoubtedly see him recognised in time as one of the world’s leading competition lawyers. David’s only weakness was the palpable irritation that he often exhibited at having to deal with strong factual witnesses. Great economists were putty in his hands, but those rare marketing managers, convenience store salespeople, bar owners and the like who managed to stand up to his cross-examination simply by sticking to their superior factual knowledge sometimes induced an arrogance and hectoring irritability that was, I think, at odds with a basically courteous, respectful personality. This case was stuffed to the gills with these ‘ordinary’ witnesses.

BATSA’s team was led by a doyen of the Johannesburg bar, Fanie Cilliers, a wily, cunning old fox if ever there was one. Fanie is no expert in competition law. But a razor-sharp mind and many decades of courtroom experience have made him a master of spoiling – with wordy courtesy, great charm and the occasional sharp and often outlandish barb – an opponent’s case. Fanie is also a chain smoker of note. I mention this because in consumer goods cases it is often difficult to avoid filtering – in a manner of speaking – the evidence through one’s own personal experience. I was reassured that, in this case, BATSA and JTI each had a chain smoker on their teams, and the Commission and the panel each had a lapsed smoker (though one is only in remission from an addiction). I often felt that we were the only four people centrally involved in the case who could really interpret the evidence, who truly appreciated what it would take for a Lucky Strike smoker to switch to Benson & Hedges Super Mild.

This lengthy, fact-laden case is difficult to summarise. Essentially, the severe limitations imposed on cigarette advertising, including previously ubiquitous sponsorship of sports events, had transformed the cigarette market from one of the most transparent into one of the ‘darkest’ of markets. The upshot, at least in the estimation of the complainants, was that the point of sale became the critical, indeed the only, site for brand promotion. BATSA stood accused of excluding its rivals – most notably the powerful brands of Camel owned by JTI, and Marlboro owned by Philip Morris – from this site of promotion. This BATSA did by incentivising stores to accord additional space and preferential positioning to its brands. Rival brands were not excluded from the displays but they were disadvantaged in space allocations and positioning. In other words, this was not a case about restricted distribution but rather about foreclosure of promotional opportunities. The incentives offered by BATSA included cash payments and the provision of various till and display furniture. Although the other manufacturers were not precluded from offering similar incentives and did so, their smaller market share meant that they had a far smaller revenue base over which to spread the incentives they offered.

When the case was filed and through much of the hearing I thought that this was a slam dunk for the complainants. And so too did the complainants themselves. After all, BATSA’s share of the total market was in the high eighties and in certain cigarette and store formats its share exceeded 90 per cent of the market. Moreover, the major conduct alleged – the payment of incentives for preferential space and positioning – was common cause. However, as the case unfolded it became clear that the complainants had neither the evidence nor the theory to support their case. And, a cardinal sin in antitrust analysis, they did not pay sufficient attention to the particularities of the cigarette market. There were essentially four points that sank the complainants.

First, cigarettes were, for a variety of reasons, never sold on open supermarket or convenience store shelves. They were sold from dedicated kiosks staffed by store assistants. This, combined with unusual levels of brand loyalty, meant that the vast majority of cigarette consumers did not browse the cigarette kiosk; they simply asked for their brand. And because distribution was not restricted – and this largely because the store owners would not have permitted exclusive distribution – the cigarette purchaser knew, without looking at the display, that his or her brand would be available. This enormously reduced the significance of the store point-of-sale as a promotional mechanism.

Second, there were simply too many points-of-sale spread over too great a variety of store formats for BATSA’s attempts at exclusion to constitute significant foreclosure from promotion at the point of sale. BATSA swamped the hearing with evidence to this effect, evidence that came from buyers from large retail chains, managers of garage forecourt stores, barmen from popular venues and vending-machine owners. Their opponents, by contrast, relied on the evidence of marketing experts who presented grand theories of ‘category management’ that simply did not stack up against the real-world experience of BATSA’s witnesses. The evidence showed that at selected points-of-sale – for example, garage forecourts – BATSA’s opponents had made significant inroads through mechanisms essentially similar to those used by BATSA. Furthermore, JTI, for reasons best known to themselves, simply chose to ignore promotional opportunities in the market for black smokers despite Camel’s evident appeal in that market.

Third, the evidence showed that there were alternative mechanisms of promotion. These ranged from mobile-phone messaging to the holding of promotional parties at selected venues. Before the evidence was actually presented, this all sounded extremely implausible. Could a serious impact be made on the market for literally billions of cigarettes by these campaigns which, at best, could connect directly with a tiny fraction of South Africa’s millions of smokers? However, when we came to understand, largely through the evidence and argument presented, the association between choice of cigarette brand and lifestyle, then the number of venues and the range of people, the lifestyle role-models, to whom direct marketing had to appeal in order to impress itself on those who aspired to the lifestyle in question, shrank significantly and the likely impact of direct marketing increased concomitantly.

Fourth, the complainants failed to engage sufficiently with the nature of retailing. One instance of this that we have already discussed is manifest in the reluctance of retailers to limit their customers’ choice of brand. This meant that the retailers were extremely resistant to excluding brands from their stores, so much so that BATSA never even attempted to incentivise this obvious and very effective form of exclusion. In addition, retailing has become, in important part, a process of selling pieces of store real-estate. This is what the retailers were doing when they accepted payment from BATSA for preferential allocations of space and position. In a competitive retail market – and we were given no reason to believe this was not the case – we surmised, with the assistance of supporting evidence from the US, that at least part of the payment received by the stores in exchange for pieces of their in-store real-estate would be passed on to the stores’ customers. This need not necessarily have taken the form of lower cigarette prices, but it may have been reflected in the prices of other products or in the introduction of other facilities that enhanced the competitiveness of the store receiving the BATSA payment.

An obvious question begged is why BATSA spent so much on incentives that, in our estimation, gave them only a marginal advantage. The answer is that BATSA is an extremely robust and aggressive competitor. It was intent on defending its massive market share, which had been achieved without a significant international brand in its portfolio (in contrast with its principal rivals who had at least two of the most powerful international brands). In fact, BATSA had good reason to believe that its dominant national brand, Peter Stuyvesant, was clearly threatened with eventual decline. Of course, that we concluded that the point of sale was not nearly so important a promotional opportunity as the complainants alleged, or as is indicated by BATSA’s willingness to pay large incentives to influence locations of space and positioning at the point of sale, suggests that we believe BATSA’s actions to be somewhat irrational, and a waste of money. That’s not our concern: BATSA, with its deep pockets and a large market share to defend, may well have overspent on this element of its marketing strategy. It is entitled to err on the side of robust behaviour. Indeed, it may well have been BATSA’s subjective intent to exclude its rivals from promotional opportunities. Our task was to decide whether it was objectively exclusionary, and whether it gave rise to cognisable competition harm. In our view, its conduct didn’t meet this test.

There is a range of lessons to be learnt from this case. But for present purposes, the one I would stress most strongly is that changing market circumstances often demand significant and aggressive responses from the firms affected. With the regulations that massively circumscribed cigarette promotion, this hitherto most transparent of markets underwent a more dramatic change than most are subjected to, even those subject to rapid technological change.

BATSA responded to this dramatic change in environment aggressively and comprehensively. Its opponents, despite their own deep pockets and considerable experience of similar developments in other markets, did not. They preferred to rely on regulatory support to limit the range of responses available to their competitor. Possibly, they did not value the South African market sufficiently to make the necessary investments that a pro-competitive response would have demanded. Maybe they calculated that the Competition Tribunal would not find in favour of an aggressive competitive strategy mounted by a firm with so overwhelming a market share as BATSA. If so, they badly miscalculated. The Tribunal has consistently taken the view that establishing dominance represents the first and lowest hurdle to be crossed in establishing abuse of dominance. More difficult is to establish the elements of the alleged conduct and the net effect on competition – that is, the necessity to establish harm to competition that is not countervailed by pro-competitive gains. Our judgment clearly asserts that position.

I should add – and this is an important consideration in many abuse-of-dominance cases – that, had we found against BATSA, a remedy would have been extremely difficult to devise. Had we found a section 8(d) contravention, we could have imposed an administrative penalty. However, an administrative penalty needs to be accompanied by a clear statement on the action the firm in question needs to take to ensure that it remains within the law. But what could that have been in BATSA?

The complainants and the Commission effectively argued for the imposition of ‘category management’ principles, which require that the display and positioning of cigarettes at the point of sale be determined by historical sales data; these would presumably have meant the termination of incentives. This, as we pointed out in the judgment, would have been a highly anticompetitive outcome, the more so if one takes the view that in a competitive retail market part of the incentives would have been passed through to consumers. The application of strict category-management principles would not even have enabled a firm like BATSA to promote one of its growing ‘drive’ brands – say Dunhill – over another in its own portfolio, for example Peter Stuyvesant, which at the time enjoyed a very high market share, but which for a variety of reasons was susceptible to long-term decline.

Of course there is far from universal agreement with our judgments in SAA or BATSA. As I’ve indicated, abuse of dominance does provide plenty of room for even reasonable people to disagree, not to mention the large number of unreasonable and self-interested people with whom we have to contend.

A small survey conducted among active, leading competition lawyers indicated, to my surprise, that several of them thought we had the SAA judgment wrong, that our conclusions were not backed by the evidence.18 In fact, the evidence was largely common cause but for the evidence on SAA’s dominance, and frankly in this part of the inquiry SAA was downright uncooperative in providing data, a formal finding that contributed to the fine it paid. In any case, we took the precaution of establishing SAA’s dominance on market share – which was clear – but we also showed that it possessed market power.

I could more easily understand a response that argued that we had got the test wrong – that it had not been shown that SAA’s conduct excluded its competitors from the market or that it resulted in harm to competition. On our foreclosure test of harm to competition, the counterfactual is whether SAA’s competitors expand in the market as successfully as they might have done, thus generating a more competitive market structure and process. I’m comfortable with inferring from the extent of foreclosure that SAA’s conduct had harmed the competitive process. The US judgment in a similar case suggests we would have had to show that the commission on the marginal ticket was predatory, that the ticket was sold below cost. Given that the marginal cost of accommodating an additional passenger must tend to zero, proof of predation is unlikely (although if the impact of the travel agent’s induced promotion of SAA necessitated putting additional aircraft on a route or even purchasing additional aircraft, it may well have been predatory).

I am less surprised at the response to our BATSA decision. It renders the provision of easy-to-provide advice to incontestably dominant clients more complex. But, I repeat, the pertinent test is not the extent of dominance but the proof of anticompetitive harm and the ability of the accused to proffer pro-competitive gains. BATSA’s actions were, to be sure, directed at enhancing the sales of its own brands, including the critical ability to take strategic action to influence the medium- to long-term marketability of its own brand portfolio. In achieving this it may well have stolen a march on its competitors, although each significant competitor was blessed with international brands far more powerful than any in BATSA’s portfolio. It also took responsibility for category management and spent considerable sums of money acquiring that ‘right’ and in undertaking the activities associated with that role. That it performed this task with an eye not merely to achieving the better promotion of the cigarette category in general – something which doubtless improved sales of a product category under significant pressure – but also to promote its own brands is not surprising. The notion, peddled by its opponents with the aid of marketing experts, that BATSA should have undertaken this costly function in a neutral manner is naive or, more likely, disingenuous. And this leaves aside the pro-competitive consequences that flow from the undoubted pass-through of some of its considerable payment to retail customers.

On BATSA I have one postscript, though it predates the case discussed here: I recall, though only vaguely, that in the dying days of the Competition Board we approved the merger between BAT and Rothmans, the Rupert-owned South African tobacco company. BAT had a very small presence in South Africa and so the accretion to Rothmans’ share in consequence of the merger was small. While I can’t recall the details of that transaction, I am certain that we didn’t appreciate the significance of even a small share of the cigarette market, the strength of a great international brand and the prospect of potential competition. Had we done so, we might well have prohibited the merger or compelled the merged entity to sell or license one or more of its brands to another manufacturer, as the Tribunal later did in Distell. I think that what BAT brought to the merger was experience of operating in international markets, and in particular in the transition to ‘dark markets’. I don’t think that a large though parochial company like Rothmans would have as easily contained the growth of the likes of Camel and Marlboro and remained within the bounds of the law.

My clearest recollection of the BAT–Rothmans merger is that, in keeping with the practice of the Competition Board, Christine Qunta and I attended a meeting with several senior executives of Remgro, Roth-mans’ holding company, at its head office in Stellenbosch, in order to discuss the proposed merger. We sat at a long boardroom table with glass holders containing cigarettes evenly spaced along it. Everybody smoked, all of the time. One of Christine’s better-founded prejudices was against smokers, but I saw her take one look at this lot and decide, remarkably for her, that a fight over this would be a bridge too far.

To return briefly to another aspect of the SAA judgment: in addition to establishing the standard for judging harm and its role in proving abuse of dominance, the SAA judgment also laid out a formula for weighting the various factors that section 59(3) of the Act mandated us to consider when determining the size of the administrative penalty. The Act itself provides very sparse guidance in this area. However, I am not sure that we were wise to refine it beyond the Act’s generalities.

Although competition law is extremely fact-intensive and so decisions must be responsive to the factual matrix that characterises each case, it is essential that judgments should, wherever possible, establish general principles and rules that apply beyond the matter at hand, thus providing a degree of certainty. However, I think that on this occasion we effectively limited our options too early and in an area where we should have retained greater flexibility.

While the Tribunal is not bound by its previous judgments, the necessity to provide certainty always loomed large in our thinking; general principles are not laid down simply to be ignored on the following day. However, I think that appropriate sentencing has to reflect, though not mindlessly ape, society’s view of the seriousness of the contravention and, with the benefit of hindsight, I think that we set these guidelines before there was sufficient public appreciation of the meaning of a Competition Act contravention. This is not to say that I think a larger fine should have been imposed on SAA, but merely to regret that we staked out a general position on this so early on in our life. It has had the effect of placing a voluntary, but nevertheless authoritative, limit on the size of our administrative penalties with unfortunate ultimate consequences. This issue – particularly as it relates to the turnover figure on which administrative penalties should be based – has predictably assumed more serious proportions in dealing with the penalties appropriate to cartel conduct.

Exclusionary Abuses: A Concluding Reflection

The prosecutorial and adjudicative record of the South African competition authorities and the substantive approach to alleged exclusionary conduct clearly do not support the notion that our Competition Act or our practice are responsible for chilling competition. We do make use of presumptions and legal rules, notably in the Act’s definition of dominance and in the characterisation of named exclusionary practices in section 8(d). However, market share can be ascertained only after a rigorous identification of the relevant market, and the requirement to prove the elements of each of the practices identified in section 8(d) combined with the necessity to prove anticompetitive effects, places significant hurdles in the way of securing a conviction for abuse of dominance largely by elevating the role of economic reasoning in our decision-making.

Indeed, I firmly believe that the provisions of our Act under-deter abusive conduct. In particular, I think the different remedial regimes that apply to section 8(c) and section 8(d) contraventions incentivise lengthy litigation around the elements of conduct specified in the various sub-clauses of 8(d), because there is so much to be gained for the defendant getting out of 8(d) in favour of 8(c). This not only serves to immunise the complainant from the possibility of an administrative penalty on a first-time offence, but it means that the Commission bears the onus of proving that the anticompetitive effects of the conduct outweigh any pro-competitive gains that may arise, whereas under 8(d) the onus falls upon the defendant.

The cleanest way of solving this would be to provide for a first-time administrative penalty for an 8(c) contravention. In addition, I think the onus for proving that the pro-competitive effects outweigh the anticompetitive effects should logically always fall on the defendants. The Commission bears the onus to prove anticompetitive effects without which it cannot sustain a finding of abuse of dominance. It is only if these are proven that the defendant is then required to prove countervailing pro-competitive effects and, needless to say, it is the defendant who is best placed to identify and prove these effects. Were this simple amendment to be effected, the incentive to litigate to death the meaning of say ‘induce’ or ‘refusing to supply’ or ‘economically feasible’ would be significantly reduced because not much would hang on whether the alleged contravention fell under 8(c) or 8(d).

A more far-reaching amendment would be to eliminate section 8(d) altogether and amend 8(c) to prohibit any conduct perpetrated by a dominant firm, of which the effect is anticompetitive. A possible downside of this amendment would be that the named conduct in 8(d) does provide guidance as to those practices most commonly associated with abuse of dominance. The rejoinder from those who live in fear of false positives would be that, precisely because each of these may give rise to pro-competitive conduct, by naming the offence, excessive caution – and hence ‘chilling’ – would be induced. The alternative view, as I’ve already indicated, is that it promotes certainty and assists law enforcement. It effectively says that, if you are dominant, then examine closely a decision to refuse to supply scarce goods to a competitor, or to price below average or marginal variable cost. But whatever view is taken, it seems that the easiest way to go would be to align the remedial regimes and onus requirements as between sections 8(c) and 8(d).

EXPLOITATIVE ABUSES

Excessive Pricing

I’ve identified the distinction between per se prohibited abusive conduct and conduct prohibited by the application of a rule of reason. The overwhelming number of practices specified in section 8 are judged by a rule of reason. Only two forms of conduct – excessive pricing (section 8(a)) and denial of access to an essential facility (section 8(b)) – are prohibited per se. That is to say, it is sufficient in respect of these practices to prove the existence of elements of the conduct in order to secure a finding that the Act has been contravened. Anticompetitive harm does not have to be established and there is no provision for a pro-competitive defence. However, proving the elements of the conduct in question represents a significant hurdle. In order to establish excessive pricing, it is necessary to prove that the impugned price ‘bears no reasonable relation to the value of that good or service’. And in order to establish a contravention of the prohibition of denial of access to an essential facility, it is necessary to prove that the facility in question is ‘an infrastructure or resource that cannot reasonably be duplicated, and without access to which competitors cannot reasonably provide goods or services to their customers’.

The other pertinent distinction to be drawn in the list of abusive practices specified in section 8 is between ‘exploitative’ and ‘exclusionary’ conduct. Section 8(a) – the prohibition of an excessive price – is the only exploitative abuse referred to in section 8; the remainder are exclusionary abuses. The distinction between the two types of abuse is self-explanatory. An exclusionary abuse refers to conduct of a dominant firm that has the effect of excluding a rival or would-be rival from the market. An exploitative abuse refers to conduct by a dominant firm that seeks to take advantage of, to ‘exploit’, its dominant position by charging an excessive price.

ArcelorMittal

For the competition authorities, this exploitative abuse assumed significance – to put it mildly – in the form of an excessive-pricing complaint filed by two gold-mining companies, Harmony Gold and Durban Roodepoort Deep (henceforth, Harmony) against ArcelorMittal South Africa (Mittal). The Commission issued a notice of non-referral and so the complainants referred the matter to the Competition Tribunal.

In a nutshell, Mittal, an overwhelmingly dominant local steel producer, charges most of its domestic customers a price based on what would be charged by an alternative international supplier, plus the cost that would be incurred for freighting the product from Europe to South Africa. The notional costs added to the international price include transport and insurance and a ‘hassle’ factor, an estimate of the additional transactions and other costs entailed in importing over purchasing from a domestic producer. The price arrived at after adding these notional costs is commonly referred to as the import parity price.

In addition, a significant proportion of Mittal’s output is sold on the international market, where it is a price taker. The consequence is a significant divergence between the price charged by Mittal to its domestic customers and that charged to its international customers, a differential achieved by factoring in the notional transport costs that would be incurred by a local customer procuring product in the international market.

There is little that offends Mittal’s South African customers more than paying the steel monopoly substantial freighting costs that Mittal does not, in reality, incur. Public policy-makers are equally offended, particularly those familiar with the circumstances that ensured that Mittal was, in its previous incarnation as Iscor, then a state-owned enterprise, the recipient of a significant tax subsidy when it built its plant at Saldanha Bay. Moreover, when the privatised Iscor’s steel and mining interests were unbundled, a strategy that enabled the recapitalisation of the then deeply troubled Saldanha Bay plant, it secured a significantly preferential arrangement on the pricing of its iron-ore requirements, an advantage that has effectively been amplified as the price of iron ore has increased. It’s authoritatively claimed that there was a ‘gentlemen’s agreement’ that this preferential iron-ore price would be passed through to consumers. However, the gentlemen’s agreement was concluded in an environment in which, as Lyndon Johnson is once said to have observed in relation to the Texas oil industry, there are no gentlemen. Government has since been reduced to the rather undignified role of the humble supplicant, attempting, without the slightest modicum or prospect of success, to persuade Mittal’s majority shareholder to provide South African manufacturers with a ‘developmental price’ for steel in order to strengthen downstream metal fabrication. Lakshmi Mittal, the patriarch of the family that owned the majority share, was for much of the time that this controversy raged – and it continues to rage – a member of former President Thabo Mbeki’s international investment advisory committee comprising international business leaders.

Harmony, a customer of Mittal, decided to take matters into its own hands by prosecuting a claim of excessive pricing against the steel producer. The eventual upshot was a mammoth trial in which the Tribunal found that Mittal had indeed contravened section 8(a) of the Competition Act by charging an excessive price.19 However, on appeal, the Competition Appeal Court decided that the Tribunal had applied the incorrect test in arriving at its decision and so remitted the matter to the Tribunal with an instruction that it apply the test favoured by the Court.20 Before that could take place, Harmony and Mittal settled their dispute in terms that have remained confidential, leaving the question of excessive pricing somewhat in limbo.

The excessive pricing issue is nevertheless alive and well and of significantly more than academic interest. It rages on in South Africa, where anxiety over excessive pricing is largely focused on our resource-extractive industries. Policy-makers are deeply aggrieved by the practice of import parity pricing that, it is argued, results in South Africa’s rich bounty of mineral resources being exported in unprocessed or lightly processed form rather than further fabricated (‘beneficiated’ in local parlance) in downstream, labour-intensive manufacturing plants. If only, or so the argument goes, these minerals and the output of their first processing stage – for example, steel and aluminium – were priced at a level that reflected South Africa’s locational advantage in the mining of critical inputs (coal in the case of aluminium and iron ore in the case of steel), we would have a thriving manufacturing sector rooted in our natural resource endowment.

I’ve long believed that the beneficiation argument is vastly overstated. Sheer luck or large subsidies are required to operate at all the levels of a value chain that begins with mining and ends with the production of manufactured consumer products. This point is starkly illustrated if one thinks of a value chain that begins with diamond mining, proceeds through diamond cutting and ends with diamond jewellery manufacturing. Though palpably far-fetched, the proposition that the presence of diamonds under South African soil should inexorably flower above ground in the shape of an internationally-competitive jewellery sector has had the South African government wasting time and money in futile attempts to regulate the diamond mining industry, all in the name of promoting diamond cutting and jewellery manufacturing despite a clearly inappropriate wage structure and an absence of the requisite skills required for jewellery manufacturing.

Nor do I believe that import parity pricing is necessarily contrary to either the laws of economics or competition rules. It is, of course, not necessarily an international trade ‘problem’, but rather one that applies to any monopolist in a market with unusually high entry barriers. In these circumstances, the monopolist’s pricing would be constrained by the price of purchasing a competing product from the nearest alternative supplier, whether that supplier is to be found in the next town, province or country. It would be a decidedly irrational monopolist that did not build into its ‘domestic’ price the additional cost entailed in a purchaser’s electing to procure the competing product from the next town, province or country. Needless to say, the greater the distance from the alternative supplier and the higher the barriers to new entry, the greater the transport cost rent that would be extracted by the locationally-privileged monopolist.

My scepticism of the general argument for beneficiation as an industrial strategy may not be as well founded for certain steel products as it clearly is for diamonds. In the case of downstream steel fabrication – where we do have extant domestic capacity – a price advantage achieved by a domestic price reflective of actual costs (excluding, that is, the fictitious freighting costs) may translate into more competitive domestic manufacturing. Indeed, our decision in Mittal specifically accepted evidence to this effect. So the question of the pricing of our mineral resources and that of its lightly processed outputs like steel will not go away.

The notion that competition law is capable of dealing with ‘excessive pricing’ is by no means uniquely South African. Certainly, as competition policy and law spread to the transition economies where price regulation has been the pillar of economic policy and management, there remains a strong expectation that this scrutiny and regulation will now be assumed by the fledgling competition agencies. Price monitoring is a commonly assumed function of the new agencies in these economies, with concepts like ‘unjustified’ price movements often central to their approach to abuse of dominance.

Nor is the excessive-pricing issue confined to the new competition regimes. It features fairly consistently in jurisprudence across the national competition authorities of Europe. Indeed, the South African Competition Act’s definition of excessive pricing – being a price that ‘bears no reasonable relation to the value of that good or service’ – is drawn directly from the judgment in the leading European excessive pricing case, namely United Brands.21

However, the issue remains fraught with complexity and controversy, and with good reason. Pricing power derives from market power. However, the mere possession of market power is not contrary to competition law. Indeed, one important source of market power is innovation and other pro-competitive conduct. The rents derived from the possession of market power will, in most circumstances, sooner or later attract new entrants, the more so if the dominant incumbent takes ‘excessive’ advantage of its privileged position. And so the effort to acquire market and, therefore, pricing power and the attention it attracts from rivals are an important driver of the competitive process.

This approach to exploitative abuses dominates judicial and scholarly thinking in the US regarding excessive pricing. I’ve earlier cited the passage from the Trinko judgment of the US Supreme Court effectively lauding monopolistic conduct because of the new entry it will beget. Indeed, scepticism of a legitimate place for excessive pricing in antitrust is shared by a great many antitrust enforcers. However, their misgivings are derived less from the doctrinaire considerations reflected in Trinko than from the overwhelming practical difficulties entailed in determining the counterfactual – namely the competitive price – and then in further determining at what point the difference between the competitive price and the actual price (which reflects the producer’s market power) is deemed to be ‘excessive’. Recall that in the earliest incarnations of our Competition Act no provision was made for excessive pricing, but it was rather inserted at a later stage at the instance of the trade unions.

However, while the mature agencies – most notably the US – and the majority of the members of enforcement and scholarly communities may continue to take the high ground and insist that excessive pricing has no place in antitrust law and practice, this objection will not be heard by policy-makers and least of all by the general public, who will continue to insist that their antitrust enforcers should retain powers to constrain excessive pricing. And they may be forgiven for adopting this standpoint. Whatever the position of the US Supreme Court, the public views antitrust’s mandate as the defence of consumers and small businesses against the power of monopolies, which is, again in popular perception, most clearly manifest in their ability to set prices unilaterally, without regard to the needs and interests of the consumers of their products.

Cognisant of this reality, we ought then to work out a credible test of excessive pricing, one capable of application by an antitrust authority, or, if such a test cannot be devised, to present a plausible explanation for prices deemed to be excessive, or to produce evidence to support the notion that the ‘excessiveness’ will be of a limited duration and will ultimately be responsible for pro-competitive new entry, or to advocate the introduction of regulatory interventions capable of producing a preferred outcome. What a competition authority should avoid is the temptation to take on the role of a price regulator. In our view, the simplistic, literal view taken by both parties in the Mittal matter and, ultimately, by the Competition Appeal Court, precisely urged the Tribunal to assume a price-regulating function.

It is also necessary to devise a test of excessive pricing capable of general application. The alternative is an endless stream of excessive pricing claims. There are countless instances in which the public may deem prices too high. And there are myriad reasons why prices may, for a time at least, deviate from the underlying costs of production or some other comparator.

For example, it is often observed – sometimes accurately, other times not – that the prices of traded goods respond sensitively to a weakening of the exchange rate but are not nearly as responsive to a strengthening of the exchange rate. Both outcomes may reflect impeachable cartel conduct and warrant the close attention of the competition authorities. Certainly, prices that respond to an exchange rate weakening by moving upward at an identical or highly similar rate and time generate understandable suspicion, given differences in the underlying costs of production among different producers. Even where a single imported item looms unusually large in the costs of production, firms will have different hedging strategies or even, more prosaically, different storage facilities that would dictate diverse price responses, at least for a time, in a truly competitive environment. Downward price stickiness may be indicative of cartel conduct or may simply reflect tepid competition where no single manufacturer or dealer – particularly if there is a single dominant price leader – moves quickly to cut its prices, thus providing the spark that will generate downward pressure on prices. Would it be appropriate for a competition authority to respond to these prima facie indications of anticompetitive cooperation by running a complex excessive-pricing case rather than by seeking further evidence of collusion?

However, it is probably true to say that prices that appear to diverge significantly and consistently from underlying production costs are generally the product of monopolistic structures or cartels. No competition authority in its right mind would choose to run a cartel case as an excessive-pricing case. So an excessive-pricing charge will invariably be levelled at a monopoly. However, these monopolies may be the temporary consequence of innovation that had successfully decreased the cost of production or had successfully brought a new product onto market and was, in consequence, able to command a price premium. The former (‘process innovation’) will generally not be patent-protected and so may be relatively easily imitated. Competition may thus be restored relatively quickly at lower price levels, thanks to the price-reducing innovation that might never have happened if the producer/innovator had been discouraged by the prospect of an excessive-pricing claim in the period when he was reaping his innovation rents.

The latter form – ‘product innovation’ – is more likely to be patent-protected. Patents do reward innovation and there are hidden costs and risks in product innovation, including the costs of significant failure, that are not revealed by a simple fixation on direct production costs. However, there is also much in the argument that patents overprotect, and that they often reward private investors who have managed to appropriate the fruits of public research. If this is so, the first-best solution is surely to repair the patent system, rather than mount an excessive-pricing claim.

But monopolies are by no means always the consequence of competition on the merits, including innovation. They are frequently the consequence of past or current state subsidy, which has enabled the monopolist to lower the investment and production costs it has to bear (this is, of course, why excessive-pricing claims are likely to be viewed with concern in countries that were characterised by widespread state ownership). And they are, in some instances, bolstered by insurmountable entry barriers and by the geographical isolation of the monopolist from an alternative source of supply. Again, it is by no means clear that the solution lies in an excessive-pricing charge. I have no doubt – as our decision in Mittal emphasises – that if the competition authority is unable to remedy the situation by introducing competitive structures or conduct, then the first-best solution will not be found in an excessive-pricing claim.

This is the background against which the Tribunal was compelled to judge Harmony’s claim of excessive pricing on the part of Mittal. Simply expressed, Harmony relied on European jurisprudence, which favours a comparison of two price points followed by a judgment call on whether the difference between the two price points – a competitive price and an actual price – is ‘excessive’. Mittal relied for its defence on an analysis of its profitability in which it attempted to demonstrate that, because it was (on its version) not earning excessive profits, it could not be charging excessive prices.

While Mittal’s defence entailed an excursion into the higher reaches of economic theory, relying, as it must do, on a calculation of economic as opposed to accounting profits, it proved thoroughly unhelpful in providing an administrable test of excessive pricing. The problems involved in deciding whether, for example, the appropriate measure of capital should be historical cost or replacement cost, in deciding whether Mittal was an ‘efficient’ producer (an inefficient producer may, in consequence of its inefficiency, be generating non-excessive profits while still charging excessive prices), and a range of other extremely complex and eminently contestable propositions, proved not merely difficult, but insurmountable. I recall my exasperation at the Competition Appeal Court’s glib reference to a passage in the UK Competition Appeal Tribunal’s judgment in the Napp Pharmaceutical excessive-pricing case to the effect that mere technical difficulty should not get in the way of the necessity to make a judgment. I looked forward to the Competition Appeal Court having to grapple with the insurmountable technical difficulties involved in determining the competitive price of steel – difficulties that were considerably greater than in Napp, where the UK tribunal was able to identify significant predation in the lower-priced market. However, because we declined to pass judgment on the profitability and pricing evidence and argument, the Competition Appeal Court was conveniently able, in a rare display of deference, to remit the judgment to the Tribunal, thus sparing itself the considerable difficulty of grappling with the highly-contested evidence.

I am open to the possibility that profitability measures may provide useful indicative tests in market analysis. I accept that they can be applied in regulatory settings where the answers to the methodological questions posed above and the necessary data are provided ex ante and constitute the agreed or statutory basis for price determination. However, these tests cannot be administered in an adversarial setting where the appropriate test, far from being specified ex ante, has to be devised and reasoned on the basis of a brief and vague definition of excessive pricing provided for in a competition statute. Certainly, the definition in our Act, a price that ‘bears no reasonable relation to the economic value of that good or service’, provides none of the guidance or the rules that would enable these questions to be answered. Two excellent recent papers that actually advocated the use of profitability measures in antitrust analysis are interesting for the caution they urge when applying these measures to the resolution of practical problems.22 These were of course conference papers; they were not prepared for submission in an adjudicative setting in which the proponent of a particular viewpoint never allows any acknowledgment of doubt to cloud an argument in favour of the client.

The Mittal argument, in particular, strongly affirmed our view that competition authorities should steer clear of price setting. Leave aside the well-founded doctrinaire objections to competition authorities assuming a price-regulating function and think only of the immense practical difficulties entailed. Price-regulating institutions typically have their own statutory foundation. The price-setting methodology they are mandated to apply and the data requirements necessary to support what is always a technically complex exercise are clearly specified in their governing statutes or in subordinate regulations. They are staffed by experts, not only in the technically complex business of price-setting, but in the very sectors they are responsible for regulating. I simply can’t credit the naivety and arrogance that have courts, competition enforcers and scholars believing that one or two sentences in a competition statute can constitute the basis for transforming a competition enforcer into a price regulator.

Harmony, for its part, did not rely on profitability measures to make its case, although it naturally engaged with the approach of its opponents in order to demonstrate its manifest shortcomings. This task did not prove to be particularly onerous, as may be expected in an adversarial setting. Bring along sufficient experts capable of identifying the holes that characterise all of these contentious methodologies and it will be impossible to prove that one methodology rather than another constitutes the correct approach to specifying the level of the ‘competitive price’, not to mention the ‘economic value’ of the product.

Harmony essentially adopted the price comparison methodology that is favoured by European jurisprudence and that appears to accord most closely with the plain meaning of ‘excessive pricing’. It is a term that suggests that two prices be compared in order to ascertain whether the higher of the two is ‘excessive’ in relation to the lower. On the face of it, this is a technically less complex task, but the difficulties are, in reality, no less insurmountable than those posed by the profitability tests.

There is, first, the small inconvenience that the statutory definition requires a comparison between the higher price and ‘economic value’, which means that to employ the price comparison methodology one must accept that the ‘lower’ price reflects the intrinsic and universally-determined ‘economic value’ of the product. Is this lower value the ‘competitive price’ and how is that determined? In fact it need not be (and arguably should not be) the competitive price because the statute specifically contemplates the prospect of market power – that is why competition law is concerned to limit abuse of that power – and so it accepts that those with market power will exercise a degree of pricing power. Thus, in most markets the actual price will naturally deviate from the competitive price. The ‘correct’ counterfactual is then the competitive price plus a legitimate degree of market power reflected in the price. So even if it were possible to identify a number that approximated a ‘competitive price’, it cannot be said that this represents the benchmark against which all higher prices are to be evaluated in order to determine whether or not they are ‘excessive’, because there is no requirement to charge the ‘competitive’ price, simply a requirement not to charge an ‘excessive’ price.

And then even if that problem is ignored or even miraculously resolved, there remains the difficulty of deciding when to ascribe ‘excessiveness’ to the higher of the two prices. Neither the South African Competition Act nor European jurisprudence offers any guidance about when a mere difference between two prices or two profit measures is deemed to be ‘excessive’. The notion that it is so at the point where the higher price ‘bears no reasonable relation to the value of that good or service’ is not helpful. Leaving aside the difficulty entailed in ascribing a numerical meaning to ‘value’, the decision whether the difference between the price and the value is ‘excessive’ is purely subjective and has varied dramatically in the European judgments.

Given these difficulties, the jurisprudence emanating from Europe is, predictably, all over the place, precisely reflecting the high level of arbitrary judgment implied by an exercise of this sort. The variability of the European judgments is exhaustively reflected in our judgment and, ironically, repeated by the Competition Appeal Court, as though it represented a smorgasbord of acceptable analytical options, rather than the outcome of endemically subjective and uncertain judgments.

I don’t take issue with the requirement for adjudicators to exercise judgment. That, after all, is what they are employed to do. But we are dealing here with the core competitive decision: the pricing of the product. A decision whether the price charged is excessive or not cannot be left purely to the subjective assessment of an adjudicator and so be free to vary indiscriminately from one case to the next. There has, at least, to be a clear, administrable test developed that will provide a business decision-maker with some degree of certainty. I’ve dealt earlier with the problem of ‘false positive’ errors arising from abuse-of-dominance cases and have disagreed with those who effectively invoke the possibility of error to eliminate all such suits from antitrust practice, a consequence that would arise from the adoption of the Trinko judgment.

However, an approach to excessive-pricing claims that is too open-ended will be particularly susceptible to the dangers of false positive error and hence to the prospect of chilling competition. This, as already intimated, would be particularly so if the pricing power derived from innovative activity that accorded the innovator market-power and that was susceptible to challenge from other innovators encouraged by the prospect of high returns. This is why it’s imperative that a test be devised that will limit, in so far as it is possible, the exercise of unfettered judicial discretion in deciding these claims.

The inevitable rejoinder from those seeking to impugn Mittal’s pricing conduct would be that this is a clear instance where pricing power does not derive from innovation. While that’s undoubtedly true, it seems preferable to design a test that identifies what will constitute excessive pricing, rather than to accept one that leaves a great deal of discretion in the hands of the adjudicator. How would that discretion be limited? Knowing the proclivity for lawmakers – including judges – to resort to terms like ‘reasonable’ and ‘justifiable’, the odds are that these are the vague terms that would have been employed to limit discretion. And so an arithmetic difference deemed ‘unreasonable’ or ‘unjustified’ would fail the adjudicator’s subjective discretionary test. See how quickly we have sunk into the language employed by the price-regulators-turned-competition-authorities of the countries of the former Soviet Union! Once a number has been designated ‘excessive’, future decisions must broadly follow that decision if gross uncertainty is to be avoided.

Moreover, given that excessive pricing is subject to a per se prohibition, I doubt whether considerations like the particular source of pricing power or an assessment of entry barriers could have limited the discretion once an essentially arithmetically ‘reasonable’ test of excessiveness had been laid down. In a per se prohibition, these ‘pro-competitive’ defences could not easily be invoked. And so any difference between two price points that failed the arithmetic test of excessiveness would, under any circumstances, potentially fall foul of section 8(a), regardless of the source of the difference.

The clear imperative to provide a degree of certainty, and the related issue of administrability, were the primary considerations underpinning our judgment in the Mittal matter. We devised a test that drew on the core categories of antitrust analysis, namely market structure, barriers to entry and conduct. We applied this to Mittal’s pricing practices and to the definition of excessive pricing in the Act and found that this was indeed an instance of excessive pricing.

To cut a very long story short, we found that Mittal was, in the words used in the judgment, ‘an uncontested firm in an incontestable market’. Thus the market structure and the presence of insurmountable barriers to entry accorded Mittal an unusual degree of market power. In order to distinguish the market power possessed by Mittal from that possessed by myriad other firms, we characterised Mittal as a ‘super-dominant’ firm. The structure of the market and the strength of the entry barriers thus enabled Mittal to sustain an extraordinary, indeed ‘excessive’, degree of market power or, what is the same thing, ‘excessive’ pricing power.

However, although we held that super-dominance is a necessary condition for sustaining an excessive-pricing allegation, it was insufficient, at least on a first offence, to ground a remedy. Although competition law does provide in specific instances for intervention in the structure of a market – most notably in merger regulation but also, in limited circumstances, in remedying anticompetitive conduct – it does not prohibit any specified market structure, including super- dominance or monopoly. Section 8 is concerned with conduct enabled by that dominance (abusive conduct) or, we held in this instance, conduct enabled by excessive dominance or super-dominance (excessive pricing). The Harmony and Mittal approach identifies that conduct as the setting of a price point, which is either deemed excessive in relation to another price point or which underpins a level of profitability that is excessive in relation to another level of profitability.

We instead asked ourselves whether we could identify the output-reducing conduct that a monopolist (a ‘super-dominant’ firm) or a cartel (both, by definition, unconstrained by competition considerations) characteristically engages in when setting its profit-maximising price. Mittal manifestly engages in such output-reducing conduct. This it does in the form of its exclusive agreement with the steel trader, Macsteel International, a joint venture owned by Mittal and Macsteel Holdings, which holds an exclusive right to market Mittal’s product in the international market and which is expressly prohibited from trading in the domestic market. So for domestic customers Mittal charges the import parity price, the highest that it can sustain, and then ensures that its excess product (partly a consequence of the imperative to attain scale economies, partly a consequence of the reduced demand generated at its domestic monopoly price level) is permanently removed from the domestic market through Macsteel’s contractual obligation to trade on the international market only.

And so we found that a super-dominant firm had engaged in a highly conditioned exclusive dealing arrangement to short the market precisely to eliminate the remaining influence that competition – through the excess supply of steel – could have exercised on the domestic price of steel. And so we voided the Macsteel arrangement and we imposed a fine of R692 million on Mittal.

Of course, we had to demonstrate the compatibility between our approach and the Act’s definition of excessive pricing. This we did in the following way:

In summary then, our examination as to the source of the pricing power is thus an examination into its reasonableness. Reasonableness in the context of a competition statute must mean ‘economically reasonable’. Economically reasonable in the context of a competition statute must mean having regard to the pro- and anticompetitive considerations that we normally apply. As we go on to argue in this decision, the occasions where one can find no reasonable relationship between a price and the economic value underpinning it are few indeed. The circumstances giving rise to Mittal SA’s pricing power in respect of some of its domestic consumers depend on the existence of a range of factual issues that we do not encounter in the market place every day, even in those markets habituated by long extant dominant firms.

Nor is there any need to dwell on dictionary definitions of what excessive means. The term is a defined one and hence it is the statutory, rather than the dictionary, definition of the word that we apply. The statutory definition as opposed to the ordinary word ‘excess’ does not require one to conclude when a particular level of differentiation is sufficiently large to constitute excess. Rather it requires one to find a relationship between a price and economic value that admits of no reasonable explanation, that is, of an explanation that does not rely upon the exercise of the degree of market power that arises from super- dominance. The finding of an excessive price is then determined not by some arbitrary measure of difference but is rather an inquiry into the rationality of pricing. It thus condemns pricing for which unchallenged and incontestable monopoly is the only explanation, as opposed to a price that may simply be high but for which innovation or even branding – that is, pro-competitive measures – provides the underlying rationale.

For this reason we find that a reading of the Act that requires us to find precise levels for the economic value and then the actual prevailing price and then to correlate them to some notional competitive price to be overly mechanistic and contextually unsupported. This reading might have some validity if we were meant to act as price regulators and to order the price back down to the non-excessive level. We have already firmly rejected the implicit contention that the sparse wording of section 8(a) is intended to convert us from an agency that promotes and protects competitive market conditions to an agency that determines price through the simulation of competitive market conditions.23

Does this sound a little strained and contrived? Possibly it does, but given that the definition would find favour with the linguistic philosophy of Alice’s caterpillar, I think we made a pretty good fist of meeting our adjudicator’s obligation to uphold the Act. We certainly gave more attention to the compatibility of our approach with the Act’s unintelligible definition – and particularly with how to deal with the concept of ‘economic value’ – than did either of the parties before us or the Competition Appeal Court.

However, the Competition Appeal Court decided that we had not applied the correct test and remitted it to the Tribunal with an instruction to apply the test they deemed to accord with the Act.24 To the extent that I understand the Court’s decision (which, given its literal reading of the Act, is bound to be only as clear as the Act’s definition of excess pricing), it appears to support Harmony’s position.

A finding on an excessive-pricing claim appears, in the view of the Court, to require the ‘factual determinations of the actual price and the economic value’. And then to determine whether the former is excessive in relation to the latter requires, in the Court’s own words, a ‘value judgment’. Having assured us that ‘legislative interpretation does not reduce to a simple recourse to a dictionary’ it then proceeds to define ‘economic value’ – which all would agree is a highly complex and contested term of art in economic theory – by recourse to … the Shorter Oxford English Dictionary! From that well-known authority on economic theory the Court predictably gleaned that ‘economic value’ is to be ‘expressed in monetary terms, as an amount of money’.

It then turns to the submissions of the court-appointed amici curiae, who confirm that the legislature must have intended by ‘the expression “economic value” an amount of money’. But this turns out to be no ordinary ‘amount of money’ such as one may find on a Mittal price list (or in the Oxford dictionary). It rather turns out to be ‘an amount of money which could notionally be the price or value of the good or service if market conditions other than those actually prevailing were to prevail’. So we are told that ‘what the legislature must be taken to have intended by “economic value” is the notional price of the good or service under assumed conditions of long-run competitive equilibrium’!

As the Act’s definition of ‘excessive price’ clearly has us all wondering what the legislature intended, the Tribunal thought that, given the categories with which the Act and all of competition law characteristically engage, the legislature might rather have intended us to identify a market structured so as to enable its participants to price without any cognisable competitive constraint. This we identified as a ‘super-dominant’ firm – a monopoly or virtual monopoly – in a market that did not permit, in any reasonable time frame, the prospect of new entry. The legislature would then have intended us to use our powers to restructure the market, or in the event that this fell outside our powers – as it did in respect of a first offence – to identify conduct in which the monopolist engaged so as to enable it to sustain supra-competitive pricing levels. It just so happened that because this particular monopolist was obliged, largely in order to operate at the lowest point of its cost curve, to produce a level of output significantly in excess of domestic requirements, it was also obliged to withhold supply of its product from the domestic market in order to attain its pre-selected price level, precisely the practice by which monopolies and cartels achieve supra-competitive pricing levels. Failing that – the one circumstance in which import parity pricing approximates the monopolist’s profit-maximising price – the state would have had to intervene in the market by regulating price. I can’t understand the consternation, on the part of the Court and some academic commentators, that our emphasis on market structure generated. Structure remains, after all, an important factor in market analysis and it’s precisely the existence of structures deemed chronically inimical to competitive outcomes – whether the result of restrictive licensing conditions or natural monopoly – that dictates the introduction of regulation.

None of this found favour with the Court. The absence of any reference in the Act to a ‘super-dominant firm’ appears to have disturbed them. I would have thought that a competition statute need make no explicit reference to a ‘super-dominant’ firm or monopoly, precisely because it implicitly represents the gravest threat, albeit rarely encountered in an unregulated form, to competition. More revealing, though, is the Court’s concern that this reference would have excluded oligopolies from the ambit of section 8(a). This was, of course, precisely what we intended. While a large firm in an oligopolistic market – a dominant firm – may have successfully engaged in exclusionary conduct, we used the concept of super-dominance precisely to convey our view that a ‘mere’ dominant firm in an oligopolistic market would not be capable of sustaining an excessively supra-competitive price without resort to a collusive agreement. An intended and, we deemed, positive outcome of this approach is that it limited the prospect of a proliferation of excessive-pricing claims, thus introducing an element of certainty into business decision-making.

However, the Court’s combination of black-letter legal interpretation and the abstractions of high economic theory predictably led nowhere. And so, many pages later, we are suddenly told that when ‘quantify[ing] things in money where only a rough estimate is possible’, it is necessary for the adjudicator to adopt a ‘fairly robust approach … particularly when account is taken that “long-run normal” profit and the conceptual basis upon which this term is predicated are notional’.

From this moment on, any attempt – never much in evidence – on the part of the Court to provide certainty to business pricing practices is thrown overboard. This is clear from several lengthy paragraphs in which the Court indicates the limitless number of instances – in reality a summary of the diverse methodologies used by various European courts – in which excessive pricing may be revealed by the necessarily ‘robust’ interpretation of concepts like ‘long-run normal profit’ which ‘cannot be employed with scientific precision’.

After this lengthy recital of the multitude of different measures – many of them inordinately complex – that may be used to determine excessive pricing, the Court strongly criticises our approach for the uncertainty it gives rise to!

In fact, our decision provides considerable certainty. It says that monopolies in incontestable markets will rationally price at the monopolist’s profit-maximising price. Under most circumstances, such a market would be, and should be, subject to price regulation or else the monopoly that inhabits the market should be dismembered. Had we been empowered to order divestiture on a first section 8(a) contravention, then that is what we would, in all likelihood, have ordered. However, in this instance, because the firm produced at a level of output significantly in excess of what would have enabled it to sustain its monopoly price, it was obliged to engage in a practice specifically designed to reduce the output available on the domestic market.

And so Mittal knew with absolute certainty what our decision required of it: it was required to terminate the arrangement with Macsteel International that underpinned the restriction of supply to the domestic output and thereby permit arbitrage between the price it realised on the international market and the price it charged on the domestic market. This is what we ordered Mittal to do. I’m convinced – and so, notably, were the Mittal witnesses – that this would have exerted significant downstream pressure on the domestic price. Mittal, as some of our detractors pointed out, could conceivably have circumvented this remedy by taking it upon itself to segment the export and domestic market – effectively by absorbing the Macsteel joint venture’s activities into its own functions. But then it too would have been faced with the imperative of segmenting its markets so as to reduce domestically-available supply. It would then have made itself vulnerable to a second excessive-pricing claim, which could have given rise to a divestiture remedy.

Alternatively, if government was not satisfied with the outcome of the remedy initially proposed, it could have taken steps to regulate the steel price or dismember the monopoly. These would no doubt have been controversial, but nonetheless credible, remedies and they remain so. If government believes the downstream benefit of a lower steel price would justify the controversy that would accompany either price regulation or enforced divestiture, then it should do so. Attempting to cajole or sweet-talk or threaten a monopolist into leaving money on the table is futile and all the evidence supports this. Using competition law is a complicated, time-consuming and, as we have seen, uncertain mechanism for resolving a monopoly pricing problem, the solutions of which are actually plain to see.

Ultimately, when explaining its decision to refer the matter back to the Tribunal so that it might apply one – I presume any one – of the many tests that it held were capable of determining whether pricing was excessive or not, the Court provided absolute certainty, at least with regard to its thinking, on the outcome. It notes that Harmony had calculated that Mittal’s domestic price was somewhere between 20 per cent and 61 per cent greater than its export prices – differences which the Court itself described, without further explanation, as ‘massive’. Clearly, then, we were instructed to use these as the comparator prices and, moreover, to find that the differences were indeed ‘massive’, which we presume means ‘excessive’, given that, as the Court acknowledged, this is a purely subjective judgment.

In the light of this, I can’t readily understand why the Court decided to refer the matter back to the Tribunal. Why it did not – as it could have done – simply substitute its decision for that of the Tribunal escapes me, as it had clearly reached a conclusion. I am of course not privy to the deal between Harmony and Mittal that eventually settled their dispute – and hence the entire matter – before the Tribunal had the opportunity to hear it. But, reading the Competition Appeal Court’s judgment, were I placed in Mittal’s shoes I would have felt under some pressure to settle. The reason is clear: the Competition Appeal Court had already effectively pronounced it guilty of charging a ‘massive’ differential between the domestic price and the export price. The upshot is that a public antitrust matter was settled in confidence between two private parties.

I’m not sure what remedy the Court would have favoured; presumably one that imposed a single price for steel. And while on the question of uncertain remedies, what would have been a less ‘massive’ differential, particularly given that there is bound to be a premium that consumers are willing to pay for local, over imported, steel? If 20 per cent was found to be the differential and that was (subjectively) deemed excessive, would Mittal, or any other firm similarly charged, be in safe territory at 15 per cent? Or if the differential had been found to be 61 per cent, would it have escaped liability under section 8(a) at maybe 32 per cent? This type of judgment makes a mockery of the requirement to provide any certainty to a firm’s critical pricing decision.

We could have covered our backs by including an additional ground for finding excessive pricing on the basis (backed up the data submitted by Harmony) that we have one of the lowest-cost steel producers in the world coexisting with steel prices that are among the highest in the world or that there was a differential between the domestic and international price for steel. The Competition Appeal Court would clearly have accepted a decision based on this simplistic reasoning.

But we have never written ‘for’ the Competition Appeal Court and it would have been the wrong decision. A monopolist geographically located as is Mittal will extract rent on the domestic market, just as it will be obliged to accept the discount that outward transport costs impose on its realised international price. So to impose a remedy, as sought by Harmony, that obliges Mittal to charge the factory gate price on all its steel output is not an efficient outcome. Nor, as we have said ad nauseam, is it the type of remedy a competition authority should impose. Our task is not to impose a price but to secure the most effectively-functioning market under less than ideal circumstances. Mittal’s excess production gave us the opportunity to impose a remedy that would have enabled market forces to exert downward pressure on the domestic price of steel.

This has probably been the most widely-discussed decision the Tribunal has taken. It is a decision that has been praised and criticised in equal measure. However, nothing that has been said or written has persuaded me we were wrong. Competition policy is rightly concerned with the pricing power of incontestable monopolies. But that doesn’t mean that competition law is the right weapon to deploy against it. However, to the extent that it is deployed, it can be done only on the basis of the powers and resources it possesses, and these do not include those necessary to regulate key economic sectors. Despite the views of the Competition Appeal Court and our other numerous detractors, I would rate this decision as the most significant contribution the Tribunal has made to international antitrust jurisprudence.

Nationwide Poles

I want finally to deal with a case that is significant because it posed the question of the interplay between competition and non-competition objectives in dealing with restrictive practices. This is the case in which Nationwide Poles alleged that Sasol Oil was practising unlawful price discrimination.25

Unlawful price discrimination is treated in the Competition Act as an abuse of dominance. However, it is not included as one of the ‘named’ abuses listed in section 8(d). It has rather been carved out of section 8 – the section entitled ‘abuse of dominance prohibited’ – and placed in a stand-alone section, section 9, entitled ‘price discrimination by dominant firm prohibited’.

Section 9 provides that prohibited price discrimination occurs when the pricing practices of a dominant firm discriminate between purchasers in the price charged, discounts extended or services offered for the purchase of goods or services of like grade and quality and when that discrimination ‘is likely to have the effect of substantially preventing or lessening competition’. However, section 9 further provides that the discriminating seller may escape liability if it can show that the differential is rooted in a ‘reasonable allowance’ for differences in the cost of manufacture and distribution entailed in supplying the purchaser charged the higher price, or where the lower price charged is a good-faith attempt to match a competitor’s price, or where a change in market conditions dictates the offer of a price differential (as in the case of obsolescence or deterioration of perishable goods).

Note, at the outset, that just as our Act accords separate treatment to price discrimination, so in the US price discrimination is not enforced through the Sherman Act, the antitrust statute, but rather through a separate piece of legislation, the Robinson–Patman Act. In Nationwide Poles, the Tribunal took the view that price discrimination is, in both the US and South Africa, accorded a separate status because its roots and purpose are to be found in the desire in both jurisdictions to support the ability of small enterprises to enter and thrive in a market. While, as I’ll elaborate, the promotion of small business is, on its own, not a mainstream objective of competition law, ease of entry into market is, and these two issues – ease of entry and the promotion of small business – interface, thus lending an orthodox competition law component to the concern with small business. Moreover, possibly because public perception, strongly and with good reason, associates antitrust with concern for the ‘small guy’ confronted with large concentrations of economic power, the defence of small business has become a fairly standard component of the ‘public interest’ element underpinning antitrust enforcement.

There can be no gainsaying that this consideration underpinned the Robinson–Patman Act. It’s also interesting to note that some 75 years on, with most US antitrust scholars and practitioners (whether of the defence bar or the Department of Justice itself) being strongly of the view that price discrimination is more likely to generate pro- than anticompetitive outcomes, a steady trickle of Robinson–Patman Act prosecutions still land up in court by way of private plaintiff actions. Similarly, in our Nationwide Poles case, the owner of Nationwide, Jim Foot, whose complaint was rejected by the Commission, elected to bring this matter to the Competition Tribunal himself.

Equally there can be no gainsaying the special status that South Africa’s policy-makers and Parliament accorded to small business in the origination and intended application of the Competition Act. Indeed, the Competition Appeal Court doesn’t appear to dispute this and it would have been hard pressed to do so. It’s plainly stated in the purposes of the Act.

Moreover, it is not only in these statements of general principle that the special status of small business is recognised. It’s a factor that the competition decision-makers are obliged to consider when considering mergers as well as applications for exemption. None of this implies that small-business considerations should, of necessity, trump the other considerations central to the administration of the Act, least of all competition considerations. Indeed, the Tribunal has, on numerous occasions, decided against small businesses. However, for reasons which I’ll explain – and which are reflected in the Tribunal’s decision – this is an instance in which small-business considerations should have triumphed. In particular, it is clear that, in this instance certainly, support for small business did not undermine competition.

I’m sure that it will immediately be pointed out that this is not the test to be applied. The test is not whether support for small business harmed competition, but rather whether the conduct of the dominant firm harmed competition. I’ll show why Nationwide Poles demanded flexibility in the identification and application of the test to be applied. However, the Competition Appeal Court favoured a rather arid adoption of orthodox approaches – largely followed by US scholars and prosecutors – without giving effective consideration to the national imperatives enshrined in the Act.

In Nationwide Poles, a small, indeed tiny, producer of creosote-coated wooden poles complained that the dominant producer of creosote, Sasol Oil, was charging it and other small producers of these products a higher price for creosote than that charged to its larger customers. There was the usual wrangle about the boundaries of the relevant market and whether or not Sasol’s share met the Act’s threshold for presuming dominance. In addition, the panel also considered whether Sasol’s pricing practices evidenced the possession of market power. While, as always, these proved to be interesting questions, I’ll not go into them here. Suffice it to say that Sasol’s overwhelming dominance of the relevant market (and its market power) was clearly established. However, the case is distinguished by the policy issues that it brought to the fore and the extent to which judicial notice is taken of these.

The essential uncontested fact is that Sasol was charging its larger customers a lower price for creosote than that levied on its small customers. And it was agreed that creosote was a materially significant input into the process of treating wooden poles.

To cut a long story short, we eventually had before us Mr Foot, the owner of Nationwide Poles, representing himself, against Sasol – South Africa’s (or is it Africa’s?) largest industrial company – represented by no fewer than three of Sandton’s finest law firms and their bevy of attorneys, as well as South Africa’s top competition economics consultant, Stephan Malherbe. I wondered at the time – and have occasionally subsequently asked myself as I do now – whether we allowed ourselves to be influenced by the David versus Goliath character of this case. I don’t think so. Our challenge was rather more prosaic and technical. We had to account for the unusual treatment of price discrimination in the architecture of the Act and for the unusual construction of section 9 itself. We also took the view that the particular facts of this case compelled us to give due consideration to the Act’s explicit support for small business. And out of all this, we were led to the conclusion that section 9 is an unusual hybrid of public interest and competition considerations, and that the public interest consideration in question was the ability of small business to enter and thrive, the one element of public interest in which there is a cognisable competition component.

I should elaborate this latter contention. While employment and the promotion of black economic empowerment constitute, when placed alongside competition considerations, ‘pure’ public interest considerations, support for SMEs, in addition to representing a public interest, also embodies cognisable competition considerations. This special status is reflected in many national competition policies and competition statutes, as well as in prosecutorial decisions and court judgments. But, this aside, it is axiomatic that when confronted, as in this case, by a market characterised by particularly low entry barriers (which appear, on the face of it, to favour small firm access) but nevertheless dominated by a small number of relatively large firms, attention, on solid competition grounds, is appropriately directed at conduct that may serve to explain this. This means conduct that erects artificial barriers to entry on the part of small firms.

One would not factor this into a case involving the steel industry, where the production process and concomitant capital barriers serve as an insurmountable barrier to small-scale entrants. But in the market for creosote-coated wooden poles, it is worthy of consideration. It was from this perspective that we examined Sasol’s price discrimination against small firms.

And attention to SMEs is bolstered by the public policy and statutorily mandated support – cast as the ‘public interest’ – for easing their ability to enter and thrive in markets where technical considerations permit of small firm participation. Of course this may involve a delicate balancing act and of course a competition decision-maker may overreach itself in supporting SMEs. An overreach in this instance would be reflected in a decision to support SMEs but which, in so doing, discriminated against efficient large enterprises and lowered their efficiency and that of the sector as a whole. While this does not render the support for SMEs indefensible on social grounds, it is not a decision that a competition decision-maker should be drawn into making, certainly not on competition grounds.

However, in Nationwide Poles, we have an instance in which price discrimination neither favoured Sasol nor promoted the efficiency of the larger firms in whose favour it discriminated. Indeed, Sasol’s market power enabled it to price as it wanted. I have no doubt that Nationwide Poles would have been as content with a remedy that raised the price of creosote for all Sasol’s customers to the level paid by its smallest customers, as with one that lowered the price for all customers to the level charged to the large customers. What Nationwide Poles wanted was a level playing field, one that enabled it to compete on equal terms with its largest competitors.

But I’m getting ahead of myself. Section 9 comprises two parts. The first describes the three essential elements that must be present if an act is to constitute prohibited price discrimination. Section 9(2) outlines the defences, those elements which, if present, will allow a firm to price-discriminate even if the elements of the first section are found to be present.

Sasol conceded that two of the elements of prohibited price discrimination were present: that Sasol discriminated between its customers of creosote as to price and discount and that the transactions between it and its customers were ‘equivalent transactions’. It did not, however, concede that its actions were ‘likely to have the effect of substantially lessening or preventing competition’. Moreover, it did not invoke the defences provided for in sub-section 2. Hence Sasol’s case rested entirely on whether or not its price discrimination was likely to lead to a substantial lessening or prevention of competition. If this was not proven, then all the elements that constitute prohibited price discrimination would not be present.

This is, of course, by no means the only time that the Act poses a ‘substantial lessening of competition’ test. Horizontal agreements – other than those that concern prices, market allocation or bid-rigging – are prohibited if they are proved to give rise to a substantial lessening of competition in the absence of countervailing pro-competitive gains, and so too are vertical agreements other than those that constitute minimum resale price maintenance. The Tribunal has determined – in South African Airways – that for a successful claim brought under section 8(c) and (d), harm to competition has to be established.

However, the section 9(1)(a) test departs from similar formulations in other restrictive practice provisions in the inclusion of the adjective ‘likely’. While the merger test specifically requires a finding of the likelihood of a substantial lessening of competition because merger analysis is, in a pre-merger notification regime, inherently predictive, restrictive practices cases are backward-looking. This means that they refer to conduct that has already occurred, so there should be no reason for a predictive test. However, there is no getting away from the fact that section 9, alone among restrictive practices, demands a predictive test of harm, if it demands a finding of harm at all.

To establish ‘likely’ harm is less onerous than the requirement to establish actual harm. In fact, it plainly suggests, as in merger reviews, that the harm has not already occurred. As in mergers, then, where ex ante action is taken to prevent likely damage to the competitive structure of the market, so too it can be inferred that section 9 is concerned with the likely impact of price discrimination on the competitive structure of the market. As I noted above, in a market with very low entry barriers, the direction from which competition is most likely to come is small new entrants, and so raising barriers to this class of firms – in this instance by obliging them to pay a premium for a critical input – is ultimately likely to harm the competitive structure of an otherwise accessible market and thus substantially lessen competition. So, if competitive harm had to be established in a section 9 price discrimination case, it was clearly likely harm. It was not the same as the section 8 test of competitive harm that we laid down in South African Airways. Here, harm that had already occurred – either in the form of direct harm to consumers or the harm inferred from significant foreclosure – had to be weighed against cognisable pro-competitive gains arising from the conduct in question. When this case ultimately made its way to the Competition Appeal Court, it overturned our decision in favour of Nationwide Poles. The Court was alive to the unusual use of a predictive test in section 9, but it held that the evidence had not conclusively established the likelihood of harm to competition arising from the price discrimination.

The Tribunal, for its part, expressly rejected the argument that in order to prove prohibited price discrimination it was necessary to prove competitive harm at all. But we still had to account for the provisions of section 9(1)(a), which clearly stated that to prove prohibited price discrimination the complainant had to prove that the pricing practice in question ‘is likely to have the effect of substantially preventing or lessening competition’.

We approached this task by attempting to account for the separate treatment of price discrimination. Why was it not simply incorporated as one of the named offences in section 8(d), given that, unlike the Robinson–Patman Act, dominance was a necessary prerequisite to establish prohibited price discrimination? Had price discrimination been included in the ‘named’ exclusionary practices listed in section 8(d) or had it been prosecuted under the general ‘exclusionary acts’ prohibited in section 8(c) (though defined in section 1(x)), then we would have had no warrant for treating the ‘competitive harm’ test differently from that of the other section 8 offences, but for the inclusion – not unimportant – of the qualifier ‘likely’.

However, the fact remains that it wasn’t included in section 8 but rather, not unlike its statutory treatment in the US, was hived off into a separate section. From this we inferred that it was to be subject to a different standard, to a different set of criteria, from those that governed the core abuse-of-dominance offences. But it had to be connected to the overall thrust of the Competition Act and its broad objectives. Jim Foot, Nationwide Poles’ owner and representative before the Tribunal, made the disarmingly simple, but utterly persuasive, observation that, given that suppliers are extremely unlikely to discriminate against their large customers, the clause was manifestly intended to support SMEs. Hence we concluded that, unlike the section 8 offences, section 9 did not proscribe a pure competition offence, but rather a hybrid offence that incorporated both competition elements and public interest elements, the latter enshrined in express terms in the Act and in the underlying policy’s explicit mandate that accords special consideration to the plight of small business.

Once we had explained the separate treatment of price discrimination, we still had to account for the apparent requirement to establish that it was likely to cause a substantial lessening of competition. While it’s unfortunate that the requirement is cast in language similar to that of sections 4, 5 and 8 – I emphasise ‘similar’, but not identical because of the predictive character of the section 9 requirement – this does not preclude the necessity to give it meaning within the specific context of section 9 (rather than section 8) and within the overall context of the Act. Here, Foot’s observation took on an even stronger meaning. As already noted, Sasol conceded that its conduct conformed with the elements of prohibited price discrimination specified in 9(1)(b) and (c) in so far as it discriminated as to price or, more accurately, rebate, and it did so in respect of equivalent transactions. It was only the element of a likely substantial lessening of competition that, Sasol argued, was not present. However, given, as Foot argued, that a large customer was unlikely to suffer discrimination, and given that a weakening of the competitiveness of a small player was unlikely to harm competition, if the section 8 meaning of competitive harm was adhered to or even if the predictive test of competitive harm used in merger analysis was used, section 9 would effectively be rendered a nullity by the requirement to prove harm to competition or even merely ‘likely’ harm to competition.

The Tribunal was also influenced by the conspicuous absence from the unusually elaborate defences provided for in section 9(2) of a provision for a pro-competitive defence, thus further distinguishing it from sections 8(c) and (d).

Why the conspicuous absence of a pro-competitive defence? One reading is that section 9 represented a species of limited per se prohibition, which is an extremely implausible interpretation. Or it suggested that, as with Mr Foot’s interpretation, this was a clause that, in part at least, was intended to defend both a public interest (as in small business) and the long-term competitive structure of the market (as in not inhibiting entry by small enterprises to a market whose barriers were, but for the price discrimination, unusually low). That is to say, no pro-competitive defence is provided for, because competitive harm is not necessarily contemplated by conduct that is about promoting a hybrid of competition and public interest concerns.

However, even though our examination of the architecture of the Act and the wording and architecture of section 9 led us to conclude that we were not required to conduct a standard ‘lessening of competition’ test, we still had to give meaning to section 9(1)(a), which identified a likely substantial lessening of competition as an element, albeit not the outcome, of price discrimination. We ascribed to this clause ‘competition relevance’ rather than ‘competition harm’. Price discrimination is, of course, ubiquitous. A house seller will negotiate a lower commission with an estate agent than the seller next door; a car purchaser will negotiate a better purchase price or trade-in price than the customer before her. This has no competition relevance and the Competition Act should not be invoked by every disaffected consumer who had paid more for a good or service than his neighbour. On the other hand, discriminating against a class of customers by charging them a price for a critical input higher than that charged to their competitors is of competitive relevance even if it does not result in harm to competition. The more so if the Act specifically seeks to promote the class of competitors discriminated against.

We also took note of Sasol’s inability to explain why it engaged in price discrimination. We fully expected it to cite differences in transaction costs, but it did not do so. In fact it did not dispute Mr Foot’s explanation that because he, like his larger competitors, was supplied with creosote by the truckload there were no cost differences entailed in supplying a large or small customer. Nor were any of the other defences provided for in section 9(2) invoked.

Although Sasol was not strictly entitled to offer up an explanation for discrimination that fell outside those provided for in 9(2), some time was spent on this issue. And its failure to provide a business reason – not to mention one that met the requirements of Section 9(2) – was telling.

Sasol appeared to discriminate simply because this is what it had always done and, not surprisingly, it is what its large customers preferred it to do. The product was, from Sasol’s perspective, peripheral – a by-product of its core petro-chemical business – and if its larger customers wanted protection from smaller upstarts, Sasol, well-acquainted with the advantages of being the largest fish in the pond, was prepared to extend this status to the biggest fishes in an admittedly infinitely smaller pond. Accord with its largest customers, even if this involved discriminating against its smaller customers, represented for Sasol the least troublesome way of marketing a product of marginal significance to itself.

Accordingly, Sasol was not segmenting markets and efficiently charging each market segment a profit-maximising price, the argument that economists would generally offer in support of price discrimination. As we have noted, it could have raised its price to its larger customers to that which it charged its smaller customers and there is nothing that the latter could have done about it. If this levelled playing field promoted the ability of small players in the downstream market to thrive, the higher input charge to the large players might well have generated a more competitively-structured downstream market, with the larger players subject to more robust competition.

Instead, Sasol chose to defend vigorously – right up to the Competition Appeal Court – a pricing structure for which it could provide no rational explanation. I have little doubt that, had this matter emerged now, Sasol, several competition convictions later, would have sought a settlement of the matter.

The Competition Appeal Court overturned us.26 It accepted the argument advanced by Sasol with respect to the question of a likely substantial lessening of competition. That is to say, it held that a substantial lessening of competition had to be given the same meaning as it had been given in the rest of the Act. And, although we had accepted that small producers were substantially disadvantaged by the price discrimination – this much is of course obvious – the court was not satisfied that the evidence demonstrated that competition, rather than competitors, had thereby been harmed (which, on Foot’s contention, is no less obvious).

Although I remain thoroughly unpersuaded by the Competition Appeal Court’s decision, it is a safe and somewhat credible decision. It doesn’t account for the separation of section 9 from the rest of abuse of dominance – indeed, it treats price discrimination as simply another named abuse-of-dominance offence. More surprising is its failure to account for the distinctive wording of the section that the Court held required proof of competitive harm, or for the conspicuous absence of a pro-competitive defence, the latter a feature of all other rule-of-reason abusive conduct. This represents a glaring omission from the Court’s decision.

On the question of evidence, while the pricing differential between those who, like Nationwide Poles, were charged the highest price for creosote and those at the most privileged end of the price spectrum is not insignificant – and nor, therefore, is the competitive disadvantage suffered by the smallest players – the Competition Appeal Court is able to claim credibly that there is no significant impact on consumer welfare or foreclosure. However, Foot’s point – one we accepted unreservedly – is that an infinitely larger difference in price would, on the conventional reading of harm to competition, make no difference because even if all the firms in the Nationwide Poles category had closed their doors (as Nationwide did, several months after the Court’s decision), it would have made little difference to competition because they were not significant competitors, at least not at that stage. The decision, even on competition grounds, might have been different had small business been viewed as a clear point of new entry into this market. This would be the case particularly as the requirement is to find a likely substantial lessening of competition, because it may well have been that the price discrimination made a difference, not to the decision whether to enter or not, but to the ability to thrive and grow.

Finally, of course, the Competition Appeal Court refused to accord any significance to the public interest mandate to promote small enterprise. Again, this is at one with antitrust orthodoxy, but, whether or not it accords with the doctrinal predilections of the Court, the Act does reserve a special place for the promotion of small business. And we have here a classic type II error, an error of under-enforcement, because the court judgment significantly reduces the prospect of small business ever succeeding in a price discrimination suit.

CONCLUDING REFLECTIONS

The judge president of the Competition Appeal Court has frequently taken issue with our insistence that the Act be interpreted purposively rather than literally. He insists that, in both Mittal and Nationwide Poles, we departed impermissibly far from the plain meaning of the statute. In Mittal we showed that the complex and confusing definition of excessive pricing prevents resort to plain meaning and this is borne out by the Court’s own interpretation of that definition. And as for Nationwide Poles, even the meaning of plain language, and sometimes even identically-worded plain language, is influenced by context; and the context within which the plain language at issue in Nationwide Poles is inserted into section 9 of the Act differs markedly from all the other contexts in which it is placed throughout the Act.

But I grant that the first resort in statutory interpretation must be to plain meaning. However – and here’s the rub – this is a statute that requires not only legal and statutory interpretation, but also the interpretation of public policy as enshrined in a number of public interest norms and an appreciation of economic theory and applied economic analysis. In smooth waters, plain meaning will take us far, but as soon as the going gets rough the adjudicator needs to put down his Shorter Oxford English Dictionary and reach for his Palgrave Dictionary of Economics and be prepared for entering a world riven with contested meanings that are anything but ‘plain’. After all, when reading the Competition Appeal Court’s decision in Mittal one can do nothing other than laugh or cry at the notion that ‘what the legislature must be taken to have intended by “economic value” is the notional price of the good or service under assumed conditions of long-run competitive equilibrium’. Unless, of course, one does not accord plain meaning to the term ‘what the legislature intended’ and it really means ‘what some writers of undergraduate economics textbooks must have intended’.

But that, I guess, is the exciting part of abuse-of-dominance cases. Reasonable people are prone to disagree on almost every aspect of the evidence and argument. I don’t think this is a reason for taking an overly cautious view of prosecuting abusive conduct. I think, though, that it may be a good reason for having economists sit as assessors on the Competition Appeal Court. In competition matters this may help introduce the necessary complexity that the admittedly uncertain science of economics must, of necessity, lend to the ‘plain language’ of the statute.

I think that, in this area, the real challenge for the Commission is going to lie in the selection of section 8 cases for prosecution. There are a limited number that can be taken each year. These are, as we have seen, cases that consume enormous resources and where the outcomes are uncertain at best. However, given the high incidence of single firm dominance in many critical markets, in significant part attributable to previous (and current) state ownership of, and support for, selected enterprises, it’s imperative for the Commission to take on enough and sufficiently varied abuse cases to keep dominant firms on their toes, to make them aware that their conduct is subject to scrutiny.

But the emphasis must be on strategic selection. My advice, for what it’s worth, is to tread carefully around excessive-pricing cases. It would be regrettable if the Commission were to be emboldened by the Competition Appeal Court’s decision in Mittal. In this decision a dangerous blend of a ‘robust approach’ to evidence and argument amounts to little more than a licence to engage in pure subjectivity. If followed, it could induce the Commission to take on cases that are filled with uncertainty and that would compromise their engagement in competition policy, in advocating to government instances where price regulation rather than competition law is the appropriate instrument. Moreover, to the extent that active scrutiny and prosecution of the conduct of single dominant firms chill competition – and I don’t think this is generally a major consideration – creating major uncertainty around pricing practices is probably the most likely candidate for chilling competition.

NOTES

1. Competition Commission v Federal Mogul Aftermarket Southern Africa (Pty) Ltd and Others (08/CR/Feb01).

2. Competition Commission v South African Airways (Pty) Ltd (18/CR/Mar01).

3. Competition Commission and Another v British American Tobacco South Africa (Pty) Ltd (05/CR/Feb05).

4. Harmony Gold Mining Limited, Durban Roodepoort Deep Limited v Mittal Steel South Africa Limited, Macsteel International Holdings BV (13/CR/Feb04).

5. Nationwide Poles v Sasol (Oil) Pty Ltd (72/CR/Dec03).

6. Netstar (Pty) Ltd, Matrix Vehicle Tracking (Pty) Ltd and Tracker Network (Pty) Ltd v The Competition Commission and Tracetec (Pty) Ltd (97/CAC/May10).

7. Leegin Creative Leather Products, Inc. v PSKS, Inc., 551 U.S. 877 (2007).

8. Dr Miles Medical Co. v John D. Park & Sons Co., 220 U.S. 373 (1911).

9. The Competition Commission of South Africa v Federal Mogul Aftermarket Southern Africa (Pty) Ltd, Federal Mogul Friction Products (Pty) Ltd, T&N Holdings Ltd, T&N Friction Products (Pty) Ltd (08/CR/Mar01).

10. Federal Mogul Aftermarket Southern Africa (Pty) Ltd v The Competition Commission of SA and The Minister of Trade and Industry (33/CAC/Sep03).

11. See The Competition Commission v South African Airways (18/CR/Mar01).

12. Ibid.

13. David Lewis, ‘Chilling competition’ in B. Hawk (ed.), International Antitrust Law and Policy (Fordham Corporate Law Institute, 2009).

14. Verizon Communications Inc. v Law Offices of Curtis V. Trinko, LLP (02-682) 540 U.S. 398 (2004).

15. This is precisely the conclusion that the Federal Trade Commissioners arrived at in their response to the Department of Justice report (p. 1): ‘The department’s report is chiefly concerned with firms that enjoy monopoly-power, or near-monopoly-power, and prescribes a legal regime that places these firms’ interests ahead of the interests of consumers. At almost every turn, the department would place a thumb on the scales in favour of firms with monopoly or near-monopoly-power and against other equally significant shareholders’.

16. As the Federal Trade Commissioners note in their response to the Department of Justice report (p. 4, our emphasis): ‘For one reason or another, it may take a long time for rivals to surmount entry barriers or other impediments to effective competition. Indeed, the monopolist’s own deliberate conduct may further delay a market correction and prolong the duration of consumer harm’.

17. Simon Roberts, ‘Administrability and business certainty in abuse of dominance enforcement: An economist’s review of the South African record’ (Paper presented at the 5th annual conference on competition law, economics and policy, University of Johannesburg, 4 and 5 October 2011).

18. Anne Njoroge, ‘Regulation of dominant firms in South Africa’ (MBA dissertation, Gordon Institute of Business Science, 2010).

19. Harmony Gold Mining Limited, Durban Roodepoort Deep Limited v Mittal Steel South Africa Limited, Macsteel International Holdings BV (13/CR/Feb04).

20. Mittal Steel South Africa Limited, Macsteel International BV, Macsteel Holdings (Pty) Limited v Harmony Gold Mining Company Limited, Durban Roodepoort Deep Limited (70/CAC/Apr07).

21. United Brands Company v Competition Commission (Case 27/76, [1978] ECR I-207).

22. Andrew Sylvester, ‘Using profitability measures in competition analysis in South Africa’; and Anthony Fleet and Thabelang Moila, ‘The use of profitability analysis by competition authorities’ (Papers presented at the fourth annual conference of the Competition Commission and the Mandela Institute, University of the Witwatersrand, 2010).

23. Harmony Gold Mining Limited, Durban Roodepoort Deep Limited v Mittal Steel South Africa Limited, Macsteel International Holdings BV (13/CR/Feb04).

24. Mittal Steel South Africa Limited, Macsteel International BV, Macsteel Holdings (Pty) Limited v Harmony Gold Mining Company Limited, Durban Roodepoort Deep Limited (70/CAC/Apr07).

25. Nationwide Poles and Sasol (Oil) Pty Ltd (72/CR/Dec03).

26. Sasol Oil (Pty) Ltd v Nationwide Poles (49/CAC/Apr07).

5. Cartels

While competition law enforcement enjoyed strong support from the ANC and its supporters, even prior to its entry into government, it’s fair to say that the penny really dropped on the morning of 28 August 2007, when Nick Dennis, the usually self-assured CEO of the giant food group Tiger Brands, appeared before the Competition Tribunal to represent his company in an application for the approval of a consent order it had entered into with the Competition Commission. On this occasion – which was to be the first of several for the group – Tiger Brands admitted its participation in a bread cartel. Until then, most of the corporate titans who had been obliged to appear before the Tribunal seemed at once bemused and not a little irritated at having to account for their business decisions. Dennis appeared no less bemused than many of his peers, but on this occasion the bemusement was accompanied by palpable nervousness, rather than irritation.

Although, as I have already recounted, the competition authorities had, through the merger review process and several restrictive practices cases, begun to acquire a reputation for both toughness and professionalism, it would be fair to say that until the bread cartel their presence had not yet made itself felt much beyond the business community. While I think that business had by then already understood and reluctantly conceded the legitimacy of competition law enforcement and merger regulation, as far as the rest of the society was concerned – and as we have seen from our earlier references to the cement dawn raid, this certainly included the judges of the Supreme Court of Appeal – we were still largely seen as the defenders of a set of ‘victimless’ statutory rules. The early restrictive practice cases in the agricultural sector were in fact initiated and won by ‘small guys’ taking on dominant firms in the sector. However, given that much of our high-profile work involved merger regulation, the victims of those who fell foul of the Competition Act remained pretty much invisible, although the appearance of some of South Africa’s leading business figures at merger hearings had already attracted considerable media and public interest. This all changed with the Tiger Brands hearing. From then on, the public and their representatives took up the cause of competition law with a vengeance, setting in train a range of important consequences for the content and practice of competition law in South Africa.

The uncovering of the bread cartel was an early product of the Commission’s corporate leniency programme. Although the existence of a Western Cape bread cartel was initially revealed to the Commission by two sales agents (the cartel involved a conspiracy both to fix the price of bread and to fix the commission paid to certain of the independent sales agents), Premier Foods – one of the largest South African baking and milling concerns – applied for and received leniency from the Commission. In its application, Premier Foods revealed the existence of a price-fixing conspiracy that extended beyond the Western Cape as well as a national milling cartel. Because of the additional information that it provided, it received leniency – that is, immunity from prosecution – conditional of course on its providing full assistance to the further investigations and the prosecution of its co-conspirators. Following Premier’s admissions, all but one of the alleged conspirators – Pioneer – entered into consent agreements with the Commission in which they too admitted their participation in national and regional bread cartels and agreed to pay administrative penalties and set up compliance programmes. Tiger Brands was one of the conspirators that concluded a consent agreement.

The Competition Act requires that consent agreements be approved on application to the Competition Tribunal. While the Tribunal is naturally not entitled to alter the terms of a consent agreement unilaterally, it may – and has, on rare occasions, done so – deny an application for approval of a consent agreement. The prevalence of consent decrees, which are predominantly the final stage in the resolution of cartel allegations, suggests that contested litigation is, in these matters and in contrast to abuse-of-dominance cases, the exception rather than the norm. Although I’ll discuss some important instances of litigation surrounding cartels, both as to procedure and to the merits of alleged cartel practice, the prevalence of consent orders speaks of the centrality played by proof of the existence of an agreement in the conviction of cartels. Once the dawn raid has uncovered the critical evidence of a conspiracy or a conspirator has applied for leniency, with the cooperation with the prosecution that this implies, the game is generally up. And so, sooner or later, a raft of consent orders, rather than lengthy trials, are concluded, implying a lesser adjudicative role for the Tribunal. However, even the consent order hearings, though for the most part uncontested, have had their fair share of drama and have played a critical advocacy role.

Consistent with the standard practice of the Tribunal, applications for the approval of consent orders are held in open court. On the occasion of the Tiger Brands application, several civil society groups as well as the complainant requested and were granted permission to make submissions at the hearing. These included the Black Sash, a long-standing and highly regarded human rights group; the South African Human Rights Commission; the National Consumer Forum; and the Congress of South African Trade Unions (Cosatu). The whistle-blower also made a submission. A number of issues that arose through these presentations and the response of Tiger Brands provide interesting insights into future developments in competition law.

In the Tiger Brands consent order hearing, the Commission, as is common practice, opened proceedings with a description of the cartel, including an outline of the leniency application. It asked us to confirm a fine based on 5.7 per cent of affected turnover. It was left to the Tribunal panel to point out that the Act permitted a fine of up to 10 per cent of total turnover. However, the Commission representative indicated that his organisation was well satisfied with the size of what, at that stage, had been, at R98 million, the highest fine ever imposed by the competition authorities and with the basis of calculation, namely the affected turnover. This – the question of the appropriate measure of turnover as the basis for the calculation of the fine – would, in time, become a highly contentious issue.

The representatives of civil society who then made submissions to the Tribunal did not, by and large, agree with the Commission on the size of the administrative penalty. Cosatu asked that the fine be based on total turnover; the Black Sash and the complainant asked for the fine to be based on the full duration of the cartel, something that the law did not permit; and, in a theme that would be repeated with increasing intensity, several of the submissions stressed that mechanisms should be found whereby the fine was used to recompense the class of consumers directly affected by the cartel. Indeed, the Human Rights Commission argued that the cartelisation of basic food products such as bread constituted not merely a contravention of the Competition Act, but a violation of the Bill of Rights. In one memorable characterisation, it stated that ‘When placing illegal corporate activities within their complete social context, anticompetitive practices become thieves at the dinner table’. The company’s presentation also took a form with which we were to become increasingly familiar. It had all the appearance of contrition – even its counsel, Jeremy Gauntlett, whom I could not easily have imagined doing contrition – managed to convey the appearance of consuming humble pie on his client’s behalf.

But again, and their apparent contrition notwithstanding, the testimony followed what would come to be a predictable path. Tiger’s CEO formally took responsibility for the actions of his company, but insisted throughout that the offending decisions were taken and the unlawful agreements concluded several layers beneath his high office. When it was pointed out that certain of the decisions involved the closing down of several significant bakeries – belonging both to Tiger Brands and its competitors – as part of a market allocation scheme, action that must have required authority at the highest level, he insisted that, while he had been party to a decision to reduce Tiger Brand’s presence in the smaller rural towns, at no stage was he aware that this was being done pursuant to an anticompetitive scheme. Like many others who would subsequently appear before us, he came across as someone either out of touch with the activities of a business that he was handsomely rewarded to manage, or economical with the truth. Either way, within 3 months of his appearance before the Tribunal, Dennis, once one of the country’s most powerful and highly remunerated business leaders, had resigned his position as Tiger’s CEO, his own and the company’s reputation severely tarnished.

At about this time, Sasol, South Africa’s leading industrial company, was hit with a massive fine imposed by the European Commission for the leading role played by one of its European subsidiaries in the establishment and maintenance of a wax cartel. And so, with fines imposed on some of South Africa’s corporate icons by the South African and European competition authorities, it was becoming increasingly clear that we had seen only the proverbial tip of the iceberg. I doubt whether anyone, even the officials of the competition authorities, had much inkling of the size of the iceberg that remained, and undoubtedly still remains, hidden from view. But let’s step back and examine this, the hanging offence in competition law.

Section 4(1)(a) of the Act prohibits agreements or ‘concerted practices’ among competing firms if the agreement has the effect of substantially lessening or preventing competition unless a party to the agreement can prove that it produced pro-competitive gains that outweigh the negative impact on competition. However, section 4(1)(b) characterises as ‘restrictive horizontal practices’, and prohibits outright, all agreements among competitors that fix prices or any other trading conditions, that divide markets and that rig bids (that is, submit collusive bids in response to tenders). Where an agreement meets the terms of section 4(1)(b), the complainant is not required to prove a negative impact on competition and the respondent is not entitled to invoke a pro-competitive defence. In the language of the trade, 4(1)(a) provides for ‘rule of reason’ offences, while 4(1)(b) offences are prohibited per se.

The remaining sub-sections of this clause provide that where firms which hold a significant interest in each other or which have a director or substantial shareholder in common engage in a ‘restrictive horizontal practice’, an ‘agreement’ will be rebuttably presumed to exist. Thus, in these circumstances, the Commission will not have the onus of proving the existence of an agreement to collude – rather, the firms will have the onus of proving that their presumed agreement did not constitute price-fixing, market allocation or bid-rigging but was rather ‘a normal commercial response to conditions prevailing in that market’. This clause was inserted to discourage the ubiquitous cross-directorships and shareholdings that characterise South Africa’s corporate structure and that undoubtedly facilitate anticompetitive conspiracies.

For the rest, section 4 provides that the proscription of conduct specified in section 4(1) will not apply to agreements between a firm and its subsidiaries or among members of a single economic entity.

A glance at section 4(1)(b) will confirm a point made earlier, and that is that the law and economics of proving cartel conduct are fairly straightforward, in marked contrast to the often mind-bending complexities involved in proving an allegation of abuse of dominance. All that 4(1)(b) requires is to prove the existence of an ‘agreement’ or ‘concerted practice’ or a decision by an association of firms – which are themselves very broadly defined1 – and then that the agreement went to the fixing of prices, the allocation of markets or the rigging of bids. Although the Supreme Court of Appeal has, in its wisdom, introduced a further layer of complexity, certainly a plain reading of the Act requires nothing more from either a legal or an economic standpoint. In all the analytical advances – not to mention the passing fads and fashions of economists – that have characterised antitrust analysis, particularly in the sphere of economics, there has been no serious attempt to question the economic damage wrought by ‘hardcore’ or ‘naked’ cartel conduct, the terms commonly used to describe the practices referred to in section 4(1)(b). These are accordingly prohibited per se – no proof of economic harm is required.

However, cartels are immensely difficult to detect. Their illegality is widely appreciated and so the agreements are generally concluded under conditions of great secrecy. But firms do make mistakes, which range from the association of South African bicycle retailers that advertised its cartel on its website, to the more common error of exposing the existence of a prohibited agreement through email exchanges or diary entries that are subsequently uncovered in the course of a dawn raid or in the process of discovery.

Fortunately, cartel arrangements are not easy for the conspirators to enforce. The incentive to cheat is considerable and it is sometimes difficult to maintain secrecy while simultaneously monitoring compliance and disciplining cheating co-conspirators. So, for example, a sudden, short, sharp price war in circumstances where prices have been reasonably stable often signals a cartel that is inflicting punishment on a deviant member or is experiencing general difficulties in maintaining its conspiracy. As we’ll see when we discuss the Pioneer case, the bread cartel was extremely unstable, necessitating action by the conspirators that eased the task of detection and evidence-gathering.

It’s also possible to construct a profile of a likely cartel or to recognise signals that may point to cartel conduct. So it’s widely accepted that the ability and, more important, the incentive to collude in markets producing homogeneous products are greater than in markets where branding and product differentiation figure prominently in competitive strategies, hence the regular incidence of collusion among cement producers.

However, economic evidence alone is generally insufficient to prove a cartel. For example, as already indicated, the outbreak of a sudden price war may be evidence of a robustly competitive market or of a cartel disciplining deviant members. Hence, the requirement to prove a meeting of minds. However, in well-established oligopolistic markets, information-sharing and other forms of signalling may well be sufficient to forge an illegal agreement. This is why there needs to be a very broad definition of agreement and the possession of the right investigatory powers to detect a carefully-constructed conspiracy.

As I write this, the morning newspaper’s business headlines report on a price war in the mobile telephony market. It inevitably brings to mind the agreement allegedly concluded back in the late 1990s in London between the members of the then mobile phone duopoly in which Vodacom and MTN participated, which was exposed by the erstwhile Competition Board and which the powerful companies successfully kept out of court. But then again, there have since then been many technological advances in telecommunications and several new entrants in mobile telephony. So does this claimed price war herald the advent of competition or does it signal a cartel ordering its affairs, soon to return to the market with offerings that are characteristically impossible to tell apart?

Judicially sanctioned search and seizure powers – the so-called ‘dawn raid’ – have been an essential part of the investigative armoury in many countries, including South Africa. Following the fiasco surrounding the Commission’s first dawn raid on the cement producer PPC in 2000, the Commission did not undertake a raid again until 2006. When it did so, the raids proceeded with considerably less public fanfare than the Commission’s initial foray, indeed as often as not with no public notice that they had even taken place. I recall arriving at the office very early one morning and finding, not without a pang of envy, a platoon of Commission officials about to mount what was, in this instance, literally a dawn raid.

However, at least as significant an element of the cartel detection armouries of competition authorities is the introduction of corporate leniency programmes. These are essentially plea bargaining arrangements that allow cartel members total or partial immunity from prosecution in exchange for cooperation with the investigation. The agreements include the provision of information that is not already in the possession of the authorities and the giving of evidence at the trial of its fellow conspirators, should a settlement not be achieved with them. The Commission introduced a leniency programme only in 2004. However, firms will not use the opportunity to apply for leniency unless they have reason to fear that they will be apprehended and that the consequent sanction will be sufficiently severe to induce them to take a step that will have them ratting on their co-conspirators and bring about the demise of a profitable collusive arrangement. Accordingly, the first application for leniency was made only some considerable time after the introduction of the programme.

Our engagement with cartels had started some time before this. Although there were several cartel cases involving open price-fixing in the medical and legal professions, our first serious entanglement with a cartel required no elaborate detection mechanisms or powers. This was Ansac, the cartel of American soda ash producers, a Webb–Pomerene association, so named because it permitted, under the terms of the US Webb–Pomerene Act, the US producers to operate a cartel in export markets, although not of course in the US market itself.

I had imagined that this would take a few short weeks: this, after all, was a group of firms that was effectively in possession of a licence to operate a cartel, albeit a ‘licence’ that did not immunise it from prosecution in affected jurisdictions outside the US. What more proof was needed? But I had not reckoned with a global cartel that had fought – and, mostly won – several high-profile battles to cartelise a range of important national markets, most famously including India, whose highest court found that the competition authority did not possess the extra-territorial jurisdiction necessary to prosecute the US-based cartel.

Ansac presented itself as an efficiency-enhancing joint venture, a group of firms associated for the purpose of achieving economies of scale in transport and logistics, without which it would not have been able to export.2 The fact that the price of soda ash had also to be fixed was, on their imaginative version, simply ancillary to their main purpose of achieving scale economies in transport and logistics. The evidence, when we eventually got to hear it, demonstrated, to the surprise of no one, that this was a flagrant contrivance – indeed, far from being an efficiency-enhancing joint venture, Ansac is more accurately described as a highly effective, tenacious, take-no-prisoners lobbying and litigating force, one dedicated to keeping a naked price-fixing export cartel in operation.

In fact, throughout the case, Ansac’s defence persistently raised a point that was indeed initially puzzling. This was the identity of the complainant and an intervener in the case before the Tribunal: Botswana Soda Ash (Botash) is a Botswana soda ash producer part-owned by De Beers, the long-time organiser of the notorious international diamond cartel (and so, presumably, well qualified to recognise a cartel when it saw one!) and the Botswana government. Surely Botash should have been pleased to participate in the South African market (which it could not supply in full) under the umbrella of the US export cartel. It certainly crossed my mind that the bitter battle between these rivals may have signalled that Botash itself had once been part of a clandestine southern African cartel with Ansac, which had broken down as one or other of Ansac or Botash attempted to expand its presence in the South African market. I recall that the evidence did suggest that, while transport costs did point to Botash supplying the inland areas of South Africa and Ansac the coastal areas, the respective geographical shares had at one point remained remarkably stable for a long time.

The eventual trial did not conclusively explain Botash’s anxiety to prevent the cartel operating in South Africa. Ansac of course attributed Botash’s motive to its desire to monopolise the South African market. But this made no sense because, apart from the fact that Botash did not have the capacity to satisfy the South African market for soda ash, Ansac’s departure simply meant that its members, each considerably larger than Botash, would enter the South African market individually. Indeed, it probably would have been wholly possible for them to maintain their transport and logistics joint venture, as long as each company marketed and priced separately. The spectre, raised by Ansac, of a Botash monopoly was in fact nothing other than a rather unsubtle threat on the part of Ansac to organise a boycott of the South African market by the individual Ansac members, a threat that was later explicitly communicated in the Tribunal hearings.

However, after listening to the evidence of Ansac’s ruthless lobbying, both in the non-US jurisdictions in which it operated and in the US itself, I concluded that, had I been in Botash’s shoes, I too would have been inspired to do whatever was necessary to get Ansac out of my neighbourhood, even if this ultimately meant more robust competition with its constituent members. I recall testimony to the effect that Ansac had successfully lobbied the US government to deny export credit assistance to a US firm that had successfully bid for the contract to build a railway line from a large soda ash deposit in Kenya to the port of Mombasa. Indeed it appears that, for a time, Ansac’s lobbying of US trade negotiators had successfully excluded Botash product from the South African market, resulting in a lengthy closure of the Botswana operations. When the consent order was heard at the Tribunal, the panel suggested to the Commission that it put the order on the website of the International Competition Network and that it consult with those of its fellow competition authorities whose economies were similarly subject to the scourge of this international cartel. I don’t, however, believe this was ever done.

I’ve already described the tsunami of litigation launched by Ansac that effectively prolonged the finalisation of this matter for little under 10 years, only for the export cartel ultimately to concede at the conclusion of the eventual hearing to precisely the charges levied against them at the very beginning of the proceedings. They then consented to pay an administrative penalty based on their turnover in South Africa in the year preceding the filing of the charges against them, and had effectively been able to continue operating as a cartel for the duration of the proceedings. The administrative penalty amounted to R10 million, almost certainly smaller than the legal fees entailed in defending the cartel and undoubtedly dwarfed by the rents derived from the prolongation of the price-fixing conspiracy.

Much of the litigation launched by Ansac was purely vexatious and obstructive, for example the risible attempt to have Botash’s attorneys disqualified on the grounds of a thoroughly contrived interest conflict, a matter that Ansac sought, unsuccessfully, to litigate as far as the Constitutional Court. However, on the simple proposition that a doublebarrelled shotgun fired into a crowd is sure to hit at least one target, Ansac did on one occasion succeed in persuading no less a body than the Supreme Court of Appeal to take a decision that, while insufficient to maintain the presence of the Ansac cartel in South Africa, may well, despite assurances to the contrary by the court itself, have compromised the operation of the per se rule. This was the case in which the Supreme Court of Appeal – overturning both the Tribunal and the Competition Appeal Court – held that it was necessary to consider evidence in order to ‘characterise’ an agreement as one that fell within the ambit of section 4(1)(b) before proceeding to apply the per se rule.

The Supreme Court of Appeal relied on the well-known US Supreme Court judgment in Broadcast Music, Inc. (BMI) v CBS.3 This case concerned the activities of two institutions, BMI and the American Society of Composers, Authors and Publishers (Ascap), both of whom were involved in the issuing, for a fee, of blanket licences to broadcast the copyrighted musical compositions of the composers who had assigned to one or other of the two institutions a non-exclusive right to sell their compositions to broadcasters. At the time this matter was heard, the repertory of BMI and Ascap comprised almost every US copyrighted composition, then some four million compositions in all. ‘The basic question presented’, wrote Justice White, ‘is whether the issuance by Ascap and BMI to CBS of blanket licences to copyrighted musical compositions at fees negotiated by them is price-fixing per se unlawful under the antitrust laws’.

I won’t delve into the details of this decision here. Suffice it to say that the US Supreme Court ruled that on the basis of ‘unique market conditions’ the arrangement in question did not fall foul of the per se prohibition of naked cartel conduct. Indeed, the court pointed out that so unique were the conditions of the particular market that, in the absence of the alleged price-fixing agreement in question, it is doubtful whether a market would have existed at all. Effective property rights in published music could not be created if it were left to the individual composer to market his or her work and to monitor how it was used. In other words, in BMI the US Supreme Court did not make an abstract, general legal ruling on the necessity for the prior characterisation of all horizontal agreements. Rather, on BMI’s facts the court decided that this specific matter should be remanded to the lower court for a decision based on the application of a rule of reason. I think the US court should rather have decided that this was not a price-fixing case at all, rather than a price fix justified by the peculiar circumstances of the market. But be that as it may, what is certain is that the case was decided on the basis of the facts pertaining to a particular market.

However, in its Ansac judgment, the South African Supreme Court of Appeal had almost no regard to the facts of the case in question. Had it done so, it would have discovered that the highly unusual facts characterising the market for copyrighted musical composition had not the remotest application in the Ansac matter, which dealt with a well-established, garden-variety commodity market. Rather, the Supreme Court of Appeal decided that because it ‘is not always simple to determine’ when ‘prohibited price-fixing [has] occurred’, it is necessary as a matter of statutory interpretation always to admit evidence that seeks to ‘characterise’ the conduct – in other words, seeks to establish whether the conduct alleged falls within the ‘ambit’ or ‘scope’ of section 4(1)(b) and, as such, is, if proven, prohibited per se.

What, if any, are the implications of the Supreme Court of Appeal decision? I can’t confidently make up my mind. On a good day, I think that the decision is of no practical consequence whatsoever. But, for the most part, I fear that the decision materially weakens the per se prohibition by introducing an efficiency defence through the back door because the Supreme Court of Appeal has decided that prior characterisation is required as a matter of statutory interpretation, as a matter of law, in order to successfully prosecute a section 4(1)(b) charge. In arriving at this conclusion, the only ‘fact’ that it has had regard to is the material context in which this matter arrived before it, namely Ansac’s assertion that it was not appropriately characterised as a cartel, but should rather be construed as an ‘efficiency-producing joint venture’.

This much was clearly known to the Supreme Court of Appeal and was acknowledged by it. However, what evidence can possibly be advanced in support of Ansac’s characterisation of itself as an ‘efficiency-producing joint venture’ that also fixes prices, other than evidence that purports to demonstrate the efficiencies generated by the cooperative arrangement? And how would this evidence differ from the evidence that would be submitted were an efficiency defence, such as that provided for in section 4(1)(a), to be permitted under 4(1)(b)? It appears that, in addition to demonstrating the claimed efficiency gains, for the cooperative arrangement to be characterised as anything other than a naked cartel it would also have to be shown that the price-fixing element of the arrangement is necessarily ancillary to achieving those efficiencies.

Hence, while the Supreme Court of Appeal emphasised that its intention was not to provide an efficiency defence in respect of those agreements characterised as naked cartels and thus subject to the per se prohibition of section 4(1)(b), it has decided that precisely the evidence that would be submitted in invoking an efficiency defence should be admissible in deciding whether to characterise a cooperative agreement as falling within the ambit of a per se prohibition. At the very least, the Supreme Court of Appeal’s judgment provides for the possibility of complex litigation involving industry and economic experts in the ‘characterisation’ phase. Even if it doesn’t necessarily enable naked cartels to win a battle over characterisation, it could add literally years and much expense to already protracted litigation in circumstances where many cartels may, like Ansac, wish to prolong litigation and, hence, the rents they derive from their unlawful cooperation.

PANDORA’S BOX

There is no end of clichéd metaphor – ‘falling dominoes’, ‘cans of worms’, ‘tsunami’, but maybe most appropriately ‘Pandora’s box’ – to describe the post-Tiger Brands period, one that is still in full swing as lawyers and well-known executives of many of South Africa’s leading corporations line up at the Commission’s door to ‘fess up to their many years of unlawful collusion, either for the purposes of filing a leniency application or for reaching a settlement with the Commission. In most instances these were initiated before and then continued into the period of the current competition legislation. The list is impressive and growing – Adcock Ingram, Sasol, Murray & Roberts, the Aveng Group, the Premier Group, Pioneer Foods, Foodcorp, South African Airways, Comair, the Reclamation Group. Several of these appear on the list on a number of occasions. And the list goes on and will undoubtedly lengthen.

The nature of the products involved in the bread and milling cartels – basic foodstuffs that disproportionately impacted on low-income consumers – attracted particular public opprobrium. However, many of the other South African cartels that have been apprehended have powerful implications for the poorest consumers. The medical equipment cartel centred on rigging bids for the supply of essential medical equipment to public hospitals. The cement products cartel involved price-fixing, market allocation and bid-rigging largely in respect of public sector tenders, frequently involving equipment for the supply of water. Soda ash – the Ansac product – is an essential ingredient of glass and detergents. Fertiliser is a key agricultural input. So from bread to soda ash, the poor are implicated and have undoubtedly been rendered all the poorer by the performance of some of our great corporations.

As striking as the association between cartelisation and the interests of the poor are the breadth and longevity of the cartels. The cement products cartel should – and probably will – become a textbook study of cartels, one that flies in the face of the argument that holds that cartels are inherently and necessarily unstable.

The cement products cartel was led by subsidiaries of two well-known South African multinational companies, Murray & Roberts and the Aveng Group. The Murray & Roberts subsidiary was first in the door with its application for leniency and so escaped prosecution. Aveng entered into a consent agreement with the Commission, which was confirmed by the Tribunal. It paid a fine of R46 million.4 Its affidavit revealed that the cartel involved price-fixing, market allocation and bid-rigging. It cartelised markets not only in South Africa but in a number of other sub-Saharan African countries. But the most extraordinary fact to emerge from the admissions contained in the papers filed for the Tribunal hearings revealed that the cartel had been in operation for 34 years! The recently appointed CEO of Aveng pointed out that he was not yet a teenager when the cartel had started its activities.

Naturally, a cartel of this breadth and longevity – and particularly one that has the rigging of public tenders as a central feature – demands elaborate organisation. Hence, an employee of one of the cartel leaders was appointed – and referred to as – the ‘banker’, who monitored compliance and who was responsible for compensating those who received a lower-than-agreed share of the market’s output.

But, above all, what the longevity of the cartel speaks of is the deep culture and widespread acceptance of collusion in the companies concerned. Over a period of 34 years, we are clearly talking about several generations of managers who participated in an unbroken period of cartelisation. One can only conclude that part of the succession planning process in the company involved the induction of successive generations of management into the workings of the cartel.

The Pioneer litigation offered some particularly interesting insights into the workings of a cartel. The bread cartel – actually two cartels, one covering the Western Cape, the other covering the rest of the country – is described in some detail in Pioneer.5 The facts don’t lend themselves to brief summarisation. However, the case clearly demonstrates the powerful imperative driving cartel formation in a market characterised by the production of relatively homogeneous products and subject to similar, although by no means identical, cost pressures. It also reveals how hard old habits die: here was a regulated industry used to arranging its affairs largely at meetings of the trade association. And after deregulation this simply continued. It probably led ultimately to the destruction of the trade association. Jannie de Villiers, the executive director of the chambers of milling and baking, whose name is mentioned frequently in the evidence in Pioneer, famously described the impact on his members of the Commission’s cartel-busting activity: ‘They hardly even talk to each other about rugby or the weather’.

In his address to the 2010 joint annual general meeting of the chambers of milling and baking, the president of the associations described the mood of the chambers’ members in the following melodramatic, but nevertheless revealing, terms: ‘Today, I feel like a citizen of Port-au-Prince in Haiti. The baking industry has concluded a year in which it had a major earthquake which caused a lot of damage and which almost destroyed the industry. Some of us are badly injured, some are wandering around in shock, some are even missing in action and some are seeking opportunities in the aftermath to look for a brighter future’.

The elimination of all forms of dialogue and cooperation in an industry is not an outcome that the competition authorities would want to achieve, but, given the history of these particular associations, there is probably no alternative to a thorough cleansing of the old.

The Pioneer judgment also highlights the incentive to cheat and, so, the persistent requirement to manage the cartel, a requirement that often compelled the pricing decision-makers to take into their confidence lower ranks of employees responsible for implementation of the cartel. In this instance, that proved particularly necessary because the critical price was not the published list price but rather the confidential discounts that were extended, often by those at the coal face, to the big retailers. This was the form that competition customarily assumed so that when the senior executives were intent upon securing a desired price increase it was necessary both to agree on the list price and then for those further down the ladder to be specifically instructed to hold the line on discounting.

Indeed, what this cartel establishes is that, far from the usual tale of innocent senior management let down by overly zealous and dishonest subordinates, the decision to pursue the cartel was taken and supported by senior management and often undermined by their subordinates who were trying to increase their sales volumes. Indeed, it is tempting to infer from the relatively light and reluctantly applied disciplinary measures imposed in a company that was ultimately subject to a swingeing fine and significant reputational damage, that the very top echelons feared that stronger discipline might have tempted those forced to take the bullet to reveal just how far up the hierarchy culpability extended. What is clear is that the senior manager whose participation in this cartel was proven always attempted to place himself at a minimum of one remove from where the offence was actually committed, the better to deny his personal involvement and that of the senior representatives of the company in the cartel.

Pioneer’s defence was weak and mendacious. As in all conspiracies, the cover-up ultimately revealed more than it hid. We were persistently cautioned by Pioneer’s counsel of the requirement to make a finding on the evidence against Pioneer and not to assume that it was guilty simply on the say-so of the leniency applicant and of the other firms that had entered into consent agreements and that were also cooperating with the prosecution. While this is obviously true, the trial graphically demonstrated the difficulty involved in mounting a successful defence against evidence of this quality. In rejecting an important element of the defence, the Tribunal clarified – following international jurisprudence – that it was not necessary to prove that the firm accused had been involved in every element of the conspiracy or in each and every interaction in order to establish a contravention.

Arguably, the most important outcome of the Pioneer case concerned the subsequent controversy over the remedy. In respect of the Western Cape cartel, the Tribunal imposed a penalty of 9.5 per cent of Pioneer’s affected turnover (that is, its turnover in the bread market) in that region and 10 per cent of the affected turnover in the national market less the turnover attributable to the Western Cape – a total of R195 million – testimony to the Tribunal’s view that there were important aggravating and few mitigating factors.

However, towards the end of the hearings the Commission amended the relief it sought from a fine based on the turnover in the affected product and geographic markets to one based on 10 per cent of national turnover in respect of each of the national and Western Cape cartels. Although the Act clearly specifies that a single year’s turnover be used as the basis for calculating an administrative penalty that may not exceed 10 per cent of that turnover, it does not specify whether the turnover figure to be used is to be the turnover actually affected by the cartel (hence the ‘affected turnover’) or the entire turnover of the group. There is clearly no statutory bar on total turnover being used as the basis.

The Commission’s amendment of the relief it sought represented a significant shift in its policy – previously it had always used affected turnover, a basis that had been accepted by the Tribunal. The shift is of enormous material significance, particularly in the many instances where a firm’s total turnover is the outcome of its participation in multiple markets, only one of which may have been cartelised. In Pioneer there were undoubtedly credible arguments for reviewing the established approach of using affected turnover. First, unlike Federal Mogul and South African Airways, which both involved vertical arrangements, we were dealing here with a hardcore horizontal agreement. Second, we were not dealing here with braking parts and air travel but with that most basic of commodities, bread. Third, Pioneer had conducted itself appallingly in the course of the investigation and subsequent trial.

After much discussion, the Pioneer panel decided to retain affected turnover as the basis for the penalty. The argument that prevailed was based on the centrality of markets in antitrust decision-making. Antitrust logic seemed to dictate that, notwithstanding the discretion permitted by the Act, the turnover in the affected markets be used.

But if logic was to be the basis for, in the Tribunal words, a ‘rational and justifiable’ exercise of its discretion, then where was the legislature’s rationality in limiting the fine to a single year’s turnover? This was thrown into particularly sharp relief by the cement products’ cartel, which had continued without discernible interruption for over 30 years. This limitation not only defies logic, but, more seriously, it is massively under-deterrent. This may well suggest that the legislature had deliberately allowed discretion in the selection of the turnover basis, having limited it in respect of the number of years of turnover. Certainly, the legislature’s recent decision to criminalise cartel conduct – about which I’ll say more later – clearly indicates that it no longer believes that the Act provides sufficient deterrence.

The Commission – driven principally by concerns surrounding deterrence – appealed the Tribunal’s decision on the penalty, arguing for a penalty based on total rather than affected turnover. However, before the appeal could be heard, Pioneer and the Commission settled the matter. The settlement initially involved an increase in the fine to R250 million plus other penalties. The nexus between the fine and affected turnover has thus been broken. Indeed, the entire percentage- based formula has been broken. There is, of course, no requirement that the fine be expressed as a percentage of turnover, merely that it should not exceed 10 per cent of some measure of turnover. The proper approach – and one that the Pioneer settlement will hopefully presage – is that total turnover be used as the upper limit for determining that the fine does not exceed the statutory limit, and that within this limit the Commission and the parties – or the Tribunal – will simply determine a fine appropriate to the offence. In so doing there is no reason to abandon the formula developed in South African Airways for weighting aggravating and mitigating factors.

The remedy imposed in Pioneer is also significant because it attempted to respond to growing public dissatisfaction that the penalty, which is paid over to the Treasury,6 has no direct positive impact on the consumers disadvantaged by the cartel conduct. This issue had been batted around the corridors of the competition authorities for some time. One view was that while the Tribunal was permitted to impose a financial penalty only in the form of a fine, it could nevertheless approve a consent agreement between the Commission and the guilty party calculated in different coin. Hence, at the time of the Tiger consent order, there was some talk in the corridors of the competition authorities of compelling Tiger to establish a school feeding scheme.

The Commission acted on this in devising the penalty imposed on Pioneer. The initial consent order imposed a fine of R250 million. A further R250 million penalty was imposed in the form of a requirement to endow a fund to be managed by the Industrial Development Corporation, which was to be dedicated to assisting new entrants and small firms active in the baking and milling markets. Pioneer was required to invest a further R150 million in additional baking capacity. Finally, Pioneer was required to reduce its margins on bread by a further R160 million over a specified period, the adherence to this requirement to be subject to an independent audit.

This agreement thus not only addressed the demand that consumers benefit directly from the penalty. It also attempted to increase the competitive temperature in the bread market by increasing capacity, by promoting new entry and by directly promoting price competition through the obligation imposed on Pioneer to lower its margins on bread. It’s revealing that in its subsequent annual report Pioneer expressed the view that the margin-reducing component of the order would not have a negative impact on the company’s fortunes because the reduction in margins would be compensated for by increases in volumes and market share. It never, it seems, contemplated the possibility that its competitors and erstwhile co-conspirators might be compelled to lower their margins precisely in order to retain their own market share. Indeed, kick-starting competition, rather than imposing a further R160 million worth of penalty, is the objective of the margin reduction. Clearly, Pioneer understands little about the competitive process (of which, it seems, it has little experience).

The order was conceived by the Commission and negotiated by the Commission. However, responsibility for competition policy – and thus line responsibility for the authorities – had recently passed from the Department of Trade and Industry to the Department of Economic Development, a newly-created department and ministry. In the nature of things political, the freshly-minted Minister of Economic Development, Ebrahim Patel, attempted to take as much credit for this innovative remedy as he could. This all took place after the end of my term of office and so I followed the saga in the media. I presume that the simultaneous announcement of the remedy by the Commission and by Patel in Parliament was part of a scheme to ensure that the minister received some of the credit for the settlement.

While I strongly supported the imaginative remedy, I simultaneously experienced a distinct pang of disquiet at the high-profile role assumed by Patel. I was concerned not at his rather transparent attempt to claim centre stage – this, as I have said, is in the nature of political office, particularly one that does not have much real power – but rather by the impression that might be gained that the minister was able to influence the outcome of competition investigation and adjudication. And while I took some comfort from the knowledge that this remedy had followed a guilty finding by the Tribunal, and had in reality been negotiated by the Commission pursuant to an appeal from the Competition Tribunal to the Competition Appeal Court, knowing well this particular minister’s proclivity for micro-management and for unbounded intervention I feared that it was but a short step to attempts at ministerial intervention in ongoing competition investigations in an effort to attain some or other industrial policy objective.

My fears were soon vindicated. The Department of Finance took issue with that part of the remedy that required the endowment of a bakery and milling competitiveness fund at the Industrial Development Corporation, another institution recently placed within the ambit of responsibility of the Ministry of Economic Development. The Department of Finance took the view that this amounted to an appropriation of public funds, a function that fell squarely within its responsibilities and prerogatives. Because of some highly publicised differences – some possibly real, others undoubtedly imagined – in policy and approach between the powerful Treasury and the fledgling Economic Development Department and their respective ministers, the media presented this disagreement as a Patel versus Gordhan dispute (Pravin Gordhan being the Minister of Finance). Hours before the consent order was to be placed before the Tribunal, and with the Department of Finance having briefed counsel to oppose the consent order, the various parties – now it seems involving the Commission, the departments of Finance and Economic Development, and a hapless Pioneer – agreed that the fine component would increase from R250 million to R500 million, the latter figure incorporating the endowment of the Independent Development Corporation (IDC) fund. The Department of Finance would still use R250 million of this to endow an IDC fund dedicated to the development of the baking and milling sectors. However, its appropriation prerogatives were not thereby challenged.

I have no view on the legal position surrounding the appropriation of public funds, although the Department of Finance’s position makes sense from a legal and fiscal policy standpoint. However, from a competition policy perspective, I think it’s important that the Commission be able to demonstrate to consumers that the fines imposed will, where possible, be used to support projects and activities that redound to the benefit of those consumers whose interests were compromised by the conduct of the cartel in question. It’s of course infinitely more important that the competition authorities should be able to demonstrate to consumers the benefits of competition, but this is no simple task, nor does it preclude the possibility of using the funds in the manner suggested.

While from a fiscal policy perspective I understand that the Treasury would not want to encourage all manner of effective appropriation, in this instance its anxiety should be assuaged by the fact that this fine is proposed by the body responsible for the administration of the Competition Act – and thus well placed to design targeted remedies – and is pursuant to an admission of guilt by the firm in question or a finding of guilt by a statutory quasi-judicial body. But be this as it may, the Treasury has demonstrated its willingness to be flexible as long as it is able to maintain its monopoly over appropriations. This issue would normally have been easily resolved between the two responsible bodies – the Commission and the Treasury. However, it was politicised by the high-profile intervention of the Ministry of Economic Development. The upshot is the apparent politicisation of the decisions of the competition authority, the notion that competition outcomes will be determined by the outcome of ‘Patel versus Gordhan’. This is, to put it mildly, extremely unfortunate and was later compounded by the Walmart fiasco, which I’ve already recounted and to which I’ll return.

CRIMINALISATION

The issue surrounding appropriate remedies for cartel conduct had taken on equally controversial – and at least as potentially damaging – dimensions earlier on when the Department of Trade and Industry decided to amend the Competition Act to introduce liability, in the form of possible prison sentences, for those individuals found guilty of involvement in cartel conduct.

While there has been unusual unanimity in the community of antitrust scholars and enforcers surrounding the damaging consequences of cartel conduct, there is significant divergence among the remedies actually imposed in the various national competition law regimes.

The US has for long been the standard-bearer for the imposition of criminal sentences, jail time, on those found to be directly implicated in cartel conduct. There are US executives as well as nationals of other countries involved in organising international cartels that have affected US markets who are incarcerated in that country.

The US arguments in favour of criminalisation are, up to a point, unimpeachable. US enforcers argue that the fines and civil damages imposed on companies – and bear in mind that a guilty finding in a US court is inevitably followed by a civil claim for damages that may be and is regularly trebled – do not compensate for the economic damage wrought by cartels and the rents derived by the firms that participate in them. Hence they argue that the imposition of fines and damages on firms is an insufficient deterrent, that only the threat of imprisonment will act as a sufficient deterrent.

Indeed US antitrust enforcers have actively promoted the criminalisation of cartel conduct in international forums and their arguments have met with little resistance, and certainly not from the enforcement community. It’s one of those rare instances of near unanimity in the antitrust community of enforcers, adjudicators and scholars: the economic consequences of agreements between competitors to fix prices, allocate markets and rig bids are economically harmful, and prison time is an appropriate remedial response. However, this unanimity notwithstanding, other jurisdictions have been slow to criminalise cartel conduct. Although in recent years an increasing number of national jurisdictions have introduced criminal sanctions for cartel conduct, the experience has not always been a happy one.

The debate surrounding remedies for cartel conduct rumbled on in South Africa but took on a new intensity following the Tiger cartels – both those in bread and in pharmaceuticals. In President Mbeki’s 2007 state of the nation address, he had made reference to strengthening the Competition Act. My more circumspect and thoughtful colleagues will justifiably remind me – as did the officials of the Department of Trade and Industry – that I had raised the spectre of criminalisation in speeches and during the hearings of one or other of the Tiger Brands consent orders. But, on closer consideration, while the competition authorities were strongly in favour of amending the Act in order to introduce stronger sanctions for a variety of offences, specifically including cartel conduct, we were extremely concerned at the unintended consequences of the introduction of criminal sanctions. However, the Department of Trade and Industry decided to introduce an amendment bill which, among other sweeping changes, introduced criminal sanctions. And in what was a marked departure from previous amendments to the Act, the responsible Department of Trade and Industry officials made every effort to sideline the competition authorities in discussions surrounding this critical amendment. This led to a stinging public row between the department and the competition authorities.

I’m not entirely sure why the department officials chose to pursue an amendment to introduce criminalisation in the face of the steadfast opposition of the competition authorities. It seems that a toxic mix of populism and personal ambition (for example, it seemed that the stature of a Department of Trade and Industry division was enhanced by the quantity of legislation it initiated) ensured that the department’s support for criminalisation was entrenched well before the competition authorities were given the opportunity to present their views on this critical issue. And the absence of leadership from a chronically indecisive minister, Mandisi Mpahlwa, ensured that the differences could not be resolved.

The upshot was the tabling of a bill that provided for the criminalisation of cartel conduct, and the highly unusual spectacle of the agencies that were being endowed with these ostensibly enhanced powers publicly opposing legislation introduced by its policy-making department. It is, of course, on the face of it, odd that an enforcement agency should oppose an apparent increase in its powers. We wanted additional remedial powers. We were concerned that our remedial powers – particularly the limitations imposed on the size of the administrative penalty we could impose – were under-deterrent. Moreover, we were firmly persuaded that personal liability, including jail time, fitted the offence. However, we were convinced that the legal and administrative context in which the Competition Act operated meant that the bill would not pass constitutional muster and that, in the unlikely event that it was passed into law, it would significantly weaken the incentive for firms to use the corporate leniency programme and to enter into consent orders, the two most important mechanisms in the apprehension of cartels.

The Competition Tribunal could clearly not impose criminal sanctions on an individual. However, section 73A(4) of the bill initially provided that the criminal court must, for the purposes of its adjudication, treat the firm’s admission in a consent order or the finding of the Competition Tribunal or the Competition Appeal Court as ‘conclusive evidence’ that the firm had participated in a cartel. This effectively meant that a finding of an administrative tribunal made on a balance of probabilities would have constituted ‘conclusive evidence’ in a criminal proceeding where the applicable standard is ‘beyond a reasonable doubt’. However, in the last-minute bargaining during the parliamentary hearings, a modicum of sense prevailed and this provision was softened. The bill that was passed now provides that the findings of the Tribunal regarding the existence of a cartel be treated, for the purposes of any subsequent criminal proceedings, as prima facie, as opposed to ‘conclusive’, evidence of the existence of a cartel. In other words, the Act now provides that the onus for rebutting the Tribunal’s conclusions rests with the accused in the criminal proceedings.

While the constitutional shortcomings of the initial formulation that sought to have the Tribunal’s findings on the existence treated as ‘conclusive’ for the purpose of securing a criminal conviction are unimpeachable, it is not at all clear that the revised provision will pass constitutional muster either. While the other reverse onuses in the Act – for example, to prove countervailing pro-competitive gains – may well prove acceptable, a reverse onus imposed on a defendant who faces the prospect of jail time may well prove a bridge too far. In any event, there can be little doubt that the Constitutional Court will be called upon to decide this.

However, even if this revised provision does pass constitutional muster, it does not mitigate the concerns regarding the likely impact of the introduction of criminal sanctions on the use of consent orders and the corporate leniency programme. It seems fairly obvious that directors and executives will be less willing to conclude consent orders on behalf of their firms, if an admission by a firm that it has participated in a cartel (which is an essential feature of a consent order) may be used as the basis – even a ‘mere’ prima facie basis – for securing a later criminal conviction against those very same directors, executives or senior managers who have made the admissions on behalf of their firms. We pointed out the difficulties of proving the existence of a clandestine conspiracy and the consequent importance in incentivising a firm to conclude a consent order. For this reason, the vast majority of cartel convictions had taken the form of admissions contained in consent orders, thus sparing the Commission the considerable cost, in both time and money, and the uncertainty of running a trial.

As to the corporate leniency programme, again the clandestine nature of cartels and the associated difficulties of detection made this plea-bargaining mechanism a critical cartel-busting weapon. Immunity is not lightly given and is subject not only to stringent requirements of cooperation but is, for the most part, available only to those first through the door and whose evidence clearly enhances the prospect of a successful prosecution of the other members of the cartel. As already pointed out, the Commission has relied on the corporate leniency programme for many of its growing list of successful cartel busts. However, since the Commission does not have the discretion to grant immunity in criminal proceedings – only the National Prosecuting Agency enjoys that discretion – the introduction of criminal sanctions will create uncertainty for would-be informants as to the extent of personal immunity that they could enjoy in exchange for information and is likely to discourage such informants from coming forward. This would undermine the Commission’s ability to investigate and prosecute cartels.

Essentially, then, the introduction of criminal liability introduces two separate proceedings subject to the discretion of two wholly independent prosecutors, the Commission and the National Prosecuting Authority. Unless the granting of immunity to the firm (by the Commission) and to the individual (by the National Prosecuting Authority) is carefully coordinated and agreed, a firm intending to apply for leniency will have no certainty that the two prosecutors will follow the same approach. A board of directors or an executive committee of a firm is extremely unlikely to apply for corporate immunity if the individual members of the board and the executive thereby expose themselves to criminal charges. It also introduces numerous technical problems about whether evidence obtained by one authority can be used by the other. Will witnesses in Commission proceedings who have not yet received criminal indemnity be permitted to refuse to testify on the grounds of possible self-incrimination? At present such a ground for refusal would not be competent because cartels are only administrative contraventions perpetrated by a firm and not an individual.

We considered possible solutions to the problems we had identified. One solution to the constitutional problem posed by this intermingling of criminal and civil proceedings is to run two parallel and separate trials, one before the competition authorities in which the firm would be tried, the other before the criminal courts in which the persons participating in the cartel would be tried. The obvious problem with this approach (apart from the serious evidentiary problems raised above) is that it carries with it the distinct possibility of two adjudicative bodies arriving at two diametrically opposed conclusions regarding the very existence of a cartel. Moreover, the separation between the civil trial and the criminal trial would not cure the adverse impact of this provision on the corporate leniency programme.

The other possible solution of the constitutional problem would be to remove responsibility for cartel prosecution and adjudication from the competition authorities and hand it over to the criminal justice system. This is essentially how the system worked under the competition statute that preceded the coming into force of the Competition Act. As already noted, under the old statute not a single cartel was successfully prosecuted and there is good reason to believe that this would be the outcome if responsibility for cartel prosecution and adjudication was taken out of the hands of the competition authorities and again placed in the hands of the criminal justice system.

This solution would also run counter to the policy position of the Department of Trade and Industry itself – one to which it professes continued adherence – that jurisdiction over competition matters should remain the exclusive preserve of the competition authorities. This is the reason why the initial bill retained the provision that a finding by the Competition Tribunal or Competition Appeal Court must be adopted by the criminal courts as ‘conclusive evidence’ of the existence of a cartel.

We then also proposed that the administrative sanctions be strengthened. In particular we proposed that the restriction of the basis of the administrative penalty to a single year’s turnover be amended to permit an administrative penalty based upon the number of years for which the anticompetitive conduct has been practised. We also suggested that certain anticompetitive practices – notably section 8(c) abuse-of-dominance offences – that are subject only to a possible administrative penalty in respect of a repeat offence be made subject to a possible administrative penalty on a first offence.

We also proposed an increase in the penalties that may be imposed for the existing criminal provisions of the Act. These offences – all provided for in chapter 7 of the Act – cover the disclosure of confidential information, ‘hindering the administration of the Act’, failure to obey a summons, lying under oath, failure to comply with an order of the Competition Tribunal or Competition Appeal Court, and a range of other offences, for example attempting to improperly influence the Tribunal or the Commission and defaming the Tribunal or the Competition Appeal Court.

We pointed out in the case of a contravention of section 73(1) (that is, failure to comply with an order of the Tribunal or the Competition Appeal Court) the penalty is a fine not exceeding R500 000 or a term of imprisonment not exceeding 10 years or both a fine and imprisonment. However, in the case of a contravention of any of the other provisions of chapter 7, the penalty is a fine not exceeding R2000 or imprisonment not exceeding 6 months or both a fine and imprisonment. We proposed an exponential increase in the penalties for these latter offences. We supported these proposals by reference to other jurisdictions in which it had been found that as the penalties for anticompetitive conduct increase, so too does the incentive for impeding an investigation. This is why the penalties for action designed to impede an investigation, such as lying under oath or refusing to obey a summons, need to be strengthened. We also proposed the introduction of a general offence for obstruction of justice that would include interfering with witnesses or destroying documentary and electronic evidence.

The amendment bill was duly passed by the National Assembly on 21 October 2008. It appears that during the brief interregnum between the Mbeki and Zuma presidencies, President Kgalema Motlanthe had been persuaded – and his office had sought and received a confirmatory legal opinion to this effect – that the reversal of the onus in the criminal proceedings was indeed vulnerable to constitutional challenge, thus confirming our view.

Accordingly, President Motlanthe referred the bill back to the National Assembly, which, in the heady atmosphere of post-Mbeki parliamentary assertiveness, sent it back to the President unaltered. On 28 August 2009, the bill was duly assented to and signed by President Zuma. However, the amendment Act has not been ‘proclaimed’, that is, the President has not formally announced the date on which the amended Act will come into force. So, more than 3 years after Parliament amended the Competition Act, the amendments have yet to come into effect.

In the meantime, the Commission’s cartel-busters may inadvertently and temporarily have the best of all possible worlds. It is now 4 years since the department published the bill containing the criminalisation provisions. It is 3 years since the bill passed through the parliamentary process and more than 2 years since the President formally assented to the amended Act. But it has still not come into effect because it has not been proclaimed. I have little doubt that the looming threat of criminal sanctions has caused some firms to examine their practices and is undoubtedly partly responsible for the proliferation of leniency applications that the Commission is receiving and, hence, the apprehension of many damaging cartels. Senior officials of the Commission believe that they are still at the tip of the iceberg, and that their most powerful weapon is the dawn raid, of which there are apparently an average of two each month. These officials believe that it is fear of what might have been discovered on these raids that is principally responsible for the proliferation of leniency applications and consent agreements. However, Commission officials still firmly believe – and I agree fully with them – that once the Act is proclaimed, the fear of criminal sanctions will lead to a reduction in leniency applications and consent agreements, and with them a falling-off in the apprehension of cartels.

While for the most part a well-directed and efficiently-executed dawn raid still triggers the leniency applications, which in turn lead to the admissions by the co-conspirators who have not managed to get first in line, there are definite signs that firms accused of cartel conduct are beginning to fight back, even in circumstances where their grubby little secrets have already been laid bare. I have outlined Pioneer’s disastrous attempt at defending itself and the mendacity and dissembling upon which it relied. I suspect that this will frequently be the outcome in those cases where a central member of the conspiracy has elected to cooperate with the prosecution and where other conspirators have admitted their role and named their fellow conspirators.

The introduction of criminal sanctions is not the only important new provision introduced by the as-yet inoperative bill. There was an attempt to add a new offence under the Act, namely ‘participation in a complex monopoly’, which has to be one of the loonier provisions introduced into competition law. This too has been softened and is now largely superfluous. The bill also introduced the concept of a ‘market inquiry’ that effectively enables the Commission to use its powers to investigate markets whose outcomes suggest underlying competition problems.

A striking feature of the fraught amendment process was the marked deterioration in the ability or even the will of organised business to represent its own vital interests robustly. The contrast between organised business’s aggressive and technically well-informed participation in the negotiations in 1998 surrounding the passage of the Competition Act and its supine, disorganised and technically inept approach to these negotiations was striking. Deciphering this is beyond the scope of this book. However, it does seem to point to two disturbing features of contemporary South Africa.

First, it points to the decline – with the clear exception of the unions – of organised civil society in South Africa. This decline is perhaps particularly surprising in respect of a well-resourced interest group like business, compelled to deal with a government that retains, for good and for bad, a residual suspicion of big business’s commitment to building an equitable and democratic society. The upshot seems to be an apologetic timidity on the part of business rather than the robust engagement that, I firmly believe, is a critical ingredient of an effective policy-making process, especially in the making of industrial and competition policy.

A second explanation – and, on the face of it, somewhat at odds with the first – is that, whereas in the early years of democratic government the social distance between white-owned business and the de-racialised, majority government compelled both to engage in public forums such as Nedlac, the distance has decreased as they have come to know each other better and as black economic empowerment has exercised its complex influence on government–business relations. And so we’re back to a form of engagement that is, once more, conducted behind closed doors and through informal channels, or else before the courts. This issue of business–government relations is complex territory, to put it mildly. However, my observation, for what it’s worth, is that the high point, the most mutually productive point ever reached, before or since, in the relationship between government and business, may have been those few years in which a wary democratic government and wary, independent organised business were compelled to deal relatively openly with one another because the informal channels of the old order had broken down and those of the new were yet to be constructed. I would not be alone in believing that among the most significant governance problems bedevilling South Africa is the relationship between politics and commerce, between state and business. As already noted in the earlier discussion of the Walmart matter and as I’ll elaborate in my concluding remarks, this smoke-filled-room type of relationship threatens to seriously undermine the independent decision-making that has characterised the workings of competition law.

COUNTER-ATTACK

It was inevitable that business, undoubtedly encouraged by its endlessly inventive lawyers, would not take the Commission’s sweeping success in apprehending cartels lying down. I was interested and surprised to learn when I participated in a recent panel discussion organised by one of Johannesburg’s leading law firms that there is serious contemplation being given, by the competition law bar at any rate, to lobbying for the introduction of a pro-competitive defence into section 4(1)(b). Given the current state of public opprobrium directed at cartels, this is highly unlikely. However, it was suggested that while the lack of a pro-competitive defence for hardcore cartel conduct may be constitutionally acceptable where criminal penalties are not available, the Constitutional Court may well feel differently when prison time becomes a possible outcome of participating in cartel conduct. So watch this space – the per se nature of section 4(1)(b) may well also become a victim of criminalisation.

Predictably, in a repetition of the early days of enforcement, and given the limited room for defending hardcore cartel arrangements, the focus of the attack has centred on the Commission’s investigatory powers and the manner in which they have been exercised. And the cartels have scored some important victories in the courts, in both the Competition Appeal Court and the Supreme Court of Appeal. I use ‘cartels’ advisedly because, as I’ll show, each of the key battles that I refer to – Woodlands, involving a milk cartel and Omnia, a fertiliser cartel – has been won by what are clearly cartels. That is to say, much as in Pretoria Portlands Cement, where the Supreme Court of Appeal took exception to the manner in which the Commission conducted its first dawn raid, in the milk and fertiliser cartels there were, and maybe still are, cartels in operation. However, as the courts have again demonstrated, considerations of fairness and procedural rectitude will inevitably trump concern over the effects of ‘monopoly’, which, as I have already cited, one eminent former judge of the Supreme Court of Appeal described as ‘a canker that eats into a free enterprise economy’.

The battleground has become the apparently technical, but substantively important, issue of the proper mode of initiation of complaints by the Commission. Because the decisions are so technical in nature, the various decisions in Woodlands and Yara are difficult to summarise, but let me try to identify the key points. I should say at once that the case law that has mushroomed around these issues is comprehensively summarised and brilliantly critiqued by the Tribunal’s decision, which Norman Manoim drafted in the South African Breweries matter.7 Here the Tribunal, obliged to follow the restrictive jurisprudence imposed by the Supreme Court of Appeal and the Competition Appeal Court, dismissed a referral against South African Breweries. But in so doing, Norman saw fit to outline precisely where the superior courts had gone wrong and his reasoning is utterly persuasive. The Commission has now chosen, appropriately I believe, to take all of these questions to the Constitutional Court. But I’ll give my penny’s worth.

In short, because of what it found to be a defective complaint initiation on the part of the Commission, the Supreme Court of Appeal in its Woodlands judgment set aside the Commission’s referral to the Tribunal and handed back the fruits of the investigation to the cartelists, ‘unfortunate’ (the Supreme Court of Appeal adds as it brings down the axe on the investigation of the milk cartel) ‘as the result may be in the circumstance’. And what is the ‘circumstance’ that renders the Supreme Court of Appeal’s judgment ‘unfortunate’? I can only interpret this rather enigmatic and uncharacteristically normative comment as judicial acknowledgement of a set of indisputable facts that strongly suggest the existence of a cartel in milk, just as the Competition Appeal Court’s Yara decision, which put a stop to the fertiliser referral, has permitted participants in a fertiliser cartel to get away unpunished. Indeed, in the milk case it seems reasonably clear that the conspiracy continues to this day. That is indeed ‘unfortunate’.

I am, of course, able to assert the existence of the cartels with such certainty because in each case there had been, in respect of certain of the alleged cartel infringements, admissions voluntarily made, either in the form of a leniency application or in the form of a consent order. In the fertiliser case, a consent order had resulted in Sasol agreeing to pay a significant fine for its participation in the fertiliser cartel and making admissions that implicated its co-conspirators, Yara and Omnia. In the milk case, Clover, a leading national dairy, had been a leniency applicant in respect of at least one 4(1)(b) infringement. However, in what has to be the height (or depth) of cynical legal stratagem, Clover contrived to admit its participation in a cartel, apply for leniency, and then join its co-conspirators in a carefully coordinated litigation-driven campaign to have the Commission’s initiation of the complaint against it and the others struck down.

Let me start with the milk case. Clover fired the first shot in the campaign. Essentially, it asked the Tribunal to set aside the Commission’s referral to the Tribunal on the grounds that the complaint had been initiated by a private complainant and not the Commission. Consequently, it argued, the year-long period in which the Commission was obliged to refer a privately initiated complaint had lapsed or, in legal terminology, prescribed. There’s not much point in spending time on this. It was thrown out by the Tribunal, which found that the complaint had been initiated by the Commission and not a private complainant and so prescription did not operate.8 We were upheld by the Competition Appeal Court.9

Enter Woodlands Dairy and Milkwood Dairy, which, despite the bucolic innocence suggested by their names, were not above a little strategic litigating themselves. Their legal team was present at the bar of the Tribunal when Clover was heard and when Clover’s counsel remarked, en passant (but did not press), that in addition to its prescription allegation, the referral might be invalid on other procedural grounds, even if the complaint had been initiated by the Commission. But they said nothing, keeping this in reserve for another day and a separate application, the better to prolong litigation and increase its costs.

Rather, in what the Competition Appeal Court graphically described as a ‘jurisprudential chain novel’, Woodlands and Clover proceeded to pen their particular chapter by filing a new application that also asked for the dismissal of the complaint referral. Their application argued that the investigation which the Commission conducted into their activities had been undertaken under the umbrella of a generalised complaint that had specified the milk industry, rather than particular named companies, as its subject, and that cited general contraventions of the Competition Act, rather than particular contraventions of specified clauses of the Act. This, they argued, was an invalid initiation and thus it and all that flowed from it fell to be struck down by the Competition Tribunal. The effect of this is to invalidate any investigation conducted by the Commission which relies on information or documentation procured following complaints that are not initiated with sufficient particularity as to both the identity of the complainant and the specific clause of the Act that is contravened. Thus, if in the course of an investigation based upon a properly initiated complaint the Commission learns of transgressions or transgressors that extend beyond the letter of the initiation, then it is required to initiate a new complaint, failing which it is not entitled to investigate the new allegations or refer them to the Tribunal.

For its part, the Commission disagreed that an initiation was required to incorporate the level of particularity contended for by the firm. Indeed, it went further and argued that, precisely in order to achieve the level of specificity that would enable the Commission to decide whether to initiate a complaint, it had to be able to utilise its investigatory power, specifically its power to summons.

The Tribunal judgment summed up the dispute in the following terms:

as a proposition of law the Commission disputes whether the Commissioner in fact needs to have an initiated complaint before him, prior to using his section 49A powers. The Commission argued that the Commissioner needs to be able to investigate before deciding whether to initiate a complaint. Without being able to investigate and thus, inter alia, utilise the section 49A summons procedure, how does the Commissioner know whether there exists a complaint to be initiated? Thus we had a debate that went to the root of the purpose of the Commission’s investigative powers. Are the investigative powers dependent, for their exercise, on a prior complaint initiation for their validity; or do they exist independently of initiation, as in many cases prior investigation may be needed to inform a decision to initiate.10

The applicants also argued that the summonses were invalid on the grounds that they were vague and over-broad. The Tribunal first examined this latter contention – much the more straightforward – and found for the applicants, namely that the summonses were indeed defective for vagueness and were over-broad. It’s pertinent that the Tribunal found the summonses to be vague and over-broad largely based on the failure to specify the section of the Act that the Commission alleged had been contravened or the part of the milk industry – that is, the broad market – which it was investigating. That is to say, the summons referred to unspecified contraventions in unspecified markets. However, the Tribunal specifically supported the Commission’s contention ‘that the identity of an accused need not be stipulated. It seems perfectly reasonable that in a cartel case the identity of potential respondents may not be an issue which the Commission wishes to disclose during its investigation phase’. (My emphasis.)11

And so the Tribunal struck down the summonses. We nevertheless did, following a recent Constitutional Court decision, order the fruits of the defective summonses to be ‘preserved’ so that it could be decided, in the context of a trial on the merits, whether the information extracted from these voided summonses was of such a nature as to be admissible in further proceedings or whether they had so grievously offended against the privacy and other rights of those summoned as to require the permanent burying of the information thus obtained.

In deciding whether the summonses were too broad and too vague, the Tribunal naturally went a considerable distance in deciding what level of specificity is required to render a summons fair or, in the term used by the Tribunal, ‘sufficiently bounded’. However, having struck down the summonses, there was no need for us to decide the wider, more far-reaching point regarding the relationship between investigation and initiation: whether the Commission could use its investigatory power – in other words, issue summonses following the precepts laid down by the Tribunal – before formally initiating a complaint or whether it could use its investigatory powers in order to ascertain whether there were grounds for initiating the complaint.

In the meantime, the Commission had initiated several other complaints against Woodlands and Milkwood that met the level of particularity contended for by the dairies. They had referred these to the Tribunal. This, they insisted, had been done as a precautionary measure, and not because they conceded that their initial complaint initiation had been defective.

However, the Tribunal’s striking down of the summonses was not enough for the dairies. They wanted the complaint initiation and not merely the summonses declared invalid on the ground that the investigations had been improperly conducted because they were not preceded by a valid complaint initiation. If they succeeded in convincing a court of this, it would mean that, even if the subsequent complaint referrals met the specifications and level of particularity for which the dairies contended, they would remain invalid as long as the ‘correctly’ formulated initiation relied upon the fruits of investigations undertaken prior to those ‘corrected’ initiations. The invalidated summonses, on the other hand, provided only a temporary respite. Aside from the preservation order, which portended the prospect of the information obtained from the defective summons making a reappearance during the apparently pending trial, the Commission could cure the defective summonses by simply reissuing them in a manner that conformed to the Tribunal’s requirements.

And so off to the Competition Appeal Court. The dairies appealed against the preservation order because it portended the possibility that the information derived from the defective summonses would be used in further proceedings. The Commission appealed against our decision invalidating their summonses.

Although the Tribunal had not decided the matter based on the validity of the complaint initiation but rather on the validity of the summonses, because a summons is a key investigative instrument and because the Act relates – though not very clearly – the process of investigation and initiation, this nexus, though not itself the subject of the appeal (because it had never been decided), remained a live issue before the Competition Appeal Court. Indeed, on the Competition Appeal Court’s reading of the Commission’s case, it appears that the Commission accepted that initiation preceded investigation, but it strenuously opposed the contention that the investigation had to be bounded by a particular, specified contravention and to parties named in the initiation. The Commission effectively contended that it was entitled to follow the course of an investigation – if it led to hitherto unidentified contraventions or suspected perpetrators, then that’s where the investigation had to go. Furthermore, the Commission contended that it was entitled to summons any party who it had reason to believe had information pertaining to the investigation.

The Competition Appeal Court did not spend much time on this issue – it was, after all, deciding an appeal on the Tribunal’s decision regarding the validity of the summonses and the preservation order. It did, though, express the view that there was, in the conduct of its investigation, no requirement to specify the provision of the Act under investigation or to frame the investigation against a specific entity.

However, in a subsequent application to the Competition Appeal Court that sought clarification of its order and special leave to appeal its decision to the Supreme Court of Appeal, the Court expressed itself in forthright terms on this question:

The Act provides that the commissioner may initiate a complaint against an alleged prohibited practice; no more, no less. A prohibited practice means a practice prohibited in terms of chapter 2 of the Act, which contains the central provisions of the Act regarding uncompetitive behaviour and structures. For example, section 4(1) refers to an agreement between or a concerted practice by firms, that is an agreement or concerted practice which involve acts such as the fixing of a price. That is what cartels do. Cartel behaviour is not found in the behaviour of one firm. It is to be found in the behaviour of a number of firms. There is nothing in the Act to suggest that an investigation against an industry, that is an industry comprising of a number of firms, is not a matter which was not envisaged by the Act nor does it appear that the wording, that I have cited, cannot bear the weight of this particular reading. To the contrary. Consider if appellants’ submissions were correct. It would be very difficult to initiate proceedings against cartels. Respondents would have to specify each firm. It could never investigate, for example, in a hypothetical case, the banking industry. It would have to specify the particular banks and, if it omitted one or two of them out [sic], it would encounter difficulty in enforcement.12

The Competition Appeal Court then continued in similarly forthright manner to lay out the objectives underpinning the appellants’ insistence that the question of the allegedly defective complaint initiation, rather than ‘merely’ the flawed summonses, be adjudicated upon. It was, the Court correctly deduced, to prevent the Commission from curing the defective summons, to prevent, for ever and a day, a hearing on the merits of the alleged contravention of section 4(1)(b): ‘Were the Court to set aside these complaints it would be impossible in the future for the competition authority to say: “we have made a mistake, we now want to start all over again.”’13

That was, of course, precisely the point. And so to the Supreme Court of Appeal from which emanated a judgment of which, I fear, we will not hear the end until the Constitutional Court has pronounced or the Act has been amended in a constitutionally acceptable manner. The Supreme Court of Appeal acknowledges at the outset the lack of clarity in the Act: ‘as to the sequence of steps that have to be followed in relation to the initiation of a complaint, the investigation, the use of power to summon witnesses to testify, and the referral of complaints to the Tribunal’.14

In the face of this lack of clarity, the court elected, as we’ll see, to clarify matters in a manner that was most restrictive of the Commission’s powers. The apparent reasons advanced for adopting this restrictive view are necessary adherence to fundamental constitutional values – the rule of law, democratic values of dignity and freedom, the rights to privacy, a fair trial and just administrative action. The Court adds that the procedural powers of the Commission ‘must be interpreted in a manner that least impinges on these values and rights’ because of the punitive powers of the competition authorities, with the administrative penalties, in the court’s opinion, ‘bear[ing] a close resemblance to criminal penalties’.

This last throwaway line is, of course, extremely significant. First, because there are a large number of administrative bodies that are entitled to impose administrative penalties, or, as the Court would have it, ‘punitive measures’. Does this mean that all of these ‘resemble’ criminal penalties? Or is it the size of the administrative penalties the competition authorities are entitled to impose that causes them to resemble criminal penalties? If the latter, it’s a fairly clear indication that the Court does not appreciate the scale of damage wrought and profit earned through Competition Act contraventions. Second, the observation is important because, when the amendment bill is finally proclaimed, actual criminal penalties will be introduced, and this is yet another signal of how the game is going to change when these become effective.

In the context of this particular matter, these constitutional principles – and the requirement to apply them with particular diligence when evaluating the Competition Commission’s powers – were held to mean that

the initiation must at least have a jurisdictional ground by being based on a reasonable suspicion. The initiation and subsequent investigation must relate to the information available or the complaint filed by a complainant.

There is in any event no reason to assume that an initiation requires less particularity or clarity than a summons. There are reasons for this. The first is that any interrogation or discovery summons depends on the terms of the initiation statement. The scope of a summons may not be wider than the initiation. Furthermore, that Act presupposes that the complaint (subject to possible amendment and fleshing out) as initiated will be referred to the Tribunal. It could hardly be argued that the Commission could have referred an investigation into anticompetitive behaviour in the milk industry at all levels to the Tribunal.

Members of the supposed cartel were in fact mentioned in the initiating statement. It was therefore not a case where no cartel member had been identified. The problem is that there were not facts that could have given rise to any suspicion that others were involved. A suspicion against some cannot be used as a springboard to investigate all and sundry. This does not mean that the Commission may not, during the course of a properly initiated investigation, obtain information about others or about other transgressions. If it does, it is fully entitled to use the information so obtained for amending the complaint or the initiation of another complaint and fuller investigation.15

In a nutshell, the Supreme Court of Appeal has clarified the hitherto uncertain relationship between, and the sequencing of, complaint initiation, investigation and referral thus. First, a reasonable suspicion of contravention of the Act. This requires the specification of the section of the Act against whose suspected contravention the Commission has initiated a complaint and the precise identification of those who are the subject of the complaint. This is then followed by an investigation which can be no wider than the complaint initiated. And then finally there is the referral to the Tribunal, which, on this argument, must faithfully follow the initiation in all material respects. If, in the course of the investigation, the Commission obtains information regarding further suspected transgressions and transgressors, then it may amend its complaint or initiate a new complaint. The Supreme Court of Appeal has erroneously conflated initiation and referral.

And so why is this particularly restrictive of competition law investigations? The court makes the point – in opposition to its characterisation of a view expressed by the Competition Appeal Court – that merely because it is difficult to investigate Competition Act contraventions, those accused should enjoy no less constitutional protection than any other accused persons. The point, though, is not about difficulty but rather about the nature of competition law offences, and particularly cartel offences. The Competition Appeal Court has already addressed this in part. When the Commission, on the basis of a ‘reasonable suspicion’, investigates two players in an industry for cartel conduct, it is entitled, or should be entitled, by virtue of the nature of the offence to suspect that others in the market are also involved in the cartel or at least have knowledge of it. The reason, given the most basic knowledge of competition economics, is quite straightforward. And so, on the basis of a reasonable suspicion that two firms in a market are involved in a cartel, it is perfectly rational to extend the investigation to all participants of any significance in the market. I wonder, though, what the courts’ reaction would be if, on the basis of a reasonable suspicion that firms A and B had agreed to fix prices, the Commission simply initiated a complaint against every participant in the same market.

Things are not quite so straightforward in respect of vertical restrictive practices and abuse of dominance. But again, with a basic knowledge of competition economics, nor are they particularly complicated. An investigation conducted following a reasonable suspicion of price discrimination (a section 9 offence) may, after investigation, conclude that the discrimination is conducted as part of a margin squeeze scheme. A simple case of resale price maintenance may turn out to be the platform for a cartel of producers or retailers. The Supreme Court of Appeal in Woodlands would have that a new complaint be initiated or, at least, an amendment be sought, as each offence is exposed and as each new alleged offender is revealed. But why? In its pleadings and witness statements the Commission will present the result of its investigations and specify who is charged and with what they are charged. A party against whom a complaint is initiated is under no compulsion to respond; it is if and when the referral to the Tribunal is filed that legal proceedings commence, and the parties identified in the referral will then have ample opportunity to respond to the charges levelled against them.

And so, the requirement that particular diligence be applied to the rights of an accused in a competition law investigation and referral is, given the nature of competition offences, particularly inapposite. The more so when both the Supreme Court of Appeal and the Competition Appeal Court have explicitly recognised and spoken out against the proclivity for strategic litigating and the consequent delay and obstruction that it occasions. This judgment is a charter for that sort of conduct. As far as the Commission is concerned, the courts’ conflation of initiation and referral perversely incentivises it to take the precautionary step of listing every firm in the market when initiating a cartel investigation and most of the conduct that could amount to abuse when initiating a section 8 or 9 complaint. And so the courts, by acting in part to limit ‘fishing trips’, have effectively incentivised ever-wider investigative forays by the Commission.

So what appears to be an attack on technical grounds is underpinned by a woeful lack of appreciation on the part of the courts of the substantive content of competition law and economics. Nowhere is this more clearly expressed than in the Competition Appeal Court’s Netstar judgment. Recall that section 4(1)(b) provides that prices that are fixed, markets that are allocated or bids that are rigged as a result of an ‘agreement’ or a ‘concerted practice’ constitute restrictive horizontal practices. Clearly all that the addition of ‘concerted practice’ is intended to convey is the unusual breadth of the different forms of coordinated action that will make competitors vulnerable to section 4(1)(b). I would not have thought it necessary to distinguish between ‘agreement’ and ‘concerted practice’ at any stage of a 4(1)(b) claim. However, the Court ruled that the initiating document had to specify whether or not the Commission intended embarking on an investigation of an ‘agreement’ or a ‘concerted practice’, thus elevating the specificity required at initiation to truly ridiculous heights. This is by no means the only basic error committed in Netstar, but let’s leave it there.

At the time of writing, the Constitutional Court is drafting its judgment of these issues. It was the first occasion that I had attended a hearing of our highest court and it was uplifting to hear a court that appeared at least as concerned with the rights of consumers to enjoy competitively-priced goods as it was with ever-more-restrictive notions of procedural rectitude. What the interrogation by the Constitutional Court judges conveyed was their refusal simply to accept the notion that the Competition Commission represented an all-powerful arm of the state moving against a defenceless citizen. They conveyed an understanding that the Commission was up against powerful, well-resourced corporations who were fully capable of undermining otherwise truly defenceless consumers, thereby restoring some balance to the scales of justice.

But if I’ve learnt anything, it’s to resist predicting the outcome of a court of law. So let’s not draft the postscript too soon.

NOTES

1. An ‘agreement’ is defined to include ‘a contract, arrangement of understanding, whether or not legally enforceable’ and a ‘concerted practice’ is defined as ‘cooperative, or coordinated conduct between firms, achieved through direct or indirect contact, that replaced their independent action, but which does not amount to an agreement’. Much has been made – far too much – of the distinction between an ‘agreement’ and a ‘concerted practice’ in the Competition Appeal Court’s decision in Netstar, which is discussed later.

2. It seems that the Webb–Pomerene Act permits firms to ‘associate’ or cooperate in markets outside the US in ways that would not be permitted in the US itself, but this does not necessarily imply that this cooperation would amount to hardcore cartel conduct – it may be cooperation in the area of transport logistics; hence the proffered defence that Ansac be viewed as an ‘efficiency-enhancing joint venture’. That having been said, Webb–Pomerene associations are generally referred to as ‘export cartels’ and it beats me why, if the activities of the association are there simply to promote efficiencies rather than to fix prices or allocate markets, they should be so specifically prohibited to operate in the US markets of the firms concerned.

3. 441 U.S. 1 (1979).

4. Competition Commission v Aveng (Africa) Limited (24/CR/Feb09).

5. The Competition Commission v Pioneer Foods (15/CR/Feb07 and 50/CR/May08).

6. The Department of Finance is also known as National Treasury or the Treasury.

7. South African Breweries Ltd and Appointed Distributors v The Competition Commission (134/CR/Dec107).

8. Clover Industries Limited, Clover SA (Pty) Ltd, Ladismith Cheese (Pty) Ltd v The Competition Commission (103/CR/Dec06).

9. Clover Industries Limited, Clover SA (Pty) Limited v The Competition Commission and 6 Others (78/CAC/Jul08).

10. Woodlands Dairy, Milkwood Dairy v The Competition Commission (103/CR/Dec06).

11. Ibid.

12. Woodlands Dairy (Pty) Ltd, Milkwood Dairy (Pty) Ltd v The Competition Commission (88/CAC/Mar09).

13. Ibid.

14. Woodlands Dairy v Milkwood Dairy (105/2010) [2010] Supreme Court of Appeal 104.

15. Ibid.

6. Competition enforcement on the world stage

For decades, much of what could be characterised as robust antitrust enforcement or the existence of a vibrant competition culture emanated from the United States and, to some extent, Germany. And so our history of international contacts begins in the US, probably still the god-head of competition law, despite the rise of the European Union as a leading enforcer and intellectual player in this field. The first workshop set up by the African National Congress in December 1992 was convened at a rather tacky resort outside Johannesburg by Tito Mboweni, then the deputy head of the ANC’s economic policy department and, later, Minister of Labour and then Governor of the Reserve Bank. The advisers to both the ANC delegation and the business delegation were two US academics – professors Geoff Shepherd and Thomas Hazlett – representing distinctly divergent traditions in US antitrust. My clearest recollection of that meeting – apart from its rather tetchy character – was that it was opened by Nelson Mandela, an auspicious start to the process of reforming competition law! I recall that when the meeting began and all of the 30 or so participants were asked to introduce themselves and their institutional affiliations, Mandela introduced himself as ‘an ANC member of the Orlando West branch’, a line that I don’t doubt he has used on more than one occasion with consistently charming effect. When I asked Professor Hazlett for his clearest recollection of the meeting, he responded:

In particular, I recall being introduced to Nelson Mandela at the very outset (he quickly left). He was so exceedingly gracious in saying, with apparent conviction, ‘Prof. Hazlett, it is a very great honour to meet you,’ that I had to catch myself. The instinctive reaction was to respond with laughter, as he seemed to be going way overboard, to the point of humour. Of course, he was, in fact, just exuding a graciousness not generally found in American politics, and I was fortunate that I maintained my composure, returned the compliment, and we went on without incident.

As noted in the introductory pages of this book, things went distinctly quiet on the competition law front for a while after that. When it was seriously placed on the ANC government’s agenda and the drafting team was set up, we did our mandatory – and, I should add, very useful and enlightening – tour of developed country competition agencies, which included London and Brussels and The Hague, the latter because the Netherlands competition authority, like ourselves, was just in the process of getting off the ground. It culminated in a trip to Washington where we attended a short – and again very useful – course at the World Bank and visited, over several days, the antitrust division of the Department of Justice and the Federal Trade Commission. On one occasion during this visit, I was somehow the only person who went to the Commission, where I had what I can only describe as a (very brief) ‘audience’ with the then chair, Bob Pitofsky, the antitrust equivalent, in my mind at least, of a rock star. He sprinkled – to mix metaphors – holy water on our venture and I happily went on my way.

This trip would prove to be of longstanding benefit to us because it was our first substantial engagement with the extremely collegial and practically helpful international community of antitrust practitioners and scholars. On this trip I met Russell Pittman, an experienced economist at the Department of Justice, who taught on the World Bank course and whom I would later adopt as a sort of informal mentor and guide through the thickets of our early cases. Grappling with the early merger decisions in the furniture retail and sugar markets, I would, at the end of my day, send draft extracts from the decisions to Russell and would open my email the following morning anxious to see whether or not he agreed with our arguments. And, without fail, his responses would be waiting for me. I am sure that this occasional correspondence loomed infinitely larger in my little world than in his. But above all, I recall how reassuring it was, particularly in those early days, to have at the other end of my email a person who was willing to share his immeasurably greater experience with us.

When we got into drafting the bill, we began to engage closely with groups of international advisers. It was through the drafting activities that I was first introduced to Eleanor Fox, a professor from New York University and a massive contributor to antitrust scholarship. Her combination of extraordinary generosity, boundless intellectual curiosity and an unceasing quest to establish a link between competition and poverty alleviation ensured that she became, and has remained, a beacon for me and many others grappling with the special problems confronting competition law and policy in developing countries. While she is not the only one who fits that description, there is something particularly striking about Eleanor’s willingness to elevate intellectual rigour above political calculation and personal ambition.

And then, when we were up and running and through the entire term of my office, the Competition Tribunal would, on at least one occasion each year, call on the best and brightest in the field to come to South Africa and participate in workshops in which they would criticise us, advise us, give us the benefit of their vast experience and, I dare say, occasionally learn from our experience. Christopher Bellamy, David Elliot, Eleanor, Harry First, Merit Janow, Fred Jenny, Bill Kovacic and Richard Whish, the last-mentioned a truly peerless teacher, all come to mind. I’m no doubt forgetting some names – and the list does not include those who worked in various ways with the Competition Commission, including staff from various national agencies who were seconded for extended periods to the Commission. But the names listed represent some of the outstanding scholars and practitioners who gave selflessly of their time to help the Tribunal.

I raise this because although I will naturally get on to discussing the institutionalised arrangements and events – from the International Competition Network to the Fordham Conference – that characterised the practice of international antitrust, I also want to reinforce an observation made in my contribution at the 2009 ICN conference in Zurich in the period that I acted as chair of the steering group, and that is the powerful personal relationships and friendships that underpinned these institutional arrangements. These are all the more noteworthy because they often coexisted with deep and robustly debated differences in national approaches towards antitrust enforcement – differences between Europe and the US and between developing and developed countries, to cite the most overt fault lines. While in the trade union world and in general anti-apartheid work international solidarity was a given, I had not expected this to be a strong feature of the competition world. It is, after all, one thing for the workers of the world to unite, but quite another matter to imagine that the suits and argumentative eggheads who stand astride the competition world would place so much store by international solidarity!

I think that this strong international network of practitioners is partly explained by the fact that, despite our institutional position as agencies within our respective national governments, in the substance of our work we frequently share more in common with each other than we do with our own governments. Although some of the national competition agencies were more cautious about expressing this than others, and many, not least of course the powerful developed country agencies, resisted acting upon this by, for example, refusing to allow national trade and industrial policies to be placed on the international antitrust agenda, the fact is that few of the national antitrust agencies and practitioners sympathised with the inevitably mercantilist character of their government’s national trade and industrial policies; they were not the representatives of some national champion or other economic interest group.

More prosaically, however, the building of a strong community of international antitrust practitioners was predicated on the rapid proliferation of national antitrust agencies and thus of national antitrust enforcement in a world of increasingly global markets. From an antitrust enforcement perspective, this conundrum was most clearly reflected in increasingly ubiquitous cross-border mergers and in the scourge of international cartels.

I’ve probably referred elsewhere in this book to the extraordinary burgeoning of national antitrust authorities in the 1980s and then particularly in the following decade of the 1990s. Go to any international antitrust conference – and there are many – and, sooner rather than later, this phenomenon will be mentioned. Estimates of the numbers differ, but we are talking about a sea change in the coverage of competition laws, from a handful of statutes that had, for decades, been almost exclusively associated with the developed industrial countries, to a web of national laws that covered most of the countries of the world.

Amidst all the doctrinal disputes of recent years, the rise of the European Commission as a major factor in antitrust enforcement, the differences that have emerged, submerged and re-emerged between the substantive approaches of the US and the European Union, the resolute pursuit of global cartels, the emergence of new international networks of national authorities, the strengthening of old networks, it is almost certainly the rapid spread of new national antitrust enforcement authorities across the world that represents the single most important development in the field over the past 25 years.

I say this because it is a development that at once raised the prospect of strengthened antitrust enforcement, of more competitive national and international markets, but also portended the prospect of significant national divergence in the procedures governing antitrust enforcement and merger regulation and the substantive standards applied to them. This would have mattered little had these divergent standards and procedures been contained within their national boundaries, but increasingly a multijurisdictional merger would be subject to antitrust rules that diverged in the various national markets in which the merger was taking place, a cartel consummated in one country would fix prices or allocate markets in many national markets, or a distribution or discounting strategy judged pro-competitive in one country would be impeached in another.

If there is an overarching theme that characterises this book, it’s one that insists upon the national particularities that underpin the adoption of national competition laws and the manner of their enforcement. But to the extent that there is a single imperative driving this phenomenon, then it’s market liberalisation, the withdrawal of the national state from direct participation in the production and the regulation of the production of goods and services, in some countries a dramatic and literally overnight occurrence, and in others more gradual. At the risk of gross oversimplification, in countries east of the Berlin Wall the withdrawal of the hitherto all-pervasive state and the rush of private investors to fill that vacuum occurred overnight, while in developing countries the gradual recognition of the limitations of the protectionist and state-centred development economics that dominated the post-colonial period ensured that the withdrawal of the state and the development of markets were, at once, both more gradual and piecemeal.

The development of the laws and regulations that provided the rules for participation in these newly ‘liberated’ markets – critically including competition rules – was anything but linear and orderly. It’s probably fair to say that the withdrawal of the state, including the privatisation of powerful state-owned enterprises, generally preceded the establishment of rules governing participation in the newly liberated markets, specifically including competition rules. The new competition agencies, when eventually set up, thus faced a myriad of problems. These ranged from hugely dominant formerly state-owned enterprises, frequently in the hands of a new business elite closely connected to the political leadership, through to deep-seated institutional deficiencies. Not only was there no experience of the complex technical exercise of administering what was often a hastily and poorly drafted competition statute, but in many instances the agency previously responsible for price regulation was simply re-designated the competition agency.

What was required, then, was not only learning to apply the complex and unfamiliar methodologies of competition law and economics, but, possibly more challenging, ‘unlearning’ old ways that were often supported by populations suffering the consequences of ‘big bang’ unregulated reforms and who longed for prices to be regulated anew. So the new competition agencies immediately confronted obstacles that the agencies in developed countries have never had to deal with. And, if many of the industrialised countries of Europe were also characterised by interventionist industrial and trade policies, they at least had the powerful imperative to create a single European market lending immeasurable weight to their respective national competition projects, and, of course, their funding and human resource capacities were vastly greater than those of the best-resourced developing and transition economies.

My first experiences of the institutionalised world of international antitrust were at the World Trade Organization and then at a meeting held in Ditchley Park in England, a meeting that would ultimately lead to the formation of the ICN.

COMPETITION AND THE WORLD TRADE ORGANIZATION

The WTO had formed a working group on the interaction between trade and competition policy to consider the inclusion of a possible competition instrument on the agenda of the ill-fated Doha ‘development round’ of trade negotiations. The ambitions of the leadership of the committee – notably Frédéric Jenny, whom I encountered for the first time at these meetings – and those, notably the European Union, who supported the inclusion of competition on the negotiating agenda, were modest and pragmatic. Certainly, a more ambitious conception which the European Union may have initially entertained was significantly pared down in an effort to secure the endorsement of their powerful opponents. The opposition to the inclusion of a competition instrument in a WTO trade round came largely from a rather unholy alliance of developing countries and the US, the former fearing that competition rules would effectively limit their ability to employ trade and industrial policies, the latter resistant to the notion that a lowest-common-denominator international antitrust agreement would weaken US antitrust action in relation to international conduct that compromised competition in domestic US markets.

As is well known, while the interaction between trade and competition did initially make it onto the Doha agenda, it was, in order to reduce the controversial side-issues and focus on the central trade issues, soon withdrawn when the negotiations began in Cancún. Given that, 10 years on, the Doha round appears to be nowhere near completion, it is probably just as well that the institutionalisation of antitrust did not tie itself to the snail’s pace of multilateral trade negotiations.

In fact the strongest impression that I took from my experience of the discussions in the WTO trade and competition committee was the glacial pace at which anything to do with international trade negotiations proceeded or, more accurately, marked time. As one experienced US trade negotiator explained, progress at the WTO is measured in ‘vapours’, certainly too ethereal for me to detect. It was not for nothing that the General Agreement on Trade and Tariffs (GATT), the predecessor of the WTO, was widely and not so fondly referred to as the ‘general agreement to talk and talk’!

I was also struck by the impressive level of preparedness of the developing countries, notably India and Malaysia, who opposed the introduction of a competition instrument. The fact is that the WTO members were generally represented in the working group meetings by their trade officials, whose antennae were acutely sensitised to anything that might limit the mercantilist ambitions and instruments that constituted the substance of WTO concerns.

Although I attended the meeting with the full knowledge and agreement of the Department of Trade and Industry and assiduously provided them with copies of any submission that I made and reported back on the meeting outcomes, I was never given a clear mandate. I was effectively at liberty to devise South Africa’s position. I don’t doubt that, had this ever impacted on South Africa’s overall position in the stalled negotiations, the trade negotiators would have taken a more serious look at the positions I was taking. My strong instinct was to support the minimalist competition agenda supported by the European Union, although the absence of a mandate from the Department of Trade and Industry meant I was never fully confident that I was on top of the implications of this position and so was at a distinct disadvantage relative to the better-prepared opponents of this position, not of course that this made any difference in the greater scheme of things.

I remain deeply attached to the view that competition principles and assessment criteria should be placed at the centre of trade negotiations and dispute resolution. There is little more than beggar-thy-neighbour mercantilism that prevents the application of well-developed predatory-pricing rules to assessing allegations of dumping, or that does not apply competition rules to export cartels, or to regulations designed to restrict access to national markets, or to selective subsidies that, as in agriculture, massively distort international and domestic markets – a game that can be won only by those with the wherewithal to pay the subsidies. However, I am not entirely unpersuaded by the US view, its self-serving elements notwithstanding, that saw the WTO as an inappropriate forum in which to deal with competition issues. I am far less persuaded, as I’ll discuss below, by the refusal, again largely led by the US, to keep trade issues off the ICN agenda.

But as frustrating as these meetings may have been, and as unsatisfactory as the parochial, insular positions of many of the developing countries were, the WTO meetings and accompanying WTO/UNCTAD (United Nations Conference on Trade and Development) seminars nevertheless transmitted an abiding sense of how important it was for national antitrust to be institutionalised on the international stage. While the WTO meetings were not in the end examples of the effective international institutionalisation of antitrust, we already had a fairly developed taste of its benefits through our participation in the OECD competition committee (also chaired by Fred Jenny). I’ll say more about the OECD later.

THE INTERNATIONAL COMPETITION NETWORK

Origins

And so when I came to learn that the South African competition authorities had been invited to participate in a meeting at Ditchley Park in 2001 to discuss precisely the international institutionalisation of antitrust and that one of the Commission officials had been designated to attend, I immediately hijacked the invitation and headed off to an event that would, I believed, ultimately change the world of international antitrust. And I make this claim because what flowed from the Ditchley Park meeting was an initiative, the ICN, that did indeed make a major contribution to institutionalising the international community and collegiality to which I’ve already referred and from which we had already derived considerable benefit.

Ditchley Park is an early eighteenth-century English country estate near Oxford. It had served in the early years of World War II as a meeting place for Churchill’s war cabinet. In the middle of the twentieth century it was donated by its then owner, a member of the Wills tobacco family, to the Ditchley Foundation, which was dedicated to strengthening relations between the US and Britain. Its remit has expanded and it is now a decidedly upmarket conference venue for the great and the good to exchange views on all the sorts of things that are of interest to the great and the good.

The meeting was convened by the patrician Canadian lawyer, Bill Rowley, himself a fully paid-up member of the great and the good and a very smart, effective, aggressive – for some, a little too aggressive – advocate of the globalisation of aspects of competition law. Bill was at the time chair of the International Bar Association’s Global Forum for Competition and Trade Policy, the capacity in which he convened the meeting.

Opening Up to the New Competition Regimes

I knew little of where I was headed for, other than that the South African competition authorities and a handful of other developing country competition authorities – the others who might have fallen into this category were the representatives of the antitrust authorities of Mexico, Turkey, Hungary and Brazil – had been singled out for participation in a weekend meeting to discuss a ‘global competition initiative’. It’s an instructive comment on the then limited membership of the international antitrust community that Ky Ewing, then chair of the antitrust section of the American Bar Association and a co-chair of the Ditchley meeting, should, in his report of the meeting, have noted that ‘Significantly, the participants included Japan, Australia, South Africa, Israel, Poland, Hungary, Turkey, Mexico, Brazil and Argentina, as well as the EC, US, Canada and the usual Western European entities – in short governments from around the world’.

In addition to representatives of the antitrust authorities and several multilateral agencies – namely the OECD and UNCTAD – there were representatives of international business associations and leading antitrust practitioners from the private bars of London, Washington, New York, Brussels and, of course, Toronto. The other co-chair of the meeting was the ever-gracious, inclusive Barry Hawk, chair of the Fordham University Corporate Law Institute, whose longstanding annual Fordham Conference is itself a major international antitrust gathering, an important bulwark against the national parochialism that often dominates antitrust discourse across international borders.

I had recalled that the invitation specified that a tuxedo was mandatory at the Saturday evening dinner. However, I took little notice of this, though I did my best and brought along my Norman Manoim-approved Tribunal hearing uniform: chinos, a blazer, and a collar and tie. As soon as I walked in the front door – on the day before the meeting was scheduled to begin – I realised that the specified dress code was indeed seriously intended. I’m not usually overly concerned about such issues but I already felt like something of an outsider in this meeting of people who were mostly well known to each other and so I didn’t particularly want to emphasise my ‘otherness’ by wearing the wrong clothes. I was slightly cheered by my first meeting with the legendary Allan Fels, the charismatic (to put it mildly) head of the Australian agency, not, I am confident he will concede, a clothes horse and not someone whom I would have imagined taking the tuxedo requirement to heart.

I spent the day wandering around the house and a small part of the extraordinary surrounding parklands. In my meanderings ‘below the stairs’ I encountered the butler, who immediately recognised my South African cadences and told me that he had served for many years as the London manservant of Sir Solly Joel, the South African Randlord. I hung out with him while he (literally) polished the silver and asked him, with some anxiety, how my flouting of the dress code would be received and he assured me that in his lifelong experience of serving ‘gentlemen’, ‘true gentlemen dressed as they pleased, and the others felt uncomfortable’. Although not very persuasive or comforting in that context, it was advice that I recalled when I occasionally ran the Tribunal in ways that did not accord with the established practices of the legal profession.

I of course survived the dress code (though with no help from Alan who, contrary to my expectations, descended the staircase resplendent in a tuxedo complete with cummerbund) and was privileged to attend an extremely interesting meeting and effectively be in at the initiation of the International Competition Network.

The Ditchley Park meeting was convened in response to a report prepared by the International Competition Policy Advisory Committee (known as the ICPAC – pronounced ‘ice pack’ – report) for the US Attorney General and for Joel Klein, then the assistant attorney-general for antitrust. Flowing from the ICPAC report, the ‘problem’ that the Ditchley Park meeting was intended to address was, drawing on the report of the meeting prepared by Merit Janow, the conference rapporteur and former executive director of ICPAC: ‘Competition policy issues, many of which transcend national boundaries, have grown increasingly complex, and the number of jurisdictions worldwide with competition laws in place, has increased’.

Drilling down, it’s clear that for the conveners of the meeting the major concern to which this gave rise was that

In today’s economy, the need is growing for expanded international cooperation in the area of mergers. More than 90 countries now have competition laws, and an ever increasing number of jurisdictions (some sixty at present) have pre-merger notification and review requirements … This proliferation of merger laws and practices is resulting in a variety of issues, including costs on corporations and new challenges to competition authorities. As a result, multijurisdictional merger review is one area where expanded international cooperation and process convergence could prove helpful to both multinational enterprises and competition authorities.1

The solution proposed by ICPAC, and effectively adopted by the Ditchley Park meeting, ‘urged competition officials to establish a new and independent forum that was broader than existing forums and more routinely inclusive of both developed and developing jurisdictions … The Ditchley Park meeting addressed as its central question the need for a forum for dialogue, and insofar as possible, for building consensus among competition professionals around the world’.2

In re-reading, 10 years on, the proposal for what, using the language of ICPAC, was then referred to as the ‘Global Competition Initiative’ (GCI) and would ultimately come to be called the International Competition Network, the centrality of drawing developing country jurisdictions into the proposed forum is striking.

The discussions focused on the need to improve cooperation between competition authorities in the context of economic globalisation and to involve developing countries more effectively in this process. There are already several existing organisations where high-quality meetings take place to facilitate dialogue between competition authorities from developed economies. What is needed is a venue where participants from developed and developing countries could consider competition law and policy issues that transcend national boundaries … It was generally noted that this proposed combination of participants represented a unique feature of the GCI and one where it could perhaps add value relative to other existing activities. Looking ahead, future GCI activities must therefore be inclusive in soliciting participation from developed and developing jurisdictions.3

And then in identifying issues on which consensus was reached and which required ‘serious consideration in the months ahead’: ‘this new initiative must attract broad participation from countries with new and developing country competition regimes and seek their active involvement. The proposed forum should make a particular effort to focus on competition issues of concern to developing countries and of those countries with fledgling competition law regimes’. While the existence of other international forums was noted, ‘Nevertheless, many participants expressed the view that existing forums, such as the OECD, UNCTAD or the WTO, each had limitations and did not consider the full procedural and substantive aspects of multijurisdictional merger review with a broad representation of developed and developing country jurisdictions and other competition professionals’.4

And so the first area of consensus noted in the report on the Ditchley Park meeting was that the intention of the new forum proposed was to ‘add value’ to the work of existing institutions, rather than constitute an alternative to these complementary forums.

The report noted that while ‘many expressed the view that it should first and foremost be a forum for competition authorities from around the world … it should consider input from other interested parties, both public (such as other international organisations) and private (such as business, legal, consumer and academic bodies); taking variable configurations, e.g., as commentators, members, participants and/or facilitators’.

The meeting expressly did not attempt to determine ‘how broadly or narrowly to determine the intellectual mandate of the GCI’; however, it is clear that problems arising from multijurisdictional merger notification and review were paramount in the minds of the organisers.

Finally, the meeting agreed that the new forum would involve ‘a minimum of permanent infrastructure, with support primarily provided by participating authorities and other experts and facilitators’.5

On 25 October 2001, fittingly at the Fordham Conference, the International Competition Network was launched with an initial membership of 14 countries including South Africa. The ICN has become a major feature in the world of international antitrust. Much has been written about it and, despite my having served on the steering group from its inception until the end of my term of office on the Tribunal about 8 years later, there is little new that I can contribute. However, reflecting on the 10 years of its existence, the three factors that stand out are, first, the gradual shift in the agenda; second, the manner in which the ICN has gone about its work; and, third, and related to this, the nature of a network.

There is undoubtedly room for cynicism regarding the underlying motive for the formation of the ICN, for its express desire to include developing countries and for its intended agenda and modus operandi.

I have no doubt that, notwithstanding the assurances that the ICN was not intended to substitute for discussion in other international forums, an important element underpinning its strong support, particularly from the US, was certainly predicated on providing an alternative forum to the WTO, which the US clearly opposed as an institutional base for discussion of competition issues. However, as things turned out, even if competition had remained on the WTO agenda, the entire round of trade negotiations has become bogged down and competition’s place on the agenda would have had little practical consequence. So, whatever the intention, the outcome is that we do now have a well-established international forum that has become a major element on the international antitrust agenda and calendar.

The same may be said of the strongly expressed desire on the part of the developed countries to embrace developing countries in a field that had hitherto been the near-exclusive preserve of a small group of developed countries. There was undoubtedly a desire, and this most particularly from the business interests that convened the Ditchley Park meeting, to promote a high level of convergence between the burgeoning national antitrust regimes, particularly where this concerned the regulation of multijurisdictional mergers. Indeed, ‘subordination’ may be a more accurate term to describe some of the notions of ‘convergence’. Among the wilder ideas that have been floated from time to time was the possibility that the decisions of the major jurisdictions simply be accepted by the smaller jurisdictions and that merger filings be made in a single language, that being English of course.

Needless to say, these extreme variants of convergence never even made it to first base, but the very existence of the ICN and the practical character of its agenda and work programme inevitably raise questions about the appropriate degree and mode of convergence in procedures, laws and analytical approaches. Or, conversely, they raise the question of the limitations of a ‘one size fits all’ approach to antitrust law enforcement and merger regulation. There are very few who would claim adherence to a ‘one size fits all’ approach, even though some may retain the vain hope. And so the modality of convergence that has effectively always dominated ICN thinking is ‘soft convergence’. ‘Soft convergence’ is difficult to define or measure but, although I think the value of much-derided ‘talk shops’ is significantly underrated, if indeed the ICN is to avoid becoming a ‘mere’ talk shop then it’s important to develop measures of its impact and performance.

At least as important, the ICN has with equal force promoted what the current chair of the steering group and a smart, thoughtful, provocative leader, John Fingleton, has referred to as ‘informed divergence’. Effectively then, while the ICN’s various working groups and conferences may not always provide the basis for convergence, they do provide the basis for identifying the peculiar national factors underpinning the divergent approaches to antitrust enforcement and merger regulation.

Broadening the ICN Agenda

As already noted, the increasing complexities in the procedures and, to a lesser extent, the substantive standards governing multijurisdictional merger review were the primary imperatives underpinning the establishment of the ICN. In my own contribution to the Ditchley Park meeting I emphasised the importance of placing anticompetitive practices in the conduct of international trade on the ICN agenda. My dog-eared notes of my contribution at Ditchley Park read: ‘When we put together our experience of international trade negotiations with our experience of antitrust enforcement, the most striking feature is the double standard employed – how difficult it is to prove a predatory pricing claim in the US or European Union, compared to the relative ease with which anti-dumping duties are levied in those jurisdictions; with the “take no prisoners” approach to the operations of cartels, domestic or foreign, in the US market relative to express condonation of US cartelisation of international markets’.

While even I recognised that the likely priority items on the ICN agenda were ‘promoting best-practice competition laws’ and ‘training to enhance the investigative and adjudicative capacities of developing country agencies’, I did suggest that the network might prove an important complement to discussions of the interface between trade and competition that were confined to the WTO, where discussion is dominated by trade rather than competition officials, and the OECD, with its more limited membership base.

This was, of course, all very naive, particularly if ICPAC and the ICN that flowed from it are viewed as initiatives driven, in large part, precisely by the desire to park discussions of international competition law issues in an arena where international trade is not up for discussion. And so it has proved to be. Early on in the life of the ICN, the phrase ‘all competition, all of the time’ somehow surfaced as the defining description of the ICN, to be restated in mantra-like fashion by leading figures in the ICN, coined (I seem to recall) by Charles James, who briefly headed the US Department of Justice’s antitrust division, but most stridently and monotonously repeated by Konrad von Finckenstein, then head of the Canadian Competition Bureau and first chair of the ICN’s steering group.

The ICN’s agenda has moved on. In line with what emerged from the Ditchley Park meeting, it was initially preoccupied with merger procedures and with the development of recommended practices in this field. It also set up an early project in substantive merger review standards. An advocacy working group was created and, early in the life of the ICN, a working group with the cumbersome title of ‘the capacity-building and policy implementation working group’ was established. This latter group was effectively responsible for dealing with developing country concerns – for example, technical assistance and advocacy.

At the 5th annual conference held in Cape Town in 2006 it was agreed, not without some opposition, to set up a working group on unilateral conduct, or, as it is more commonly referred to, abuse of dominance or monopolisation. While this group has predictably proved to be fairly controversial, there are effectively no longer any barriers to what the ICN can take on within the confines of competition law. If there were those who thought to confine discussion to commercially pressing topics such as procedures for regulating multijurisdictional mergers and to relatively uncontroversial topics like cartels, then the actual ICN agenda has long since passed them by.

One of the most important new items to become a regular feature of ICN conference agendas and working groups, and one that has been elevated to an issue of first-order importance, is the question of agency effectiveness. This was initially a matter confined to the newer agencies when suddenly – partly inspired by the questions posed by the financial crisis but also somewhat preceding this – developed country agencies started asking questions about their own effectiveness. And so the work of the capacity-building and policy implementation working group, which had, arising from its focus on new agencies, been preoccupied with strategies for building effective new agencies, was suddenly ‘main-streamed’, as its concerns were quickly transformed into the most important item on the ICN agenda. Strategic planning, prioritisation, and evaluation criteria became common ICN themes, instead of matters that were of interest only to the newer agencies.

However, the ICN has rarely strayed beyond the confines of competition enforcement and merger regulation. In particular, it has steered clear of the interface between competition policy, on the one hand, and trade and industrial policy, on the other. In the report of the steering committee chairman to the 2009 ICN annual conference, I recommended that we revisit the ‘all competition, all of the time’ mantra:

This was always a restrictive principle on which to base ourselves because it was a tactical consideration rooted in caution and in a lack of familiarity with one another, of a fear that one nation would begin interfering with the internal economic policies of another. It is not a positive, expansive and forward-looking statement of our mission and vision. Competition law, as a US Supreme Court justice once famously reminded us, is the Magna Carta, the fundamental law, of the market system. As such, it cannot be confined to an island where its relationship with every other branch of economic and social policy – particularly with trade and industrial policy – is studiously ignored. And if anybody seriously believed that this was ever possible, then the financial and economic storm that has battered us all since the latter half of 2007 and the policy responses to these events should have put paid to any such illusions. With industrial policy ascendant and protectionism in the air, can we afford to remain aloof? Must we not rather engage with the concerns of policy-makers to ensure that their interventions – many of them necessary – respect and preserve that which is dynamic and creative and democratic in a market system? There can be little doubt that the unusual effectiveness of the European Commission’s competition directorate rests in significant part on its authority over national state aid, thus effectively empowering it to deal with both competition law and policy and the critical instruments of industrial policy. We cannot all aspire to this level of institutionalised authority over industrial policy, but we can use the forum of the ICN to discuss the appropriate interface of our work with industrial and trade policy and with the social and developmental needs of our citizens. This is the least that we can do.

I’ll elaborate in the concluding chapter why I believe that in their second decade the South African competition authorities are going to have to expand their focus to take on public restrictions on competition. The same imperatives that will compel the South African competition authorities in this direction – in essence, more aggressive state intervention to confront market failures or, simply, market outcomes that are politically inconvenient – will dictate a similar pattern in the programmes of competition authorities across the globe. And, of course, it will profoundly influence trade and investment across borders. The ICN is an important forum at which this should be discussed, at which to understand the different imperatives that will drive divergent national policy stances, and at which to workshop appropriate responses from those tasked with defending, by enforcement and advocacy, competition principles. And this presupposes that the increasingly anachronistic requirement that the ICN eschew involvement in any other policy fields, notably trade and industrial policy, despite their manifest importance for competition principles, be dropped.

I have no doubt that the constraints imposed by ‘all competition, all of the time’ will ultimately be dispensed with. These discussions are, after all, commonly held in the OECD competition committee, and even the painfully diplomatic UNCTAD (whose acronym is parodied as ‘under no circumstances take any decisions’) – which, given its membership of the UN family of institutions, makes its conferences possibly the most representative gatherings of competition enforcers – allows for the cross-fertilisation of trade and competition issues.

This self-imposed constraint will ultimately be dropped because the contrast between espousing high-minded competition principles at the ICN and the egregiously mercantilist conduct of national governments will at some stage become, like the proverbial elephant in the room, impossible to ignore. Take Canada, that least offensive of nations, and the oldest competition law regime of them all. In recent times, the Canadian government has blocked the acquisition by an Australian mining company of a Canadian potash producer, the largest potash producer in the world. Moreover, it appears that the Canadian government will block the acquisition of the Toronto Stock Exchange by the London Stock Exchange.

The global competition family has recently been joined by China and India, the latter already a member of the ICN. Both countries have become major actors in international trade and production and both have a marked proclivity for economic nationalism. In fact there is already some level of discomfort about the allegedly nationalistic criteria that the Chinese competition authorities are bringing to bear on merger decisions involving foreign acquirers. By explicitly eschewing discussion in the ICN of the interface between trade and industrial policy and competition policy, does the ICN not signal that anticompetitive competition policies are acceptable, as long as they do not emanate directly from the practice of competition law? In the name of realpolitik it is always possible to live with some level of hypocrisy, but at some stage it will severely call into question the integrity of international antitrust and those institutions that, like the ICN, are mandated to uphold its principles.

This does not necessarily mean that the ICN should single out particular national trade and industrial policies for public analysis and criticism. But surely the ICN should discuss these issues, the better to prevent their wholesale adoption in the rest of the world? Discussion of the US soda ash cartel discussed in an earlier chapter may better enable national competition authorities to join together in taking on a practice that all in the ICN profess to deplore. As already noted, when the Tribunal approved the consent order that effectively ended the operation of the US soda ash cartel in South Africa, the Commission was explicitly requested to post the order on the ICN website in order to show some of its fellow ICN members who are severely disadvantaged by the cartel that it can be and should be opposed through the use of national laws. This was, to my knowledge, never done.

Frédéric Jenny, whom I, and I imagine many others, consider to be the leading international antitrust activist and advocate, is sceptical of the ability of national antitrust authorities to take on export cartels and other cross-border contraventions of competition principles. He insists that a multi-country antitrust authority is necessary to achieve this. I, on the other hand, think that the South African Ansac case shows that, in some if not all circumstances, it is indeed possible for national action to take on cross-border contraventions of antitrust principles and laws, albeit at considerable cost. However, there is no doubt that even in these cases a platform for international coordination would be extremely useful. Had the range of countries in which Ansac operates simultaneously used their national laws to confront the cartel, I have little doubt that it could have been buried once and for all. However, if these issues are not even open for discussion at the world’s leading network of national antitrust authorities, then there is little possibility of that coordination taking place. Again, I emphasise, I don’t necessarily expect the ICN to be the platform on which action against specific export cartels is organised. However, I do insist that the topic should be discussed at the ICN, and that instances of export cartelisation should be documented, as should national action against these be highlighted. In the absence of an international forum at which these issues are at least raised for discussion, the likelihood of coordination among affected national agencies is significantly diminished.

I look forward to the day when export cartels, anti-dumping measures, tariff and non-tariff barriers to trade, trade-distorting subsidies and the like find their way on to the ICN agenda and ‘recommended practices’ are developed around these competition-reducing measures as well. And the day will come, or the ICN will increasingly marginalise itself … or run out of topics to discuss.

Building a Virtual Network

For anyone concerned with institution-building, the truly outstanding feature of the ICN is the sheer volume of mostly high-quality output – in the form of working-group reports, recommended practices, steering-group meetings, workshops and conferences, online conferences, and, recently, training films – that the ICN has managed to produce with the slender resources at its disposal.

As is well known, the ICN is a virtual network. That is to say, it does not have any employees or budget to speak of or any real estate. To the best of my knowledge it doesn’t have a single full-time employee: none of the officials who work on the ICN, even those from the best-resourced international departments, is employed to do ICN-related work alone. Essentially the ICN is a website and lots of telephones and email accounts that assume corporeal form at an annual conference in one part of the world or at a rare workshop somewhere else altogether. Mostly, it consists of lots of people participating in international teleconferences, exchanging emails and preparing material that is, after much toing and froing, posted on the website.

I was extremely sceptical about the possibility of sustaining an international network on this extremely slender infrastructure. And as each highly successful conference passed – each dependent on the quality of work achieved by the working groups and other contributions between annual conferences – my anxiety heightened because of the increasingly high standards that were being set. Would we be able to maintain the quantity and quality of work achieved by this group of essentially part-time workhorses?

My fears have proved groundless. In fact, I have finally come to understand that it is precisely the ICN’s reliance on the commitment of its members that accounts for the unusually high level of output each passing year produces. It is because there is no professional, full-time secretariat on which to lean that ICN members are willing to step up to the plate in the manner that they do.

This is, of course, not without its shortcomings. The effort required to ensure the year-on-year success of the ICN is truly herculean. And while the effort put in by some of the smallest agencies is extremely impressive, the lack of a dedicated infrastructure does leave the ICN somewhat hostage to the largest agencies maintaining their level of commitment. Should even one of the US or the European Commission or Canada significantly downscale its level of commitment, the ICN may well find itself in serious trouble.

And, of course, it is the weakest who generally rely most heavily upon the work of a full-time secretariat – to prepare and circulate the background papers, to chivvy up the submissions, to prepare the agendas. But on the other side of the scale, I repeat, it is in significant part the absence of a full-time secretariat that has called forth an exceptional level of commitment, including that from some very poorly-resourced agencies.

While the lion’s share of the secretariat-type functions, including the management of the working groups, has undoubtedly been borne by the developed country agencies, the developing countries are making major contributions at every level of the ICN. As is already clear from the Ditchley discussions, the inclusiveness of the ICN, specifically defined by the participation of developing countries, was viewed as the key argument for its establishment.

As I’ve already indicated, there were a variety of factors underpinning the strongly expressed desire to incorporate developing country agencies and new agencies in general into the world of international antitrust engagements and institutions. However, whatever the reasons, the preferred modality for achieving these objectives was inclusiveness as opposed to the modus operandi of much multilateral interaction at the time, one characterised by the exclusive gatherings of the major developed European and North American nations (and Japan) and by their domination of the key multilateral economic institutions. Again, it is possible that some of the key drivers may have viewed the new agencies as playing little more than a passive role, one that essentially served to do little more than legitimise the views of the older, developed country agencies.

However, if this was indeed the intention, it has not been realised. First, it’s my impression that the initial intention, at least of the conveners of the Ditchley meeting, was that what came to be called the ‘nongovernmental advisers’, or NGAs – who were largely, although not entirely, representatives of large North American and European law firms – would play a role in the decision-making structures of what was to become the ICN. This was not to be. As early as Ditchley, it was made clear that the institution to emerge from these discussions would be controlled by the antitrust agencies and that this would not be shared with the NGAs. I may be wrong about this, but my recollection is that it was the refreshingly blunt Alex Schaub, then head of the European Commission’s competition directorate, and his perfect foil, the gracious, charming and courageous Mario Monti, who played the leading role in asserting the dominant position of the agencies. This alone immediately made the ICN more open to genuine participation by the newer agencies.

And then, second, this was sealed once certain of the newer agencies – South Africa, Brazil, Mexico, Korea, a number from the former Soviet bloc countries – started, of their own volition, to play an active role in the various structures of the ICN. For a long time – and still today – it is these middle-income bridges between the developed and developing worlds who are, predictably, the most active representatives of the developing world in the work of the ICN. But this is not without significant exception – for example, I think that El Salvador’s hosting of the 2007 cartel conference should be counted as one of the ICN’s proudest moments – and my impression is that the smaller agencies have become increasingly active participants.

However, if the ICN wants to maintain or, preferably, increase the level of involvement of developing countries in its work, then it has to be thinking consistently about specific approaches, modes of working, that optimise the contribution of its membership base amongst developing countries. This has largely been viewed as a content-driven issue – that is, what content is of particular interest to developing countries? It was this sort of thinking that led to the formation of the capacity-building and competition policy implementation working group, which was, in reality, the group where developing country issues were discussed – technical assistance principal among them. Interestingly, because this group was also the natural home for talking about agency effectiveness, which has in recent years become an issue of major concern to developed country regimes, this has had the effect of ‘mainstreaming’ what were previously conceived of as topics most pertinent for new agencies.

I think, as I’ve already made clear, that the content of the ICN agenda is an issue of vital importance. I have indicated why and in what direction I think it should be broadened. My limited but growing experience with competition issues in low-income developing countries indicates that trade-related issues may be of particular importance to developing countries. And I don’t simply mean the hardcore trade issues like developed country agricultural subsidies, but rather issues like the continued existence of export cartels in markets of particular significance to developing countries. I am convinced, however, that if one scratches beneath the surface of these, they will emerge as issues of concern to developed countries as well.

In fact, as the agency effectiveness issue illustrates, it is not easy – nor is it desirable – to pigeonhole topics along developing and developed country lines. All agencies are required to be effective; and all agencies require a technically capacitated staff, among other attributes, if they are to be effective. So effectiveness and technical assistance, though the forms may differ, are matters of vital concern to all agencies, wherever they hail from and whatever their longevity. The ICN should be thinking about mechanisms of inclusion rather than what, in effect, becomes a mechanism of division along issue-defined grounds. There have been examples of this. The early decision to ‘twin’ a developing and developed country in the leadership of every working group – a commitment not always honoured – is one good example.

The induction session for new members which the two US agencies, the Korean Fair Trade Commission and the Japan Fair Trade Commission hosted at the Zurich conference is another example. This is something that should be repeated at regular intervals with all small and new agencies, thus enabling them to make an informed choice about where, in the vast spectrum of ICN activities, to devote their energies. Some of the choices may be surprising. For example, one may find that agencies from small countries that are inevitably obliged to confront dominant firms may be more interested in grappling with the technical complexities of unilateral conduct than with the more nationally particular techniques of advocacy. Careful analysis of the problems confronting low-income developing countries may well reveal a special interest in issues related to the interface between international trade issues and competition enforcement. However, I have no doubt that, like agency effectiveness, once this was placed on the agenda it would emerge as a topic of concern to developed and developing countries alike. So to repeat: the question of inclusiveness should be dealt with not as a delineation of separate topics but rather as a question related to the internal functioning of the ICN.

Be that as it may, I have little doubt that the objective of inclusiveness has, in significant part, been met. It has been met partly as a result of the enlightened self-interest of the large agencies and the influence of the NGAs, partly as a result of the active goodwill of many of the older agencies, and partly because a number of the middle-income, and later the low-income, developing and transition economies seized the opportunity to engage. The level of engagement may well have exceeded both the expectations and even the wishes of those who were intent upon drawing developing country agencies into the workings of the ICN.

However, in addition to introducing measures that specifically ease the ability of developing countries to participate meaningfully in the ICN, it is a ‘right’ that has to be continually reasserted by the newer agencies themselves. It was won, as I’ve already noted, partly because some developing countries enthusiastically embraced, even insisted upon taking, the opportunity to engage with their more experienced peers in the developed agencies. This does not mean that a developing country is likely to win a point, particularly one concerning substantive antitrust doctrine, against a united North American–European Commission view. But these blocs are frequently not united and it is possible to oblige the larger agencies to bend somewhat to achieve consensus in order to maintain what has become the spirit of the ICN. But more important even than winning contested doctrinal ground is ensuring that issues are put up for discussion. It is less difficult to insist that an issue be placed on the ICN agenda for discussion than it is to prevail in a doctrinal dispute.

There are multiple reasons for this, but in the end it is more difficult to argue that an issue does not even bear discussion than it is to support one or other position on the merits of the issue in question. I recall strenuous objection to the Japanese proposal that the Kyoto Conference include a session on the question of ‘fair trade’ or, as I think it was termed, ‘unequal bargaining power’. But the Japanese prevailed, partly because they are the Japanese, partly because they were exercising what had become a host’s prerogative to put up a topic for discussion, and partly because many on the steering group were uncomfortable with putting beyond the bounds of discussion an issue that was clearly of importance to some ICN members. The upshot was that we had one of the more interesting – because contentious – discussions that have been held at an ICN conference and, in the process, we came to better understand the somewhat heterodox East Asian view on competition, a view that the international competition and business world is, for obvious reasons, well advised to start taking more seriously.

These are then the mechanisms for enhancing the participation of newer agencies in ICN work: keep thinking about specific modalities of inclusion; keep fighting for expansion of the agenda; and maintain the virtual character of the ICN because it has somehow become apparent to all of its members that the ICN does rely, for its very survival, on their active participation. It has also demonstrated that the creative use of fairly straightforward electronic communications technology can go a long way – I’m confident that the current effort to develop interactive DVD and internet-based training material will not only help improve national enforcement, but will ensure that the ICN is taken into the national authorities below the leadership level that attends the annual conferences or even the periodic workshops. The ICN may not be able to maintain quite so lean a structure for all time. But, as in so many other walks of work and life, lean has proven not only to be mean but it has induced higher levels of productivity and innovation. And, of greatest importance, it has strengthened the network by ensuring that its success is dependent on the efforts of its members.

The NGAs are also an important element in the composition and character of the ICN. They are composed overwhelmingly of senior members of large European, US and Canadian law firms. In addition, there are a smattering of lawyers from the newer jurisdictions and a small, but very high-quality, number of scholars. As already noted, my impression was that those NGAs who attended the Ditchley meeting – mostly the aforesaid lawyers – expected to play a more central role in the decision-making structures of the ICN. However, although this was not to be, the NGAs have played a role of varying degrees of intensity and influence in the working groups and in the panels at the annual conferences and working groups.

I have little doubt that the active involvement of the NGAs, reflected partly in the important contributions they have made to the work of certain of the working groups and other ICN output, is an important aspect of the character and effectiveness of the network. However, beyond their direct contribution, a strong, regular interaction between business and antitrust enforcers is essential. After all, with the exception of cartel conduct – and the NGAs have been excluded from participation in the cartel working group – antitrust regulators are required to have a deep understanding of business thinking and practices if they are to provide the necessary certainty and achieve the requisite degree of legitimacy in their dealings in merger review and non-cartel enforcement matters. As it is, the leadership of the working groups is always in the hands of agency representatives and final drafts of working group reports are approved at the Steering Group from which NGAs are excluded. But I have some misgivings about the composition of NGA representation.

Defence lawyers from Europe and North America are significantly overrepresented. For better or worse, few of the newer jurisdictions are as comfortable ‘fraternising’ with defence lawyers as are our developed-country counterparts. This certainly changes as the enforcement agencies grow in confidence. However, while we have benefited to some extent from the cross-fertilisation of private practice and public service, we will never have the same, apparently smoothly-functioning revolving door that characterises US practice. So while the relationship between the South African enforcers and the members of the private bar may be cordial and respectful, it is never likely to be quite as chummy as that between, say, the US enforcers and their counterparts in the private sector.

I also entertained the hope – which I now acknowledge will never be realised – that the lawyers may occasionally introduce a real live businessperson to the ICN proceedings. I don’t expect them to take part in working groups or attend entire conferences, but I persist in thinking that an engagement between antitrust officials and business executives – particularly at the nuts-and-bolts level of the ICN – in a panel discussion or breakaway session, may be more enlightening than the engagement with the same group of lawyers year in and out. The only time that I can recall an actual businessperson participating in an ICN panel was at the first conference, where we invited a South African Breweries executive to participate in a panel discussion. I am assured, however, that no US lawyer worthy of the name would ever allow a client to participate in a conference dominated by enforcement agencies, an approach that reveals how shallow all the chumminess really is.

On the other hand, scholars and consumer groups are extremely underrepresented in the working groups and conferences of the ICN. I have never fully understood why the ICN has never succeeded in interesting consumer groups in its work. And most scholars apparently find the ICN boring. While I sympathise with those who find the tendency to avoid controversy and sublimate differences less than riveting, I can’t agree with those who are uninterested in the ICN, because of its intensely practical, nuts-and-bolts character. There are exceptions of course: those who can discern the intellectual interest in understanding how competition agencies think and who are willing to try to influence and assist in the development of that thinking. However, this does not include those who earn vast fees for appearing as expert witnesses before these bodies, much (I fear) to the detriment of their clients, who may benefit more from an understanding of the considerations that drive the thinking of the enforcement agencies, and rather less from the most recent fad in academic economic thinking.

Is the ICN Worth the Candle?

An enormous amount of effort and – when the cost of the conferences, workshops and employee time is added up – a fair amount of public money go into the ICN. So, too, does a not-insignificant quantum of political and reputational capital, both from the leadership of the antitrust agencies and from the lawyers who were instrumental in initiating it. Has it all been worth it?

Much energy has been dedicated to devising measures of effectiveness of the ICN. I think that the strongest pressure to measure the impact of the ICN has come from the NGAs who, you may recall, encouraged and assisted in the formation of the ICN primarily in order to solve a particular problem, namely the diverse procedures for regulating multi-jurisdictional mergers that followed in the wake of the mushrooming of new competition authorities. This issue, although not unimportant, is relatively unusual in the extent to which it lent itself to technical solutions and recommendations.

However, it has established a modus operandi within the ICN that is most clearly reflected in the overriding concern with agreeing relatively easily-specified recommended practices and with judging the organisation by the extent to which these agreed practices have been implemented. While this is an approach that remains valid and important, it clearly lends itself to dealing precisely with the relatively straightforward, technical questions raised by merger notifications and procedures. The US agencies have responded most energetically to the pressure from the NGAs to measure the adoption of recommended practices, in large part because they are the agencies most susceptible to NGA pressure. However, this approach to assessing the effectiveness of the ICN is at once too ambitious and too narrow to deal with the more nuanced and controversial questions that increasingly characterise ICN work.

It is too ambitious because the transmission belt from ICN recommendation to national legislation is bound to be extremely slow and imperfect and so, by this criterion, the network will always be judged to be falling short of its own targets and stated expectations. In any event, the evidence is that ‘best practices’, even in relatively uncontroversial areas, usually have to be tailored to national circumstances and so implementation will always be uneven at best. Nor is it clear that even these relatively simple measures have, on their own limited terms, accurately reflected the extent to which the ICN recommendations have been taken up. It was, for example, reported with some fanfare that Brazil had modified some of its merger notification procedures so as to conform with ICN recommended practices, only for it to be revealed some time later that the Brazilian legislature had rejected the changes proposed by its competition authorities. On the other hand, I have no doubt that, in other instances, the recommendations have influenced legal drafters and legislatures but have not been attributed to the ICN simply because the ICN has become a leading public good, a leading authority on the way things are done in competition law. In my estimation, this represents a significantly more powerful endorsement of the ICN than does the active ICN agency that attributes its changes directly to the influence of the ICN’s recommendations, because the former represents the spread of an ICN ‘culture’. However, it is obviously extremely difficult to measure the latter type of response.

Adherence to recommended practices is also too narrow a measuring rod, because if the ICN has chosen to adopt an organisational form and character distinct from that of an organisation responsible for devising binding rules, then it should not judge itself by precisely the same norms and standards appropriate to organisations with those powers.

In other words, the ICN has to judge itself by a more qualitative and nuanced set of norms. For example: are the ICN’s component national parts developing a deeper understanding of each other’s key drivers and practices? Is the ICN providing a framework where the better-resourced and more experienced agencies are able to transmit the learnings from their successes and failures to the newer agencies? Is the ICN providing the space for the full participation and voice of the newer agencies? Is the ICN a voice for competition on the international stage and in the various national policy debates?

A glance at the ICN’s range of activities will reveal that it is extending its work into areas that do not easily lend themselves to textual consensus and harmonised implementation. For example, the expanding work on agency effectiveness and market studies will be difficult to measure by adherence to recommended practices or degree of convergence. The ICN is also fully immersed in the muddy waters of unilateral conduct, and this has predictably proved to be such an area, one where debate and controversy are endless, not simply for the sake of argument, but because we are dealing with a dynamic environment and with diverse national circumstances and policy orientations. Moreover, unilateral conduct inevitably deals with very large companies that, multinational though they may be, inevitably fly one or other national flag, are somebody’s ‘national champion’. They will seek the protection of their flag, just as others will seek to protect their own would-be champions, and this undoubtedly impacts on the way that enforcement actions are conceived and received.

However, the ICN is clearly in the business of making ‘soft law’, and this is what it must continue to do. But ‘soft law’ is not just a less binding, pale version of ‘hard’ law. It is different. It is rooted in consensus, rather than majority; in persuasion through shared experience, rather than coercion; in understanding the varying bases for divergence, rather than suppressing them; and inevitably it is about achieving similarity in approach, rather than agreement on identical rules. Where a ‘soft law’ institution attempts to eradicate differences in hard rules, it is likely to meet resistance in achieving agreement over broad approaches for fear that they will be translated into hard rules. Alternatively, it will encounter the sort of largely unintended passive resistance that, for the sake of form and maintaining the apparent coherence of the organisational form, accepts ‘hard’ form agreements that will simply not be implemented.

This is, I believe, precisely what is happening. Judged simply from the perspective of the range of work it undertakes, the quality of that work, and the extent of participation, the ICN is an undoubted success. However, by the level of adherence to the recommended practices, its achievements are limited.

So how should it be judged?

I have already suggested the criteria that I consider important in evaluating the performance of the ICN.

Is the ICN providing the space for the full participation and voice of the newer agencies? I have already indicated that I think the ICN has made considerable ground when judged against this criterion. Could it do more? Certainly. It could, with relative ease, institutionalise mechanisms that keep the newer agencies better informed of the range of activities in which the ICN is engaged, and so be better able to select from this range of activities those it deems to be of greatest utility and interest. It could expand the range of topics – which would probably entail dropping the ‘all competition, all the time’ constraint. The smaller agencies themselves could go the extra mile in asserting themselves, in demanding greater voice. They could organise themselves better through regional forums and engagements that enhance their collective voice. The timing of initiatives of this nature is auspicious because the developed countries have reluctantly come to concede a certain decline, across a range of fronts, in their ability to impose their preferred practices and outcomes on a number of institutions and engagements they have traditionally dominated. Were the Indian authority to get off the ground and the Chinese authority to join the ICN, the power balance would change, admittedly in ways that are unpredictable, but certainly against the degree of dominance exercised by the older agencies.

Are the ICN’s component national parts developing a deeper understanding of each other’s key drivers and practices? I think so. This is exemplified by the discussion of ‘fair trade’ at the Kyoto Conference and by some of the debates in the unilateral conduct group. I think that this will strengthen, and ‘informed divergence’ will be advanced and better appreciated, once the newer agencies themselves gain more enforcement and advocacy experience and develop a better understanding of their own particular requirements and limitations.

Is the ICN providing a framework where the better-resourced and more experienced agencies are able to transmit the learnings from their successes and failures to the newer agencies? By complementing the fairly indigestible agendas that characterise the annual conferences with the more focused workshops, the ICN has shown sensitivity to the need to address national competition authority staff below the leadership level. The response to these workshops has been universally positive. More creative use of electronic communications technology will also assist. The need for a more user-friendly website has been acknowledged and progress has been made on this front. The recently established ICN curriculum product, which will see the production of a series of training movies – a brainchild of the endlessly creative international department of the US Federal Trade Commission – is particularly exciting. Endless opportunities arise from this project. The movies could be used as a complement to existing ICN material. Where necessary, new manuals could be produced. The possibility of providing an online tutor to facilitate discussion among those watching the movie is likely to prove particularly useful. The ICN’s efforts in their training programmes are bedevilled by language issues – the language of the ICN is effectively limited to English, with the host country providing translations into its own national language at the annual conferences. However, it should not be beyond the ken of the well-connected ICN leadership to procure funding for translation in order to undertake training programmes of this sort.

Is the ICN a voice for competition on the international stage and in the various national policy debates? This is an area in which the ICN has not succeeded. I think that this is largely due to the unwillingness of the leadership to engage with anti-competitive national policies and practices, particularly with respect to practices impacting on international trade. This does not presuppose getting into undiplomatic rows with particular national governments, not to mention national member agencies of the ICN. But it does involve the ICN placing issues such as export cartels and the abuse of anti-dumping rules and other common anticompetitive international trade practices on its agenda and then reporting the fruits of those discussions in the form of media releases or op-ed articles. One person – Frédéric Jenny – has done more for promoting competition on international markets than has the entire ICN network. These are areas in which the ‘old guard’ of the ICN is ripe for challenge.

Paradoxically, for an organisation so little involved in public advocacy, some of the best work that the ICN has done is on how to conduct advocacy campaigns. It should take note of its own work. A simple suggestion: many agencies run active public relations departments. Those employed to do this should be drawn into the discussion of how to advocate in favour of the ICN. This is, after all, their area of expertise. But, of course, in order to attract media attention, the ICN has to be willing to take on some controversial issues. I’m confident that it is able to do this without alienating any of its member agencies or their governments.

As I write this and contrast these evaluation criteria that, I am convinced, better fit a ‘soft law’ network like the ICN than a ‘hard law’ network like the WTO, I can almost hear the pejorative ‘talk shop’ applied to my suggestions. Certainly a ‘talk shop’, if that is what the ICN is, is better than nothing at all, which is what the WTO produced after much talking. Talking is also far preferable to the vain attempt to impose hard law-type criteria as the measures of success and then to come up with some agreements on procedure that, even if implemented, few would recognise as representing significant progress.

On the other hand, the volume and quality of material on antitrust enforcement, advocacy and merger review that has been produced by the ICN and the clear evidence of its utility for newer agencies represent substantial progress. The intensive discussions on agency effectiveness represent progress for all of the member agencies. It assists the ICN member agencies to become better antitrust enforcers. However, I repeat, to be recognised by the public, by national governments and by other multilateral institutions as an organisation making substantial progress, these fruits of the ICN’s progress need to be injected into public international debate and, where possible, it needs to be seen as a force opposing anticompetitive national practices, most appropriately, though not exclusively, to promote and maintain open international markets.

The OECD is an organisation that is also often derided as a ‘talk shop’ – indeed as the ‘organisation for excellent cocktails and dinners’. In fact the cocktails are, at my level anyway, fairly Spartan and I don’t recall ever being offered dinner. However, I know, even if I can’t turn this assertion into a measurable quantity, that we have benefited enormously from our non-member status on the OECD’s competition committee, from its annual global forums where participation is broadened to include a large number of non-member countries, and from the training programmes it has conducted inside South Africa. We would benefit even more if South Africa actually joined the OECD.

We have benefited because we wanted to learn from our more experienced peers and also because we thought it important to explain why we chose, in certain key areas, to do things our way. Hence we became the first non-member country to submit voluntarily to an OECD peer review; we ensured that we submitted papers, as participants in the competition committee are required to do, on the huge range of topics discussed at these meetings; and we took full advantage of the expertise that the OECD has to offer in delivering training programmes in South Africa. Recently, OECD officials have run a series of training programmes with the competition authorities, and, more important, with the procurement officers in various government departments and agencies in order to assist them in identifying and combating bid-rigging. If this is a ‘talk shop’, then we need more of them!

But to return to a theme raised at the beginning of this chapter, one of the great advantages of these dreaded talk shops is the depth and breadth of personal contact they enable. This is even more true of the ICN than it is of the OECD, largely because the absence of an ICN secretariat makes it so imperative to engage intensively with one’s peers. In the OECD it is possible to have your relationship with the other participants in the competition committee mediated through the exceedingly competent secretariat. This is what talking does. It builds networks, and in the ICN and the OECD it has underpinned two highly successful networks of committed institutions and individuals. And if you do want to hear the one about the two ducks, well then you would have to meet Allan Fels at an OECD or ICN meeting!

CONCLUSION

Increasingly, multilateral institutions will come to look more like the OECD and the ICN networks, and less like the WTO treaty-making institutions. The countries of the world may be growing more interdependent, bound together by trade and investment, migration, security, climate change. You name it – there are not many issues of central significance to prosperity and security in either the long term or short term that can be solved by unilateral national action. And gone are the days when the US and Europe (with Japan in tow) meet around a single table and reach agreements that would have to be accepted by the rest of the world. There are economic powerhouses, countries with massive actual and potential impact on peace and security, and countries whose actions have overwhelming implications for medium- and long-term environmental outcomes, all of which command a place at the decision-making table.

But while the countries of the world may have become more interdependent, their historical legacies remain intact, their economic circumstances remain vastly divergent, and their governance systems vary dramatically.

This dichotomy is starkly present in the world of international antitrust, where increasing connectivity through the globalisation of trade and investment coexists with hugely divergent national histories and contemporary circumstances, including institutional capacities, which impact dramatically on national competition laws and practices. The former set of factors requires greater convergence, while the latter dictates national specificity.

This was graphically brought home to me in two projects that I have undertaken since leaving the Tribunal. The first was a peer review of the Armenian competition authority, undertaken on behalf of UNCTAD. The second involved working on the team appointed to assist in the establishment of a forum for African competition authorities.

I was privileged to undertake the Armenian peer review with Bill Kovacic, a long-time official and former chair of the US Federal Trade Commission. Bill has an extraordinary ability to maintain the highest standards of scholarly integrity and rigour while remaining centrally engaged in the always scrappy and politicised world of US antitrust enforcement. And so he is an active player who is nevertheless capable of a refreshing outspokenness. He is the most articulate, thoughtful and passionate of antitrust advocates, and unquestionably one of the great figures in the field.

And so there we were in Yerevan, Armenia’s capital city, whose otherwise unprepossessing skyline is dominated by the peculiarly unsettling presence of Mount Ararat (of Noah fame) and a massive, truly massive, statue of ‘Mother Armenia’, whose major claim to fame appears to be that in about 1990 she replaced an equally monstrous statue of Stalin.

If this alone doesn’t serve to encapsulate the difficulties and peculiarities of promoting competition in Armenia, then think about this: Armenia is, largely because of the heritage of extreme regional specialisation that was a feature of Soviet central planning, a highly open economy. Vital food and other products are imported while the major export, machine tools largely destined for Russia, disappeared overnight with the demise of the Soviet Union. However, intractably hostile regional relations between Armenia and its major neighbours, Turkey and Azerbaijan, ensure that Armenia’s only somewhat cost-effective access to the world economy is through a narrow corridor where Georgia and Armenia share a border. This border is effectively occupied by highwaymen – otherwise referred to as oligarchs – who have captured the Armenian customs and excise department, which means that they do not pay duties, largely through the expedient of massively under-declaring the extent of trade. Needless to say, the advantage of not paying customs duties helps secure the dominance by the oligarchs of key Armenian markets – sugar, rice, petroleum and many other basic products.

And so entry barriers are raised both because potential investors are unable to know the true size of the Armenian economy (which is understated by a significant order of magnitude), and also because to enter a market when the incumbent has captured the tax authority is extremely difficult. These barriers to entry are not to be found in the standard industrial organisation textbooks. And put it together with a very hostile and powerful telecommunications, gas and electricity regulator, and a woefully under-resourced competition authority, which is grappling with ‘unlearning’ its past incarnation as a price regulator, and a competition strategy for Armenia faces obstacles that may or may not be greater than conventional wisdom assumes but are certainly different.

This is not to say that Bill and I came away believing the Armenian situation to be hopeless. Far from it. A spirited, literate, numerate competition authority, governmental and non-governmental support for competition enforcement, strong support from the treasury for introducing competition principles into the public procurement process, and international support, all revealed viable strategies for strengthening competition enforcement in Armenia. But these strategies would have to be deeply rooted in Armenia’s concrete circumstances. The ICN’s recommended practices in merger review may, like chicken soup, do no harm to the competitive process in Armenia, but neither are they likely to do much good.

I’ll say little about the difficulties confronting the various competition authorities of Africa, which vary greatly from region to region and country to country. I will just recall that when, in the process of setting up the African Competition Forum, a questionnaire was sent to all the countries of Africa in search of information about the state of competition enforcement, we received the following response from the Liberian government representative: ‘Thank you for your quick response to my request, and I want you to be informed that because of the fourteen years of war, mail coming and out delay. I have already posted the form to the address and it will take may be three weeks. I am from the legal section of the Ministry of Commerce & Industry of the Republic of Liberia presently serving as a legal Aid’.

None of this is intended to counsel despair. I can easily imagine that with a little good fortune, clever strategising and appropriate international support for the right strategies, the Armenian competition authority may well become a robust agency. That the Liberians, after 14 years of war, are even thinking about competition enforcement provides room for optimism. However, each country is going to have to follow an approach that will undoubtedly draw on a century of antitrust enforcement, but that is certainly going to have to be tailored to fit each peculiar set of circumstances.

Those determined to secure hard convergence may be tempted to conclude that Armenia and Liberia don’t count for much in the greater scheme of things. But each country has its particular issues, its own war stories, that defy hard convergence. And so too do most, if not all, countries, developed and underdeveloped. And very soon the exceptions will add up to a significant portion of world trade and investment and a large number of merger jurisdictions.

So, in summary, the governance of world affairs has to become more democratic at least in the sense that the few can no longer take decisions for the many. But, as is most graphically represented by the Doha WTO round and, dare I say, supported by some of the stories recounted here, conclaves of the many are not conducive to reaching binding agreements.

Enter then the world of networks, of informed divergence, of frank talking and patient persuasion, and, if lucky, of ‘soft convergence’. In the field of competition law, the ICN and the OECD are leading representatives of the value of, and the possibilities for, very different kinds of networks, but networks nevertheless. These are, I believe, the sorts of institutions and international arrangements that we should be strengthening and reproducing.

NOTES

1. Report of the Rapporteur (Professor Merit Janow), Ditchley Park meeting on global anti-trust, 2001.

2. Ibid.

3. Ibid.

4. Ibid.

5. Ibid.

7. Conclusion and a postscript

In the introductory paragraphs to this book I identified the ‘big story’ of the first decade of the implementation of the Competition Act as its success. That having been said – and restated – it’s extremely difficult to measure, with any degree of confidence, the quantitative impact of competition law enforcement and merger regulation. There are reasonably widely-accepted rules of thumb that purport to measure the impact of hardcore cartel conduct on price. It’s possible to undertake retrospective analyses of merger decisions. Some competition authorities, most notably the UK’s Office of Fair Trading, have purported to measure in considerable detail the quantitative economic impact of its activities, even its advocacy functions.

While the OFT’s measures of its impact involve some pretty heroic assumptions and reasoning, it’s important that the South African Competition Commission makes every effort to develop quantitative measures of its own impact. The Commission has in fact worked with the OFT to develop measures that it is beginning to use. Although I am sceptical of aspects of this empirical measurement – for example, I fear that it will inevitably understate what I think of as the institutional impact as well as the deterrent effect – if it does help counter a sometimes widely-held view that the enforcement activities of the competition authorities do not have a sustained impact on price levels, then it will have played a useful role.

The primary impact of robust competition enforcement, including merger regulation, is indirect in nature. It is concerned with the defence and promotion of institutions, indeed of the most pervasive and powerful institutions of the economy, namely markets. However, these critical institutions are not defended for their own sake: powerful evidence establishes that, the failures and shortcomings of markets notwithstanding, they do, given the right circumstances, underpin relatively efficient resource allocation and promote economic dynamism and opportunity. So we are measuring the impact of markets, the smooth functioning of which competition law and competition policy play an important role in ensuring. But there are many factors that impact on the quality of markets or, indeed, in many developing and transition economies, on their very development.

Loosely speaking, competition law acts on private restraints on the effective functioning of markets, while competition policy is focused on public restraints. It’s certainly possible to demonstrate the positive impact of relaxing the restraints imposed by market power or state power on important markets – the UK’s Department of Trade and Industry, in particular, has commissioned some interesting work on the impact of competition in selected markets, which demonstrates dramatic results.1 But, in general, the impact of competition law and policy is slow and long-term. It is concerned with ensuring that a series of diverse institutional arrangements function effectively, and its default view is that they will generally, although by no means always, function best when unrestrained by overbearing private or public power. So, I repeat, the impact of competition law and policy is, for the most part, diffuse and indirect and its modes of intervention – or what lawyers may term its ‘remedies’ – are often counterintuitive. For these reasons, it is famously difficult to create reliable interest-group support for competition law and policy. In turn, this is also why it does not easily lend itself to short-term electoral cycles and thus serve the objectives of ambitious politicians. And it is also why, like it or not, even as more rigorous empirical methodologies for measuring impact are developed and achieve widespread acceptance, we are probably going to have to be satisfied with using largely qualitative measures of the impact of enforcement and merger regulation.

This conclusion is underlined when we think of competition law as the foundational law of the economy. It establishes the key rules that effectively regulate critical aspects of the relationship between market power and the citizenry in much the same way as the Constitution itself regulates the relationship between state power and the citizenry. Just as competition law seeks to limit market power and regulate the manner in which it is used by those who possess it, so the Constitution regulates the use of state power and the manner in which it is used. And similar problems would confront those who sought to quantify the impact of these laws. How does one quantify the impact of a Bill of Rights that defends the rule of law – a critical feature of a successful market economy?

By the same token, it is extremely difficult to quantify the impact of enforcing the rules that make for the effective functioning of institutions like markets. It is implicitly assumed that the impact of an economic law would lend itself more easily to quantification than would a basic civic right. But competition law is, at its heart, an instrument that protects and elaborates property rights against overbearing market power and state power. And while it is widely held that secure property rights (the utilisation of which are subject to rules that prevent their abuse, competition rules principal among them) are a precondition for sustainable development, the empirical problems associated with measuring the precise impact of these rights and their associated rules are near intractable. There are a myriad factors that account for the performance of markets. It’s extremely difficult to isolate the impact of competition, even more so the impact of competition enforcement on the functioning of markets and, from there, on economic performance.

Drawing then on qualitative measures, there can be little gainsaying that the bodies responsible for the administration of the Competition Act, the Competition Commission and the Competition Tribunal have developed a reputation for a high degree of professionalism and independence in the deployment of their powers and in the quality of their decisions. They have earned the respect of important centres within government, of the business community, of the media, of the public, and of their international peers.

Arguably, then, the most important direct impact of the competition authorities has been in what I can best term the broad area of governance. They have embedded themselves as a central element in the business decision-making process. This they have done by holding to account important institutions whose practices have hitherto been invisible and certain of which have increasingly come to be accepted as contrary to acceptable business practices. Hardcore cartel conduct, in particular, has been conclusively demonstrated to be extremely widespread and longstanding and is now widely perceived by the public, by all the key interest groups, by the government, the media and the business community itself, to be contrary to acceptable business practice and deserving of robust sanction.

While this certainly doesn’t mean that cartel activity has ended, I would find it hard to believe that the diffusion of public knowledge of cartels and the increased awareness of their negative consequences have not made a major contribution to better-functioning markets and the associated benefits. In much the same way, we have, in line with our peers elsewhere, prohibited a small number of mergers and limited the market power to which they would have given rise. I also have no doubt that certain mergers that might otherwise have taken place have not been attempted because of the certain opposition of the competition authorities. Again, I am confident that this has contributed to effectively-functioning markets.

In short, in a relatively brief period, the activities of the South African competition authorities have instilled a society-wide view that may reasonably be characterised as a ‘competition culture’. Or, more accurately, we have now in South Africa a culture of respect for competition law. But this does not mean that we live and work in an environment in which competition is accepted as one of society’s fundamental social values. In particular, this more deep-seated support for competition has not impacted on the courts or on the custodians of industrial policy, with debilitating feedbacks into the practice of competition law itself.

I have outlined, at some length, the threat posed to competition enforcement by the courts as evidenced in the decision of the Competition Appeal Court and the Supreme Court of Appeal in Yara and Woodlands, respectively. Although in certain aspects the Competition Appeal Court’s Yara decision is even more restrictive of the Commission’s powers than Woodlands – namely the Court’s ruling that holds that the Commission is not entitled to amend a defective initiation – there would be little point in taking Yara on appeal to the Supreme Court of Appeal, which would doubtless simply confirm its Woodlands ruling. However, the Woodlands decision and subsequent decisions in Yara, South African Breweries and Senwes are effectively all based on what the Supreme Court of Appeal considered in Woodlands to be an injudicious and oppressive use by the Commission of its powers – in other words, for violating the constitutional rights of the accused. Hence, the Commission has elected to take Yara to the Constitutional Court. The other alternative that has been mooted is an amendment of the Act. However, should the Constitutional Court ultimately side with the Supreme Court of Appeal, then it is difficult to see how an amendment extending the Commission’s powers of initiation and investigation would pass constitutional muster.

But, in the face of the courts’ attack on the powers of the Commission, these are the short-term alternatives available to the Competition Commission: a Constitutional Court decision effectively overturning the Woodlands principles or an amendment to the Competition Act that would be able to withstand the inevitable constitutional challenges it would encounter.

However, whatever the outcome of a Constitutional Court challenge, there are longer-term responses that should be considered. And this goes to the question of instilling a competition culture, of instilling respect for competition on the benches of our courts. While I don’t intend traversing this ground again, the Woodlands and Yara decisions do have bearing on the question of competition culture in so far as they represent a failure to persuade the courts to treat competition as a foundational value. This is not to say that competition is sacrosanct or that it should be treated as such in either law or economics. Our own competition statute provides for a balancing of competition with other values and socio-economic objectives; and the concept of market failure is well established in economic theory and practice. But competition and the property rights that it protects and whose use it regulates are not uniquely constrained values. However, not even those values already formally enshrined in the Constitution are absolute. The exercise of these rights must also be balanced against other, equally weighty foundational social values. In one of the Competition Appeal Court’s Woodland judgments it cites an extract from a judgment of Justice Kriegler, a former judge of the Constitutional Court, which bears restating:

In any democratic criminal justice system there is a tension between, on the one hand, the public interest in bringing criminals to book and, on the other, the equally great public interest in ensuring that justice is manifestly done to all, even those suspected of conduct which would put them beyond the pale. To be sure a prominent feature of that tension is the universal and unceasing endeavour by international human rights bodies, enlightened legislatures and courts to prevent or curtail excessive zeal by state agencies in the prevention, investigation or prosecution of a crime. But none of this means sympathy for the crime and its perpetrators. Nor does it mean a predilection for technical niceties and ingenious legal stratagems. What the Constitution demands is that the accused be given a fair trial.2

I know that we are not (yet) dealing with alleged criminal offences, although the Supreme Court of Appeal’s casual reference to the criminal character of the Tribunal’s fining powers may well have influenced its decision in Woodlands. But the same balance confronts us: we want to avoid even the appearance of not extending justice to all, we want to curb excessive zeal by powerful state bodies (although there is reason for thinking about where precisely the balance of power actually lies as between the Competition Commission – not to mention consumers – and the firms whose conduct it is charged with supervising), but we also do not want those who may have perpetrated a serious offence to walk away scot free by seeking and finding refuge in ‘technical niceties and ingenious legal stratagems’. This surely is the balance that needs to be struck. However, I have no doubt that when called upon to decide competition matters, our courts accord the first of these elements due importance; but I am far from convinced – and the Woodlands decision is the clearest example of what I mean – that competition offences are correctly weighted despite their characterisation in a previous Supreme Court of Appeal case as a ‘canker’.

So exactly how cancerous do anticompetitive practices have to be in order to overcome ‘technical niceties and ingenious legal stratagems’? Surely, following Judge Kriegler’s argument, the court would have wanted to suspend its judgment until it was able to see whether or not the accused were indeed accorded a fair trial. Nipping in the bud an investigation where there is a strong suspicion of cartel conduct is tantamount to removing the cure because of a relatively minor adverse side-effect, thereby allowing the cancer to flourish. And bear in mind that when we talk of an unfair or abusive investigation we are talking about the investigation of the actions of a corporate citizen – no one is being tortured or deprived of liberty. At worst, at the very beginning of a process, one in which the company will, if eventually charged, enjoy all the rights accorded a litigant, a corporate officer is, in the presence of his legal representatives, being subjected to a range of questions that may stray beyond the black letter of the initiating document. ‘Ag shame’ (loosely translated as ‘Oh, pity on you’) would appear to be the appropriate South African response to this level of abuse, if abuse it ever was.

I recall, but never fully appreciated the reason for, the strong reaction when French President Nicolas Sarkozy – not someone who enjoys having his rampant nationalism and mercantilism limited by competition considerations – recently succeeded in removing a reference to ‘un-distorted competition’ from the European Treaty. The place of competition in foundational laws such as the European Treaty or the South African Constitution is not simply an important political statement. It also has powerful legal implications. So instead of trying to amend the Constitution by eliminating the Supreme Court of Appeal’s role in competition matters – a path favoured by some but likely to come to naught and, even if successful, to have little positive consequence – it may be preferable to open up the debate by arguing for a constitutional amendment that elevates the right to compete for the support of consumers to the level of a constitutional right. Our courts may find this harder to ignore than they do the current public outrage directed at cartel conduct, which, on at least three occasions, has, in the face of fancy legal footwork, gone unpunished.

However, arguably an even graver danger to robust enforcement than that posed by our courts stems from our economic policy-makers’ uncritical adherence to tired old formulations of industrial policy and, particularly, to the thoroughly anticompetitive modes of economic governance that generally accompany these policy choices.

I’ve long held that if the first decade of the competition authorities’ regime was characterised by action directed at private anticompetitive conduct, then the challenge of the second decade would be to take the fight against anticompetitive conduct to the state itself. And this for two principal reasons. First, the state is undoubtedly an important source of barriers to entry and anticompetitive conduct. Second, it is difficult to see competition law continuing to thrive in a thoroughly unsupportive policy and administrative environment.

The state’s role in erecting and maintaining entry barriers – or, at very least, according little priority to the requirement to dismantle these impediments – is well documented. A report by the Organisation for Economic Co-operation and Development (OECD) has exposed the increasing regulatory barriers that impede the formation of new business entry.3 The dismal underperformance of new investment in our mineral-rich economy, and this in the face of the most sustained commodities boom in living memory, is another stark reflection of state-erected barriers to entry. I readily recognise that in these and other areas there is evidence that the responsible policy-makers and officials share these concerns and are making considerable efforts to overcome these entry barriers. Where this is the case, the competition authorities should support state efforts by identifying entry barriers and advocating their elimination.

The Competition Commission has already established a solid track record in assisting the state in supporting pro-competition objectives in selected, important areas. For example, it has actively sought to align its strategic enforcement priorities with the pro-poor policies of the state and with the role of public infrastructure spend as a major instrument of economic policy. To this end it has focused its enforcement activities on food and health markets and on construction markets that are bedevilled by ubiquitous bid-rigging and other forms of cartel conduct. It has extended this beyond enforcement into proactive programmes designed to prevent bid-rigging, as opposed to merely apprehending existing cases of bid-rigging. A prime example of this is the cooperation between the Commission and the OECD in a major programme initiated by the Commission, which is directed at training public procurement officials in the evaluation of public tenders, with a particular eye to assisting these officials in recognising bid collusion. The profile of this sort of intervention should be raised, the better to enhance the reputation of the competition authorities for promoting competition through mechanisms other than enforcement and, thereby, to instil among public officials a culture of respect for competition.

There are other instances where the state is contemplating new and extensive regulation. The fraught area of health care is one. While the terms of this policy review are not yet clear, its outcome will significantly impact on the very nature of competition in major markets covering pharmaceutical production and retailing, hospital care, emergency services, health insurance, the rules of professional bodies, and the entire interface between the public and private provision of health care. The competition authorities have significant accumulated knowledge of the workings of these markets and should assume a major role in the design of more effective markets and regulations. There will undoubtedly be a variety of modalities for intervention by the competition authorities in these public sector initiatives. These may range from the Commission participating, where appropriate, in government task teams, through to undertaking independent investigations that will provide advice in an area of vital importance to those concerned with promoting competition and positive social outcomes, through to using, if at least the relevant amendment to the Competition Act is proclaimed, the powers to mount a fully-fledged market inquiry.

And then there are those areas where government is consciously promoting what appear to be ill-advised, anticompetitive policy initiatives and programmes. In my estimation the outstanding example of this centres on the holy grail of ‘beneficiation’. Although this issue has, at one time or another, surfaced in all of South Africa’s many commodity markets and that of their potential downstream ‘beneficiaries’, it has been pursued with most vigour and with the most substantial competition-distorting effects in the iron ore–steel–fabricated metals value chain. Essentially, government insists that the iron-ore producers and, in particular, Kumba, the largest of them, provide the steel producers, and Mittal in particular, with iron ore at heavily discounted prices. Second, it wants Mittal to pass through this benefit to metal fabricators in the form of a discounted steel price. Without going into the whole steel-pricing saga again, I just note that Mittal has always received iron ore at a discounted price but the government has never succeeded in persuading the steel monopoly to pass this on to its domestic customers. The reason why this has come into sharp focus again is that Kumba’s obligation to provide Mittal with a discount on iron ore is now threatened by a dispute over the ownership of mineral rights over a portion of Kumba’s largest mine.

From the perspective of the competition authorities, there are two aspects to this. First, and this is emphasised in our decision in Mittal, if government is determined to intervene in price-setting in various parts of the value chain, then unless there is a reasonably clear cause of action in competition law, it should not expect the competition authorities to take upon themselves the responsibility for price-setting unless, of course, it is willing to establish the statutory framework and the resources necessary to take on this massive additional responsibility. The Tribunal – as I’ve already extensively documented – thought that its remedy in Mittal would have both remained within the remit of competition law and depressed the steel price below current levels. However, this was rejected by the Competition Appeal Court. Unless the Commission or another complainant can bring another case and persuade the Tribunal to reverse its previous ruling, a ruling that would then have to be upheld by the superior courts and that will at best take several years to resolve, the government cannot hide behind the skirts of the competition authorities. By continuing to do so, all that it does is create thoroughly unrealistic expectations of what competition law can achieve and, so, disrespect for its actual achievements.

If, however, government believes that its social objectives cannot be achieved through market-based outcomes, then it should have the courage to use its powers to regulate, instead of constant vacillation and fruitless attempts to persuade market players to regulate themselves. Hence, if government firmly believes that health care markets should be regulated, then it should focus its energies and resources on developing a regulatory framework and institution, instead of trying, in vain, to toy with the creation of a complicated voluntary bargaining framework, which it hopes will result in the outcomes it desires. By the same token, if government is determined that the prospect of greater social returns dictate that iron ore and steel should be priced at sub-competitive levels, then it should take the bull by the horns and establish a sound regulatory framework, instead of engaging in a long, fruitless charade that alternatively attempts to jaw-jaw the steel monopoly into pricing below what it can achieve in the market or threatens the iron-ore producers into voluntarily discounting the price of its product.

However, if recent events are anything to go by, it appears that government’s (or at least the Economic Development Department’s) preferred course is to ensure that the application of competition law produces the outcomes it desires.

A recent outstanding example of ill-advised and poorly executed government intervention is in the area of direct foreign investment. As I’ll elaborate below, government clearly views foreign direct investment with great suspicion – or at least the policy-makers in the Department of Trade and Industry and the Economic Development Department do – so, again, it tries to have the competition authorities take its decisions for it. Or worse, it attempts to influence the competition and public interest outcomes that are the statutory responsibility of the competition authorities by conducting parallel and often conflicting negotiations with the firms before the competition authorities, thus returning us to the pre-1999 era of the smoke-filled room. This is exemplified by the recent Walmart and Kansai cases.

In short, we have policy-makers who appear to want to regulate key markets, but who are unwilling to assume responsibility for taking that step. And so the state looks to the competition authorities to impose the outcomes it desires. Time and again all that it establishes is that the competition authorities are not sectoral regulators. All that is achieved by attempting to use the competition authorities for this purpose is to undermine their independence, a key element in the respect that they have earned.

This does not mean that the competition authorities will always oppose regulation. But, by dint of their mandate, the competition authorities are duty-bound to examine the consequences of regulation and, if appropriate, to point out the costs of doing so. For example, in the case of iron ore and steel, government has not yet made out a solid case that supports the proposition that iron ore extracted from miners will be passed on by Mittal to the downstream fabricators. It has not even made out a persuasive case that the competitiveness of downstream metal fabrication would be materially advantaged by discounted steel prices. In this instance, the Commission should be investigating the case for regulation. If there proved to be a strong case in favour of regulation, then it should be investigating alternative modalities of regulation – the Competition Commission would want a form of regulation that simulated competitive outcomes as closely as possible. If the Commission could not establish a sound case for regulation, it would use its advocacy powers to oppose it. This would first take the form of advising government against such a step and, if this failed, then informing the public of the consequences of that step. For example, in the iron ore–steel-pricing debate, one argument suggests that because the other small steel producers do not use iron ore as their feedstock, depressing the price of iron ore would advantage Mittal relative to its few, small competitors, thus further inhibiting whatever slight prospect there is of a competitive steel market. It may also lead to the underexploitation of new, marginal iron-ore reserves.

Undoubtedly the most openly conflictual interaction between government policy-makers and the competition authorities has come over the recent amendments to the Competition Act, and then in recent attempted acquisitions by foreign firms of large South African firms.

I have dealt with the most significant amendment to the Act at length and will not go into this again. I would just emphasise that some 3 years after assenting to the bill, the President has still not proclaimed the date on which it will come into effect, and none of the extensive regulations required for the implementation of the amendments have been discussed, much less drafted. This may be due to simple incompetence, although I suspect that if the Minister of Economic Development had wanted the bill proclaimed then it would have been done. I can only speculate that, to some extent at least, he must share some of the competition authorities’ misgivings regarding the content of the amendments.

The recent ministerial interventions in two major mergers are, if anything, more sinister. In asserting its pre-eminence in amending the Act, the Department of Trade and Industry could at least lay claim to ‘ownership’ of the policy framework and hence to any amendments to the legislation. The dispute over the last set of amendments was never about the locus of final decision-making authority over amendments placed before Cabinet and Parliament, but rather about the bloody-minded reluctance of Department of Trade and Industry officials to take any meaningful account of the agencies responsible for the implementation of the Act.

However, in its attempts to influence the outcomes in two recent mergers as well as the remedies in Pioneer, the Ministry of Economic Development (the ministry now responsible for the competition policy framework), with some help from the Department of Trade and Industry, has sought to intervene directly in decisions that are the statutory province of the competition authorities. The mergers are those involving the acquisition of two South African firms, Massmart and Freeworld, by the US retailer Walmart, and the large Japanese paint manufacturer Kansai, respectively.4

Is there anything wrong with this? After all, as far as the merger review process is concerned, the Act specifically provides that the merging parties are required to notify the ministry (and the representatives of their employees) of an intended merger precisely in order to give them the opportunity to intervene. And I’ve already indicated how frustrated we were in the preceding 10 years at the apparent inability of the Department of Trade and Industry to participate in merger hearings, particularly when policy and regulatory issues were under consideration. The ministries have, in fact, attempted to present their intervention as representing a reconsideration of policy – the then-acting director-general of the Department of Trade and Industry explained the intervention in the Walmart transaction as an attempt to expand the definition of the public interest in the Competition Act.5

In so far as the claimed policy shift is concerned, while the competition authorities and the courts frequently make policy choices in the process of interpreting the Act, once a position is settled upon and consistently applied, policy-makers have to take care when they wish to change that settled policy lest they generate significant uncertainty and undermine the independence of the regulatory body and the courts. Policy doesn’t get changed on the hoof, with, as in these cases, retrospective consequences for a merger (a perfectly legitimate business decision) in which the merging parties were entitled to expect that their transactions would be assessed in a manner consistent with some 11 years of regulatory practice and jurisprudence. Indeed, once an approach to the interpretation of a statute has become accepted practice, in all but the most minor cases a change in that policy would require an amendment of the legislation.

And these are not minor changes in policy. After all, ‘public interest’ has a long and not-so-glorious lineage in the application of competition law, largely because of its vagueness. As already elaborated, a very specific policy decision was taken to delineate the definition of public interest precisely in order to limit the uncertainty created by the use of that potentially open-ended term. And the Competition Tribunal has attempted to honour that policy decision by limiting interventions in the name of public interest to those expressed policy intentions as reflected in the Act. The approach to public interest has been tweaked in order to respond to changed circumstances, as evidenced in Momentum and other recent decisions. However, if the policy-makers now want a significant expansion in that definition, then coherent, fair governance demands that government tables its intended change in policy, provides the opportunity for public discussion, amends the Act accordingly and then applies the new policy.

Nor is it easy to divine the desired policy change from the position taken by government in these two mergers. In the Kansai matter, government wanted conditions that sought to impose commitments to promote ‘industrialisation’ in the form of new investment and the location of R&D facilities. This would represent a change in the definition of public interest, although it is difficult to imagine how it would be drafted and applied.

In Kansai, the government also wanted to impose a condition that required the merged entity to introduce a BEE shareholder. This is already accounted for in the existing definition of public interest. In passing, it’s interesting to note that the target firm, Freeworld, a South African firm of long standing, with a number of BEE luminaries on its board as well as a number of business people with an established reputation for supporting transformation, had never introduced BEE shareholders into the company. The Japanese acquiring firm readily agreed to introduce BEE shareholders.

For the rest, in both the Kansai and Walmart transactions, the government wanted to impose employment-related conditions. Both acquiring firms have clearly indicated that there will be no merger-related employment losses. In fact, both are confident that employment in the merged entities is likely to increase. For 11 years the approach to employment has focused on direct employment loss, and both firms were prepared to accept conditions that effectively supported their contention that there would be no direct employment loss. However, what government seems most exercised about is the prospect that Walmart will increase the proportion of foreign procurement by the merged entity, thus indirectly impacting on employment in South African manufacturing. Accordingly, government wants conditions imposed that would place upper limits on the proportion of imported goods procured by the merged entity.

I have, in the merger chapter of this book, footnoted a reference to a transaction involving two clothing retailers, Edcon and Dawn Trading, in which Ebrahim Patel, then general secretary of the South African Clothing and Textile Workers Union and now Minister of Economic Development, wanted similar procurement-related conditions imposed on the merged clothing retailer. In that instance, the Tribunal rejected his arguments. We pointed out the strong, direct trade-off between producers and consumers implied by this sort of condition. In Edcon we also pointed out that a procurement condition along the lines sought could not be applied to a single firm. It represents a significant trade policy and should be treated as such and would be viewed by the international trade authorities as a strongly interventionist trade-related measure. Moreover, the condition, particularly when applied to a general retailer like Walmart, would be a nightmare to monitor. For example, the imported content in every item among the stores’ thousands of products would have to be measured and monitored.

So the changes sought by the Economic Development Department are significant. First, to add the promotion of ‘industrialisation’ to the public interest criteria almost certainly requires a legislative amendment. Were the merging parties to resist this, then neither the competition authorities nor the ministry would be entitled to impose these sorts of conditions. And, I should add parenthetically, not only is this bad at law, but it represents industrial policy of the worst sort. It is widely accepted that to apply industrial policy through a single targeted firm, be it in the form of a subsidy or, as in the imposition of the desired merger conditions, an effective tax, is the most retrogressive, discredited form of industrial policy, not least because of the unhealthy relationship that it sets up between business and government. Patel has publicly lauded Kansai because ‘it took to heart the government’s New Growth Path strategy’.6 In fact, to the extent that Kansai acceded to conditions that it had not already unilaterally decided to implement, all that it ‘took to heart’ was the gun pointed at its head.

Second, the substantive change that the Department of Economic Development’s position represents appears to discriminate against acquisitions of South African-owned firms by foreign-owned firms. All that the Kansai and Walmart transactions have in common is that they are foreign-owned. As already noted, we would not, by any stretch of the imagination, be the first country to apply special vetting mechanisms to foreign acquisitions. For example, Canada appears to have a review mechanism and has used it to block the acquisition by BHP Billiton of a large national potash producer, while the US has, in recent years, used and strengthened its review mechanism for foreign investment in the US after blocking an acquisition by Dubai Ports World of port management businesses in the United States.

A recent Financial Times op-ed article reviewing these arrangements proposed that to the extent that these review mechanisms are introduced they should be confined to national security questions (yoghurt?! cheese?!) and should be free from political interference. The authors specifically caution against a ‘national interest’ review standard, precisely the standard used in effect by the Economic Development Department in the Kansai and Massmart cases: ‘The risk with a “strategic” sector approach, such as that implemented by France, is that it in fact serves as a pretext for other objectives. No one could seriously debate a country’s right to promote domestic industrial growth. Yet it is one thing to utilise governmental tools to develop industries, but quite another to limit access to those industries. Allowing foreign investment tends to promote sectoral growth, not retard it’.7 The Financial Times article also stresses the importance of certainty and expeditious decision-making and accountability. In the US, for example, the decisions of its foreign review mechanism, while not subject to judicial review, must be reported to Congress.

The Economic Development Department’s intervention in the Kansai and Massmart cases violates every one of these standards. The standard used is a ‘national interest’ standard – but ‘national security’ is not remotely at issue; the intervention has created uncertainty for the two acquiring firms concerned as well as for future potential foreign investors; both interventions have dramatically lengthened the process of merger review; and the department is completely unaccountable for its interventions.

Essentially, the mode of intervention has been characterised by closed-door lobbying, actively encouraged by the Economic Development Department, which has attempted to cut deals surrounding the conditions for its support, with, it appears, the implicit undertaking that it would then use its position of claimed influence to ensure that the competition authorities supported these.

I’m not optimistic that a foreign investment review along the lines suggested would be introduced. Both the departments of Trade and Industry, and Economic Development have, not surprisingly, made it explicit that ‘national interest’ – expressed as industrialisation – rather than ‘national security’ is at the centre of their concerns. It’s a particularly haphazard approach to industrial policy because it’s not possible to dictate which foreign firm will decide to invest and thus develop a clear, strategically targeted industrial policy-based intervention. All that the intervening body – whether the Economic Development Department or the Department of Trade and Industry – can do is hang around, highwayman fashion, at the narrow pass and wait for someone to try to get through. It may be a poor straggler, or it may be a gold-laden stagecoach. This, and the strength of the desire to get through the pass, will determine the size of the highwayman’s pickings. So industrial policy is the objective, and opportunistically executed industrial policy at that.

The only certainty is that certainty is not an option from interventions of this sort. First, industrialisation is a broad and undefined objective and the nature of the intervention is to get whatever the opportunity presents. Hence, in the Walmart case the objective is to impose procurement conditions on a single firm in the retail trade and to secure a variety of good labour practices. In Kansai the Department of Trade and Industry proposed a number of competition-related conditions and then a variety of public interest conditions that ranged from employment conditions through BEE conditions and conditions aimed at promoting industrialisation, which, as I’ve pointed out, are not even provided for in the Act. In the Kansai case it seems that certain of the conditions were offered by Kansai itself, others were imposed by, or devised in, closed-door discussions with the Economic Development Department and the Department of Trade and Industry, while yet others were imposed at the instance of the Commission.8

Moreover, the favoured mode of intervention is predicated on a lack of transparency and accountability. In the Walmart case, the Commission has recommended an unconditional approval because, or so it stated, it was aware that the Economic Development Department was in negotiations with Walmart. Because these negotiations had not concluded, the Commission recommended an unconditional approval in the expectation that the Economic Development Department would place the conditions it recommended – and to which the firm may or may not have agreed – before the Tribunal. At that stage, the Commission would have an opportunity to make its views on any proposed conditions known. So the whole process is characterised by opacity rather than transparency.

In fact, my take on the Walmart fiasco is that, in that instance certainly, the last thing that the Economic Development Department wants is transparency. It would have preferred the Commission and ultimately the Tribunal to impose the decision that it, the department, wanted imposed. And this for precisely the same reasons that it’s so reluctant to regulate the steel industry or the health-care sector. Should it feel compelled to permit the Walmart transaction and drop the unattainable procurement condition, it would like to be able to say to the unions that the competition authorities are the decision-makers. Or should it feel compelled by union pressure or its own poorly-specified industrialisation objectives to prohibit the merger or to impose conditions that will cause Walmart to withdraw, it would then be able to hold out to Walmart and other potential foreign investors that it, the Economic Development Department, was hostage to an independent quasi-judicial process of competition review. At the same time, it would want its interest in the outcome to be known.

The big losers – apart, of course, from South Africa’s consumers and the country’s general economic prospects – are the competition authorities. Their decision-making powers and their independence have been attacked in the most destructive manner. Rather than having the courage or honesty of purpose to amend the legislation and assume clear responsibility for the merger decision, by leaving the process opaque the Ministry of Economic Development has in effect held out that it, rather than the competition authorities, is the arbiter of mergers, but has nevertheless left itself the ‘out’ to pass the buck to the competition authorities when necessary.

The media reported Minister Patel as saying that government had secured the sort of conditions that allowed the Kansai merger to go through.9 I would be somewhat comforted if I believed that this simply represented a minister wanting to claim centre stage for an outcome that he believed, rightly or not, to be positive. However, I think that his claim has a more damaging intent and outcome: he is effectively signalling that his office is open to approaches by merging firms, whereupon his office door will be closed and the lobbying and horse-trading will begin. Needless to say, this represents the antithesis of the independent and transparent process that has characterised the modus operandi of the competition authorities. It is, in fact, distressingly reminiscent of the smoke-filled-room process that was the hallmark of the pre-1999 competition regime.

Under these circumstances, the competition authorities should advocate for the establishment of a foreign investment review body that takes responsibility for vetting acquisitions by foreign firms out of the hands of the competition authorities. The foreign investment review body could then take its decision, impose the conditions that it saw fit, including a prohibition on the transaction, and then, should it permit the transaction to go ahead, leave it to the competition authority to do the statutory review of the competition issues and the legislated public interest conditions, within the framework imposed by the initial foreign investment review, and leave it to the minister to take responsibility for industrial and trade policy.

Of course, the really distressing prospect is that the interventions of the Economic Development Department may not ‘merely’ represent discrimination against foreign firms. They may, rather, represent the minister’s views on the appropriate locus for competition decisions and so may extend beyond foreign acquisitions or beyond mergers. Indeed, in the Pioneer matter the department has, as earlier elaborated, already intervened in the imposition of remedies that the Competition Commission thought fit to recommend be imposed on a firm found guilty of cartel conduct. This is potentially a short step away from intervening in the investigation or prosecution stage of the Commission’s enforcement activities. Why should the department not approach a firm charged with a section 4 or section 8 contravention and undertake to lean on the Commission or even the Tribunal in exchange for attaining some or other industrial policy objective?

Were this to occur, the competition authorities and the Ministry and Department of Economic Development would be in open conflict. Fortunately, both the Commission and the firms they regulate have the Act on their side, as well as the various statutory measures that promote administrative fairness, including the Constitution itself. The Economic Development Department would have inserted itself into the process of administrative decision-making in a manner clearly at odds with the requirement for fairness and transparency. In fact, I think that, should a firm placed in the same predicament as Kansai or Walmart be so minded, it may well find the courts willing to review the intervention of the Economic Development Department in the decision-making process of the mergers in question. Unfortunately, most firms place store, and understandably so, on a sound and friendly relationship with the state, either because they aspire to be good corporate citizens or, more likely, because they fear the consequence of an antagonistic relationship with the state. However, should the Economic Development Department continue with this manipulative and non-transparent form of intervention, it may well find a firm calculating that the cost of doing business had effectively been increased to such an extent as to warrant taking on the minister.

Does this all serve to change my mind regarding the insertion of public interest criteria into the Act and, in particular, extending the responsibility for decision-making on the application of these criteria to the competition authorities? On balance, I think not. The problem lies not with the public interest criteria or the locus of decision-making regarding these criteria. As the Kansai and Pioneer cases have shown, the Economic Development Department and the Department of Trade and Industry have not confined their interventions to the public interest matters as defined in the Act. And the Walmart case shows that the Economic Development Department is willing to stretch the interpretation of the public interest criteria to breaking point. Indeed, senior officials are on record confirming their intention to ‘expand’ the definition of public interest criteria, although they have not indicated any intention to amend the substantive or procedural provisions of the Act to give effect to this intended ‘expansion’.

So the problem, I repeat, is not with the public interest, but with a state department that has unilaterally, and without following the requisite statutory and administrative processes, decided to depart from a policy which has enshrined, in statute, the independence of the competition authorities.

Nor, I should emphasise, does the problem lie with the Economic Development Department or any other department of state intervening in the processes of the competition authority, whether at the investigatory or the adjudicative stage. I still firmly believe that these interventions would strengthen and assist these processes, provided they conducted themselves in an open and transparent manner. In the Mittal hearings, a senior Department of Trade and Industry official and a former director- general of the department gave evidence in the hearings. They were cross-examined, the media reported their contributions, and, in the process, they greatly assisted the Tribunal in arriving at its decision and raised the level of public discourse surrounding a very complex and important issue. This, rather than the procedures reminiscent of the previous competition regime, is the quality and character of intervention that should serve as a role model for the new regime.

In fact, when the Economic Development Department does finally get around to deciding what it is going to do with the amended Act, it may find that it – and I mean the state broadly rather than merely the competition authorities – has a powerful new instrument for informed policy-making in the power to initiate market inquiries. Of course, this power can only be used effectively by a state that is willing to hear, if not necessarily to accept, views that conflict with its own prima facie views, and is in turn willing to subject these to the scrutiny of the various interest groups and, more importantly, the public of consumers and citizens who have entrusted them with power.

So when, at the conference celebrating the tenth anniversary of the competition authorities, I predicted, in the presence of the Minister of Economic Development, that the next 10 years of the competition authorities’ life would be more closely involved than hitherto in the anticompetitive conduct of the state, I envisaged the Competition Commission using its advocacy powers and its market inquiry powers, bolstered by its reputation for fairness and independence, to advise a generally receptive state on how to reduce entry barriers and how to strengthen regulation where appropriate, or to dismantle it where it did not serve the interests of economic development.

While I was not so naive as to believe that this would always be a smooth and amiable process, I did not imagine that the competition authorities would, some 15 months later, be on the cusp of what looks like a long and hard battle with its policy department. But this is what inevitably happens to an administrative body that is not charged with advancing the interests of a selected group of the powerful, but is rather mandated to advance, in a manner that is fair, the interests of the broad swathe of consumers, most of whom have very little voice at all.

So, it was a dark and stormy night. The dangers lurking out there have shifted, even abated somewhat, but there are still dangers aplenty. No one relishes or invites a conflict with powerful interests, be they large monopolies or an over-zealous state. But if it must be engaged, then with careful strategising, confidence in a well-deserved public reputation, a willingness to listen and a good Constitution, it is possible to envisage that disagreement will lead to, in fact may be responsible for, constructive outcomes.

WALMART: A POSTSCRIPT

Much of significance has recently taken place in the world of South African antitrust. One of the least significant consequences of this dynamism is knowing where to place the final full stop on this effort. Questions raised here will have been answered by the time this book becomes available. Not least are the matters currently before the Constitutional Court, the decisions of which will dramatically impact upon the investigatory powers and practices of the Commission. I’d be inclined to put my money on the Commission to win back its investigatory powers. But I’m not a betting man, especially where court decisions are concerned, and so the Constitutional Court’s views will be debated by others, elsewhere.

However, the Competition Appeal Court has laboured long over the appeal against, and review of, the Tribunal’s decision in Walmart. It handed down its decision a few weeks ago, in March 2012.10 For the most part the decision is clear and predictable. The vexatious review application filed by the three ministers has been denied by the Court. Reading the Court’s decision rekindled all the disquiet I’ve felt for the destructive role that the ministers have played in this saga. That three cabinet ministers could take a decision of an administrative tribunal on review, alleging that they have not been accorded a fair hearing, is, on any reading, questionable. It has always been my understanding that the courts are reluctant to adjudicate intra-governmental disputes, although a judicial review may fall into territory that the courts cannot but entertain. All parties that participate in an administrative process are entitled to be treated fairly and I presume that this extends to members of the executive.

However, the grounds of review clearly reveal that it has nothing to do with fairness, and everything to do with the wilful obstruction of a statutorily prescribed investigative and adjudicative process. I’ve already suggested that the ministers’ central objective is to transform the decision-making process into a bargaining process and their obstruction is an attempt to gain advantage, essentially through regulatory hold-up, in this transformed process.

The ministers have firstly argued that the Tribunal should have granted them more extensive discovery. This is not an uncommon gripe, although the extent of discovery demanded by the ministers is breathtaking. However, the other ground for review – that the schedule of hearings determined by the Tribunal was somehow prejudicial to the ministers’ ‘right’ to a fair hearing – plumbs new depths in bloody-minded obstructiveness. And, of course, should it have succeeded – a long shot if ever there was one – it would have severely impacted on the Tribunal’s ability to govern its own processes, particularly with a view to promoting expedition in its merger regulation function. The Commission, quite correctly, opposed the review on this ground.

The appeal mounted by the trade union was partly upheld. The argument for imposing a procurement condition on the merged entity was rejected. Some of the employment conditions that the union wanted imposed and that the Tribunal had only partially acceded to were upheld by the Court.

However, the most significant finding by the Court concerned the Tribunal’s ‘investment condition’. This is the R100 million fund, initially proposed by the merging parties, that would be deployed to strengthen the competitiveness of prospective South African suppliers to the merged entity and to the Walmart global supply chain. The Court was of the view that the Tribunal erred in simply accepting the merging parties’ suggestion that the fund should total R100 million and it held that the Tribunal should have been more prescriptive in determining the uses to which the fund should be put.

In order to plug these perceived gaps in the Tribunal decision, the Court has imposed a process that is intended to impart greater rationality to the investment condition. It has ordered the merging parties to commission, at their cost (why the merging parties should have to pay the government’s costs is entirely beyond me), a study to be undertaken by three experts, with the union, the government and the merging parties each entitled to appoint one of these experts. The results of the study are to be submitted to the Court within 4 months of its judgment. The three parties will also be entitled to submit responding affidavits to the Court that will, on the basis of the report and the affidavits, impose a more reasoned and elaborated version of the Tribunal’s condition.

We’re in for some interesting times. The Court clearly contemplates that the three experts will submit a single study – its order refers to ‘a study’ and ‘the study’. It’s possible – though unlikely – for three genuinely independent industrial policy and supply chain experts to agree on a single report. However, most allegedly independent experts I’ve encountered are simply advocates of their own client’s cause, which makes it extremely unlikely that a single report will emerge from the process. And then there will be the affidavits of the three parties to consider.

At the end of this process, the Court will probably be confronted by several highly contested studies. And for all the science that the experts will invoke, the differences will inevitably centre on divergent distributional interests rather than on dispassionate economic expertise, not to mention legal rights. It’s not clear whether those appointing the experts will then be entitled to appeal the decision that the Court ultimately makes.

In this process, the Court may begin to understand why antitrust adjudicators shy away from elaborate court-supervised remedies and why the Tribunal preferred the high-level remedy that it initially imposed.

NOTES

1. Stephen Davies, Heather Coles, Matthew Olczak, Christopher Pike and Christopher Wilson, ‘The benefits from competition: Some illustrative UK cases’ (Department of Trade and Industry Economics Papers No. 9, July 2004).

2. Key v Attorney-General Cape Provincial Division and Another (1996 (4) SA 187 (CC)).

3. OECD, Economic Assessment of South Africa 2008.

4. I should declare that I acted as an adviser to the Kansai attorneys in this matter.

5. Business Day, 5 April 2011.

6. Business Report, 21 April 2011.

7. Financial Times, 21 April 2011.

8. Kansai agreed to divest of its newly acquired auto coatings plant and to reinvest in a new auto coatings plant. However, it reserved its right to have this condition – apparently imposed on competition grounds – reviewed by the Tribunal. The Commission appears to have then reconsidered this condition and decided that it would not oppose Kansai’s application to have it removed.

9. Business Report, 21 April 2011. This was before Kansai had the auto coatings divestiture condition overturned by the Tribunal.

10. 110CAC/Jul11; 111CAC/Jun11.

Index

abuse of dominance 49–53, 73, 130–33, 142–4, 187–9, 266

exclusionary abuses 144–51, 159–61

British American Tobacco (BAT) case 142, 146, 150, 152–9

South African Airways (SAA) case 142, 143, 146, 147, 151–2, 157, 159

exploitative abuses 143

ArcelorMittal case 161–78

excessive pricing 161

Nationwide Poles case 178–87

resale price maintenance (RPM) 132, 134–5

Federal Mogul case 132, 135–42

Adcock Ingram 20, 202

AECI 20, 22, 106

African National Congress (ANC) 7, 8, 228

African Oxygen 113

Afrox 82, 83, 84, 113

agriculture 48

Altech 85

Amalgamated Beverage Industries (ABI) 116

AME 85

American Society of Composers, Authors and Publishers (Ascap) 199–200

Anglo American Corporation 2, 5, 6

Ansac 53, 197–202, 244

apartheid system 4, 6, 8, 16

ArcelorMittal 49, 131, 133, 142, 146, 149, 150, 187, 268–9, 270, 278

excessive pricing 161–78

Armenia 73, 258–9, 260

Ashton Canning 119, 120

Asmal, Kader 40

Avalon Cinemas 21–2

Aveng Group 202

AZ Friction Products 136

Baldwin Steel 107

banking sector 55–7, 59

Banks Act 56

BCX 80, 95

Bellamy, Christopher 230

beneficiation 268

BHP Billiton 274

Bidco 113

black economic empowerment (BEE)

public interest test and impact of mergers 109, 112–17, 272

Black Sash 193

Bonakele, Thembinkosi 84

Botswana Soda Ash (Botash) 198, 199

Brazil 236, 252

Brimstone 113

British American Tobacco (BAT) 132

abuse of dominant position 142, 146, 150, 152–9

Broadcast Music Inc (BMI) 199–200

Brooks, Pierre 14, 20

Business Day (newspaper) 15

Canada 104, 110–11, 243, 274

Carrim, Yasmin 90

cartels 11, 16, 52, 53, 73, 191–201, 263, 266

counter-attack 217–26

criminalisation 209–16

‘Pandora’s Box’ metaphor 201–209

Caxton 84–5

cement 54

Chalmers, Robyn 18

China 243

Cilliers, Fanie 153

Clover 218, 219

Cockrell, Alfred 101

Coleus Packaging 116

collegiality 90

Comair 202

Communications Workers Union (CWU) 119

Competition Act (1998) 2, 3, 5, 12, 21, 22, 24, 263, 268, 272

abuse of dominant position and 131, 133, 134, 142–4, 147, 150, 152, 159–60, 161, 164, 172, 174, 175, 178–86, 188

amendments 270–71

cartels 191–5, 200, 201, 208, 210, 211, 213–20, 222–6

institutional structure 41–3

building institutional culture 43–7

first period 48–60

see also individual institutions

merger review 75, 77, 78, 91, 104, 108, 111, 113, 121, 126, 127

National Economic Development and Labour Council (Nedlac) process and 30–39

Parliamentary process 39–40

policy document 25–30

Competition Appeal Court 42, 43, 61–8, 263, 280, 281

abuse of dominant position and 133, 152, 170, 173, 175, 177, 178, 179, 186, 187, 189

resale price maintenance (RPM) 140

cartels and 207, 211, 213, 217, 218, 219, 221, 222, 224, 225

merger review 77, 79, 80, 85, 86, 92–3, 97, 101, 102, 103, 125

Competition Board 12–23, 38, 41, 45

Competition Commission 2, 23, 41, 42, 62, 263, 265, 267, 269, 270, 275, 276, 277, 279, 280

abuse of dominant position and 132, 149, 188

resale price maintenance (RPM) 138

cartels and 192, 193, 197, 202, 205–208, 212, 213, 215, 217, 219–23, 225

first period 48–60

institutional culture 45–6

measurement of impact 261

merger review 70, 74–5, 79, 80, 84, 94, 114, 117, 118, 122, 124

staff members 4

competition culture 264

competition institutional structure 41–3

building institutional culture 43–7

first period 48–60

see also individual institutions

competition policy 1, 8, 10–11, 16, 262, 272

policy document prior to Competition Act (1998) 25–30

Competition Tribunal 2, 9, 23, 28, 41, 42, 61, 62, 230, 263, 268–9, 272, 276, 280, 281

abuse of dominant position and 132, 146, 173, 174, 176, 180

cartels and 192, 199, 202, 204–207, 211, 213, 219, 220, 221, 223

first period 48–60

institutional culture 46–7

merger review 70, 78, 79, 80, 82, 85, 97, 101, 102, 103, 111, 114, 118–22, 124, 125, 126

staff members 4

confidentiality in merger review process 81–2

Congress of South African Trade Unions (Cosatu) 193

Constitution of South Africa 52, 262, 266

Constitutional Court 86, 211, 220, 223, 226, 264, 279

constitutional law 3

cooperatives 48

coordination failures 10

corporate structure 5–7, 8, 10, 14

Creamer, Kenneth 33–4

criminal offences 265

cartels 209–16

Crown Cork 116

culture

competition culture 264

institutional 43–7

Davies, Rob 122

Davis, Judge 66

Dawn Trading 273

De Beers 34, 39, 198

de Villiers, Jannie 203

delays 51

Dennis, Nick 191, 193

Department of Economic Development 207, 208, 209, 269, 271, 273, 274, 275, 277, 278

Department of Finance 208

Department of Trade and Industry 20–21, 24–5, 122, 126, 210, 213, 269, 271, 275

discriminatory pricing 58

Distell 116, 141, 158

Ditchley Park 235

dominant position, abuse of see abuse of dominance

Edcon 273

efficiency 27

merger review process and 104–108

El Salvador 247

electoral cycles 262

Ellerine, Eric 72

Ellerines 9, 72, 93

Elliot, David 230

employment

public interest test and employment impact of mergers 109, 117–27, 272–3

Employment Equity Act 116

enforcement of competition law 2, 20, 133, 261, 262–3

international 228–33, 257–60

World Trade Organization (WTO) 233–5

see also International Competition Network

Engen 80, 83, 84, 96, 112–13, 141

entry barriers 267

Erasmus, Koos 136, 137

Erwin, Alec 21, 22, 23, 25, 40, 122

European Union (EU) 194, 228, 231–4, 266

Ewing, Ky 236

exclusionary abuses 144–51, 159–61

British American Tobacco (BAT) case 142, 146, 150, 152–9

South African Airways (SAA) case 142, 143, 146, 147, 151–2, 157, 159

exploitative abuses 143

excessive pricing 161

ArcelorMittal case 161–78

Nationwide Poles case 178–87

Federal Mogul 132, 135–42, 205

Fels, Allan 236

Financial Times 274

Finckenstein, Konrad von 241

First, Harry 230

Foodcorp 202

Foot, Jim 179, 180, 185

Fordham Conference 230, 239

foreign direct investment 269, 274, 275, 277

Fourie, Frederick 25

Fox, Eleanor 39, 229, 230

France 110, 274

Freedom Charter 7, 8

Freeworld 271

Gauntlett, Jeremy 87, 193

Gear policy framework 31

General Agreement on Trade and Tariffs (GATT) 234

Germany 38, 110, 228

Global Competition Initiative (GCI) 238, 239

Global Forum for Competition and Trade Policy 235

Goldfields 52, 77, 78, 86, 88

Harmony 52, 77, 78, 86, 88, 132, 162, 163, 167, 169, 172, 173, 177, 178

Hawk, Barry 236

Hazlett, Thomas 228

health insurance 96–103

High Court 43, 62, 86

Highveld Steel 116

Hirsch, Alan 32

hostile mergers 85–8

Hungary 236

Independent Communications Authority of South Africa (Icasa) 39

India 197, 234, 243

Industrial Development Corporation 206, 208

industrial policy 21, 28

Industrial Strategy Project 27

information asymmetries 10

innovation 49, 166–7

institutional culture 43–7

insurance

health insurance 96–103

interim relief 50–51

International Bar Association Global Forum for Competition and Trade Policy 235

International Competition Network (ICN) 199, 230

broadening ICN agenda 240–45

building virtual network 245–51

opening up to new competition regimes 236–40

origins 235

value of 251–7

International Competition Policy Advisory Committee (ICPAC) 237, 241

international enforcement of competition law 228–33, 257–60

World Trade Organization (WTO) 233–5

see also International Competition Network

International Health Care Distributors 52

international trade policy 21

interpretation issues 63

Iscor 49, 162

Italy 110

James, Charles 241

Janow, Merit 230, 237

Japan 249

JD Group 9, 72, 93, 94

Jenny, Frédéric 230, 233, 244, 255

Joel, Solly 237

Johnson, Lyndon 162

JTI 152–3

Judicial Service Commission 63

Kansai 270–78

Katz, Michael 18, 58

Khaya FM 85

Klein, Joel 237

Knight, Phil 32, 58

Kovacic, Bill 230, 258, 259

Kriegler, Justice 265, 266

Kumba 268

Labour Relations Act 32, 118

Langa, Pius 52

Leegin 134–5, 137–40, 149

Liberia 259, 260

Liberty Life 5

London Agreement 17

McGregor, Robin 14, 20

Macsteel Holdings 172

Maduna, Penuell 112

Maintenance and Promotion of Competition Act (1979) 17

Malaysia 234

Malherbe, Stephan 33, 34, 35, 37, 180

Mandela, Nelson 8, 228

Manoim, Norman 32, 44, 58, 77, 78, 90, 91, 107, 140, 217, 236

Manuel, Trevor 1, 2, 25, 32

market power 10, 12, 164–5

Marketing and Agricultural Products Act 48

Massmart 123, 126, 271, 274

Mbeki, Thabo 163, 210, 214

Mboweni, Tito 32, 228

MCG Ltd 116, 117

media 55

mergers and 80, 81

Mediclinic 82, 83, 84, 113, 114

Medicross 96

merger regulation 11, 14, 19–20, 22, 57, 70–74, 261

review process 72, 74–6, 79–81, 271

change of control 76–9

confidentiality 81–2

efficiencies 104–108

hostile mergers 85–8

public interest test 99, 100, 104, 109–28, 272

substantial lessening of competition 89–103

substantive review 88–128

third-party interventions 82–5

Mexico 236

Midas 136, 137

Milkwood 221

Mittal, Lakshmi 162

Moerane, Marumo 91

Momentum 120, 122, 272

monopolies 167, 176

mergers creating 107, 108

regulation of 10

see also abuse of dominance

Monti, Mario 247

Moosa, Moosa 21

Motaung, Lerato 13

Motlanthe, Kgalema 214

Mouton Commission 17

Mpahlwa, Mandisi 122, 210

MTN 17, 85, 196

Murray & Roberts 202

Mvelaphanda 113

Naidoo, Jayendra 32

Napp Pharmaceutical 168

Naspers 84–5

Natal Wholesale Chemists 51

National Consumer Forum 193

National Economic Development and Labour Council (Nedlac) 14, 24, 30–31

Competition Act (1998) and 30–39

National Prosecuting Authority 212

national security 274

mergers and 110

nationalisation 7, 8, 9

Nationwide Airlines 152

Nationwide Poles 66, 133, 142, 178–87, 188

Naudé, Stef 15, 33

Nedbank 118

Nedcor 55–6, 59

Netcare 82, 83, 113

Netherlands 39

Netstar 133, 225, 226

non-governmental advisers (NGAs) 246–7, 250

Nu Metro 22

Old Mutual Group 5, 6, 56

Omnia 217, 218

Oppenheimer, Nicholas 34

Oppenheimer family 5, 34

Organisation for Economic Co-operation and Development (OECD) 235, 236, 238, 241, 243, 256–7, 267

ownership of conglomerates 5–7, 8, 10, 12, 14

Parliament

Competition Act (1998) and 39–40

Patel, Ebrahim 122, 207, 273, 276

Pattison, Grant 126

Pharmacare 20

pharmaceuticals 22, 51, 53, 210, 268

Philips South Africa 16–17

Pick n Pay 39

Pioneer Foods 196, 202–207, 277, 278

Pitofsky, Bob 229

Pittman, Russell 229

Posner, Richard