Brief Summary

Competition Law — Abuse of Dominance — Exclusive Supply Agreements — Competition Tribunal determining whether Uniplate Group (Pty) Ltd abused its dominant position by requiring customers to purchase number plate blanks exclusively from it when acquiring embossing machines — Tribunal finding that such agreements contravened sections 8(d)(i) and 8(d)(iii) of the Competition Act — Uniplate's conduct deemed to have foreclosed the market and deprived competitors and customers of choice.

Comprehensive Summary

Summary of Judgment

1. Introduction

This matter was a prohibited practice referral in the Competition Tribunal concerning an alleged abuse of dominance by way of exclusive dealing. The proceedings were brought by the Competition Commission of South Africa (the complainant and referring party) against Uniplate Group (Pty) Ltd (the respondent).

The dispute arose from complaints lodged with the Commission by (i) New Number Plates Requisites CC (NNPR), a competitor of Uniplate, and (ii) JJ Plates and Signs CC (JJ Plates), a customer (an embosser) that had contracted with a Uniplate distributor. The Commission consolidated the complaints, investigated, and referred the complaint to the Tribunal on 27 November 2015 in terms of section 50(1) of the Competition Act 89 of 1998.

The hearing ran over multiple sittings (November 2017, December 2017, and May 2018), and the Tribunal issued its order and reasons on 27 June 2019. Although the Commission’s referral initially sought a declarator, a voiding order, an interdict, and an administrative penalty, the Tribunal recorded that during final argument the Commission appeared to abandon the interdict and voiding relief. The case therefore proceeded to determination primarily on (i) a declaratory finding of contravention and (ii) an administrative penalty.

The general subject-matter of the dispute was the legality, under the abuse of dominance provisions of the Competition Act, of Uniplate’s business model in which customers who acquired a Uniplate embossing machine were contractually obliged, through exclusive supply agreements, to purchase all their number plate blank requirements (and certain related consumables) exclusively from Uniplate for long periods.

2. Material Facts

The Tribunal situated the dispute in the value chain for number plates, comprising manufacturers of embossing machines, manufacturers of number plate blanks, embossers (who produce finished number plates from blanks using machines), and end users. It was accepted that embossers are typically small businesses with limited access to capital, and that there are approximately 1000 registered embossers operating in regional and local markets.

It was common cause that there are different types of embossing machines and blanks (Acrylic, Type A, Type B), and that machines are generally compatible with a corresponding blank type. The Tribunal recorded evidence that although a newer machine could theoretically emboss different plate types, substitution between Type A and Type B was practically limited due to significant adjustments and costs, and that machines were generally sold for their relevant type use.

Uniplate and NNPR were identified as the only two manufacturers of embossing machines, and also as the only two firms manufacturing both embossing machines and blanks. Other firms (including Arga Plates & Signs (Pty) Ltd and Naicker Toolmaker and Metal Pressing) operated in the blanks market, but not in machine manufacture (at least during the relevant period as described by the Tribunal). The Tribunal also noted the role of Uniplate’s distributors, Teqplate (Pty) Ltd and Baleka Plates (Pty) Ltd, which exclusively on-sold Uniplate machines and blanks.

A central factual premise relied upon by the Tribunal was that “mixing”—using one firm’s machine to emboss another firm’s blanks—was technically possible and occurred in practice. The Tribunal relied on the evidence of the witnesses (including Uniplate’s managing director) that there was no technical impediment to mixing, and on a letter from the South African Bureau of Standards indicating that SANS 1116 does not prevent mixing and that accreditation focuses on the output (the compliant number plate) rather than matching machine and blank provenance.

The material contractual facts concerned Uniplate’s agreements with embossers, including cash sale (including instalment sales), rental, and loan agreements. A total of 514 contracts covering the period 1994 to 2014 were produced during the Commission’s investigation. The Tribunal relied on the Commission’s analysis that 99% of Uniplate’s contracts over the period contained an exclusivity clause requiring customers, for as long as the agreement subsisted, to purchase solely and exclusively from Uniplate all their requirements for blanks and related inputs, subject only to a limited exception where Uniplate could not supply and where written permission was given.

The Tribunal treated as significant that the exclusivity obligation extended beyond blanks compatible with the machine purchased, requiring purchase of all blank types from Uniplate regardless of the type of Uniplate machine acquired. The Tribunal also relied on evidence that Uniplate described its own model as a “bait and hook” system, where the machine is the bait and the supply contract for blanks is the hook to secure ongoing sales.

Duration and exit constraints were central factual features. The Tribunal accepted the Commission’s evidence that a substantial majority of contracts endured for 120 months (10 years), that many contracts lacked termination provisions (particularly older agreements), and that many provided for automatic renewal for further lengthy periods. Although Uniplate asserted that customers could terminate in practice (including via buy-back mechanisms), the Tribunal relied on evidence that Uniplate could not provide figures of actual terminations or buy-backs during the relevant period, that a “spreadsheet” promised to the Tribunal was not provided, and that testimony suggested cancellations were rare.

The Tribunal further relied on the evidence of JJ Plates (through its owner, de Lange) that attempts to cancel were refused and that the distributor’s response was that the customer was bound for the ten-year term. The Tribunal accepted that the customer ceased purchasing from alternative suppliers after receiving threatening letters, and it relied on documentary evidence (including notices and letters) indicating that Uniplate and/or its distributors threatened legal action to enforce exclusivity against customers who purchased rival blanks.

An additional factual feature relied on by the Tribunal was that, even in cash sales where ownership of the machine passed to the customer, Uniplate still imposed exclusivity obligations. The Tribunal also relied on evidence relating to dies (consumables) which wear out sooner than machines, and on testimony that purchasing replacement dies could require the customer to sign up for a further 120 months of exclusivity, effectively prolonging lock-in beyond the initial term.

In relation to market outcomes and competitive conditions, the Tribunal relied on evidence that exclusivity reduced the contestable demand available to rivals and that staggered contract end-dates further limited rivals’ ability to secure scale. While market share evidence was debated, the Tribunal treated it as supportive in context of other evidence, noting Uniplate’s enduring high share in blanks and the relative stability of NNPR’s share in a lower band.

The Tribunal also relied on evidence of discouraged or failed entry, including evidence that certain international firms explored entry but did not sustain it. The Tribunal accepted that there was at least evidence of failed entry and discouragement, while noting limits on direct proof in relation to some entrants.

3. Legal Issues

The central legal questions for determination were whether Uniplate, accepted to be dominant for purposes of section 8, engaged in an exclusionary act prohibited by section 8(d)(i) of the Competition Act by requiring or inducing customers not to deal with competitors, through exclusive supply provisions tied to the acquisition of embossing machines, and whether that conduct had the requisite anti-competitive effect.

A further set of issues concerned the analytical framing of the relevant market(s). The dispute involved an assessment of whether the relevant product market was best understood as two markets—one for embossing machines (a primary product) and one for blanks (a secondary product)—or as a single integrated “systems market”. Although dominance was not ultimately contested as determinative, the market definition question remained material to the Tribunal’s assessment of the competitive theory of harm and the plausibility of foreclosure.

The issues were primarily ones of the application of law to fact within an effects-based statutory test. The Tribunal was required to make evaluative determinations as to whether the exclusive dealing was substantially significant in foreclosing the market (or otherwise harmful to consumer welfare), and whether any claimed technological, efficiency, or other pro-competitive gains outweighed anti-competitive effects once the Commission had discharged its effects burden.

If a contravention were found, the Tribunal was also required to determine an appropriate administrative penalty under sections 58 and 59, applying the statutory factors in section 59(3) and the Tribunal’s penalty methodology in prior decisions.

4. Court’s Reasoning

The Tribunal first addressed the market definition dispute, as it affected the assessment of the nature of the conduct and foreclosure. It rejected Uniplate’s contention that the relevant market was a single “embossing systems” market and found instead that there was a primary market for embossing machines and a secondary market for number plate blanks. The Tribunal’s reasoning relied on (i) the accepted technical possibility of mixing and evidence of mixing occurring without SABS accreditation being revoked, (ii) contractual terms that contemplated alternative blanks (albeit with written permission), (iii) the existence of blanks-only suppliers, (iv) evidence suggesting a secondary market for machines (including buy-back/first refusal provisions and the possibility of resale), and (v) the commercial logic of the “bait and hook” model itself as reflecting two distinct sources of revenue (low-volume machines and high-volume blanks).

Having accepted that Uniplate was dominant (as agreed by the experts irrespective of market definition), the Tribunal assessed the conduct under section 8(d)(i). It characterised the Commission’s theory of harm as foreclosure, namely that Uniplate leveraged its position in machines to tie up demand in blanks by requiring customers not to deal with blank competitors.

The Tribunal treated section 8(d)(i) as effects-based, and applied the test articulated in the Competition Tribunal’s decision in Competition Commission v South African Airways (18/CR/Mar01) and confirmed by the Competition Appeal Court in South African Airways (Pty) Limited v Comair Limited and Another (92/CAC/Mar10). It adopted the approach that anti-competitive effect can be shown either by evidence of actual harm to consumer welfare or, alternatively, by showing that the exclusionary act is substantial or significant in foreclosing the market to rivals, permitting an inference of anti-competitive effect. It also relied on subsequent Tribunal decisions indicating that complete foreclosure or exit is not required and that impeding rivals’ ability to expand may suffice.

Applying these principles to the facts, the Tribunal focused on the breadth and duration of Uniplate’s exclusivity obligations. It accepted that the exclusivity clause was pervasive, that it applied to all blanks and embossing-related inputs irrespective of machine type purchased, and that it operated across contract types, including cash sales. It found that contract durations were excessively long in practice, given the prevalence of ten-year terms, automatic renewal features, lack of express termination clauses in many agreements, and the evidential weakness in Uniplate’s claim that customers could freely exit.

The Tribunal also treated as probative the evidence of aggressive enforcement, including threats of legal action and notices warning embossers against purchasing from NNPR or others. While acknowledging that dominant firms may legitimately enforce contractual rights, it considered the evidence sufficient to show that the enforcement stance constituted harassment such that customers desisted from dealing with rivals. The Tribunal further treated the evidence on dies and contract extensions as reinforcing lock-in, particularly where customers needed replacement dies before the initial exclusivity period elapsed and were required to sign further long exclusivity undertakings.

On foreclosure, the Tribunal concluded that Uniplate’s contracts increased rivals’ costs and impeded competition. It reasoned that competitors seeking to sell blanks were effectively forced also to supply machines to compete for locked-in customers, thereby raising barriers to entry. It accepted that rivals in blanks-only remained small and that new entry was discouraged in a market where demand was tied up for long periods and contract end-dates were staggered and uncertain. It found that customers were harmed through reduced choice and higher prices, as suggested by evidence that alternative blanks could be cheaper and that customers were constrained from switching.

Once the Tribunal found the Commission had shown likely significant foreclosure, it addressed Uniplate’s efficiency defences under the statutory framework, emphasising that the evidential onus to justify the exclusionary conduct lay with Uniplate once anti-competitive effect was established.

Uniplate advanced three efficiencies: recoupment of R&D and investment costs through subsidised machine pricing; prevention of free riding and protection of quality/reputation; and cross-subsidising maintenance services. The Tribunal rejected these justifications on the evidence.

On recoupment, it found that Uniplate did not substantiate that machines were supplied below cost or genuinely subsidised in the manner claimed. It relied on evidence from Uniplate’s own prior submissions to the Commission indicating only limited profit, if any, on equipment sales and on concessions in cross-examination that selling prices were higher than costs in relevant instances. It found Uniplate’s supporting margin analysis unhelpful because it was outside the complaint period, covered a short period, and lacked transparency on underlying costs and pricing. It also found that the recoupment rationale did not explain why cash sale customers, who owned the machine, were bound to extensive exclusivity, nor why replacement dies should trigger further lengthy exclusivity.

On free riding and quality control, the Tribunal found the logic inconsistent with Uniplate’s own conduct. It reasoned that if Uniplate’s concern were truly preventing rivals’ blanks from being used on Uniplate machines, it was inconsistent to require customers to buy all types of blanks irrespective of machine type, and inconsistent with the fact that mixing was permissible and occurred. It further relied on evidence that SABS standards did not prevent mixing, that accreditation focused on the final number plate, and that JJ Plates’ certification was not revoked for mixing, undermining the contention that exclusivity was necessary to ensure compliance and protect reputation.

On maintenance, the Tribunal accepted the Commission’s point that maintenance is contractible and could be priced and specified, and noted that Uniplate’s contractual warranty period was six months, with the claim of longer “in practice” maintenance not clearly reflected contractually. It treated the maintenance justification as insufficiently established as a basis for the breadth and duration of exclusivity.

In the result, the Tribunal concluded that Uniplate had not shown pro-competitive gains that outweighed the anti-competitive effects. It therefore found a contravention of section 8(d)(i) for the period 2010 to 2014.

On penalty, the Tribunal applied sections 58 and 59 and adopted the six-step methodology developed in Competition Commission vs Aveng (Africa) Limited t/a Steeledale and others (84/CR/Dec09). It identified 2014 as the relevant year for affected turnover but accepted Uniplate’s submission, in part, that certain revenue streams should be excluded. It excluded exports from affected turnover on the basis that the conduct would likely have affected only the domestic blanking business, while rejecting Uniplate’s submission to exclude Type A blanks from affected turnover because exclusivity included Type A blanks and the penalty is based on affected turnover rather than dominance in a sub-market.

The Tribunal determined an affected turnover figure of R108 871 419 (excluding exports). It set a base percentage of 5%, reasoning that Uniplate’s reliance on comparisons to a cartel matter involving a once-off incident did not reflect the pervasive nature of the conduct. After multiplying for a five-year duration, it reached an amount exceeding the statutory cap, accepted the 10% cap calculated on 2016 turnover as R23 131 879, and then applied a 30% discount at the mitigation/aggravation stage, treating Uniplate’s listed factors as at best neutral but nevertheless applying the discount. This yielded a final administrative penalty of R16 192 315, below the statutory cap.

5. Outcome and Relief

The Tribunal made a declaratory finding that Uniplate contravened section 8(d)(i) of the Competition Act 89 of 1998 during the period 2010 to 2014.

It ordered Uniplate to pay an administrative penalty of R16 192 315, payable within 90 business days of the order.

No order as to costs was made.

Cases Cited

Competition Commission v South African Airways (18/CR/Mar01) was relied upon for the formulation of the anti-competitive effects test under section 8(d). South African Airways (Pty) Limited v Comair Limited and Another (92/CAC/Mar10) was cited as confirming that anti-competitive effect may be established by proof of substantial foreclosure potential from which harm may be inferred.

Nationwide Airlines and Others v South African Airways (Pty) Ltd and Others [2001] ZACT 1 (5 January 2001) was cited for the proposition that complete foreclosure is not required and that preventing or impeding expansion may suffice. Competition Commission v Telkom SA Limited (11/CRFeb04) was cited for the proposition that harm can be established by conduct likely to prevent or lessen competition, including impeding competition, without proof of exit or loss of market share.

ITT Promedia NV v Commission (Case T- 111/96 ITT Promedia NV v Commission [1998] II-2937) was cited, together with Competition Commission v. Telkom (Case No. 11/CR/Feb04), in the Tribunal’s discussion that litigation conduct by a dominant firm may in some circumstances be abusive.

Competition Commission vs Aveng (Africa) Limited t/a Steeledale and others (84/CR/Dec09) was cited for the penalty methodology. Competition Commission vs Giuricich Coastal Projects (Pty) Ltd and Another (CR162Dec14) was referenced in the parties’ penalty submissions by way of comparison on base percentages.

Legislation Cited

Competition Act 89 of 1998, in particular sections 8(d)(i), 8(d)(iii), 8(c), 50(1), 58(1)(a)(i), 58(1)(a)(vi), 59(2), and 59(3), provided the core statutory framework for the prohibited practice assessment and penalty.

National Road Traffic Act 93 of 1996 was referenced as part of the regulatory context governing number plate standards and compliance with SANS 1116.

Rules of Court Cited

No rules of court were cited in the Tribunal’s reasons.

Held

The Tribunal held that Uniplate’s pervasive exclusive supply provisions, which contractually required customers who acquired a Uniplate embossing machine to purchase all their number plate blank requirements and certain related inputs exclusively from Uniplate, constituted an exclusionary act in terms of section 8(d)(i) and had a likely substantial foreclosure effect on competitors in the blanks market during 2010 to 2014.

The Tribunal held that the Commission had discharged its burden to show significant foreclosure, relying on the breadth and duration of the exclusivity, the lack of practical exit routes, the staggered lock-in of demand, and evidence of aggressive enforcement that impeded customer switching. It further held that Uniplate failed to establish efficiency or other pro-competitive gains that outweighed the anti-competitive effects.

The Tribunal held that an administrative penalty was appropriate and, applying the Aveng methodology and the statutory factors, imposed a penalty of R16 192 315, with no costs order.

LEGAL PRINCIPLES

Section 8(d)(i) of the Competition Act is effects-based. Anti-competitive effect may be established either through evidence of actual harm to consumer welfare or through evidence that the exclusionary conduct is substantially significant in foreclosing the market to rivals, in which case an inference of anti-competitive effect may be drawn, consistently with Competition Commission v South African Airways (18/CR/Mar01) and South African Airways (Pty) Limited v Comair Limited and Another (92/CAC/Mar10).

Proof of complete foreclosure, competitor exit, or loss of market share is not necessary to establish harm. It may be sufficient to show that the conduct prevents or impedes a rival from expanding in the market or is likely to prevent or lessen competition, as reflected in Nationwide Airlines and Others v South African Airways (Pty) Ltd and Others [2001] ZACT 1 (5 January 2001) and Competition Commission v Telkom SA Limited (11/CRFeb04).

Once the Commission has established anti-competitive effects under section 8(d), the evidential burden shifts to the dominant firm to show technological, efficiency, or other pro-competitive gains that outweigh the anti-competitive effects. Where such justification is not substantiated on the evidence, the conduct remains prohibited.

In determining an administrative penalty, the Tribunal must consider the factors in section 59(3) and ensure the penalty does not exceed the 10% turnover cap in section 59(2). The Tribunal applied the structured penalty methodology developed in Competition Commission vs Aveng (Africa) Limited t/a Steeledale and others (84/CR/Dec09), including determination of affected turnover, a base percentage, duration, application of the statutory cap, and mitigation/aggravation adjustments.

Competition Commission of South Africa v Uniplate Group (Pty) Ltd (CR188Nov15) [2019] ZACT 61 (27 June 2019)