Alexandros Constantinou and Sifiso Mhlaba

In South African competition law, dominance is usually associated with high market share in the relevant market. While this is intuitive, it may not necessarily always be the case. This is because even with low market share, if a firm can be shown to possess market power, it can also be defined as ‘dominant’.. This article explores the potential usefulness of inferring dominance frommarket power.

Abuse of dominance offences fall under section 8 and section 9 of the Competition Act no. 89 of 1998 (“the Act”). Section 8 involves prohibited practices such excessive pricing, exclusionary conduct, refusal to supply scarce goods, amongst others. Section 9 entails price discrimination which essentially means the charging of different prices for goods or services of like grade and quality to different purchasers.

However, abuse of dominance cases cannot proceed unless dominance itself can be shown. To prove dominance, the tests are laid out in section 7 of the Act, which is reproduced below for ease of reference:

“7. Dominant firms:

A firm is dominant in a market if –

(a) it has at least 45% of that market;

(b) it has at least 35%, but less than 45%, of that market;
unless it can be shown that it does not have market power; or

(c) it has less than 35% of that market, but has market power.”

We reviewed all 42 local instances[1] in which abuse of dominance allegations were put before the competition authorities (either as a main allegation or mentioned ad lib). This includes cases that were granted or dismissed and the aim was to observe whether dominance was tested against section 7(a), 7(b) or 7(c).

All reviewed abuse of dominance cases were tested against section 7(a), or dismissed because dominance could not be shown. There is no case where 7(b) or 7(c) was fully tested; with the exception of Pharmaceutical Wholesalers vs Glaxo[2] where section 7(c) was argued by the applicants, but this argument was ultimately dismissed by the Tribunal.

In other words, all prosecuted abuse of dominance cases proceeded on the basis that the offending firm’s market share was in excess of 45%, i.e. section 7(a) of the Act. For example, in the following abuse of dominance cases, CC vs SAA^[3], CC vs Telkom^[4], Nationwide vs SAA^[5] and CC vs Senwes^[6]; the respondents all exceeded the 45% threshold, so the question of dominance was mostly academic. To this end, an extract from the CC vs SAA case states: “…because we find that SAA is presumptively dominant [i.e. market share exceeds 45%] we need not deal with evidence raised by SAA…to the effect that it does not have market power.”[7]

What this means is that by showing the offending firm has a high market share, in excess of 45%, it is deemed to be dominant without need for further deliberation on that leg of the test. However, as indicated above, section 7 allows for the possibility to infer dominance from market power even if a firm has less than 45%. And in terms of the definitions in the Act, “‘market power’ means the power of a firm to [i] control prices, or [ii] to exclude competition or [iii] behave to an appreciable extent independently of its competitors, customers or suppliers.”[8] [own emphasis]

Of particular interest in this article, is the power to exclude competition. This implies that market power can be established by demonstrating the ability to exclude competitors from entering and/or participating in a market, such that competition is stifled. In such instances, it is important to recognise the difference between a competitor and competition. For example, excluding one ineffective competitor may not necessarily amount to excluding competition. To actually exclude competition means to exclude a significant effective rival(s) from the market.

Once market power has been firmly established, section 7(b) and 7(c) allows for the inference of dominance, regardless of the market shares of the power-wielding firm. For example, this may be applicable in cases where parties have entered into exclusive lease agreements with other market participants with the sole intention of excluding competitors (for example, the retail market). We believe this ability to exclude competitors confers market power on the offending firm, and therefore implies dominance.

The problem with relying only on section 7(a) is that the calculation of market share relies on the determination of a robustly defined market. This is because failing to define the correct market can lead to misleading and/or inflated market shares. Indeed, this is highlighted in CC vs Sappi[9]:

“… the Commission’s allegation of dominance cannot be sustained -the market share may be exaggerated; the relevant market may be incorrectly identified.”[10]

Thus defining the relevant market can itself be both contentious and difficult. It follows that defendants will almost always challenge market definition at Tribunal hearings. Typical strategies employed can include broadening the product and geographic scope of the market so as to include more competitors in order to decrease the share below the 45% threshold. Once shares dip below the threshold, it typically becomes more difficult to sustain an abuse of dominance case. This is especially prevalent in cases where the boundaries of the market may not robustly withstand scrutiny during the Tribunal proceedings.

Making use of section 7(b) or 7(c) can counter some of these challenges to market definition. This was recognised by the Tribunal in the FFS/Eskom[11] case wherein it advised that:

”It might be possible for a complainant, relying on section 7(c), who is not certain of the boundaries of the market e.g. whether the market is for blodgets alone or for widgets as well as blodgets to allege instead that the respondent has market power. Yet even on this approach the complainant would still need to plead the facts that support the allegations of market power and at the very least offer some permutations of a possible market where that power is exercised.”[12]

There are other cases including Nationwide vs SAA[13], and York Timbers vs SA Forestry[14] where market power vis-à-visdominance was briefly discussed, but alternative options were not tested and ultimately the Tribunal reverted back to section 7(a); seemingly to avoid the complexities involved with proving market power.

Looking beyond the 45% threshold typically used to prove market dominance might appear more complex, but we believe that in cases where market definition is difficult to sustain, inferring dominance from market power may be a credible option worth exploring. This could lead to an increased number of successful abuse of dominance prosecutions.