By Lerisha Naidu and Angelo Tzarevski
Digitisation has ushered in an era of hyper-connectivity, marked by disruptive digital platforms that operate on a global scale.
According to Refinitiv data, 37 cross-border merger and acquisition (M&A) deals in Africa were announced in the technology, media and communications (TMT) sector in the first half of 2021, valued at USD 768 million.
The growth of the digital economy across the continent has naturally been accelerated by the COVID-19 pandemic and this unabated demand for technology has caused extensive cross-sector disruption in the financial, energy, transport, retail, health and agricultural sectors.
There is an appreciation that digital markets are characterised by, among other things, multi-sided platforms, large returns to scale and complex network effects. As a result, competition authorities are increasingly presenting novel theories of anticompetitive harm, which, unlike those in the more traditional markets, are yet to be tested.
From an African perspective, this dynamic evolution of markets presents an opportunity to drive structural transformation and development, as market participants integrate to reach consumers and suppliers that would otherwise be inaccessible.
To achieve this, competition authorities must balance the importance of upholding the regulatory process with the promotion of innovation and investment. The common themes related to merger control, abuse of dominance and cartel conduct in Africa point to the nexus between competition regulation and the digital economy.
The effectiveness of merger control as a means of furthering competition policy objectives is dependent on the competition authorities’ ability to avoid two types of errors – false positives, which occur when a merger that should have been permitted is blocked; and false negatives, which occur when a merger that should have been prohibited is approved and consequently, implemented.
Competition authorities increasingly perceive false negatives as being a more probable eventuality in the context of digital transactions, suggesting that there has been inadequate enforcement in this sector. The question is whether competition authorities are adequately equipped and resourced to consider mergers in digital markets.
Market definition is becoming more intricate in the evolving digital era, especially in relation to so-called zero-price markets where users of products or services do not pay money for their use, such as social networks.
Further, in market environments with two-sided platforms, the question is whether the relationship between the platform and the respective market sides can be considered separate markets or whether there is a single market.
There is also the issue of whether there are circumstances under which a market can be viewed in isolation of the other side or whether the interplay between both sides ought always to be taken into account.
One of the emerging views is that because market boundaries are difficult to define and change rapidly in the case of platform markets, less importance should be placed on market definition in the competition assessment and more emphasis should instead be placed on the theories of harm and identification of anticompetitive strategies.
This view is compounded by the methodological problems associated with applying traditional economic tests when defining digital markets.
It is commonplace for mergers to be notifiable and subject to evaluation only where the merging parties meet certain financial thresholds, usually in terms of turnover figures and asset values or market share thresholds.
An unexpected consequence of the use of financial thresholds is that mergers with meaningful effects in digital markets may, in certain circumstances, fall well below the prescribed monetary thresholds, with the result that market-altering transactions are able to escape scrutiny by the competition authorities.
Compounding this concern is the threat of “merger creep”, where numerous small start-ups are acquired through transactions that may appear relatively inconsequential on an individual basis but, when considered collectively, may have significant competition implications for the market.
Competition authorities argue that the traditional financial threshold-based approach to merger notifiability may need to be reconsidered and, perhaps, replaced in light of the dynamics of the digital market.
South Africa is considering a combination of deal value and market share metrics in this initial assessment around whether the transaction should be compulsorily notified.
A key attribute of digital markets is the acquisition of small start-ups by large firms. Start-ups often need to be acquired to access the capital required to scale-up, leading to procompetitive effects.
Africa has the fastest growing tech start-up ecosystem in the world – going forward, competition authorities will likely pay close attention to determining and distinguishing between procompetitive acquisitions intended to expand or improve product offerings from those that have the object of eliminating competition – also termed killer acquisitions.
In South Africa, this concern has prompted the competition authority to apply greater scrutiny to digital transactions that would ordinarily not warrant notification, and its competition authorities recently published proposed amendments to the Small Merger Guidelines, which call for notification of small merger transactions involving digital market players based on deal value and/or the parties’ market shares.
Abuse of Dominance
Competition authorities have identified conduct that, if undertaken by dominant firms, may result in harm to competition. In the context of digital markets, the issue is whether existing theories of harm apply to digital markets or whether new theories should be considered. In addition, it is not clear how certain abusive conduct arising in digital markets will be assessed.
In terms of the existing framework, certain conduct is automatically deemed to constitute a breach of competition with no room for the advancement of procompetitive justifications, while others are analysed by reference to the effects of the conduct on competition.
Self-preferencing is the act of giving preference to your products or services over those of your rivals. Dominant firms operating in two-sided markets may leverage the market power they possess on one side of the market, to gain an advantage in the other.
Competition authorities have identified self-preferencing as potentially harmful competitive conduct that has the effect of entrenching dominance and excluding competitors.
Acquisition of Data
The ability to acquire, process and analyse large volumes of data gives dominant firms a comparative advantage in the digital market. Competition authorities are concerned that firms may look to exploit user data to exclude rivals.
Given its importance in the digital market, it has been debated whether data can constitute an “essential facility” and, if so, to what extent the refusal to grant access to large datasets may constitute anticompetitive conduct.
In South Africa, a dominant owner of an “essential facility” would risk abusing its dominance if it refuses to grant access to such facility to its competitors where it is economically feasible to do so. There are several difficulties associated with treating data as an “essential facility” and forcing data owners to share it with competitors such.
Data is ubiquitous, replicable and varies in its value and usefulness. Also, it cannot be guaranteed that the data held by one entity is essential for the market participation of another entity.
Further, placing an onerous obligation on data-rich firms to share data may also enable competitors to reverse-engineer proprietary algorithms and, in so doing, encourage free riding. Ultimately, this will deter investment in large-scale data-collection and innovation into data driven platforms. Additionally, obligations to transfer data to competitors may give rise to data privacy concerns.
Use of Algorithms
There is growing concern that algorithms can result in exclusionary anti-competitive conduct and consumer harm. Potential theories of harm include the use of ranking algorithms to, inter alia, manipulate consumer and limit choice, apply different pricing/terms to different categories of consumers; and manipulate platform ranking to exclude rivals.
It is not clear how anticompetitive algorithms can be detected or assessed within the existing framework. In addition, competition authorities are faced with conceptualising whether algorithms that result in more automated processes, can constitute “unilateral” conduct for purposes of the assessment of abusive conduct.
Competition authorities are concerned that digital markets have altered the nature of interactions and are questioning whether the use of algorithms can facilitate agreements or coordination on price and other trading conditions in a more efficient way than traditional human interactions.
One of the challenges for antitrust authorities would be demarcating cartelistic flow of virtual information resulting in tacit collusion from mere market transparency and machine learning adaptation to such detected market trends.
Convergence of Competition and Social Policy
Governments around the world have decisively shifted away from the purely economics-based origins of competition regulation, turning instead towards a model that acknowledges and, to an extent, caters to the broader needs of modern society.
With digital innovation opening up the economy to many individuals and businesses that were, until recently, excluded from meaningful economic participation, it is likely that public interest imperatives will play a crucial role in the development and implementation of competition law in the digital space.