Kenyan Competition Policy After Ten Years of the Competition Act: A Progress Report

Abstract

With the coming into force of the Competition Act (2010), Kenya entered the world of modern competition policy. The 10 years that has passed since has seen a great level of activity in the competition space in Kenya, including: the staffing up of a modern competition authority; the rolling out of an ambitious advocacy campaign; the drafting of key guidance documents; the passing of a number of key amendments to plug gaps and open up new avenues for cases, and of course the pursuit of numerous cases including some related to cartels, abuse of dominance and mergers. Through all this, Kenya has become a leader in competition policy in Africa. This paper provides a critical review of many of these developments—with a particular focus on current challenges and opportunities. It also offers reflections on how competition policy may need to be different in a developing country, based on Kenya’s experience.

1 Introduction

With the passage of the Competition Act in 2010, Kenya entered a new era in its evolution toward a more market-oriented economy. A modern competition law, an ambitious and resourced competition authority, and support from its government have made Kenya a leader in competition policy in Africa.

It is clear to all in the field that the passage of a new competition law—in any country—is just the first step toward creating a new competition policy regime. There will inevitably need to be a great deal of education, advocacy, skills development, statutory amendments, guidance document development and, of course, enforcement actions including judicial decisions. With the passing of 10 years under the Act, we believe that this is an appropriate time for a review and stock-taking.

These last 10 years have been a remarkable period for Kenya, and it is fair to say that the Kenya of 2020 is different from that of 2010. The country is absorbing enormous governance changes after the introduction of a new Constitution in 2010, which will devolve significant powers to county governments—a process that began in 2013. Expanded demand for government spending has increased the public debt to GDP ratio from 41% in 2013 to 62.4% by December 2019.¹ At the same time, the economy has continued to grow and achieved “low middle income status” (up from “low income”) in 2014. The average growth rate from 2010 to 2019 was approximately 5.8%; the gross domestic product (GDP) in 2019 was approximately US$95.5 billion.² The country also benefits from a relatively young population: 70% of Kenya’s population of approximately 47.5 million is under the age of 30.³ Despite substantial overall growth, however, income inequality remains very high.⁴

We make two additional observations that have relevance for the design and implementation of a competition law: First is the importance of agriculture and informal work in Kenya. In 2018 56% of employment was in agriculture.⁵ The World Bank (2019a) also reports that 84% of workers are employed in the informal sector. These are areas with large numbers of small enterprises and from which data can be extremely difficult to secure—yet they represent large fractions of the Kenyan economy.

Second, on the positive side, there is evidence that the Kenyan business environment has improved in important respects since 2010. In the “Ease of Doing Business” rankings of the World Bank (2020a), Kenya’s position has improved from 106th in 2010 to 56th in 2020. The key factors that contributed to the higher ranking included improvements in the ease of: getting construction permits; securing electricity; accessing credit; paying taxes; and resolving insolvency; and improved protections for minority investors. All represent promising developments toward providing the structural conditions that are needed to support competitive markets.

Significantly, Kenya is currently negotiating a Free Trade Agreement with the United States. The Kenya-US Free Trade Agreement is expected to be based on the text of the United States-Mexico-Canada Agreement (USMCA) of 2018 and is expected to include an important chapter that is related to competition policy.⁶ This illustrates the increasing status of Kenya in Africa and the perception of Kenya as a gateway to Africa in the global context. Deeper preferential trade agreements recognize the need to account for competition promotion as a fundamental tool to achieve the benefits of market access through trade.

Our purpose in this paper is to evaluate the progress that has been made in Kenya with respect to its competition policy regime and to assess some of the challenges that lie ahead. We do this from the perspective of competition economists, and we readily admit that legal issues—for example, with regard to legal processes and constitutional alignment—will get little attention in this essay. We will describe the successes that we have observed and offer suggestions for improvements in legislation and enforcement that would further strengthen and modernize Kenya’s new approach. We believe that there are important insights to be derived from the Kenyan experience that can inform our understanding of how competition policy might need to function differently in a developing country context. As such there are lessons here for other countries in earlier stages of this process.⁷

While our ambitions and approach here are related to those of the “peer review” studies that have been conducted by, most famously, the Organization for Economic Cooperation and Development (OECD) and the United Nations Conference on Trade and Development (UNCTAD), they differ in a number of ways:⁸ Principally, we believe our perspectives as outside academic researchers provides a somewhat different lens through which we evaluate the success and challenges of Kenyan competition policy; this is less of an insider’s or practitioner’s view perhaps but one that is possibly more wedded to current research and thinking about the appropriate design and enforcement of competition policy.⁹ We are also much more focused on economic aspects of competition policy here and less on legal and procedural elements.¹⁰

We will not review the history that led to the 2010 Act, or provide a great deal of detail about its original contents. This has been covered in earlier work.¹¹ We will also not consider in any detail the consumer protection elements of the Competition Act—they deserve their own treatment. The next section will focus on the evolution of the Act from its beginnings to the present day—including amendments made as recently as 2019. Many of the early amendments were made to correct what were likely typographical errors or drafting oversights, while others have introduced important new elements: for example, new powers for the Competition Authority of Kenya (CAK) and new provisions on the abuse of buyer power.

Section 3 examines the key competition policy intuitions that were created by the Act: the CAK and the Competition Tribunal. As the Tribunal has only very recently issued its first decisions, most of the attention here is on the CAK: We study changes in its size and budgets, and examine the activities it has undertaken. Chief among these activities are the cases that it has pursued; and, while space does not permit a detailed review of many cases, we discuss some important themes that have emerged.

We review the evolution of the law and institutions in Kenya—not merely to record these histories but also to suggest directions for future progress, which will involve new methods of enforcement and possibly further legislative changes. In making such recommendations, however, we realize that we must always be aware of the different contexts in which competition policy works in different countries. Kenya’s experience suggests a number of ways in which competition policy might be called upon to be different in a developing country than in a developed one. Section 4 examines some of these differences.

As will be clear from these earlier sections, there are a number of ways in which Kenya’s competition policy regime might be strengthened and brought into closer alignment with what competition economists might see as best-practices. Section 5 offers our suggestions. They do not generally represent criticisms of the current activities of the CAK; indeed, addressing most could require further legislative amendments.

Section 6 offers a few concluding remarks.

2 The Legislation: The Competition Act of Kenya 2010 and Its Amendments

The story of the evolution of the original Competition Act of 2010 to its present form after the most recent amendments of 2019 contains the familiar themes: the borrowing of best practices; innovations that suit local circumstances; drafting missteps; learning; and adaptation.¹² We conjecture that any jurisdiction that adopts its first competition law will follow a path with similar steps, and we hope that detailing the Kenyan experience here might provide some guidance.

Borrowing from the established competition laws of South Africa and the European Union (among other countries), the original Competition Act of 2010 contained the major elements of a modern competition law In Part I, at Section 3, it laid out the objectives of the law which, pleasing to economists, placed a heavy emphasis on efficiency: with language that suggests goals of enhancing productive, allocative, and dynamic efficiency.¹³ In Part III on Restrictive Trade Practices, Section 21 condemned anticompetitive agreements (both horizontal and vertical) and Section 24 prohibited the abuse of a dominant position.¹⁴ Part IV laid out the elements of a merger review regime. The two major competition policy institutions were established, and their powers and functions were described, in Parts II (for the Competition Authority) and VII (for the Competition Tribunal).¹⁵

Two, less common, elements were included: Section 22 essentially applies the agreements provisions of Section 21 to trade association activities, and Part V (Sections 50–54) addresses the “Control of Unwarranted Concentration of Economic Power”. These are interesting additions that address the different needs of some developing countries; we discuss them further below.

While there is no disputing that the original law was a major step forward, some problems with its drafting appeared fairly quickly¹⁶: A typographical error in Section 21 made a reference to “exemptions” in Section C when, in fact, the exemption provisions were contained in Section D.¹⁷

Less obvious, and perhaps more of a drafting issue: Provisions that were necessary for the enforcement of the abuse of dominance sections were omitted. Specifically, while the provisions that described how investigations were to be conducted and decisions rendered began in Section 31—clearly addressing both anticompetitive agreements and abuse of dominance matters—from Section 34 onward references to abuse of dominance matters were omitted. This omission would not allow, for example, the CAK to take an administrative action after investigating an abuse case. The process was eventually extended to cover abuse of dominance cases through amendments to Sections 34, 36, and 37 in the Competition Amendment (2016) Act.

Finally, a less obvious problem arose with the drafting of the Section 42, subsections (5) and (6) which specifies punishments for firms that violate the merger provisions. Subsection (5) provides for a criminal penalty. Subsection (6) provides for an admirative penalty; but the initial wording of (6) began with “In addition, the Authority…” Subsequent legal opinion concluded that this meant that an administrative action and sanction could take place only after a criminal process had been concluded. As the CAK wanted, reasonably, to use administrative processes in merger cases (at least in the large majority of cases), this was problematic and held up the review of a particular merger case. The same Competition Amendment (2016) Act fixed this problem by deleting the words “In addition” from the beginning of subsection 42(6).¹⁸

Beyond these amendments that were focused on restoring the original intent of the law, a number of new and interesting amendments have been made to the Act in the aforementioned Finance Bill (2014) and Competition Amendment Act 2016, as well as the recent Competition Amendment Act 2019. We characterize these as being of three types: (1) rewordings and new definitions; (2) new enforcement provisions; and (3) the addition of a new substantive section.

2.1 Definitions

Two alterations that were related to definitions were introduced in the 2014 amendments, with potentially significant implications. First, as originally introduced in 2010, “Part V—Control of Unwarranted Concentration of Economic Power” could have been read to provide substantial powers to address entrenched market power; power perhaps a product of earlier periods of heavy state-ownership and/or control.¹⁹ Such conditions are recognized to be important impediments to vigorous competition in developing economies.²⁰ The 2014 amendments (at S. 31 of the amendments bill) narrowed the application of Part V dramatically, by adding a definition to Section 2 of the Competition Act:

“unwarranted concentrations of economic power” means the existence of cross-directorship between two distinct undertakings or companies producing substantially similar goods or services and whose combined market share is more than forty per cent";

While it is certainly understandable that Kenyan business would like some sort of definition of “unwarranted concentrations of economic power” so it might gauge the scope of these provisions, this version reduces a potentially powerful section to a (still useful) directive on cross-directorships.

The other important change that we describe as definitional is really more than that: The original definition of dominant position in the 2010 Act was purely structural—in Section 4:

4(3) A person has a dominant position in a market if the person ―

• (a) produces, supplies, distributes or otherwise controls not less than one-half of the total goods of any description that are produced, supplied or distributed in Kenya or any substantial part thereof; or

• (b) provides or otherwise controls not less than one-half of the services that are rendered in Kenya or any substantial part thereof.

This indirect approach—inferring dominance from market shares—is to be contrasted with definitions that more directly focus on the possession of market power. For example, as defined by the European Court of Justice, in the Hoffmann-La Roche case, dominance is:

"a position of economic strength enjoyed by the undertaking, which enables it to prevent effective competition being maintained on the relevant market by affording it the power to behave to an appreciable extent independently of its competitors, its customers and ultimately the consumers".²¹

Two amendments that were made as part of the 2014 legislation altered the Kenyan definition. First, a definition of market power is added to the definitions list in S. 2:

“market power” means the power of a firm to control prices, to exclude competition or to behave to an appreciable extent, independently of its competitors, customers or suppliers;”

Hence, there is now a more direct definition—but of market power itself rather than dominance. Somewhat differently, this definition of market power resembles definitions of dominance from other jurisdictions where dominance would be seen to be a position with a significant amount of market power. This raises the question of whether, in the Kenyan context, any market power is enough to be found dominant or, alternatively, whether a firm is seen to have market power only if it is dominant (in the usual sense).

However, the structural approach to dominance was not abandoned. Section 23 provides a definition of “dominant undertaking”. It is not clear why there are two different sections that seem to be trying to define the same thing²²; the 2014 amendments confused things a bit more by expanding the definition in the latter section, with the addition of a new subsection (2)²³:

23. Criteria for determining dominant position

• (1) For purposes of this section, “dominant undertaking” means an undertaking which—

  • (a) produces, supplies, distributes or otherwise controls not less than one-half of the total goods of any description which are produced, supplied or distributed in Kenya or any substantial part thereof; or

  • (b) provides or otherwise controls not less than one-half of the services which are rendered in Kenya or any substantial part thereof.

• (2) Notwithstanding subsection (1), an undertaking shall also be deemed to be dominant for the purposes of this Act where the undertaking—

  • (a) though not dominant, controls at least forty per cent but not more than fifty per cent of the market share unless it can show that it does not have market power; or

  • (b) controls less than forty per cent of the market share but has market power.

The addition of subsection (2) greatly expands the potential reach of the dominance provisions. It also brings in a more direct test (market power) of dominance in 2(a) and 2(b). This has the potential to be confusing: Subsection 23(2)(b) seems to suggest that a firm with any market power at all can be viewed as dominant. Of course, given the definition of market power in Section 2, perhaps this section will be applied only to firms that would typically be seen as dominant. Perhaps the main take-away from this is that some clarity about the definitions of market power and dominance would be welcome.²⁴

With these definitional issues taken as given, the expansion of the definition will help to avoid some of the under-inclusiveness of the original legislation: A firm can have considerable power with a market share that is less than 50%—particularly if its competitors are very small and it is protected by barriers to entry.

However, there remains the danger of over-inclusiveness: If a firm operates in, e.g., competitive international markets but holds a significant share of the Kenya market (or some part of it), the firm could be caught by the purely structural definition in 23(1)(a), which has no market power requirement at all.²⁵

One final change to a definition is worth pointing out: In the 2016 amendments, the definition of “undertaking” in Section 2 was amended to clarify that it does include “trade associations”.

While Section 22 already contained special provisions that were directed at trade associations, they might have been read as rather specific and limiting; this amendment made it clear that the provisions of the Act apply generally to trade associations as they would to other undertakings. This change, combined with the addition of stronger market study powers, as will be explained below, gave the CAK greater authority when it came to investigating powerful trade associations in important sectors.

2.2 New Enforcement Provisions

A number of new and very valuable enforcement tools and powers were added in amendments, with subsequent elaboration and associated processes that are spelled out in Guidelines and as part of Rules that were promulgated in 2019.

The 2014 amendments added S. 89A, which explicitly authorized the CAK to operate a leniency program to encourage the disclosure of information that is related to the existence of agreements or practices that are prohibited under the Act. Subsection (2) of this new section declared that the details of the leniency program were to be spelled out in guidelines that are issued by the CAK. These Guidelines appeared in 2017.

The 2014 amendments also added a new subsection 30(2) that authorizes the CAK, with the approval of the Cabinet Secretary, to “exclude any category of decisions, practices or agreements by or between undertakings from application of the provisions of this Part.”²⁶ This “block exemption” provision is a useful tool to provide guidance to business and to lighten the load of competition authorities that are overburdened by applications for individual exemptions in cases where there is little chance of competitive harm.²⁷

While the original version of S. 36 of the Act indicated that the CAK could impose financial penalties for violations of the agreements provisions of the Act, it did not provide any instructions as what the magnitude of the penalty could be—and there was no provision at all for financial penalties for abuse cases (as noted above). The amendments in 2016 fixed this, revising S. 36(d) to read:

(d) impose a financial penalty of up to ten percent of the immediately preceding year’s gross annual turnover in Kenya of the undertaking or undertakings in question;

and amending S. 36(a) to make it clear that the actions that are available to the CAK—including the imposition of penalties—apply in cases that involve abuse of dominance as well as restrictive agreements.

Finally, the amendments of 2016 added a significant power to the CAK when it conducts market studies: The “Power to Hold Inquiries” S. 18 had added to it new subsection (6), which puts an obligation on individuals, firms, and organizations to “provide information requested by the Authority in fulfillment of its statutory mandate for conducting an inquiry or sectoral study regulated by this section.” These are important powers for a competition authority; they are arguably more important in developing countries where new authorities may have a strong need to conduct market studies in certain problematic sectors and where public sources of information might be very limited.²⁸

2.3 A New Substantive Provision

With its amendments in 2016, Kenya joined the relatively small but significant group of countries with specific prohibitions against the abuse of market power by powerful buyers. Njako (2019) describes the Kenyan approach to buyer power and lists a number of other countries with specific buyer-power provisions.²⁹ The new section that was added in 2016 is included under Part C (Abuse of a Dominant Position). The new section 2(A) was supplemented by Rules that were issued by the CAK in 2018, and then the section was modified, expanded, and renumbered as S. 24A in the amendments of 2019 (which incorporated some of the 2018 Rules).³⁰ Briefly, S. 24A now prohibits (at (1)): “Any conduct that amounts to an abuse of buyer power in a market in Kenya or a substantial part of Kenya”. In S. 24A(5) a number of examples of conduct that amounts to an abuse of buyer power are enumerated, including:

• (a) delays in payment of suppliers without justifiable reason in breach of agreed terms of payment;

• (b) unilateral termination or threats of termination of a commercial relationship without notice or on an unreasonably short notice period, and without an objectively justifiable reason;

• (c) refusal to receive or return any goods or part thereof without justifiable reason in breach of the agreed contractual terms.³¹

These examples of the abuse of buyer power help to reveal that part of the impetus for this new provision was to help suppliers—particularly small businesses—enforce contract terms on larger customers. Widely publicized allegations against, in particular, large chain retailers for mistreating small suppliers helped propel these amendments forward.

We offer a few observations about this new section: First, it appears that the motivation for these provisions are not of the classical competition policy variety. In fact, the standard approach to market power on the buyer side in competition policy is very cautious: condemning buyer cartels, perhaps, but not very aggressive with respect to power that is exerted by single large buyers. This is at least in part because theory is not clear on the conditions under which power on the buyer side is socially harmful.³² There certainly are conditions under which the exercise of market power by buyers can be socially harmful, but they are not as easy to satisfy as those for market power on the seller side. That competition law would be invoked to address what, to some extent, are really contract enforcement problems is an interesting development to which we return below.

Second, though the section is included in the Part that is devoted to “Abuse of Dominant Position”, there is no requirement for the CAK to establish that the buyer has a dominant position.³³ Indeed, it is not clear what, if any, market power the buyer would need to hold to qualify under this section.

The definition of “buyer power” in S. 2 simply refers to the “influence exerted ...to... obtain from a supplier more favourable terms…”. All negotiations involve attempts to secure more favourable terms, so this is potentially a very broad definition that could catch many firms. When we couple this with example (g) of 24A(5): “reducing prices by a small but significant amount where there is difficulty in substitutability of alternative buyers or reducing price below competitive levels” we have the potential to significantly constrain the negotiating position of a buyer—which, after all, may not itself be large.

Balancing this potential for over-reach, and our third observation, is that the section does seem to suggest that the CAK should use persuasion rather than administrative sanctions to address problems of buyer power. Section 24A confers on the CAK powers to impose reporting requirements and require industries and sectors to develop binding codes of practice (to be developed with the relevant stakeholders, relevant Government agencies and the Attorney-General). Importantly, administrative penalties are still provided for (under S. 36). In addition, Subsection 24A(9) further provides for criminal sanctions; however, the CAK has shown an understandable reluctance to pursue competition matters using criminal law, and it would be hard to believe that this is an area where they would choose that path.³⁴

3 Kenya’s Competition Policy Institutions

The 2010 Act created both the Competition Authority (in Part II) and the Competition Tribunal (in Part VII). While the former performs the usual advocacy, education, investigation, prosecution and adjudication functions, as in other administrative systems,³⁵ the latter hears appeals from decisions of the CAK. There is less to report here on the Tribunal as it has just rendered its first two decisions in April 2020; hence we will focus largely on the performance of the CAK.

3.1 The Competition Authority of Kenya (CAK)³⁶

The CAK was created as a new agency in the Kenyan public service. While some staff were transferred from the predecessor Monopolies and Prices Department (of the National Treasury), the leadership immediately embarked on capacity building with the support of its government sponsors. Over the decade the size of the CAK has grown: in staff complement and also in budget. As is revealed in Table 1, the total budget of the CAK increased by over 150% in the 5-year period from 2012/2013 to 2017/2018 (the most recent year that is available).

Table 1 CAK Budget Breakdown (in million Kshs)

The greatest portion of that increase came from the CAK’s exchequer allocation, but other significant contributions came from the retention of various fees and fines (e.g., merger filing fees), and contributions from development partners. The total number of staff almost doubled over this period, from 32 to 60.³⁷ This expansion in terms of financial and human resources—and the willingness of development partners to engage with the CAK—suggests a certain level of commitment from the national government, which we noted above in terms of the government’s willingness repeatedly to provide amendments that have improved the legislation.³⁸

It is always difficult to compare staffing and funding levels across different jurisdictions as agency mandates may differ (e.g., some may have consumer protection and/or regulatory responsibilities) as may the costs of securing highly qualified professional staff. It can also be unclear what the relevant measures for comparison might be.

Nevertheless, some simple comparisons—between Kenya and South Africa, which is arguably the most professional and successful competition agency in Africa—may be suggestive: If we imagine that resources that are devoted to a competition agency should be roughly proportional to the size of that nation’s economy, we might compare the number of employees in an agency relative to the national gross domestic product (GDP). Measuring GDP in purchasing power parity (PPP) terms (in international dollars) and using the most recent data reported by the agencies, we observe that the CAK has roughly 0.27 employees per billion dollars of GDP, while South Africa’s Competition Commission has only slightly more at 0.29.³⁹

Also important is the agency’s budget per employee as a measure of its ability to pay higher salaries for more qualified staff and to provide employees with enforcement resources: The CAK budget (again in PPP terms) per employee amounts to approximately $277,000, while that for South Africa’s Competition Commission is approximately $245,000.⁴⁰

Rough as these comparisons are, they suggest a level of financial support in Kenya that is comparable to that in South Africa.

3.1.1 Cases

The CAK has been busy on many fronts. Table 2 provides data on the number of cases that were examined by the CAK over 6 years as reported in published annual reports. Not surprisingly, the need to approve almost all mergers before they may proceed has given the CAK its largest number of cases—though some of these will likely not have required considerable resources.

Table 2 CAK—number of cases

However, a number of these matters were truly substantial: They presented the CAK with important challenges—for example, with respect to market definition—and compelled the CAK to take on some of the most successful and powerful firms in the country. For example, merger reviews included those of the Brookside–Buzeki dairies and the Tuskys–Ukwala supermarkets. Abuse of dominance cases included actions against the dominant wireless provider—Safaricom—and allegations of predatory pricing against Kenya Airways and Uber. Agreements cases that were reviewed included agreements entered into by Coca-Cola (vertical) and the Association of Kenya-Reinsurers (collusion).

Reviewing all of the important cases involving the CAK is beyond the scope of this article, but a few themes emerge:

• (1) The CAK has taken on a number of trade associations that violated the law by suggesting (sometimes fixing) prices for their members. The origins of much of this behaviour predated the passage of the Act and, in some cases, represented standard, accepted practice when it started. The cases include those that involved: the PHP Consortium (a private health-care consortium); the Association of Kenyan Reinsurers; and the Outdoor Advertising Association of Kenya.⁴¹

  The CAK generally took a compliance-first approach with associations—though there were financial penalties that were imposed in some cases. To engage more directly and educate more widely, the Authority instituted a “Special Compliance Process” with a number of trade associations in the agriculture/agro-processing sectors and the financial services sector.⁴² The purpose was to increase awareness of the Act among the associations and to address and resolve inadvertent contraventions. Subsequently, some of these associations applied for exemptions.

  These battles have not always been easy for the CAK, as some associations have pushed back. For example, after the CAK in 2015 conducted a detailed review of the tea sector—a very important export product for Kenya—the Kenya Tea Development Agency (KTDA) went to court to prevent the release of the CAK’s draft report, which included recommendations for reform that could potentially help small tea farmers. Court battles to secure the release of the draft report continue, but recently the Agriculture Cabinet Secretary has promised his own audit of tea factories and to push for reform in the sector.⁴³

• (2) Many mergers have been approved with conditions, and the proportion of mergers approved with conditions has been growing. In the 2013/2014 year, of the 26 approved mergers, conditions were imposed in only one. In the 2017/2018 year (the most recent available) of the 42 approved mergers, conditions were imposed in 11 cases (over 25%).⁴⁴ Most often these conditions involve the protection—typically for a specific period of time—of at least some of the employment that might have been lost as a result of the merger.

  Notably, there were a number of mergers in which employment conditions were attached even though the CAK had not found that the transaction would harm competition. For example, after studying the acquisition of the Real Insurance Company (REAL) by the British-American Investments Company (BRITAM), the CAK concluded that the transaction would not lead to a substantial lessening or prevention of competition in the relevant market (non-life insurance). Nevertheless, the CAK was worried about job losses and approved the merger (in 2014) on the condition that BRITAM retain at least 85% of the staff of REAL. In another example, the acquisition of the Unilever baking, cooking, and spreads business by Sigma Bidco BV was not seen to raise any competition issues (the acquiror was not previously in the market); but the CAK’s approval (in 2018) came with a condition that the acquiror retain all of the employees of the target.⁴⁵

  Making merger decisions conditional on satisfying public interest tests (e.g., on employment) separately from competitive effects tests is possible in Kenya because the legislation does not prioritize competition effects. The merger provisions do not prohibit mergers that lessen competition and then allow for some exceptions for (e.g.) public interest reasons. Instead, the provisions merely require that mergers (above a threshold) be approved by the CAK before they may proceed. The Authority is given a number of factors to consider in deciding whether to approve mergers—including the potential competitive effects and various public interest considerations, which include effects on employment or on a particular industrial sector or region.⁴⁶ The CAK’s merger guidelines make it clear that its competition and public effects assessments are conducted separately and that a merger’s approval could be made conditional on the parties’ taking steps to address concerns with respect to either or both dimensions.

  While competition policy experts have often been critical of the addition of public benefits tests to merger review, they are part of the review processes of many countries.⁴⁷ Concerns that have been raised about public interest tests have included that the competition authority might not be the most competent body to balance various public interest effects and, more commonly, that the introduction of such tests makes merger review less transparent and predictable (and perhaps more vulnerable to outside influence). However, others have expressed the view that—particularly in developing countries—a competition authority that is charged with approving mergers will be compelled to take into consideration a wider set of merger effects if it is to retain credibility and public support.⁴⁸

  The Kenyan approach to including public interest tests—similar to the South African approach—appears to take something of a middle path: While there are public interest assessments to be made, the general aspects of public interest to be considered are defined and are relatively few. And, to this point at least, they seem to be invoked to fashion remedies that allow transactions to proceed and have not been used to block mergers that pose no threat to competition.⁴⁹

• (3) The CAK has shown sophistication in its ability to define markets and fashion remedies. With respect to product markets, the Brookside–Buzeki dairy merger involved finding that a transaction that would have raised significant competition issues in a market of processed milk was much less problematic in a market that was defined (with evidentiary support) as fresh milk including that provided through informal channels.⁵⁰ In retail mergers the CAK has required divestiture of retail locations in particular areas in which it determined that there could be competitive harms.⁵¹ In a number of cases the CAK’s conditions have involved the protection of contracts between the merging entities and other parties.⁵²

  We offer one additional observation about conditions in merger review: In a number of cases the CAK has accepted merging parties’ claims about important post-merger plans without those claims’ being made into formal commitments in the mergers’ approval.⁵³ For example, in the 2019 merger between Commercial Bank of Africa and NIC Group, the CAK seemed to accept the parties’ claims that there would be branch closures only where there were overlaps in branch networks and that, where overlaps existed, the parties intended to open new branches in other locations. Neither of these claims were enshrined in conditions imposed on the transaction.⁵⁴ It is not clear to us why this was the case—were the actions not really material to the CAK, or perhaps the CAK was convinced that the parties would follow through out of self-interest, which would make conditions unnecessary?

• (4) The CAK has been—appropriately in our view—cautious in its approach to predatory pricing; this is an area of antitrust that carries with it dangers of over-enforcement that can chill legitimate competition. In both the 2017 Uber case and the 2016 Jambojet case the authority rejected complainants’ claims of predatory pricing.⁵⁵ In the Uber case, the CAK went so far as to collect evidence on Uber’s average variable cost and concluded that it was below its prices and therefore not predatory.

3.1.2 Non-case Activities

The CAK has been very active beyond its case activity: From the beginning it saw the importance of education and advocacy activities. It launched a number of initiatives, including: information campaigns; programs to train legal and economics practitioners on competition policy; and programs to educate journalists on the goals and methods of competition policy. It has also engaged in teaching and research activities with universities in Kenya and South Africa.⁵⁶

An important early step for a new competition agency is to inform itself about the competitive conditions in key markets and sectors—something the CAK has done with the use of its market studies powers. It has conducted inquiries or sector studies—sometimes in cooperation with other authorities in the region—of several areas, including those that involve construction, fertilizer, cement, banking, leasing, and sugar.

The CAK has also been active in its advocacy and cooperation across government. For example, in 2015 it produced a report that was intended to influence other policy makers and regulators with respect to the design and implementation of various regulatory policies. The goal was to encourage greater respect—across government—for competition as a driver of efficiency and creator of wealth in the economy.⁵⁷

The Authority also successfully engaged directly with other regulatory authorities to fashion more pro-competitive reforms to regulatory structures, including in: mobile financial services (with the Central Bank of Kenya); tea (with the Tea Board of Kenya); and retail wireless communications services (with the Communications Authority).⁵⁸ To facilitate open communication and cooperation the CAK has signed Memoranda of Understanding (MOUs) with a number of key regulatory authorities, including: the Central Bank of Kenya (2015); the Communications Authority (2015); the Insurance Regulatory Authority (2016); the Kenya Civil Aviation Authority (2018); and the Public Procurement Authority (2018).

To provide further advice, to Kenyan business and their advisors, as to how it interprets and enforces the Act, the CAK has also produced a number of guideline documents. Some of these documents have evolved, including through consolidation into broader guidelines, however currently there are public guidelines covering: (1) the leniency program; (2) search and seizure processes; (3) fining and settlements; (4) buyer power; (5) merger review (now consolidated and including public interest test guidelines); and (6) relevant market definition.⁵⁹ Each of these represents an important and substantial effort—which typically involves broad consultation—to help stakeholders understand what is needed to actualize an effective competition law. They are particularly important with a new competition law when there is no history of enforcement for those stakeholders to draw on to predict how the Authority will enforce the law.⁶⁰

Finally, the CAK has been very active internationally with high levels of participation in multinational competition bodies such as the International Competition Network and the Organization for Economic Cooperation and Development. It was also a founding partner, and first host authority, of the African Competition Forum. It has also been engaging on a bilateral and community basis with other national and regional authorities. The CAK has signed MOUs with authorities in: South Africa; Japan; the Common Market for Eastern and Southern Africa; (COMESA); and China.⁶¹

3.2 The Competition Tribunal

The Competition Tribunal was created, and its powers given to it, by Part VII of the Act. Its principle purpose is to hear appeals against orders issued by the CAK. According to S. 74(3), the Tribunal may, in its determination of any appeal, “confirm, modify, or reverse the order appealed against, or any part of that order.” S. 75 also provides that the Tribunal may refer appeals back to the CAK for reconsideration.⁶²

After earlier delays in launching the Tribunal—for example with respect to appointing members—the Tribunal recently issued its first two major decisions in cases that involved the East African Tea Trade Association,⁶³ and the merger of Airtel Kenya and Telkom Kenya wireless businesses.⁶⁴ The former case involved the application for an exemption by the Tea association under the Special Compliance Process launched by the CAK. While the parties agreed on some elements of the exemption, the CAK was not prepared to allow the Association to fix brokerage fees or warehouse prices. The Tribunal, on review, allowed the fixing of brokerage fees (for 2 years only); but it upheld the CAK’s refusal to allow the fixing of warehouse prices.

In approving the Airtel/Telkom merger the CAK had imposed a number of conditions that are related to, for example, what the merged firm could do with its spectrum and whether it could sell all or part of itself to other investors. The Tribunal relaxed many of the conditions that the CAK had imposed. The Tribunal found (in some cases) that the actions that the CAK would have prohibited would in any case be subject to regulatory review (by the CAK and/or the Communications Authority) should they actually arise.

A detailed review of these cases is beyond the scope of this article; but we recognize these decisions as key milestones as Kenya’s competition policy regime matures. The CAK did not get everything it wanted in these decisions. Though the Tribunal did show respect and deference to the CAK’s expertise—as one would hope it would—it also took a long, careful look at both matters and came to independent, reasoned judgments. Notably, the decisions are much richer in detail about the parties’ positions and arguments than we see in CAK case summaries. This additional information and transparency can only be helpful as competition law continues to evolve in Kenya.⁶⁵

4 Competition Policy in a Developing Country: Lessons from Kenya

Though Kenya’s Competition Act is very modern in many respects, it does differ from what some might see as international best practice (or at least common practice). However, there is a growing consensus that—even if one embraces the general goal of competition policy to be the protection of competition to create more efficient markets—there is a need for competition law to be adapted to its national context. And a big element of that national context is the country’s state of development.⁶⁶

Recognizing that competition policy might need to be different in developing countries generally does not (of course) mean that there will be a single model for developing countries. The best policy for a country will clearly depend on its own situation: for example, levels of concentration; barriers to entry; inequality in wealth and income; legal institutions; governmental resources; and its business and political history. And, as in any country, a law and its enforcement will need to evolve with learning: both to address original goals better and to take on new challenges.

Many of the differences that we observe today between Kenya’s competition policy and that of more developed countries can be viewed as stemming from particular challenges that are faced in developing countries and in Kenya in particular. We describe a few here.

First, for a new authority to build credibility—with the public and its political leaders—it may be useful for the authority to take on additional responsibilities that are very visible and beneficial. These may be responsibilities that, in developed countries, are assumed by other public bodies. However, in some developing countries if the competition authority does not address them, they may go unaddressed. For example, in Kenya it might be necessary for the CAK to be seen to:

• (1) be protecting employment levels, even if the job losses that are due to a merger might be a key part of the efficiencies that motivated the transaction. In countries with weaker safety nets for unemployed workers and high rates of unemployment, it may seem unconscionable for a merger review body to approve mergers that lead to extensive job losses;

• (2) allow potentially anticompetitive actions if some of the resulting profits are directed to “good purposes” with social value⁶⁷;

• (3) be dealing with excessive pricing. Such provisions may be more important to give the authority the power to deal with very high prices under pressure from a public that may not care that simply having and exploiting a monopoly is not typically an offense in developed countries. With markets not competitive and prices relatively high, the protection of the Authority’s credibility may demand that it acts⁶⁸; and

• (4) be dealing with aggressive behaviour by big buyers (e.g., large retail chains) against much smaller (SME) suppliers—particularly if this behaviour (e.g., not paying bills on time) is already a violation under contract law. In this case the competition authority can be seen to be standing in for costly or ineffective contract enforcement mechanisms.

In each of these cases the CAK was (or would be) filling in “gaps”—providing for what other public bodies might do in a more developed country—and this might help build greater public appreciation for its role in the economy. There is a concern, however, that the CAK could be asked to do too much with the resources that are at its disposal. And there is the danger that its decisions could be seen to be unpredictable if it becomes difficult for other parties to sort through all of the priorities that the Authority is addressing.⁶⁹

Second, funding a professional agency to undertake detailed and time-consuming reviews is not easy for any government, but will be particularly difficult for a developing country that faces tremendous unmet needs: for example, in education and health care. For this reason, it is to be expected that the CAK should be authorized to collect fees for services—for example, with respect to merger approvals—as is done in many countries, developed and developing.⁷⁰ It might also not seem unreasonable then to allow the CAK to also keep the fines (pecuniary penalties) that it levies for violations of the Act, (as opposed to the fines’ going into general government revenues) as is provided for in S. 78(c).⁷¹ Keeping its fines may also provide strong incentives for the CAK aggressively to enforce the law and select good cases, which will generate resources that are commensurate with its level of activity. However, this is more unusual, and perhaps problematic, as it could be seen to put the CAK in a conflict of interest position as the body that both imposes a penalty and receives the penalty. Of course, fined parties do have recourse to appeals to the Competition Tribunal, which certainly provides some protection from the abuse of these powers.⁷²

Third, provisions may need to be added to the law to deal with particular aspects of the national economy that might not be common in other countries. In the Kenyan case we see substantial provisions in the Act that are devoted to activities by trade associations, for example. To some extent these provisions might seem to be duplicative of other general provisions; but it appears that they were added to make very clear to trade associations—in a country in which they had historically played a major role in price-setting for their members—that this new law was changing the rules under which they would be operating. A second Kenyan example involves the inclusion of Part V of the Act—on the control of unwarranted concentrations of economic power—which might have been seen to grant powers to the CAK to deal aggressively with entrenched dominant firms, perhaps even by breaking them up.⁷³

Fourth, a lack of resources—financial and human—can challenge a new authority’s ability to review complex cases on an “effects” or “rule of reason” basis. For this reason, it could be preferable to make greater use of per se regulations than might be the case in a developed country with an established, well-funded authority. For Kenya, in addition to the likely treatment of hard-core cartels as per se violations (under an “object” test, as in Europe) which is common around the world, the abuse of dominance provisions at S. 24(2) provide a short list of practices that are viewed—if adopted by dominant firms—as violations without the need for the CAK to establish effects.⁷⁴

Further, while Sections 25–26 provide for the possible exemption of anticompetitive agreements for efficiency or public interest reasons, they do not apply to actions that are defined to be abuses of dominance. From the standpoint of economic theory, this is unfortunate, as some of the listed practices—such as tied selling and price discrimination—may well have pro-competitive (or competitively neutral) effects.⁷⁵ The desire to simplify enforcement under such conditions is natural; but our hope is that the CAK will consider potential effects in its case selection and that, in time, the law might be modified to provide for a test of competitive effects (and/or allow for exemptions).

The Kenyan experience also provides lessons with respect to how a country might implement a new competition law where there had not been one previously. The CAK began cautiously, as it staffed up and trained its new teams. Their early focus was on education and communication about the new rules; they did not launch into a significant number of major enforcement actions too quickly. And a number of the early actions that they did take were resolved without punitive measures’ being taken: The point in the early days was to educate.⁷⁶

5 Kenyan Competition Policy Moving Forward

There can be no doubt that competition policy in Kenya has come a very long way in the last 10 years. While this progress was ably kick-started by the passage of the original Competition Act in 2010, a great deal is also due to the careful and professional enforcement that has been provided by the CAK and to the Kenyan government’s willingness to adopt amendments addressing problems with the legislation when they appear.

With this all recognized, we nevertheless respectfully offer some views—based on arguments that we made above—about paths that we would like to see competition policy in Kenya take in the next 10 years. Most of these may require (or benefit from) further legislative amendments, and we will describe them this way; however, there may be alternative approaches possible without amending the legislation.⁷⁷

Cartels: This part of the law and its enforcement are in solid shape. We have two thoughts looking forward:⁷⁸

• (1) In our view, it will be important for the Tribunal (and possibly higher courts) to adopt the view that naked price-fixing is a prohibition “by object” as it is under EU law. This will provide as close to a per se treatment of cartel behaviour as possible (subject only to possible—but unlikely—exemptions).

• (2) It may be desirable to increase the maximum possible financial penalty (under S. 36(d)) beyond the ten percent of gross turnover—for just the most recent year—in Kenya.⁷⁹ This is not a particularly high ceiling—particularly for large multinational corporations. It can be contrasted with the approach in other jurisdictions that do not limit the number of years (e.g., the U.S., Canada) and/or base fines (with a similar percentage rate) on global turnover (e.g., the European Commission, France, Norway).⁸⁰ Understandably, the CAK has wanted to be careful and cooperative with early prosecutions, but as Kenyan business learns more about the law and its enforcement, the CAK might need the power to impose more serious punishments.⁸¹

Abuse of dominance: Here, we have somewhat more pressing suggestions for change over the coming years. Much of our reasoning for these suggestions can be found above, so we are brief here:

• (1) We believe that greater clarity is needed with respect to the definition of dominance under the Act. The need here is spelled out above in some detail. S. 23 is now awkwardly worded: It states that a firm “though not dominant” nevertheless “shall be deemed dominant”.And in S. 23(2)(b) it indicates that even a firm with a smaller market share can be deemed dominant if it “has market power”—seemingly equating any market power with dominance, which would normally require something more like substantial market power. However, S. 2 does offer a definition of “market power” (“power of a firm … to behave to an appreciable extent, independently of its competitors, customers or suppliers”) that reads more like a definition of dominance.⁸² At best then, the Kenyan law is using non-traditional definitions for market power and dominance; at worst, the law is confusing the two concepts. Either way, clarity would be valuable.

• (2) In our view, in time Kenya should move toward a more effects-based regime for abuse of dominance cases. This could include two changes: First, a definition of abuse of dominance could be added to clarify that abuse involves an anticompetitive effect.⁸³ Second, the possibility for exemptions that are granted to restrictive agreements such as cartels (Sections A and B) in S. 25 could be extended to cover abuse matters (Section C). This latter change would allow dominant firms that want to employ one of the listed practices (e.g. tied selling) with expected pro or competitively neutral effects to apply to be allowed to use the practice.

Abuse of Buyer Power: It is a bit early to gauge the effect of these provisions and it is likely that they will be enforced more through the cooperative development of codes of practice than enforcement actions. That said, we believe it would be useful to consider adding some sort of a dominance or market power test in time so that the CAK is not compelled to respond to what are essentially contract disputes between relatively small players.⁸⁴

Mergers: Here we have one high-level suggestion, and one suggestion that relates more to enforcement powers:

• (1) Though it might not be easy to draft, we believe that some language that clarifies the primary objective (or objectives) of merger review would be helpful in providing greater certainty to business. As written it is not clear what the main priority for the CAK should be. S. 46(2) indicates that the CAK can use any criteria that it considers relevant and then provides a list of factors that it “may” consider. For example, an amended law could state something like: “mergers that threaten to prevent or lessen competition substantially are prohibited”, with provision for other considerations (e.g., for efficiencies and public interest considerations) that are applicable in some cases. Of course, this could demote public interest considerations below competition effects, so we understand that alternative wording may be preferred: perhaps language that describes the dual competition and public interest reviews that are currently conducted. The purpose here is to clarify for potential applicants in merger review that the “any criteria” is really two main criteria: competition, and the public interest (with the latter fairly narrowly interpreted).

• (2) Unlike the powers that it has with respect to investigations into prohibited practices under S. 31, the CAK cannot compel third parties (e.g., customers or competitors) to provide information when it is conducting a merger review. Currently the provision of information to an investigator in a merger matter is voluntary under S. 46(5). While possibly not needed often, we believe that the ability to compel participation would be helpful to the CAK in breaking through codes of silence in an underdeveloped competition culture.⁸⁵

Other process and enforcement issues:

• (1) As was explained above, a number of drafting issues arising from the original Act have been addressed through amendments. We believe there is at least one left. S. 36 describes the actions the CAK can take after an investigation. The provided actions are:

  • (a) declare the conduct that is the subject matter of the Authority’s investigation, to constitute an infringement of the prohibitions contained in Section A, B, or C of this Part;

  • (b) restrain the undertaking or undertakings from engaging in that conduct;

  • (c) direct any action to be taken by the undertaking or undertakings concerned to remedy or reverse the infringement or the effects thereof;

  • (d) impose a financial penalty of up to ten percent of the immediately preceding year’s gross annual turnover in Kenya of the undertaking or undertakings in question; or

  • (e) grant any other appropriate relief.

• This drafting suggests that finding an infringement (under (a)) is an alternative to taking actions in response to that infringement (e.g., stopping and/or punishing the conduct). It seems very likely to us that a declaration under 36(a) should be—and likely was intended to be—a precondition for any action under subsections 36(b) to 36(e).

• (2) We have observed that many of the key sections of the Act can be enforced under criminal law or administrative law. While this does provide the CAK with a great deal of flexibility to deal with matters though the use of processes that are best suited to the overall objectives of the Act, there is a risk of something akin to double-jeopardy. In our reading, nothing in the law precludes the CAK from following up with an administrative action on the same matter after a criminal action has concluded, or for the Director of Public Prosecutions (DPP) to decide to prosecute a case after the CAK has conducted its own administrative evaluation. Specifically, even after a party has settled a matter with the CAK (and maybe even paid a financial penalty), it could find itself facing criminal prosecution by the DPP.⁸⁶ This weakens the CAK’s ability to offer full resolution of a matter as part of its settlement negotiations.

  A number of approaches could be helpful here: One option would be an amendment that states that if one route (criminal or administrative) is taken, the other cannot be. A second would empower the CAK to provide assurance—from the DPP—that a civil resolution will end the matter and that criminal actions will not be initiated. Failing this, formal guidelines perhaps jointly prepared by the CAK and DPP could explain how those bodies would normally handle their respective responsibilities under the Act.

• (3) Though we know it takes time and resources, we would also encourage the CAK to provide as much information as it can about its enforcement activities. The contrast in information that has come from the two recent Tribunal decisions and from CAK decisions is sharp. Not only will more information on decided matters educate legal counsel and facilitate compliance by other parties, it will provide the material for critical review and evaluation by researchers in universities and policy institutions.

We conclude this section by mentioning briefly a set of challenges—which are present now to some degree—but which are possibly set to grow. These involve the complex interplay between national and regional competition authorities when they are all relatively new (particularly the regional ones) and when an individual country may be a member in more than one regional community.

Kenya—through its membership in COMESA and the East African Community (EAC)—is now bound by competition rules in both treaties. The two communities have some overlapping memberships (including Kenya), but there are members of each that are not members in the other. This presents to business a complicated set of rules that will vary depending on which countries are involved in the activities being investigated.

COMESA launched its Competition Commission, based in Malawi, in 2013. It is reviewing mergers and conducting investigations into anti-competitive business practices. The EAC's Competition Authority opened in 2016 in Tanzania but to this point has largely been devoting itself to advocacy and outreach as it builds its professional capacity.⁸⁷

This is a situation that clearly requires some negotiated, detailed, and transparent guidance so that firms that operate in a particular region know which set of rules applies to what agreements and conduct. Early reviews of the COMESA Competition Commission’s work have found that it is underfunded, with its work dominated by merger reviews.⁸⁸ The relatively slow launch of the EAC’s Competition Authority suggests that it might also be dealing with funding challenges.⁸⁹

Regional authorities like those of COMESA and the EAC present particular challenges for more advanced national authorities such as the CAK, because some member countries either do not have a competition law in place or have a law but not a funded authority to enforce it. Fostering cooperation—though required by community agreements—can then be very difficult.⁹⁰ And jurisdiction may not always align with resources—in some cases the regional bodies will have the authority but not the funding to pursue certain matters, which leaves the more experienced CAK watching from the sidelines. It is also possible that over time enforcement and decisions could move the laws of the various national and regional bodies apart such that very different decisions will be made based on which community/country is reviewing a matter. This increases uncertainty and risk for business, which can discourage investment.^91,92

6 Conclusions

In a relatively short span of time—less than 10 years—Kenya has built a credible, effective competition policy regime that is based on a modern law with Kenyan characteristics and is enforced by a resourced, professional authority. None of this happened instantaneously, and along the way there was a great deal of learning that could undoubtedly be of value to other new agencies—particularly those in developing countries.

The first law was not perfect and, to its credit, the Kenyan government has been prepared to revisit and improve the legislation repeatedly. The current Competition Act includes all of the key elements that one would expect in a competition law today, but it is far from a copy of any developed country’s law. It is clear that the law and the CAK have been called on to provide services that are not usually asked of competition authorities in developed countries—specifically to think about aspects of the public interest that are not captured by the idea of consumer welfare. This includes, for example, concern over losses of employment, the provision of resources for valuable projects, and the support of small businesses that face contractual challenges with large buyers.

Even with all of these accomplishments, competition policy in Kenya will nevertheless continue to evolve, and we have offered several suggestions in this regard. Some involve clarifying terms or approaches to certain problems—which might be accomplished through the provision of more information or even formal guidance—while others argue for substantive changes to parts of the law. To be clear, our perspective has been that of economists, who are recommending changes that align with advice from economic scholarship. We fully recognize that our legal colleagues might present a very different set of suggestions (and think less highly of some of ours.)

With a strong law, a professional authority, and a skilled private bar, Kenya is well-positioned to continue to be a leader in Africa in competition policy. If there is one part of the Kenyan competition policy framework that needs building, it is the academic/research side. We hope Kenya will continue to see growth in the number of post-secondary courses that are devoted to competition law and economics and in the number of researchers in universities and research institutes who are engaged in the field. To make the next 10 years as bright as the last will require a great deal of human capital that is knowledgeable about competition policy and also about the distinctive Kenyan context in which it operates.