SACF Submission on the 2nd INFORMATION MEMORANDUM User Manual

THE SACF’S COMMENTS ON THE 2ND INFORMATION MEMORANDUM

Introduction

1. The South African Communications Forum (SACF) is an industry association that represents a broad cross-section of members across the ICT value chain. Some of our
   members are licensees and others are in the ICT ecosystem that either provide the equipment to enable the network rollouts, or their businesses are dependent on the spread and quality of the networks.

2. Therefore, the impact and consequence of the licensing of the high-demand IMT spectrum is arguably the most significant project to face this Council. Against the the backdrop of failed starts, it must take this process to its successful conclusion which is the licensing of high demand IMT spectrum on a long-term basis.

3. While the SACF strives to limit commentary to areas of the agreement only, this may not be the case on all aspects for this submission. This is partly as a result of the limited time to comment on the submission.

4. The SACF would like to participate in any further processes in this regard.
   Context

5. The SACF is of the view that high-demand IMT spectrum must be licensed urgently. The licensing of the spectrum has been delayed for far too long which has constrained incumbents and prevented new entrants from entering the market despite having acquired licenses close to a decade and half ago.

6. There are no creative solutions left on which licensees can depend to navigate the spectrum crunch. The dire spectrum crunch is exacerbated by the phenomenal increase in demand in data traffic since the beginning of the COVID 19 pandemic. Access to the temporary or provisional spectrum helps but is far from a long-term solution. Access to spectrum forms the foundation of addressing several national priorities, including reducing the cost to communication. The SACF believes that spectrum must be licensed urgently. Our comments are submitted in the context of ensuring an efficient licensing process that can continue to completion.

7. Our written comments have a significant focus on the process. We are of the view that a sound process is essential to the successful conclusion of the licensing process. We
   acknowledge that a perfect process is implausible however, a sound process is important, to mitigate potential challenges.

8. An urgent licensing process must not be confused with a haphazard, ill-conceived one that has a significant potential of again being halted through litigation. ICASA should rather adopt a more pragmatic approach which may be a slower but more considered approach. South Africa cannot sustain yet another setback in the licensing of high demand IMT spectrum.

9. The SACF is gravely concerned by the following
   a) self-imposed impractical timeframes for the licensing of the spectrum. The timeframes are so shortened that it is difficult to comprehend how comments have been properly considered, and incorporated into the 2nd IM and that Council has adequately had the opportunity to apply its mind to as the arguments were complex and substantive. It is evident from the 2dn IM and the Reasons Documents that ICASA has not considered all the comments and the evidence placed before it by interested parties.
   b) Both the 1st and 2nd IMs fail to adequately address the issues raised in the litigation that stopped the licensing process. The failure to address the issues raised may endin the outcome.
   Process

10. The SACF’s comments on the first IM appears to have been misunderstood so we will clarify them due to our commitment to the successful conclusion of the spectrum licensing process. We welcome the articulation of the process in section 1 of the 2nd IM.
    Each process is separate and distinct

11. The processes are separate and distinct. Paragraph 2 of the High Court Order of 15 September 2021, states that “ICASA’s decision to publish the invitation to apply for licensing process for International Mobile Telecommunications in respect of mobile broadband wireless access services for urban and rural areas using complementary bands IMT700, IMT800, IMT2600 and IMT3500 spectrum frequency through an auction published as Government Notice 535 of 2020 in Government Gazette No. 43768 of 2 October 2020 (“the Auction ITA”) is reviewed and set aside and the matter is referred back to ICASA for reconsideration.” We understand this to mean that the 2020 ITA has been withdrawn and therefore no longer exists. . The decisions taken by ICASA in the 2020 ITA process were never ventilated in court and the courts did not pronounce on the legality of the process. Therefore, ICASA should take a cautious approach and ensure that it does not fail to take into consideration all the issues raised by interested parties. ICASA must go back to begin and start the process again considering the substance of the legal challenges that stopped the licensing process.

12. Consequently, the IM published in 2019 and the ITA published in 2020 are of no consequence and may only serve as historic documents but may not be treated as documents that form part of or referenced in this process as they have been set aside. Equally, we are of the view that the competition assessment must be treated in the same manner. Therefore, as we pointed out in our submission of 2 November 2021, it is difficult to understand how ICASA could reference documents that form part of the 2019/2020 ITA process which has been set aside.

13.  ICASA should publish an updated assessment and not rely on the MBSI which is based on 2018 / 19 data.

14. In Section 4 of the Reasons Document, it appears to conflate the licensing process that began in 2019 with the current process and links the period that the documents were out for consultation. The SACF is of the view that this approach is incorrect, as the High Court Order of 15 September 2021, set it aside and therefore the 2020 ITA no longer exists. Consequently, all consultations including the timeframes must be viewed separately and in isolation from the now defunct licensing process of 2020.

15. Further, the arguments advanced by active litigants during the legal process, cannot be considered to form part of the licence consultations as purported in the reasons document. The SACF is therefore of the view that this is an incorrect interpretation.

16. This interpretation is patently incorrect because it fails to recognize the number of
    respondents against the proportion that were active litigants in the process. This does not even consider the number of interested stakeholders that were not respondents in the legal process but have a vested stake or interest in the licensing process.

Highly Truncated Timeframes

17. The SACF highlighted its concerns about the highly truncated processes raising the challenges of an inconceivably shortened timeframe that in our view severely  compromises the process. It limits ICASA’s ability to consider submissions resulting in the superficial treatment of significant and complex issues as seen in the reasons document. Stakeholders have been given an inordinately shortened period to review, comprehensively consider, consult, and comment on the 1st IM and later, more detailed IM.

18. ICASA published the first IM for comment on 1 October 2021. ICASA planned to close comments on the 1st
    IM on 1 November 2021 which changed to 2 November 2021.
    ICASA did not resultantly amend the publication date for the 2nd IM. This gave stakeholders less time to consult and comment.

19. ICASA stuck to its truncated timetable to publish the 2nd IM on 15 November 2021, not considering the amended timetable. ICASA had less time to consider the comments on the 1st
    IM. Please refer to the figure below demonstrating the truncated timelines.

20.  The reasons documents lists the 10 stakeholders that submitted comments. We are aware that some of our members’ submissions were accompanied by a detailed economic analysis. It is difficult to conceive how such complex submissions could have been properly and thoroughly considered in the extremely shortened timeframes. In fact, it is obvious from the Reasons Document that these economic reports were not properly considered. In addition, ICASA offers no evidence to dispute the findings in
    the reports but seems to largely ignore them. This is extremely important as it increases the risk of an administrative challenge.
    Figure: Timeline for consultation from the publication of the 1 st IM to the ITA
    Days are calculated based days as defined in the ECA. Licence Milestone| Stakeholder Consultation| ICASA Consideration of Stakeholder Commentary| Commentary
    ---|---|---|---
    1st IM| 21 days| 9 days| 10 submissions
    2 nd IM| 9 days| 9 days|
    Final ITA| | 7 days from commentary to publication  Consideration of comments Analysis and inclusion
    Review + approval by + Council + publication|

21. While ICASA has repeatedly indicated that the 2020 licensing process went further than any previous process, ICASA also published the reasons document on the eve of the litigation and after the fact. The previous ITA process followed a similar timeline and was stopped for 191 days due to a flawed process and included content that was not consulted on. We cannot afford any further delays in the licensing of high-demand spectrum. Therefore, we urge ICASA to strike a balance between proceeding with speed and caution. This in our view will include reasonable timeframes and a solid process. The SACF would therefore suggest that ICASA conduct a legal audit of the process and conduct a thorough risk assessment before publishing the final ITA.

22. The opportunity costs to licensees and prospective licensees in compiling bids cannot be underestimated. Some of these applicants are the stakeholder base that ICASA has sought to include, yet these actions relegate some to the annals of history.

23.  Our comments are made against the backdrop of the 2nd IM Reasons Document which appears to be superficial and thin in relation to the written commentary received from interested stakeholders.

24.  Stakeholders were effectively given 8 days (17 – 30 November 2021) to consider and comment on an extensive document. This is not reasonable, nor does it meet the test for the expedited timelines or reasons to argue that this approach is in the public interest. This is an ICASA self-imposed truncated timeline. The brevity and apparent the unfairness of the process has the potential to cause unnecessary delays and result in the lack of administrative fairness.

25. Again, it is difficult to conceive that ICASA would have been able to analyse the submissions and apply its mind on a significantly longer and more complex document in a mere 8 days and incorporate stakeholder comments into a final ITA as ICASA has insisted. This again implies superficial and inadequate consultation. Should ICASA indeed stick to its irrational deadline of publishing the Final ITA on 10 December 2021, it seems implausible that stakeholder comments would have duly been considered.
    In the public interest

26. We understand that the ECA empowers ICASA to shorten timeframes in the public interest. For an action to be in the public interest we understand that the action must comply with Section 3 (4) b of the Promotion of Administrative Justice Act (PAJA, 3 0f 2000).

27. The most critical question that ICASA would in our view be required to satisfy is whether such shortened time frames are indeed in the public interest. For ICASA to satisfy the test of the shortened timeframes being in the public interest it would have to satisfy each of the tests below:
    a) Have stakeholders been given a reasonable period to consider the contents of the IM.
    b) Would the shortened timeframes make the process efficient and promote good governance?
    c) And finally, is the urgency beneficial to stakeholders.

28. We would argue that the highly truncated process fails to meet each of the above tests as,

29. Stakeholders have not been given a reasonable period to fully consider the contents of the IM in a mere 8 days. Despite the 2nd
    IM largely mirroring the 2020 ITA, there are significant changes that require a detailed and complex review for stakeholders to understand the impact to develop relevant positions. An example of this is the changes to the opt-in round of the auction. The changes are significant and complex.

30. The highly truncated timelines deny stakeholders the opportunity for this consultation and especially prejudice the smaller operators that ICASA seeks to include and compromise the soundness of the process.
    The SACF would argue that the highly truncated timeframes do not promote an efficient process because:

31. Stakeholders would not have been able to properly consider the contents, nor would ICASA have sufficient time to review, analyse, consider, and incorporate the commentary comprehensively and exhaustively. A potential consequence is that there may be a higher margin for error and dissatisfaction which is likely to increase the risk of legal challenges. Regardless of the success of a legal challenge, it inevitably delays and often derails the process. The last process was stopped for 191 days. While a sound and comprehensive process cannot eliminate the risk of a challenge but is a good mitigation strategy.

32. Therefore, an all too familiar approach of reckless speed of the process cannot be considered efficient. In each case despite ICASA’s dogged commitment to the litigation, it abandons its position on the eve of the matter being heard.

33. The SACF would further argue that the current process does not promote good governance as it is incomprehensible that because of the complexity and volume of the content this could have been properly considered.

34.  According to the United Nations, Good Governance is measured by the eight factors of
    – Participation,
    – Rule of Law,
    – Transparency,
    – Responsiveness,
    – Consensus Oriented,
    – Equity and
    – Inclusiveness,
    – Effectiveness and
    – Efficiency, and
    – Accountability.14 Oct 2021Finally, is the urgency beneficial to stakeholders? Urgency is beneficial to stakeholders, but a hasty and imprudent process is not.

35. The process was halted by litigation for 191 days due to several crucial issues. Progress and the successful completion of the licensing of high-demand spectrum is dependent on adequately resolving the basis for the previous legal challenges.

36. Therefore, the highly truncated timelines are not in the public interest. The SACF would support an approach that enables a comprehensive consultation with reasonably shortened timeframes. Consequently, the SACF would strongly urge ICASA to reconsider its incongruous and highly truncated timetable. South Africa cannot afford
    any further delays to the licensing of high-demand spectrum.

37. As explained above, the 2019 consultation on the IM is a distinctly different and separate process from the current process as the 2020 ITA was set aside.

38. Therefore, for the reasons set out above we are of the view that ICASA has not met the criteria set out in section 3 of PAJA of the public interest. The SACF is therefore of the view that ICASA therefore cannot use the public interest grounds as set out in the ECA for the highly truncated process. Fairness of the process

39. The 1st IM was published on 1 October 2021 with an original deadline of 1 November 2021 at 12h00, later extended to 2 November 2021. The process followed was patently flawed and prejudicial resulting in the rejection of submissions by some of our members for apparently being out of time. The SACF in our letter dated 9 November 2021, argued that the reasons for the inclusion of the comments by our members as it could only enhance the process. The SACF sought condonation which was denied.

40. During the workshop on the IM on 15 October 2021, ICASA indicated that the deadline had been extended to 2 November 2021, as 1 November 2021 had been proclaimed a public holiday. At the hearing ICASA did not indicate the applicable time of 12h00 for submissions. ICASA in Government Gazette No. 45416 published on 29 October 2021, which was only available on 2 November 2021, prescribed that the submission deadline 12h00. This is patently unfair for the following reasons:

41. The Gazette was late and ICASA did nothing to create awareness of the submission time as it is a considerable deviation from the regular submission times. ICASA typically alerts stakeholders of such issues through media releases. This was not the case in this instance.

42. This is a consultative process, where the more input received can only enhance the process through the submission of each additional perspective tabled. As it is not a
    competitive process the only prejudice suffered is to the process and the collective being deprived of the additional input. This is exacerbated by the fact the submissions rejected were by our smaller members whose perspectives can only enhance the process. This is important especially as ICASA has consistently and steadfastly indicated its pursuit of increasing the number of competitors in the market. This appears to be a counter-intuitive approach.

43. Had this been a competitive process such as a license process or licence amendment, licence transfer, or transfer of control application, the applicant may have been prejudiced by late submissions. In such an instance it would be appropriate for ICASA to reject late submissions.

44. In addition, on 9 November 2021, the SACF wrote to ICASA seeking condonation for submissions submitted after 12 on the 2nd of November. And we set our reasons as explained above, the fact that ICASA published the extension so late, it was conceivable that stakeholders could have missed the extension gazette.

45. ICASA wrote to the SACF on 12 November 2021, rejecting the SACF’s application for condonation indicating that in its view stakeholders would suffer no prejudice as they would have an opportunity to submit comments with the 2nd IM.

46. The SACF respectfully disagrees with this view, as we are of the view that key insights have unnecessarily been omitted.

47. We note with concern that none of this had been included in the reasons document which in our view should have included this for the sake of completeness, fairness, and transparency so that stakeholders are aware of the quantum of stakeholders affected by this decision and the faux pars in extending the submission deadline.

48.  In the interests of transparency, ICASA must include a list of submissions rejected. Two draft IMs vs the inclusion of a draft ITA

49. To reiterate our position, the SACF is of the view that critical high-demand IMT spectrum must be licensed as a matter of urgency following a thorough and sound process, as South Africa cannot afford yet another delay to the licensing of the spectrum.\

50. We noted ICASA’s rejection of our submission in first IM for the need for a draft ITA. Due to the significance and importance of the point we will again explain the rationale in our view for the necessity of a draft ITA.

51. The SACF welcomes the articulation of the process in the IM and the acknowledgment that the IM equates to a discussion document. Although the publication of the IM isn’t required it is welcomed. Equally a draft ITA isn’t required but it is good practice and the process outlined by ICASA has demonstrated that there is indeed a precedent for the publication of a draft ITA.

52. In addition, the truncated timelines for the WOAN indicated that ICASA would publish a draft ITA. We agree and support that approach.

53. An ITA is the specifications or terms on which interested licensees or stakeholders may apply for a license or a resource such as a spectrum. There are precedents set by  ICASA where draft ITAs have been published for comment. That in our view allows stakeholders an opportunity to highlight areas of concern and equally allows ICASA an opportunity to correct unfortunate errors. The SACF’s comments are geared toward reaching a successful conclusion to the licensing of the high-demand spectrum.

54. During the 2015/16 attempt to licence spectrum, ICASA adopted a process that included an IM for consultation followed by a final ITA. Accordingly, ICASA over its numerous false starts to licence high demand spectrum has adopted different processes creating varied precedents. During the 2010 /2011 process it published a draft ITA. Therefore, it follows that ICASA could follow either process as there are precedents for both. However, as almost all our members who have participated in the consultative processes for the licensing of spectrum have on more than one occasion requested a draft ITA, it would seem to be a pragmatic approach to allay the apparent concerns of key stakeholders.

55. This in our view establishes the precedent. Precedents, where ICASA has published a draft ITA for comments, includes the following:
    I. Annexure B of ICASA’s truncated licensing timeframes indicates that ICASA will publish a consultation document in respect of the Draft ITA for the I- ECNS and Radio Frequency Spectrum Licences for the purpose of
    operating a WOAN.
    II. Draft Invitation To Apply to solicit public comments on the award / granting of Radio Frequency Spectrum Licence to provide mobile broadband wireless access service for urban and rural areas using the complimentary bands, 800 MHz and 2.6 GHz in General Notice 912, Government Gazette No. 34872 of 15 December 2011. The Draft Invitation to Apply was subsequently withdrawn. We note that ICASA on 19 November 2021 instead announced the review of its tactical approach to the licensing of the WOAN, which temporarily stayed the process.

56. We, therefore, disagree with ICASA’s argument for why it cannot publish a draft ITA inviting commentary from stakeholders. There is indeed precedent and considering
    the number of false starts for the licensing of the spectrum, it would be prudent for ICASA to publish a draft ITA for comment for a reasonable period that allows stakeholders sufficient time to interrogate and comment on the draft ITA as well as allow ICASA sufficient time to consider all comments.

57.  Again, our objective is for the high-demand spectrum to be permanently licensed on an urgent but well-considered basis.

58. The 2nd IM even though it is an update of the 2020 ITA, it remains a discussion document with tentative views and positions. As evidenced in the 2020 defunct ITA, there was a significant difference between the IM and the ITA, which created a myriad of challenges.

59. The SACF has continued to ask for the publication of a draft ITA, as we are of the view that this would provide all interested parties with an advanced view of the ITA. This we expect will address and we are hopeful eliminate the potential challenges in the process.

60. We acknowledge that this may be viewed as slowing down the process, but it our view does mitigate the risks, finally contributing to the successful conclusion of the licensing of spectrum.  SPECTRUM AWARD
    Principles underpinning the Licensing
    SACF Submission on the 2 nd IM 30 November 2021

61. The objects of the ECA include the promotion of competition, innovation, and the efficient use of the spectrum. Spectrum is a key enabler of competition in a mobile environment therefore access to spectrum on a fair and equitable basis is important. Therefore, the spectrum must be assigned on a fair and equitable basis so that it does not give any licensee an unfair and exaggerated advantage over any other licensee.
    Therefore, access to each band should be equitable.

62. 5G rollouts are predicated on larger assignments, typically between 80-100MHz per licensee. Most licensees in their current assignments and in new assignments will not
    be able to access similar assignments. The assignments are typically significantly smaller.

63. For example, Telkom is currently the only licensee in the IMT 2300 band, with the largest contiguous spectrum which gives it dominance in a critical 5G band. Granting it further spectrum in the same band will create absolute dominance in this band and an unassailable advantage.
    Inclusion of critical high-demand bands

64. We note the exclusion of the IMT 2300 band from the current licensing process. The reasons for its exclusion have not been set out, so the SACF can only speculate on the reasons for its exclusion. The 2015 assignment plan had set a feasibility study as the prerequisite for the licensing of IMT 2300 band. However, in Telkom’s recent application to amend its spectrum licence, Telkom had indicated that it had migrated its fixed wireless links out of the IMT 2300 band thus eliminating the need for the feasibility study. As a result, the current usage in the band is limited to IMT only.

65. Therefore, the SACF is of the view that the available spectrum in the IMT 2300 band ought to be included in the current auction, as it will give all interested licensees a fair opportunity to access the spectrum. During the temporary spectrum licensing process, we have noted the interest of other licensees in the IMT 2300 band.
    Spectrum Pricing

66. Spectrum is an input cost to extending coverage in support of connectivity. Therefore, the pricing of the spectrum must be balanced to be a fair reflection of the accurate value of the spectrum and attract serious bidders who can use the spectrum but not be prohibitively expensive that it retards the rollout of infrastructure.

67. The inclusion of high-demand bands is anticipated to attract competitive but fair pricing.

68.  While licensees may be determined to access the spectrum and are likely to bid aggressively on the more competitive lots, it, however, is less realistic to expect that licensees would be willing to pay an unrealistically high price for the spectrum. Licensees will have to rollout after acquiring the spectrum and the IM significantly increases the overall price of spectrum for Tier 1 licensees because of the differences in coverage obligations, the opportunity cost of exclusion during the opt-in round and then the acquisition of lots still available are expected to attract a higher price.
    Pro-Competitive Measures

69. The ITA process includes pro-competitive measures for smaller licensees to access spectrum more easily through the introduction of the opt-in process. If applied incorrectly will enable certain licensees categorized as Tier 2 licensees, the first option on access to spectrum which may prejudice licensees that these pro-competitive measures are designed to help.

70. The competitive assessment has focused on historical information, almost at the expense of current data on the current growth of operators and the respective market share and revenue growth. A comprehensive review of the significant changes to the market must be conducted before applying pro-competitive measures.

71. The opt-in round and the MSPs are mechanisms that ICASA has included as pro-competitive measures to increase the spectrum assets of licensees towards making them credible licensees. However, the table below demonstrates the overall spectrum assignments and raises questions on ICASA determinations for the MSPs.
    Operator| 900| 1800| 2100| 2300| 2600| 3500| 3700| Total
    ---|---|---|---|---|---|---|---|---
    Telkom| | 24| 30| 60| | 28| | 142
    Rain| | 34| | | 20| | 80| 134
    Liquid| 10| 24| | | | 56| | 90
    MTN| 22| 24| 40| | | | | 86
    Vodacom| 22| 24| 35| | | | | 81
    Cell C| 22| 24| 30| | | | | 76
    Available| | | 140| 86| | | |

72. When the spectrum holdings are considered, together with market share and operator growth, ICASA may be giving some licensees an unfair advantage as a result of being
    mistaken for a small operator or a new entrant to whom pro-competitive remedies should legitimately be apply. This is part may be attributed to the considerably out of
    date information that has been used. There have been significant changes in the market since the competition assessment was completed.

Auction process

Opt-in

73. The opt-in round of the auction has been contentious in the 2020 ITA for the absence of consultation and the prejudice and exclusion of some operators. Therefore, it is important that this carefully and correctly remedied and does not create grounds for a new challenge.

74. It is important for ICASA to consider the competition assessment of the market and current spectrum assignments when considering eligibility and Minimum Spectrum Portfolios.

75. In creating the MSPs, ICASA must be careful to not give some licensees an unfair advantage over other licensees. MSP 2 is of concern as it appears to give some licensees an unfair advantage over others because of the current spectrum assignments. The SACF, therefore, urges ICASA to relook at the eligibility criteria for MSP2.

76. After the Opt-in Round, the Authority will announce the number of Bidders that submitted valid Opt-in Bids, the identity of the winning bidders for each MSP, the spectrum that they will be assigned, and the prices that they must pay for that spectrum. The Authority may also publish the same information on its website. This should not be discretionary – may change to must publish the information on ICASA’s website Hybrid Auction Methodology

77. The introduction of a hybrid auction methodology is a pragmatic approach as it has the potential to mitigate potential access challenges of the pandemic, especially as it allows teams to operate from different and presumably multiple locations if required. We note that mock auction will serve as the dress rehearsal for the successful peration
    of the platform. Despite the successful operation of the platform during the mock auction, there is no guarantee that there will be no technical glitches during the actual auction. The reasons document and the IM do not set out the mitigation strategies that would apply in the event of technical challenges during the actual auction. We urge ICASA to include the mitigation and processes that will apply during the auction in the event of technical glitches including latency to provide comfort to licensees.

78.  The rationale for the main auction being online but not the opt-in is unclear. Therefore, we would appreciate the clarification of whether this is a deliberate decision or  an omission and the rationale for its exclusion from the online option.
    THE LICENSING OF IMT 700 AND IMT800
    Licensing of IMT 700 and IMT 800 now regardless of digital migration

79. The SACF is of the view that IMT 700 and IMT800 must be included in the current licensing process. Excluding it from the current licensing processes is impractical as it would deprive licensees of access to critical coverage spectrum bands, and it would be difficult to envisage a scenario where the WOAN would be able to efficiently rollout infrastructure in the absence of access to IMT 700 and IMT800.

80. We are aware that there are legal challenges underway and that the SABC during the recent hearings on the Draft National Frequency Band Plan have indicated on its  lack of readiness to migrate. This position is further exacerbated by the critical global shortage of chipsets. As a result of the above the SACF is of the view that the
    completion of digital migration by the end of March 2022, however determined may not happen as envisaged.
    Increase spectrum usage as more of the bands become available

81. We recognize that access to these bands will be delayed. Notwithstanding, the challenges to accessing these critical coverage bands, the SACF is of the view that these bands must be included in the current licensing processes regardless of whether digital migration has been completed or not.

82. While access to these bands is not optimal and is unlikely to be until digital migration has been completed, it has nevertheless proved to be useful when licensed under the temporary spectrum framework.
    Proportionate Licence Fees until all IMT 700 and IMT800 is available

83. Temporary and provisional access to spectrum has created a licence regime that attracted proportionate license fees which we anticipate will continue to apply until all spectrum in these bands are available. We believe that this contributes to the efficient use of these bands rather than potentially lying dormant and squandering potential economic value.

84. However, obligations, especially rollout obligations must be deferred until all the spectrum in the band is available for IMT services.
    Principles underpinning obligations

85. In the 2020 IM and again in the 2021 IM, ICASA has set an obligation for operators to rollout services from the outside in. The SACF is of the view that this is not a practical approach as it fails to recognize current consumers as a result the rollout of temporary spectrum or the need for licensees to recover investments to expand on the rollout.

86.  The consequence of pressing ahead with this obligation is that licensees will no longer be able to provide services in urban and metropolitan areas and consumers currently
    being served may no longer be, this can never be ICASA’s intention. Our members submit regular reports to ICASA which includes that of the usage of spectrum, in support of these arguments.
    In addition, we operate in a difficult economic environment where South Africa has suffered significant downgrades increasing the cost of capital which is fundamental to
    spectrum acquisition and rollouts. The cost of access to capital is dependent on the strength of the borrower’s balance sheet. Licensees must be able to recoup capital laid
    out in acquiring spectrum and rolling out infrastructure. Investments are more easily recouped being able to rollout in a manner that matches demand.
    Spectrum Sharing

87.  The SACF recognizes that ICASA’s approach of use the assigned spectrum or shared access is part of current policy best practice. However, this is premised on licensees having full access to the spectrum assignment and the licensee being able to fully use the assignment. This is unlikely to immediately be the case in the IMT 700 and IMT 800 bands. Therefore, such as obligation can only be implemented once the licensee has full access to the spectrum assignment in the IMT 700 and IMT 800 bands. Anything else would be unfair and prejudicial.
    The cumulative impact of Obligations

88.  Considering the current obligation that the licensees have to the USAF, i.e., an annual contribution of 0.2% of their turnover derived from licensed activities, the SACF finds the additional IM obligation to provide broadband services to 97%/99.8 of the identified underserviced areas onerous as it translates to high capex which will be financially draining for the licensees. This, if not curbed and/or revised at the time of licensing may impede the objective of reducing the Cost to Communicate as the licensees will have no option but to transfer the cost burden to consumers.

89. In addition to the cost of spectrum through the auction, the IM seeks to impose additional licence obligations, which includes coverage and social obligations. These obligations cannot be viewed separately from the reserve price and instead must be viewed collectively in determining the reserve price and the rollout obligations.

90. The must be imposed in a manner that does not distort the market.

Open Access Obligations for the Industry

The WOAN is intended to promote service-based competition

91. The policy framework enables the licensing of the WOAN to create a Wholesale Open Access Network that can drive service-based competition. The various examples of WOANs globally which ICASA has based the WOAN have floundered and failed.

92. Therefore, the licence must promote the successful implementation of the WOAN. ICASA’s current approach in the licensing of high-demand IMT spectrum through the auction process appears to be at odds with the ongoing sustainability of the WOAN.

93. We note ICASA’s announcement on 19 November 2021 where it considers additional aspects of the WOAN for its licensing. We are concerned about the delays i   licensing and the model.

94. The WOAN as a wholesale open access network is premised on other licensees buying wholesale capacity from the WOAN and therefore imposing an MVNO obligation on Mobile Network Operators is at odds with the object and model for the WOAN as MVNOs are bulk resellers who sit on the networks of others and should therefore be a key target market for MVNOs.

95. Instead, ICASA seeks to place an MVNO obligation on the MNOs rather than the WOAN. This is in our view incongruous and at odds with the sustainability of the WOAN. Incongruous MVNO obligation

96. Mobile Virtual Network Operators(MVNOs) are designed to buy bulk capacity, repackage, and resell. Service is the differentiator.

97. It seems incongruous to the policy objective of establishing a WOAN that competitors (MNOs) would have an obligation that seeks to drive wholesale capacity to MNO
    competitor networks rather than drive the traffic to the WOAN.

98. The sustainability of the WOAN has always been questionable, which is why MNOs have the obligation to buy wholesale capacity from the WOAN for a period of 5 years.
    However, this is a pro-competitive measure that is only intended to help a new entrant. The objective ought to be to create an environment to encourage and sustain traffic for the WOAN. Therefore, imposing an obligation on the MNOs that deliberately detracts from this objective is contradictory and illogical.

99. During the 2019/2020 consultations, most stakeholders raised concerns with the imposition of the MVNO obligation, yet ICASA has doggedly pursued this.

100.  Before ICASA imposes an MVNO obligation on any licensee it must first conduct an in-depth review to understand the critical success factors of MVNOs. The case studies exist and some of our members have conducted their own research into the success and sustainability of MVNOs and will elaborate on them in their submissions.

101. Firstly, it is difficult to understand how ICASA envisages an independent firm that provides a service or enables access to its platform even at preferential rates can ensure the sustainability of another independent firm.

102. Our members for example have enterprise development programs, where they support, mentor, and train the owners and managers of these entities. Even in that environment, they do not have a 100% success rate with all the entities, although they have a high success rate. It simply demonstrates that one entity cannot guarantee the success and sustainability of another.

103. If an MVNO fails within the three years what does ICASA intend to do?
     a) Does ICASA envisage sanctions against licensees should an MVNO fail? If licensees will assist on a best effort basis, what will be considered the best effort?

104. The current IM requires licensees assigned spectrum through the auction process to provide open access to MVNOs. There is no number specified. We would like clarity on the following:
     a) Is the number of MVNOs on each network left to the discretion of each licensee?
     b) Are there criteria to determine how many MVNOs that must be attached to a licensee? If so, what are the criteria?

105. The criteria for MVNOs and the obligations must be explicitly stated as these factors are contributory to determining the cost of access to the spectrum.

Uplink and throughput obligations for the Industry

Social Obligations for the Industry

106. The cost of the obligations must be viewed collectively with the price of the spectrum as they cumulatively are input costs that ultimately impact the national project of reducing the cost to communication.

107. The obligation to zero-rate websites was first introduced under the COVID 19 regulations which provided valuable lessons. The COVID regulation extended to the zero rating of educational content and medical institutions. The number of sites grew exponentially since the implementation of the obligations and have been and continue to be the subject of considerable abuse.

108. Extending such a benefit on such a vast and open-ended scale across the economy in the absence of cost analysis to understand the costs of rollout would be negligent and unfairly burden licensees.

109. It is not feasible to Zero rate an unlimited amount of urls. There need to be a limit as operators need to balance capacity and also free traffic vs commercial traffic so that paying clients are not put at a disadvantage or subjected to poor quality of service.

110. Operators have reported high incidents of data tunneling – unscrupulous users tunnel through zero-rated URLs to URLs that are not free and use large amounts of free data on services that are not zero-rated. Best way to mitigate – a daily cap and a monthly cap. Also, no dynamic IP’s and no cloud-based URLs must be zero-rated as
     this exposes us to fraud and tunneling.

111. It is important for ICASA to understand the costs of the implementation of the zero-rating of the specified websites under the COVID regulations as well as the quantum of abuse.

112. Such an obligation would unfairly burden the licensees and therefore should not be included as an obligation for the licensing of the high-demand spectrum. It is in our
     view this is a significant impediment to the national priority project of reducing the cost to communication.
     Empowerment and Transformation Provision for the Industry

113. The SACF believes that an aggressive approach to transformation is imperative, but the Authority must be cognisant of the following:
     a) The Codes are outside of the jurisdiction of the Authority. Historically, the Codes have had radical amendments which were implemented immediately. The consequence was that all our members dropped several levels due to the absence of the lead time to prepare for the implementation of amendments. They then clawed their way back, attaining progressively higher levels of compliance.
     However, compliance with the Codes comes at a significant cost.

114. This context is critical, particularly as we understand that the compliance levels would be included as licence obligations. Accordingly, we are of the view that the
     Authority would need to provide for changes that are beyond the control of licensees.

115. Therefore, we would propose an additional clause that makes provision for a reasonable transition period for compliance, with a minimum of 12 months. However, it is critical that the transition period is aligned with the materiality of any amendments.
     The evidence of the materiality would be demonstrated by the level by whic compliance levels fall. Such a transitionary period is essential due to the applicable costs associated with achieving BBBEE compliance.

116. The objective is to promote meaningful transformation without being too onerous and prejudicial to licensees.

117. While most of the SACF members participating in the licensing of high-demand spectrum are currently at a Level 1 or 2. Therefore, in the short-term this obligation is largely achievable for our members. However, it is high barrier to entry for licensees who have achieved lower levels of compliance with the B-BBEE Codes and currently only the minimum entrance requirements of Level 4 compliance. Attaining Level 1 within 12 months may be unachievable and exclusionary for such entities.

118.  It is also important to note that licensees must set aside an annual budget for compliance to attain the Level 1 status. This is an annual commitment and not as straightforward an obligation as it may seem.

119. This obligation would apply for the term of the licence as the market conditions change and market and economic conditions could also affect this level of compliance. Consequently, ICASA should build in a level of tolerance for potential fluctuations and consequent changes over the 20-year licence period.

Transformation through procurement

120. It is important to recognize that the ownership of larger licensees is more restrictive due to the associated capital requirements and the number of licensees of scale in any given market which limits significant levels of ownership by Black people. This does not mean that our members do not support ownership by Black people, it simply means that there should not be a single and unnuanced approach to transformation.
121. Procurement is an immediate and valuable tool that has the potential of empowering small businesses and black businesses in a relatively short time and potentially becomes a feeder mechanism for more graduated and capital-intensive empowerment.
122.  All licensees procure significant components of network elements. There are many Black-owned companies able to provide such services, if there are not, it must be incumbent on network operators to develop these suppliers. This in our view will engender more meaningful and inclusive transformation across the value chain also increasing the value of annual investments within the country.
123. Our recommendation for the inclusion of procurement targets but does no negate the Authority’s proposal for the progression of compliance with the B-BBEE Codes.

Conclusion

124.  The SACF welcomes ICASA’s intention to move with speed to begin the licensing of the high-demand spectrum and its attempts to consult. However, it is essential that the consultation is not a superficial exercise.
125. Instead, it must be comprehensive and thoroughly and meaningfully interrogate the arguments of stakeholders and ensure a fair and transparent process. The consequences of this licensing process are far-reaching and will have an effect for two decades. It is also a signal to investors that are essential to infrastructure sectors.
126. The SACF therefore strongly urges ICASA to abandon its highly truncated and deficient timetable for the licensing of the high-demand IMT spectrum.
127.  Thoroughly consider the submissions of stakeholders and properly incorporate these comments into a draft ITA, which in our view will significantly contribute to lessening errors and reducing the risk of successful legal challenges.
128. South Africa cannot afford any further delays in the licensing of the high-demand spectrum, nor can ICASA withstand the same legal challenge as it weakens and undermines its credibility as the industry regulator.
     SACF Submission on the 2 nd IM 30 November 2021

Read User Manual Online (PDF format)

Download This Manual (PDF format)

Download this manual  >>