| Comment | Reply comment |
|---|
| Telecom Namibia Limited |
|---|
| The 2013 General Policy Guidelines on Universal Access and Service in Communications (“the 2013 guidelines”) state that, a formula for charging universal levy is to be determined by the Regulator from time to time, following a consultative rule-making and market review process. We do not agree with the proposed formula and recommend that a percentage should be set to apply to all equally because the award of the tender will be equal to all irrespective of contributions made. We also do not see how the Authority has meaningfully considered the market review process in the determination of the levy because, if one is to consider the Gap analysis report provided by the Authority, 5 regions are without the required 4G coverage. The Government has provided funding for purposes of universal services for the next 3 years and for year one, the money allocated is in excess of what was initially required for year one of the universal funding by the Authority. In what way has this funding been considered in the proposed levy? We submit that the proposed levy will result in over-recoveries and such over-recoveries should in our view be considered from the first imposition due to the fact that the funding is already committed by the government before file levy is imposed. The inflow from licensees will therefore not be the first funds to be received for universal services and as such the consideration for overrecoveries is relevant in file circumstances. We further recommend that since the government has made a 3 years commitment towards universal services, the Authority should consider postponing the charging of universal services from operators and utilise the government funds for now.In terms of the 2013 Guidelines, the expenditure from the Universal Service Fund should be undertaken in accordance with implementation plans developed by the Regulator from time to time via a public consultative process. We are not aware of the implementation plan agreed upon.Furthermore, The Authority awarded 703-788 MHZ AND 790-862 MHZ spectrum last year with mandatory roll-out obligations aimed at attaining 80% 4G coverage in specific regions which would in turn reduce the Capex required to roll out universal services and hence reduce the proposed fee. We now see that some of the regions appearing both in the draft regulation and concerningly in the bid document for universal services are the same regions allocated to M TC and Loc 8 under 703-788 MHZ AND 790-862 MHZ spectrum obligations. Under the 2013 guidelines, the Authority is required to publish the progress of the rollout of these universal service obligations on its website upon receipt of the progress reports from the operators. Since the Authority had advised that these mandatory sites were imposed for purposes of universal access, the regulator needs to maintain transparency with universal service obligations and reports and publish the 703-788 MHZ AND 790-862 MHZ mandatory sites rollout on its website, in the absence of an agreed universal service implementation plan, the publishing of this report will assist all stakeholders in avoiding duplications and establishing which sites are already covered under 700 800MHz mandatory obligations and which sites are yet to be installed under license conditions. Furthermore, the proposed levy does not take into account other rollout obligations that licensees have been subjected to do over the years e.g. Telecom was expected to roll out fixed and mobile services to unprofitable areas. | The formula proposed for the Universal Service Levy differs fundamentally from that used for purposes of the Regulatory levy. This levy is different from the Regulatory levy in that the levy proposed charges all telecommunications licensees the same percentage of the telecommunications revenue, namely 1%. It is not a glidepath levy formula which charges more as the revenue increases. All telecommunications licensees except non-profit licensees will pay the same.The Authority analysed the market in line with two policies, namely the General policy guidelines on universal access and service in communications 2013 and the Broadband Policy of 2009. At the same time all the proposed towers that are planned by all licensees that were granted spectrum in the 700 MHz and 800 MHz band were taken into consideration. Furthermore, the Authority used a demand stimulation methodology to ensure that the towers would become financially viable in the future. Therefore, subsidies either towards capital or services will be paid to licensees. The Gap Analysis Report attached shows that there are still many areas of which only 119 were identified that will still not have services even with the roll-out of the 700 MHz and 800 MHz spectrum obligations.Consequently, the money that was provided by the Government towards the Universal Service Fund will only be enough to cover the capital cost of 42 of the 119 towers required. There will thus be no over-recoveries and if such over-recoveries should be realised it will result in a lower levy once the levy is reviewedThe overreaching goal of the universal service as per policy statement 6.3.3 of the General policy guidelines on universal access and service in communications 2013 states as follows “to achieve a 98% level of universal access in respect of telephony, broadcasting and broadband services within a period of 10 years or less for individuals, households, businesses and marginalised groups in all cities, towns, villages, settlements and nomadic communities; clinics, schools, libraries and public facilities; small firms, mines, farms and nature reserves; and places where poor households are unable to afford access” However, due to the litigation matter that was on going, this goal was not realised.The first Universal Gap Analysis Report was published in 2021. Consequently, this Report was updated in 2023 and 2024, respectively, as discussed with all licensees at several different engagements.All the proposed towers that are planned by all licensees that were granted spectrum m the 700 MHz and 800 MHz band were taken into consideration in the Gap Analysis Report and calculation of the proposed levy. Even with the roll-out obligations, regions such as Kunene and Omaheke will not reach 80% broadband coverage due to small population numbers in many of the areas. A subsidy would therefore be required. |
| We submit that the proposed levy is too high and does not consider other fees, levies and charges paid to the Regulator by the licensees. In terms of the 2013 guidelines, contributions to the universal levy by licensees should not be unduly onerous. The proposed levy should therefore not be more burdensome than necessary on the licensees taking into account the affordability of licensees and the impact of such a levy on the sustainability of the licensees. For Telecom, some of the fees paid to the regulator per annum are as follows: Regulatory levy about N$ 14 million Number resource -about N$ 500 000 Spectrum fees — above N$ 7 million Universal levy— may be as high as N$ 14 million Looking at the above and considering Telecom’s financial position, it is no secret that Telecom is unable to afford the regulatory costs to the extent that we have multiple payment arrangements with the Authority for payment of the regulatory levy, for which we continue to be exposed to exorbitant interest charges. When regard is being had to the total fees paid to the regulator in any given financial year, we note with concern that the price of Regulation that licensees pay to the regulator is high in some cases more than our profits in any given year. | The Authority has taken all the comments into consideration and a levy of 0.5% will be implemented to reduce the short-term impact on licensees. |
| In terms of aligning the proposed levy with regional and international best practices and avoiding placing an unreasonable burden on licensees, we considered the levy applicable in the region. In South Africa, licensees seem to pay 0.2% of turnover for universal levy and in some countries like Botswana, the government subsidies the operators, so the universal levy charged does not make a good comparison case. | In response to the different benchmarked rate of levy in comparison with the amounts levied in other jurisdictions in the region such as South Africa and Botswana, the levy is relatively low due to the maturity of those funds in comparison to our infant Fund. As a result of the length of service of these funds, these levies have been regularly reviewed to factor in the role of the fund in comparison of the services that the funds are aimed at delivering. In those countries for instance, the fund is not utilised towards services and capital but to fund other obligations such as computers to schools, ICT teachers, etc. However, the expectation is that as the fund matures and obtains its objective, this levy may be reduced. A table on oher jurisdiction's levies were added to the document. |
| It is common cause that the Authority made N$ 28,500,000.00 (twenty-eight million live hundred thousand Namibian dollars) in the spectrum auction of the 700 800 MHz, We maintain that the Authority should fund the universal levy from the proceeds of spectrum auctions as well because the funds generated from these auctions should not form part of the money required to defray the Authority’s expenses or to cover the spectrum management costs because these two items are covered by the regulatory levy and the spectrum fees respectively. The Authority has also advised that the spectrum management costs are paid for by the spectrum fees hence the reason why the Authority increased the spectrum fees again recently. The money received from spectrum Auction can therefore be used for the universal levy. | The funds generated from the spectrum auctions will be exclusively allocated to defray the expenses associated with spectrum management and cannot be redirected to the Universal Service Fund or used to subsidise other services. Additionally, Telecom Namibia spectrum fees were reduced by 31% following the review conducted in 2023, which took into account the outcomes of the spectrum auction. |
| We propose that the draft regulations should have a provision to the effect that universal levy should be reviewed annually to address any over-recoveries timeously, as opposed to every three (3) years. We consider it necessary to review the levy annually to prevent taking more money from the industry unnecessarily, which money could have been directed toward technological innovations, commercial investments and network expansions as opposed to just addressing coverage. | In terms of section 23(9) of the Act the Authority is mandated to cany out a regulatory review for a prescribed regulatory levy inclusive of the Universal Service Levy every five years to ensure that there are no under or over-recoveries. Simultaneously, Universal Service Report with audited financial statements, will be published annually. This report will provide a comprehensive overview of the utilisation of the fund. Consequently, conducting annual reviews of the levy is unnecessary and it poses a risk which can impede licensees’ ability to engage in effective financial planning and hamper the objective of the Fund. |
| We further encourage the Authority to be more efficient and faster with its pace of designing the bidding process as well as the implementation of a Universal access programme to ensure that money allocated by the government goes into benefiting the industry in the fastest possible manner and to ensure that the current Universal levy strategy is not overtaken by market developments because telecommunications sector is ever-changing. | The Authority released the first bidding document for comments to the industry to ensure the release of funds to successful bidders and at the same time ensure roll-out to unserved communities. The final Bidding Document was released on 12 August 2024. |
| Finally, in terms of the Gap analysis study by the Authority, we note that the study still omits communities and customers stated in Regulation 5(2) of the Universal levy obligations of 2018. Our interpretation of this provision is that the fund is also supposed to subsidize the provision of telecommunications services to communities and customers in urban informal areas and households in urban areas where there is no coverage, and who are equally in dire need of telecommunications services. | The Universal Service Policy guidelines provide for the priorities in terms of roll-out of which educational and health facilities are the main priorities. By ensuring that these facilities have service will lead to many of the other groups to automatically also have services. Due to the limited amount of funds available the policy guidelines have to be utilised to determine who would benefit first. |
| Mobile Telecommunications Limited |
|---|
| Given the crucial role of connectivity and the externalities associated with connectivity providing universal service to households and communities is important. Internationally, governments and regulators have intervened in order to provide structures and regulations to incentivize the build out of infrastructure in underserved areas. Historically, this has often been in the context of expanding access being a licence requirement for fixed-line network providers ate a time when telecommunication was considered to be a “natural monopoly”. However, incentivizing infrastructure build out in this manner has been met with certain challenges and there have been mixed experiences and results in terms of the operation and efficiency of universal service funds. Furthermore, markets have evolved substantially as new forms of technology and competition emerged. As a result, the international best practice for designing and funding universal services has evolved. At present there is a range of models used to expand coverage. This includes traditional infrastructure subsidization in some countries, but also demand side subsidies for users. In considering the approach used by CRAN there are a few key considerations and principles we discussed in the section below. Universal service funds could shift spending that would have improved coverage, quality and cost. Many countries in Africa have established universal service funds. Where this fund is funded by industry levies it ultimately is a form of cross-subsidy between existing customers of the MNOs and new customers in underserved areas. As such, it could lead to higher communication costs or less investment as the funding earmarked for other uses is shifted to universal service projects. For example, where the MNO has a roll-out plan and strategy, which may include upgrading of sites in existing locations from 2G to 4G, migrating to a different spectrum band to improve coverage in the vicinity of existing sites and funding promotions or special offers, a compulsory levy may constitute an opportunity cost in terms of the company’s own expansion plans, impact on quality, coverage and costs of a different group of customers. As such, there is a concern that if there is insufficient consultation the imposition of a particular universal service plan may crowd out private investment or spending in other areas that would have been beneficial. This is particularly the case if a company is still in the process of expansion and the universal service plan priorities sites that may be less effective than expansion plans which would have been engaged in by operators absent the fund or priorities areas that would have been covered in terms of the long-term infrastructure rollout on a commercial basis in any case.As a result, infrastructure rollout plans by the USF should be designed after significant consultation with operators about (i) their existing rollout and expansion plans, (ii) the most cost-effective means of increasing coverage given the existing network and topology of the area. | In accordance with the discussion document, it has been determined that there are only fourteen (14) sites in Namibia that remain commercially viable. Consequently, the under-served and unserved areas cannot be serviced by licensees. Furthermore, the Authority has taken into account all roll out plans submitted by licensees for the forthcoming three years, which formed basis for the development of the Gap Analysis Report. The Authority employed a demand-stimulation methodology to ensure that the sites to be rolled out in under-served and unserved areas attain commercial viability. Should any of these sites fail to achieve such viability, the Fund’s purpose is to subsidise the sites until they become commercially viable. |
| Universal service funds should not collect more funds than are required. The financial model used to estimate needs should be based on accurate data and financial information. Mechanisms should be built into the design so that additional revenue that is not used leads to adjustments in contributions. | In response to the concerns raised regarding universal service fund collecting more funds then required, it is important to note that in accordance with Section 58 of the Communications Act, 2009 (Act No. 8 of 2009), the Authority is required to publish an annual Universal Service Report. This report includes audited financial statements that provide a comprehensive overview of how the funds are being utilised. The Authority had engaged with various licensees, including MTC, to obtain estimates for the costs of towers and services. The calculations for the levy were based on this information. Additionally, pursuant to section 23 of the Amended Communications Act, levies must be reviewed within a five-year period. Should there be any over-recoveries or unspent funds, these will be considered and offset against the proposed budget for the next cycle. |
| An important consideration in planning for extended coverage is the extent to which the target population is sufficiently ready and enabled to utilize the infrastructure being provided. This requires assessing gaps in adoption, uptake and use, skill gaps and SME development and innovation gaps. Research ICT Africa found that access to devices was a greater barrier to the Internet for consumers (28% of survey respondents) than the availability of mobile services (6%). MTC data shows that 36% of devices on its network are basic phones and 32.9% are 2G feature phones. This means that two thirds of the network comprise devices that do not support internet services. Smart phones only make up 25% of devices. There are a number of initiatives that the regulator could undertake to improve device affordability and thus increase uptake, including by advocating for the reduction in the exceedingly high taxes on smartphones. While the current proposal includes the provision of solar kits and Wi-Fi routers at schools it is not clear that a full analysis of the device availability and demand in these areas within and beyond schools has been engaged in. Provision on the supply side will not have an effect unless there is complementary investment and resourcing on the demand side. While CRAN has considered anchor demand through schools by providing uncapped 4G, it is not clear that the complementary strategies to provide these schools with skills, devices and training is also being considered and costed so that it is adequately utilized. A demand-side analysis to consider device penetration and affordability needs to be engaged in for each site. This can include assessing utilization and penetration in surrounding areas. Investment in suitable skills and devices are a pre-requisite to ensuring uptake. | The Authority had conducted extensive engagements with all telecommunications licensees since 2021. In response, the Authority engaged the Ministry of Finance and Public Enterprises to secure approval to secure approval for tax exemptions on the importation of smart devices, aimed at ensuring that all Namibians have access to the necessary devices. Additionally, the Authority is actively collaborating with the line Ministry to enhance digital literacy across the country. The Authority also completed a Pricing Study to evaluate the affordability of services and is currently reviewing the regulations prescribing the quality of services standards applicable to service licensees to guarantee that all Namibians have access to affordable and quality services. The implementation of the Universal Service Fund Levy will be aligned with the demand stimulation model and not earned out in isolation. As stipulated in section 56(4) of the Communications Act,2009, the funds may only be utilised to cover the administrative costs for managing the Fund and to subsidise capital expenditure and/or services. Consequently, the Fund cannot be used to subsidise other goods and services such as devices and framing which is why these costs were not factored into the considerations. |
| A full suite of tools to improve coverage should be considered. Apart from Universal Service Funds there are other tools used by regulators including spectrum coverage obligations. In Namibia, spectrum obligations have been successful in expanding coverage in line with national obligations to 88% of file population by allowing operators to engage in their own planning and prioritization. Additional policies can include cost reduction policies, for example, by reducing duties and taxes on network equipment for the purpose of expanding services, rebated on licensing and spectrum fees, and reduction in tax on devices on the demand side. Demand side subsidization can also create sufficient stimulus for private provision while limiting competitive distortions. Public investment in open access infrastructure is also another tool for enhancing coverage. Reducing barriers to building new infrastructure should also be considered to expand universal service. In particular, there are very long delays in approval for new sited of between 3 and 8 years. A range of policy options should be considered on both the supply and demand side with the aim of minimizing distortions. | The Authority followed a holistic approach to allow for roll-out of services through licensing obligations especially on high-demand spectrum. Local Authorities, especially the City of Windhoek was engaged to ensure faster approval for sites, and other barriers have been addressed to assist licensees with the provision of services and we will continue to assist and engage wherever possible. |
| Different financing options should be considered.Consideration should be given to alternative sources of financing so that the cost is not entirely born by the operators alone. | CRAN takes note of the recommendation and engaged the Government who provided N$ 115 million towards the Universal Service Fund. |
| Technology neutrality should be observed in the use of USF funds. CRAN proposes one specific technology, in particular 4G mobile services, in its universal service plan. However, there are a variety of technologies that can be used to provide broadband services in addition to 4G mobile, including fiber to the premises, satellite and fixed-wireless access services. A technology neutral approach ought to be considered in which quality might be specified (such as average download speeds experienced by users) rather than the technology, and thus the full range of available technologies should be considered by CRAN, rather than only 4G services. | CRAN sets 4G technology as a minimum requirement, based on the assessment of the quality-of-service results for 3G which indicated that in many areas 3G does not meet the broadband requirement as set out in the Broadband Policy of a download speed of at least 2Mbps. Successful bidders are free to deploy any additional technology on the sites or use faster technologies. |
| Comments on methodology Various assumptions are made in the model. However, our assessment suggests that certain assumptions are incorrect and that this is leading to an overstatement of costs.That building new RAN sites are the most effective model for universal service in outlying areas. It is not clear that this is a correct assumption. New RAN sites are not necessarily the most cost-effective way to cover areas with limited coverage. Changing spectrum usage of adjacent sites would be more cost-effective in many cases. MTC data suggest that 15 out of the 49 sites identified by CRAN and not covered at present by MTC could be covered with an upgrade in spectrum to 800MHz. This would be significantly cheaper than building new sites.Other technologies should be considered. In other areas (such as Kunene) satellites including low earth orbital satellite for example can also provide alternatives. In many areas fixed options direct to a school may provide better-quality alternatives for schools. As such, calculating a levy using a model based on new mobile sites instead of also considering alternative means of extending coverage and alternative technologies could overstate the investment needs. While the document does note that some gaps could be closed using community-based approached and satellite services this does not seem to have been assesses more specifically.That a radius of 8km is appropriate. Coverage is calculated based on using a model with a radius of 8km, stated to be based on GSMA parameters. However, this radius may not be correct for all frequencies with 800-900MHz typically allowing for higher penetration and 2100-2600MHz allowing for lower penetration. A raft version of the document based on reaching universal access and service objectives through the assignment of 800MHz and 700MHz spectrum states that it uses 12km for 900MHz, 8km for 1800MHz and 4km for 2600MHz, but it is not clear form what has been presented how this has ultimately been modeled as the final document only refers to the 8km radii across the board. While MTC data does not use a 8km split we see a large number of connections are a radius of more than 6.6km with a non-trivial share being 14km, particularly for LTE in 900MHz which has 14.3% of connections. MTC data therefore suggests that sites in nearby areas have at times attained coverage that is narrower or broader than 8km in many instances. This calls into question the accuracy of the estimates of the number of people that would be covered by a site. The sites selected by CRAN in their model are incorrect and do not appear to be selected in terms of suitable criteria. MTC suggest that there are big differences between their actual coverage maps and those used by CRAN. In particular, 30 sites that CRAN assumes are uncovered are actually already covered by MTC network. Furthermore, a many of the sites in the model are already covered y 800MHZ spectrum obligations. For example, this includes several sites in Kavango’s such as Ncamagoro and Mpungu. The model selects sites in regions that meet the coverage target of 80% ad fails to prioritizes regions with lower coverage. | The first phase of UAS centres around the rollout of new RAN sites. Other technologies, including satellites, may be utilised to reach schools that cannot be reached by mobile broadband in a later phase. The 800MHz spectrum allocation was incorporated into the radio propagation model, based on the submissions of licensees regarding the spectrum use per cell. The UAS require a minimum of 4G coverage for an area. If licensees are able to provide this using existing towers, by upgrading backhaul and different spectrum use, then this will be considered by CRAN. Section 23, as amended, makes provision that the levy should be recalculated at least every five years. The Authority will regularly review the levy as per the methodology where over recovery is subtracted and under-recovery added to the revenue requirement of the Fund to ensure that the levy is not too high. |
| The calculation of unmet demand: The calculation for unmet demand assumes that 30% of the population in those areas will adopt broadband and that they will have an ARPU of N$ 60. It is not clear what the estimates are based on. It can be noted that many sites have coverage in surrounding areas and some people in these areas are already subscribers with SIM cards that they use when they go into an area with coverage. As such there is not necessarily new uptake from those areas by rather incremental usage by some existing subscribers who are able to use their SIM cards ate the location where they currently use it. Secondly, it means that the ARPU of N$ 60.00 is likely overstated for these customers as there may be an incremental increase if they can access a network from their current location of use, but this is not necessarily equivalent to the added revenue from a new subscriber. Thirdly, MTC data based on nearby sites suggests that population coverage in sites in the area of those identified by CRAN is likely lower than is estimated in the model. | CRAN hosted four public engagements on the Universal Service Access Report between 2021 and 2023 and one engagement with MTC’s technical team where the methodology for the radio propagation model and Gap Analysis was shared. The recommendation from the MTC technical team was to reduce the coverage from a 12km radius to 8km radius for coverage. This immediately increased the number of sites required and therefore also the costs. CRAN did a site verification exercise in July 2024 and the this indicated that the 8km radius suffice. With every bid this will be done to verify if a site, identified, has coverage and if there is coverage it will be indicated as such and the site removed. The RAN site locations were based on population not covered with sustainability being the main objective. Underserved areas in Omaheke and Kunene will be addressed through different Universal service obligations of which one is tied to the recent allocated of 800Mhz spectrum. |
| Finally, the model assumes that customers will be broadband customers using 4G. It is not clear from the MTC data that in many areas identified customers currently have the 4G devices. This is a critical aspect of the model. | The NAD 60 ARPU per month is based on the cheapest 7-day bundle available in Namibia, times 4 for monthly use. The 30% adoption rate is somewhat arbitrary but conservative as it is below the actual penetration rate for Namibia of 65%, active broadband subscriptions, for the population. The conservative estimate is in favour of bidders as it results in a higher subsidy. |
| The policy on schools as an anchor tenant is insufficiently developed. While school connectivity has very important social benefits it si unclear that the current model truly provides the connectivity needed to support learning. However, there is no provision made for devices for learners to use in order to learn these computer and Internet skills. | CRAN engaged the Ministry of Finance and Public Enterprises to obtain tax exemption for 2 million smart devices since this issue was previously raised as a concern. We await feedback for the Ministry. |
| It is unlikely that entirely uncapped data can be provided to these schools and some cap will be necessary. This is because CRAN’s proposal for uncapped services at a download speed of 20Mbps provided to schools would not leave very much bandwidth for other LTE users in the surrounding area. | The Universal Service Fund can only pay for services and/or infrastructure in terms of the Act. Currently the provision of devices for schools falls under the Ministry of Education, Arts and Culture that is rolling out a programme to provide all schools with the necessary devices. Uncapped services are currently available from Telecom Namibia and capping services could lead to limited access. |
| Assessment of commercial viability: CRAN assumes that 1500 uncovered people within 8km makes a sit commercially viable. It is not clear what this is based on or how it relates to operators’ profitability. | 1500 * 30% * NAD60 = Monthly revenue of NAD 27,000. With adoption rates increasing over 7 years to national average the expected monthly revenue would increase to: 1500 * 65% * NAD60 = NAD 58,500. This would make a RAN site profitable. |
| CRAN assumes that for sites with more than 700 people uncovered a direct subsidy would be sufficient to cover buildout with a 7-year subsidy period. This does not accord with internal calculations. The maximum subsidy provided by the fund is projected to be N$ 3,220,126. This is less than the cost of building certain sites, even if the cost of provision of Wi-Fi to schools etc. was incorporated. | The cost of providing uncapped WIFI is incorporated in the total subsidy calculation, which includes demand side stimulation (WIFI routers and 7 years monthly at NAD 999) as a direct subsidy. The model is designed so that all investments are recovered over 7 years. The seven years were obtained from the financial statements and discussion with operators. Setting the recovery of towers to 20 years would result in a lower subsidy. |
| The model is incomplete. Firstly, it does not consider customer affordability. There is no evidence provided that CRAN assessed the cost to end-consumers of devices and whether they can in fact afford them. Research has shown that device uptake is important. | The anchor-tenant model means that teachers and pupils can use 4G speeds with any WIFI enabled devices. In addition, teachers and pupils can use the Internet for free, addressing the poverty aspect. They further get their devised charged at school. |
| Additional cost to the company such as operating expenses are not clearly considered. This includes administrative costs. There is also no consideration of spectrum availability. It also does not consider the potential impact of on crowding out of other investments and the downstream effect that could have resulted if additional funds were used to lower the costs of communications. Understanding the impact of the levy on the sustainability of business is a requirement of the Communications Act. | CAPEX and OPEX were considered in the modelling of average cost of establishing a RAN site. Conservative figures were used to safeguard that licensees providing infrastructure and services do not make a loss. The bidding process will follow a reverse auction so that the bidder with the lowest subsidy requirement wins the bid. This allows for the use of potentially higher cost for establishing RAN sites. The subsidy requirements given by the model is maximum amounts. |
| Thirdly, there has been no impact assessment on the affordability of the levy. There is no attempt to actually quantify what this means for operators. | A 1% levy is expected to have a 1% impact on the revenues. However, the collected funds are immediately disbursed back into the ICT sector for network rollout and are thus deemed to be virtually neutral. CRAN has added a paragraph on affordability to file study document. |
| At present the full amount of subsidization is being borne entirely by the operators. It is not clear that this is appropriate from either factual or theoretical perspective. CRAN’s approach fails to recognize other existing sources of financing including funding earmarked in the medium-term expenditure framework. The money earmarked from Government should serve as a basis to reduce the proposed universal service levy for the next three years. While operator levies are utilized in certain models for universal service, in this instance it is not clear that the levy is required nor appropriate in the context of existing budgetary allocations. The advantage of tax-funded budgetary allocations is that they are more completely neutral and do not affect investment incentives and competition in the market. While the gap study covered the cost of providing the next tranche of universal service to the population it has serious flaws in that it then simply apportions the cost to operators in its entirety. It does not consider the following:1.The impact of the universal service levy on the sustainability of the business and ensuring that it does not have an unreasonable negative impact on such sustainability.2.Alignment with regional and international best practices.3.To avoid income in access of what is required.4.The necessity to manage risks in the communications industry associated with the imposition of the universal access levy.5.Any other fees, levies or charges which the providers of communication services are required to pay.At 2023 revenue the levy would cost MTC N$ 25, 278,025.99. The amount is significant and ought to be carefully justified. Recent publications from bodies such as ITU provide a range of options for funding universal services. | Section 56(2) prescribes that the licensees have to fund the Universal Service Fund. The levy is collected from licensees and also disbursed to licensees for network rollout. The levy will further be reviewed regularly to safeguard that there is no over collection. This is international best practice. CRAN, however, engaged Government and an amount of NAD 115 million over three years was allocated to the Fund. This will grow the ICT Sector and ultimately lead to socio-economic growth. |
| At present the model used by CRAN to justify a levy of between 0.5-1% appears to have various inaccurate assumptions which is likely to lead to an over-estimate of costs and therefore the amount required to be raised from operators. Furthermore, its calculation is based on inclusion of a range of sites that are already covered by MTC or form part of the obligations under the 800MHz spectrum, and furthermore, a range of areas that can be covered using existing sites with a lower cost spectrum migration. It is clear that the costing model used by CRAN does not represent the lowest cost means of covering areas and is likely to lead to over-collection from operators. This is inefficient. Furthermore, there is no evidence that other components required for universal service including a consideration of demand, devices and skills have been considered. In addition, impact assessments have not been undertaken from the perspective of operators to understand how it could crowd out their own investment and roll-out plans. | The Authority has engaged extensively with telecommunication licensees throughout the process to ensure adherence to international best practises. Licensees were equally given the opportunity to provide in out to technical parameters and the costs of infrastructure, which informed the development of the current model. All sites provided to the Authority, whether presently rolled out or envisaged planned under spectrum licence conditions were considered to avoid double coverage. Spectrum allocation for specific sites remains at the discretion of licensees, recognising that not all have been assigned low-band spectrum. The Universal Service Fund is technology neutral, allowing licensees to propose the most cost-effective technologies for the deployment at specific sites, within the quality-of-service framework established by the Authority. The demand side factors have been carefully considered, with multiple engagements held with licensees to address concerns and incorporate feedback. As part of ongoing efforts to refine the model, a reassessment will be undertaken, with a revised levy of 0.5% proposed to support the Universal Service Fund. The model will be reassessed and a new levy of 0.5% is proposed to fund the Universal Service Fund. |
