SA's ICT status quo has to be upended - TechCentral

SA’s ICT status quo has to be upended

joe-mjwara-180When the editor of TechCentral promised to cover extensively the recently published national integrated ICT policy white paper — and provide analysis of it — we were excited.

In the past two weeks, that excitement has turned into bewilderment as TechCentral’s pages have been filled with endless articles whose sole purpose is to bash the new policy and call for the status quo to remain.

There’s been a steady stream of single-track contributions to TechCentral lauding what is an unworkable and unsustainable market structure, one that has led to high communication costs and stunted the entry of new and innovative players in the sector.

There are public interest issues, beyond the selfish interests of the dominant incumbents, which need consideration in shaping the development of the sector for the future.

The views that have been raised, bashing the policy, were adequately debated during the thorough consultation process we followed. They did not carry the day in view of the competing interests for access and use of scarce resources. These views are not held by the entire sector, but just some players in it.

The suggestion is that there is something wrong with wholesale open-access networks. We need to correct the impression that wireless open access is by itself unworkable either because it hasn’t been done in the past or because it differs technologically from fixed networks.

It is important to remind everyone, once more, that ours is a technology-neutral regime that does not discriminate in accordance with the technology used. What walks and talks like a digital network is a digital network that must be regulated like all other networks.

Open access

The case for exceptional treatment of the wireless mobile networks is long gone because of technological developments that have made the sharing of networks by different providers possible. If the wireless open-access model has not been applied in the past, the real reason is because technologies did not permit it. Recent technological developments have allowed for the wireless networks to be shared.

For example, the handset was the biggest obstacle to the development of the open-access wireless networks as handsets contained only one radio front end. Now they contain a large number of front ends targeting different technologies and frequency bands enabling them to work across Wi-Fi, 2G, 3G and 4G.

Software-designed networks are a recent feature that has made it possible to implement network functions without the need to install new equipment. This has opened the way for agile networks that can support open access.

Most of the critics of government’s choice of open access deliberately ignore the competition dimensions

Interestingly, while the reality of technological advancement that enables the building of networks to facilitate and service different competing entities is never disputed, some flimsy arguments are advanced to spread doubts as to the technical soundness of the choice for wireless open access. None can dispute that the technology is here to allow for the entry of more players, which in the past were denied entry because of the scarcity of spectrum and the network technologies that could not support a multiplicity of users.

If the current market structure is maintained, it would not be because of the unavailability of the technologies to foster more competition and innovation but a desire to protect the incumbents and to maintain the status quo that has excluded small and medium-sized businesses. Such exclusions need to be corrected as early as possible.

Most of the critics of government’s choice of open access deliberately ignore the competition dimensions and the need for public policy to take into account the needs of all sector players rather than the few who have risen to the top. In reality, most of what has been written on this website is a repeat of statements by the dominant players who themselves were enabled to enter the market through concerted government policy actions and regulations, including forcing the then monopoly of Telkom to supply network services at asymmetric rates. It is estimated that the two biggest mobile operators benefitted to the tune of R50bn through this regulatory- and policy-sanctioned indirect support.

Access to spectrum

It is a fact that there are more than 400 licensed entities that require access to spectrum in order for them to effectively participate and compete in the ICT markets. Add to this number Internet services providers, over-the-top players and others for whom access to the mobile networks is critical for their success. These are the people whose rights need to enter the equation and whose needs should be balanced against the traditional players for the sector to expand and embrace innovative services. This is the real policy dilemma that must be resolved, not the fanciful notion of that what works best is what suits the dominant operators best.

For this to happen, policy must mandate the sharing of critical and scarce resources like spectrum instead of exclusively assigning it to those who are already privileged.


Another critical step is that access to infrastructure that cannot easily be duplicated must be opened up to more players. Owners of this infrastructure would be able to charge for the services rendered in a non-discriminatory and transparent way to competitors. If there is anything worth pondering, it’s this: why are the people who were at the forefront of demanding that the local loop of the fixed network be unbundled and opened to competition have now suddenly turned opponents to open-access principles?

What is crystal clear is that access to spectrum is the biggest barrier to participation in the growing broadband market. What also cannot be denied is that access to the infrastructure using the spectrum is critical for effective competition for the new entrants. If there is to be any growth in GDP as well as rising employment, it’s new entrants, broadband and Internet companies and small, medium and micro enterprises that will make more of a contribution compared to vertically integrated oligopolies that place insurmountable barriers to entry to new players.

Spectrum is a limited resource in a situation in which demand for it is increasing as a result of new bandwidth-hungry services

With the technological developments enabling open-access networks, the new discussion no longer focuses on past approaches to market structure and the assignment of resources. The reality is that spectrum is a limited resource in a situation in which demand for it is increasing as a result of new bandwidth-hungry services that have been introduced by the digital revolution. More services will require access spectrum as the Internet of things connects the millions of devices that are yet to be connected. It is inconceivable, as well as impractical, that anyone that wants to participate in the ICT evolution needs to deploy their network and fork out billions of rand in spectrum auctions that would in any case not provide a solution to their needs as there is no spectrum to satisfy everyone.

In many countries, current discussions are increasingly focusing on the shared use of the spectrum and networks from technical, business, policy and regulatory perspectives because it is seen as providing a solution to the scarcity problem.

Lessons from abroad

In Europe, there is ongoing work on “licence shared access” that allows spectrum that has been licensed for IMT to be used by more than one entity. This complements directives that imposed shared access of the infrastructure of dominant operators by mobile virtual operators.

The US president’s advisors on science and technology made proposals that led recently to the development of Federal Communications Commission rules to support the private sector’s sharing of spectrum with government agencies.

Mexico has set aside the entire 700MHz band for an open-access wireless network whose licensing is currently underway.


Rwanda currently runs a wireless open-access network in partnership with Korea’s KT Corp.

Kenya chose a wireless open access model that was sabotaged by the dominant operators.

Private-sector players who have built and managed these open access networks include Russia’s Yota, which operated a wholesale LTE network that covered 180 cities and 70m people. Czech Republic’s PPF is another example. The developed and developing worlds are embracing open-access networks, and so should we as we look into the future.

Whatever the approach is in today’s discussion on open access, the sharing of the spectrum and the allocation of the spectrum on non-exclusive terms is the norm as opposed to the antiquated single exclusive-use principles those opposing open-access principles advocate.

The debate is no longer whether open access is the right approach, but rather how to take into consideration the different socioeconomic profiles of countries where it’s implemented.

There is a growing expectation the world over that resources need to be pooled or shared so that they can be used more efficiently and effectively to serve growing user demand and the needs for new, competing innovative services.

The open-access network approach envisaged in the white paper is futuristic and factors in the reality that exclusive licensing arrangements for high-demand spectrum are unworkable.

Some past contributions to TechCentral have gone beyond challenging the technological soundness of opting for open access and wondered if shared networks can foster service-based competition, produce high-quality innovative services, bring about affordable rates and drive broadband into underserviced areas. The effects of the duplication of the networks in a situation in which the existing capacity is not fully utilised in all the networks and consequently the cost to the operators that get passed on as communication costs is never discussed as an obvious saving were the networks to be shared.

The reality is that the cost of operating multiple and separate networks will always be higher than operating a shared network. This truism is not challenged in any of the networked sectors of the economy, ranging from transport to energy. In the telecommunications environment, this cost impact is glaringly obvious if all four operators would be required to extend network coverage to reach all areas of the country, including the areas that are not economically sustainable.

Cost savings

At a conservative estimate, it would cost in excess of R15bn for each operator to build a network that has national scope. There are obvious capital spending and operational cost savings that cannot be denied in the efficient utilisation of a common network.

Also, the operators will be able to deliver more services to the entire population, instead of being limited to people who live in the areas covered by their networks.

Much has been said about mandated and non-mandated open access. The beauty of the open-access approach detailed in the white paper is that it is premised on a voluntary association of willing partners. The principle of non-exclusivity is embedded in the requirement that no one entity must have control of the operations of the network.

Current operators as well as other interested parties are required to come together and operate the network on open and non-exclusive way. Those who feel that they cannot participate in the open-access network will not be disadvantaged as they will still be entitled to acquire network services from the open-access network on open and non- discriminatory terms.

This is a win-win solution fit for the future.

  • Joe Mjwara is acting director-general of the department of telecommunications & postal services


  1. There is a large gap between theory and practical implementation. Free market systems are not efficient because they optimally allocate resources in static view — they are efficient because they constantly adapt to forces which are sometimes only visible in hindsight.

    Despite the best of intentions this will set us back 20 years. Competing on the service level is akin to the ISPs that “competed” with each other in the 90s and 00s whilst Telkom strangled the market. Experience from history shows us what a bad idea this is.

    Besides — there are other ways and means to achieve better outcomes. We already have MVNOs. We already have some level of infrastructure sharing. These mechanisms could be used to achieve the outcomes required without gambling away a large chunk of South Africa’s GDP. Small policy and regulatory interventions could make a big difference without driving us off a cliff.

    And did somebody calculate the number of job losses as result of this madness? Building one network instead of four means you need a quarter of the people. Please pick up the phone to CWU and tell them that it is for the good of the country.

    Also good luck with attracting foreign investment by effectively nationalizing something that was part of the private sector domain.

  2. I am afraid Mr Mjwara is living in a bubble, completely isolated from the industry in which he has been involved for quite a few years. To say that an extensive consultation process has been undertaken is nonsense. I have been a participant in the industry for many years now and none of the businesses or industry bodies on which I have served have been able to secure an audience with any minister or deputy minister for the past five years. They simply don’t interact with the private sector who are responsible for the ONLY successful infrastructure initiatives over the past five years – I challenge Mr Mjwara to contradict this statement! This is appalling considering the billions that have been squandered by the likes of government entities such as Broadband Infraco, Sentech, BeWired etc. The fact that the DOC has the temerity to propose solutions for the industry given their track record beggars belief. I also challenge Mr Mjwara to name a single DOC employee whose name has ever featured in the real world of telecommunations (ie the private sector).

    The ICT white paper is a combination of vanilla statements that have been “cut and pasted” from elsewhere. It is clear from the content that the authors (presumably external consultants) have no practical understanding of the drivers of the investment dynamics in the industry.

    The cynic in me interpretes the real motive behind this document to be the expropriation of spectrum (possibly initiated by Atul Gupta who sat on the same DOC advisory committee as Mr Mjwara in the Dina Pule era?).

    Open access has been a great success in the fixed line fibre environment, but the technical challenges of doing the same with spectrum have clearly escaped the attention of Mr Mjwara. Despite his half-hearted efforts to describe this as a new trend, there are NO successful examples of open access wireless networks in the world. That includes Rwanda and no, Kenya’s failure was not due to sabotage by incumbents.

    He demonizes the mobile operators who helped SA achieve one of the highest penetration rates in the world by 2005. The fact that wireless broadband penetration is where it is is not their fault but rather that of government who have starved the private sector of new spectrum allocation.

    The DOC’s desire to be more interventionist will only serve to retard the market and broadband penetration.

  3. “This is the real policy dilemma that must be resolved, not the fanciful
    notion of that what works best is what suits the dominant operators

    Give this man a Bells.

  4. What about the foreign companies who want the auction halted because it effectively keeps them out? Nationalisation is not always the right solution, but it isn’t the wrong solution either. There are many examples of public utilities taken from private hands and still managed to attract investment.

    We need a middle ground, not just a free market approach. The last fifteen years is plenty of reason why privatising spectrum is not in the best interest of the country.

    Perhaps this policy paper is not ideal. But we are going nowhere if we keep polarising the topic instead of looking at alternatives. I do not want ten more years of the same treatment we have been getting from the incumbents. Do you?

  5. Hang on – high penetration rates do not reflect high usage rates. No sense if everyone has a car, but nobody can afford fuel. He is not demonising the networks. They did that themselves.

  6. “In reality, most of what has been written on this website is a repeat of statements by the dominant players who themselves were enabled to enter the market through concerted government policy actions and regulations, including forcing the then monopoly of Telkom to supply network services at asymmetric rates.”

    If this is correct, then why is Cell C denied the same request for asymmetric rates against Vodacom and MTN? What’s good for the goose surely? It seems that govt has created a duopoly which now refuses to enable the very systems that launched them into the stratosphere. See below the diagram relating to how these two operators function.

  7. Greg Mahlknecht on

    > We need a middle ground, not just a free market approach. The

    Yes – I don’t see why we can’t do an auction, and then if the government is so confident their plan is awesome, they start a company, buy a chunk, and become the open access operator.

  8. I notice with interest that this discussion is back again and this time it’s taking an interesting turn with a clear and solid motivation of the need for the open access network by Mr Mjwara. The fact is that the mobile broadband market is currently held ransome by two MNOs and these MNOs determines the level of competition,cost of services, coverage, etc they directly and indirectly control the market. Telkom and Cell C still roams on their networks and roaming charges are not regulated, until recently only Cell C hosted MVNOs on their network, ISPs and other resellers relies largely on their services, they determine their prices across all elements of the pricing value chain and final bill goes to consumers. This essentially means that all the service providers in the value chain of providing mobile communications services entirely relies on Vodacom and MTN to directly or indirectly set prices for them, largely as a result of how the market is structured. This is a real problem in SA mobile market and we know that we haven’t had effective regulations to adress this.

    Is Open access network the solution to open up the market to other players and increase competition? I certainly believe it is. The issue is whether it is doable or not, and I think the fearea is that the two MNOs may not play along. The policy however doesn’t exclude anyone from partipating on this network and neither does it threaten to close anyone’s shops should theye choose not to participate in this network so I don’t know what the uproar is about other than protecting the financial interest of the two MNOs at the expense of the entire industry. It’s time for change so let’s embrace it and move on!

  9. Here’s the issue with an auction. So, let’s take two baseline requirements for a successful bid in the current auction process:

    – You need R3 billion minimum
    – You need an LTE-ready network

    Of our three current major networks, only Vodacom qualifies comfortably. Not only that, but it has more than enough money to dominate the auction and outbid everyone else.

    The lots up for auction are not clear and are not equal. ICASA won’t release crucial details and specifics around the various lots. I don’t quite understand how that works, but experts assure me this is a problem. You essentially have a car boot sale, where the best gear goes to the company with the deepest pockets.

    (This is also why your idea of the government buying a chunk won’t work, since it won’t have the best spectrum to host what are several hundred telco companies in SA. Spectrum is a bit like a crowd: if you have 1000 people together, you have a large crowd. But if you split them into 10 groups, you don’t have a large crowd – you have 10 small crowds. Spectrum effectiveness works exponentially – again, not too sure how this works and relying on expert views. But pooling spectrum is apparently quite important and past spectrum license allocations in SA may be partly to blame for both high prices and iffy performance.)

    Outbidding is fine, right? Not really. So far the only conclusive method in the industry to reduce data and other costs is through competition. It has been getting far less press, but a lot of telco players are very unhappy about the auction. It not only excludes most of them thanks to the R3 billion minimum, but even established players like Cell C have no chance in bidding against Vodacom for the best spectrum.

    In short, what an auction will do is entrench the current players, using the profits they made by charging us a bundle, and make it even harder for new companies to compete. Consequently we will not see prices fall. Every argument I have read against this statement assume some sort of altruism from Vodacom and co, even though their track record suggest the complete opposite.

    Vodacom and MTN have been actively confusing this issue, even suggesting high prices are due to a shortage of spectrum (even though we have been paying ridiculous data prices long before spectrum became an issue).

    I’m not saying shared spectrum is ideal. But we are dangerously assuming an auction is the only alternative because we are paranoid about the shared spectrum concept. The biggest supporters of the auction – ICASA, Vodacom and MTN – also stand to gain the most and yet there is little to suggest the rest of us will gain anything.

  10. Greg Mahlknecht on

    – You need R3 billion minimum
    – You need an LTE-ready network

    I don’t see these as problems – wouldn’t a new player be building an LTE network anyway? It would be silly to think you could compete long-term by building a last-gen network these days. So scratch that off your list as a problem. And with that having been said, the R3B becomes a relatively small amount. The R3 billion seems to spook so many “analysts”, but when taken in context, it’s not that excessive.

    >we are paranoid about the shared spectrum concept.

    I think the history of government’s involvement in projects like this shows they are uttely incapable of running them effectively. It’s not paranoia, it’s just being realistic.

  11. Andrew Fraser on

    If there is to be an open access network, then it should be managed by a consortium of the telecoms industry and government should be a minority shareholder. A new (section 21?) company to hold the spectrum rights, no two network operators to control more than 50% of that company. That way, the spectrum can be managed to the benefit of most. Ideally MTN and Vodacom should hold no more than 20% each in the consortium, with the other two smaller MNOs and government holding 10% each. Balance of shareholding to be split between emerging telcos and other network operators. All surpluses from allocation of spectrum to go to the fiscus.

© 2009 – 2021 NewsCentral Media