Broadband Networks through the Infrastructure Sharing Route

I had mentioned earlier that we should perhaps be concerned about the current trend of state funding for broadband roll outs. It is often presumed that private sector will not roll out high capacity Optic Fibre Cable (OFC) networks at the speed or with the spread required for desired levels of broadband penetration.

It is true that private sector may need various incentives or even subsidies to venture into less lucrative markets or uneconomical areas. However, in my view, a variety of measures can be taken  that still fall short of state funding or state ownership.

The Universal Service Fund of India (USOF) had initiated two excellent schemes for the remote and relatively backward North Eastern states of Assam,  Meghalaya, Mizoram,  Tripura (N.E I Telecom Circle) and Nagaland, Manipur Arunachal Pradesh (N.E II Telecom Circle) that involve high capacity OFC backbone networks being laid out in rural areas (from district to block level) through Output Based Aid projects. These were bid out (reverse bidding) after a painstaking bench-marking exercise to arrive at the upper limit of subsidy, keeping in view possibility of renting out existing OFC from incumbent operators, apart from laying fresh cable. The resultant network is to be shared by the lowest bidder i.e. designated Universal Service Provider or Host Operator  on non-discriminatory, open access basis with other service providers. The tariff  offered by the USP has to be at a specified rate of discount vis-a-vis the Telecom Regulatory Authority's  (TRAI's) ceiling rates for leasing OFC. Discounts were worked out keeping in view capital cost subsidies, revenue projections and operating cost requirements. The bid for the states of Assam was won by the incumbent fixed line operator BSNL. However interestingly for the N.E states Railtel won the bid even without BSNL's  advantage of ownership of majority of OFC networks. There is a strong possibility that it has relied on back-end agreements for renting OFC from private operators rather than laying fresh cable to achieve its obligations in a cost effective manner. This is permitted by the USOF tender. 

The above Private Public Partnership model could have been successfully replicated for block to village level roll outs too. Given that OFC as a technology/broadband platform is here to stay, adeqaute subsidies on reverse bidding basis could have attracted private capital in many (if not all) bidding units (states/telecom circles). This model was rejected during decision making on the National Optic Fibre Network (NOFN) on the debatable grounds that bench-marking takes too long. Personal experience with the above mentioned schemes tells me that this is not correct and that the benefits of involving a large number of market players in laying of the nation's OFC backhaul far outweigh the effort involved in tendering individual bidding units.  I have mentioned earlier relying on public ownership or funding the incumbent is perhaps more attractive in the short run in terms of  relatively less time and effort required to commence roll outs. However the long term impact of monopoly ownership of even open access networks (on competition and accompanying aspects such as innovation/customer service/technological neutrality) and regulatory burden involved in ensuring open access on continuing basis, merit consideration.

It is interesting to note that Indian telecommunications players are looking at voluntary sharing of OFC networks and setting up joint ventures to invest and manage shared networks as the way forward. This may be happening only in cities and towns at present, but it is a moot point whether this trend would not have been replicated eventually in rural areas if the PPP approach to network roll out had been followed.

As of now 2.5 lakh village panchayats (local government centres) are to be connected through NOFN or the public sector SPV called Bharat Broadband Network Ltd. This roll out would take high speed broadband  to rural India and hopefully revolutionize rural telecommunications. It is hoped that the roll out is achieved on time and  that the resultant network is effectively regulated to ensure open access and a level playing field between participating Public Sector Units (PSUs) and various private entities involved in the broadband eco system. needless to say these supply side initiatives must be accompanied by measures to address other aspects of the rural broadband value chain.   

Another important, not entirely unrelated development is the forthcoming creation of a Telecom Finance Corporation to provide capital to telecom operators in India at internationally competitive rates. This should give a fillip to network and service expansion and will hopefully be used to fund not only infrastructure but also content and capacity building  related projects.

National Broadband Plans-The Largely Un-examined Competition Debate

I recently came across a very interesting post on the subject of competition in OFC roll outs. This well written post by Paul Budde argues that (in the Australian context but extrapolating through examples to the international context) either we do not really need infrastructure competition in OFC infrastructure or at least it is not a very practical possibility. He cites USA and Europe as examples of lack of nation-wide fixed line competition.

It would take much more than a blog post to analyse his arguments but I would like to make one simple counter argument. Why must we have a nation wide network? In vast countries like India, USA and Australia even regional or sub regional fixed networks would be a feasible option. In non viable areas, competitive service provision may be seeded by Universal service funding. Please see my post on the Indian USOF model at Broadband Networks through the Infrastructure Sharing Route. This model did succeed in creating potential competition to the incumbent with USOF subsidy even in a remote region of the country. Other posts on infrastructure sharing could also be viewed. 

Perhaps the inability to fathom such a model comes from historical reasons wherein in almost every country the incumbent managed to protect its monopoly by harping on the economies of scale issue and the best option with the state was to regulate prices etc. Regulating monopolies cannot solve inefficiency and lack of drive to innovate that plagues all monopoly service provision. Readers are invited to read my previous posts on NBN and NOFN. Today both networks are delayed and mired in roll out problems. There is a news item about NOFN planning to impose heavy penalties on its vendors who are delaying roll out. Need I say more. I have written earlier cautioning against the faddish nature of national broadband plans and the fact that they are likely to recreate monopolies with the usual set of associated problems.

Also, unlike Mr Budde, I am not so sure that mobile networks can ever be considered perfect substitutes for fixed lines. European regulators seem to agree with me.  

I do agree that service level competition is very critical, but as far as competition in broadband goes, if it is there at every level-all the better. 

NOFN-Do we need PPP or plain USOF subsidy?

I reproduce below a news item regarding TRAI recommendations on NOFN 

I have previously pointed out that the simplest and fastest route to funding optic fibre roll out in rural areas would have been to faithfully follow the USOF model of bidding out service areas based on reveres auctions with open access conditions. This was done by USOF for the North Eastern states. That would have been akin to PPP based on BOO model rather than BOOT.

There is absolutely no reason for the ownership of the network to be with/transfer back to the government unless it is to justify the huge paraphernalia that has been created by way of BBNL.  The whole  idea of Universal Service Funds is to provide a minimal smart subsidy and let markets take over. 

Issues such as fair open access and Right of Way cannot necessarily be solved only by public equity participation. NOFN/BBNL  at present under public ownership has failed to deliver for past 4 years (including solving the RoW problem) and its cost has trebled. 

I have examined this debate earlier in my post "Broadband Networks through Infrastructure Sharing Route"  (also placed under the label NOFN). 

Its time we dusted the departmental files containing the original idea of universal service funding based on international best practices and allowed the Indian USOF to deliver as per its own original rules of competitive neutrality. 

The news article:

The Telecom Regulatory Authority of India (Trai) has recommended a public-private partnership (PPP) model for BharatNet, an ambitious project involving setting up a broadband network in rural India.

A model with private incentives and long-term service delivery similar to the build-own-operate transfer or build-operate-transfer models of implementation would be the preferred means of implementation, Trai said in its recommendations announced on Monday.

“PPPs seek to combine the private sector’s capacity for delivery with the Government’s role as an enabler and regulator to overcome market failures. PPPs must be viewed as not just an instrument for easing finance and capacity constraints, but as an effective tool towards ensuring competition in service delivery and improvement in quality of service,” Trai said.

A special purpose vehicle, the Bharat Broadband Network Limited (BBNL), under the telecom ministry is now handling rolling out the optical fibre network being executed by BSNL, Railtel and Power Grid.

The previous government had approved a project cost of Rs 20,000 crore for laying optical fibre network in 2011 but progress has been poor. It is expected that BharatNet will be completed by 2017-18, after missing many deadlines. Even the project cost has increased to about Rs 70,000 crore over the years. The project was earlier named the national optical fibre network but later renamed BharatNet by the current government.

Trai said the concessionaires should be given the job of deploying the optical fibre cable and other network infrastructure as well as operating the network during the period of contract. The contract period should be of 25 years which can be further extended in block of 10, 20 or 30 years.

The national optical fibre network (NOFN) project had failed in achieving its original objective of increasing broadband subscription in the country. The task of rolling out a broadband network should be given to a concessionaire selected through reverse bidding. Funding should be done to bridge the loss incurred due to higher operational expenses and lower commercial accruals, Trai said.

It can be safely concluded that the NOFN has failed in achieving its original objectives, the regulator said. Focusing on the design of the finance and investment model for future roll-out of broadband is critical.

The National Telecom Policy of 2012 (NTP 2012) envisaged broadband on demand by 2015, and 175 million broadband subscribers by 2017 with a minimum speed of 2 Mbps and up to 100 Mbps on demand. As of September 2015, the total number of broadband (defined as download speeds >=512 Kbps) subscribers stood at 120.88 million (largely concentrated in Andhra Pradesh, Delhi, Karnataka, Kerala, Maharashtra and Tamil Nadu), with only 27.20 million rural subscribers. This “internet divide” between rural and urban India has become more relevant as the scope of activities carried out on the Internet has expanded beyond what was previously imagined, Trai said.

Moreover, rural broadband access will help address multiple service deficits that arise due to other infrastructure related constraints widespread among the rural population. The potential gains from increasing such access are tremendous — the Report of the Committee on NOFN in its projections of the economic benefit from BharatNet estimated that an additional 25 million Internet users by 2018-19 would result in economic benefits of Rs 66,465 crore due to the direct, indirect and spillover benefits of Internet access.

It also recommended that the central and state governments become anchor clients of this project and purchase a bandwidth of 100 megabytes per second at market rate.

To ensure that the concessionaire does not discriminate between service providers in granting access of optical fibres, Trai has recommended arm's length relationship between concessionaire and service providers, adding that 50 per cent of the optical fibre should be reserved for telecom and cable service providers.

Besides, the government should become a minority partner of the concessionaire with 26 per cent stake as this would lower financing cost and risk. "In addition, this can help the government check monopolistic behaviour on the part of the concessionaire," Trai added.

South Africa's Plans for a National Broadband Plan

It has been reported that at a recent industry event with the theme of "Broadband – A Catalyst for Sustainable Economic Development and Promoting Digital Inclusion" the need for better policies, "collaboration between stakeholders," vertical separation and demand side measures like "ICT Skills Development, digital literacy programs for students and adults, IT resources and training"  and the" need to move to impact and creating an ecosystem … and mesh together supply side and demand,” have been emphasized by participants from government and industry.

This echoes much of what has been agreed internationally as posted earlier under National Broadband Plans and Broadband Networks. 

An earlier report about the Government's plans for broadband expansion and reactions of the industry may be see here.

http://www.zdnet.com/can-south-africas-grand-broadband-plan-succeed-7000007165/

An extract as below indicates that the 3 options being considered are similar to those which may have been considered by many a nation and certainly same as those considered in India. (Please see post titled "Broadband Networks through the Infrastructure Sharing Route"

"The government currently owns a number of assets in the telecoms market – including long-distance infrastructure provider Broadband Infraco and a 39.8-percent share in South Africa's fixed-line incumbent Telkom. The state now wants to work with the private sector to build a wholesale national broadband network along open-access principles. With around 3.5 million PC broadband connections and 10 million smartphones between South Africa's population of more than 51 million, the country is far from achieving its goal of universal access by 2020.

Though there are many broadband expansion projects underway, they are fragmented, and a comprehensive, centrally planned strategy is essential to boosting broadband in South Africa .., three funding options for the national network [are]:

• Financing a state-owned enterprise.
• Incentives for operators to offer services in economically unattractive rural areas.
• Equity and incentives provided by government could be ring-fenced in a special purpose vehicle."

The public consultation paper on National Broadband Policy suggests that for OFC backbone the incumbent (Telekom) will play the lead role in providing whole sale access even though service competition will be encouraged in service provision to customers . The document lays a welcome emphasis  on developing the broadband ecosystem.

Ensuring Affordability of USF Supported Services

A query from an esteemed telecom expert and colleague made me feel that this may be a good topic to cover today.

Part of the reason for the Actual Access Gap referred in my previous post is the non-affordability of services for certain segments of the population. This could be because they have lower than average paying capacity in absolute terms (say the urban poor) and/or relative to cost of provision of the services (on account of geography/ own disability etc.). In developing countries, the bulk of the population in rural areas could easily fulfil both the absolute and relative criteria making it essential to provide services not only at par with urban tariff rates but at times below urban rates in spite of higher costs of provision.

The underlying rationale of many an output based aid (OBA) USF project is that if  subsidy can be provided to help the USP break even and cover CAPEX and OPEX for a finite period, demand will eventually pick up enough to make services profitable (even at a lower tariff rate.) In any case ideally in an OBA based USF project, once the USF contract comes to an end, the USP should be  free to revise traiffs as per market conditions and other (non USF related) regulatory restrictions. In the  interest of  protecting its investment the USP would not like to drive away customers by charging unreasonable tariffs.

In India, the USOF mostly follows the OBA route. USOF projects are bid out. Reverse bids are floated  with a maximum permissible subsidy level based on a detailed benchmarking exercise. Tariffs in India are regulated by the Telecom Regulatory Authority (TRAI) and beyond the jurisdiction of USF. Thus, USF may in its tenders/contracts refer to TRAI regulations on say fixed line tariffs in rural areas and require rental and call charges to be at par or lower than the same. Alternatively, as explained in my previous blog titled "Broadband Networks through the Infrastructure Sharing Route" the USOF tender/contract may require that during the contract period the USP offers subsidized infrastructure/services at a discounted rate with reference to TRAI's ceiling rates. In USOF's mobile infrastructure and services scheme, the static infrastructure was required to be offered rent free by the wining infrastructure provider to the three winning mobile service providers who would share the towers. However, mobile services themselves could be offered at any rate to end users. With their rentals costs being nullified and given the competition between three players it could be assumed that they would vie with each other to provide attractive tariff plans to the served rural population. In fact my own experience with monitoring of this scheme has shown that in this case it was customer services (such as regular supply of recharge vouchers for pre-paid connections and QoS  which distinguished the more successful USPs from the laggards). In USOF's Wire line Broadband Scheme a couple of very affordable entry level broadband tariff plans were arrived in consultation with the USP (selected by nomination in this case on account of incumbent owning 99.9% of rural wire lines). These were required to be offered along with any other tariff plans (as per USPs choice) to rural customers being served through subsidized infrastructure. (Broadband tariff is on forbearance). Significantly, and as predicted when the broadband scheme was first introduced, the entry level tariff plans formed the bulk of the uptake but over time, the higher value tariff plans offered by the USP in parallel gained popularity. As on April 2012, entry level packages constituted 32% of the total broadband subscriptions under the scheme whereas initially their share was up to 90%. Thus, the decision to discount tariffs is always a considered one based on the characteristics of the market and the gap that we need to address.

In each USOF scheme, the benchmark subsidy is modeled on the basis of projections of CAPEX, OPEX, estimated demand and paying capacity of subscribers separately for each bidding unit. This could be a state, selected individual districts, group of districts etc.The tariff assumptions/prescriptions form part of the subsidy model and benchmarking exercise.

ITU's ICTs Regulation Toolkit explains this approach in terms of 'Smart Subsidy':


‘A  smart subsidy is the term used to describe an initial subsidy (usually given on a once-only basis) that is designed to be results-oriented, does not distort the market, and encourages cost minimization and growth of the market. It helps to kick start a project or service, with the ultimate objective of the programme becoming commercially viable, whereas without the subsidy investors might otherwise have been reluctant to invest. Investors’ reluctance could be due to perceived risk or general lack of capital for the kind of service opportunities that are considered by government to be essential for socio-economic development. The important element of the smart subsidy zone is that an initial subsidy to private sector providers will make the project commercially viable on an ongoing basis by filling the financial gap with a one-time subsidy, which increases the operator’s rate of return and reduces his risk. No further subsidies are needed if the service targets are set realistically, with medium term commercial viability in view. Targeted interventions are usually implemented using a Universal Access and Service Fund (UASF).’

State Speared Fibre Roll outs-NBN

Australia's NBN is almost always in the news for political reasons. An article titled, "Quest for 21st Century Broadband: A Tale from Down Under" also mentions  problems of slow roll out and slow uptake apart fro  NBN being the subject of "political football."

Thus it is written therein that,

 "The project had only reached 207,500 homes at the end of June, well below its target of 341,000. And only 70,100 of those with access to the network had signed up as paying customers......

 ....Work had to stop earlier this year when deadly asbestos was found in the pits where workers were laying new fiber cabling, and some sections of the network had to be redone in the nation’s capital Canberra because of poor workmanship.  The latest blow came on Monday, when the the Government-owned NBN Co. announced it would have to find new contractors to install fiber cabling in homes in South Australia and Western Australia."

Perhaps the problem lies in the size of the venture being implemented by a single telecom operator. Readers may also like to see "Broadband Networks through the Infrastructure Sharing Route" where I have described projects for state/regional OFC network roll outs by USOF India where the implementing USP is selected through bidding. Also see posts on Broadband Networks

Delay in NOFN Roll Out-As Expected

The Economic Times today reports yet another delay in roll out of NOFN by BBNL as the PSUs are unable to award contracts worth Rs 6 billion for cable laying and trenching. 

I would invite readers to review my post titled, "National Broadband Plans-The Largely Unexamined Competition Debate" under the label NOFN. I have already covered in previous posts, my reasoning as to why  India should have hesitated before venturing to roll out a country wide network using the nomination route involving Public Sector Incumbents. When various option were being examined as to which methodology to choose for NOFN, there was an explicit impatience with the usual USOF method of first arriving at subsidy benchmarks and then bidding out a scheme on a regional/sub-regional basis to all eligible operators. This was frowned upon as too tedious and a source of delay. 

It was decided that creating an SPV of PSUs would be the better way forward especially as BSNL already owns the chunk of rural OFC networks.

I have examined this debate in my post "Broadband Networks through Infrastructure Sharing Route"  (also placed under the label NOFN). An alternative model has been presented to readers. One that is based on bidding.

 The right way in  my view would have been to encourage/mandate  BSNL to share its OFC capacity with the region wise winning bidder and to include the leasing plus incentive cost in the subsidy benchmarks. With this arrangement the network could have been rolled out by multiple USPs thereby creating the required  non-discriminatory open access  OFC backbone in rural blocks  with no adverse impact on competition. The facilitation extended by USOF (Central Government) by way of coordination with state governments for right of way clearances could have been done in this model too. This would probably have gone faster and ensured that at least  a good proportion  if not all villages would be reaping the benefits of high capacity OFC backbones connectivity by now.