Diversifying Participation in Network Development moving beyond the

Diversifying Participation in Network Development moving beyond the market Study of India’s Universal Service Instruments Preliminary Findings Harsha de Silva and Payal Malik LIRNEasia, 20 May 2005, Colombo

Outline n Overview of the Regulatory and Policy Developments n Status n Access Gap n Universal Service Instruments n Universal Service Fund: Progress and Issues n Conclusions on USF n ADC: Status and Issues n Conclusions on ADC n Discussion

Regulatory and Policy Developments of the Indian Telecom Sector: Diminishing Market Efficiency gaps Future • Comprehensive spectrum policy License Policy: • Sharing of backbone • Tax Policies: Onerous license fees • Number portability • Connectivity of Wireless operators to carry inter-circle calls • Unified 2003 -05 • Unified 1999 -2002 access license regime introduced to enhance competition and create a level playing field • Transfer of Wireless licenses allowed among operators • Intra-circle Wireless mergers allowed • IUC regime implemented • Lowering of ADC from 30% to 10% of the revenue • New 1994 -98 • TRAI Telecom Policy introduced • Entry of third and fourth operators in Wireless services • Free competition allowed in Wireline: WLL Introduced • NLD & ILD opened up to competition • First round of tariff rebalancing done: TTO • Operators moved from fixed to revenue-sharing license fee established as an independent regulatory body • Wireless licenses allotted to private operators • Wireless services opened up to competition Industry deregulation and liberalization Declining tariffs and handset prices Prepaid offerings Implementation of CPP regime

Status n Telecom Sector: benchmark for other infrastructure sectors n Teledensity 2 percent in 2000 now close to 10 percent n Urban teledensity 26. 2 vs. rural teledensity 1. 74 n Increased focus on cellular mobile infrastructure deployment: 68. 81 percent growth vs 6. 6 percent n Rural DELs installed by BSNL through license fees relief n Roll Out Obligations failed

Access gap n 70% of population is rural: GDP per capita US $352 n High costs of extending network to uncovered areas n Current ARPU’s/EBITDA’s inadequate to fund capex required n 5000 urban agglomerates: Mobile coverage 50% n Growth will be driven not so much by falling tariffs: increasing geographical spread essential n Additional investments: mobilized through intervention n Market Failure Arguments

Universal Service n Efficiency Vs. Equity Grounds n USO a special case of redistributive pricing: Tariff Policy n Policies can be optimal in a second-best setting: more efficient policies like direct transfers n Traditional funding: unworkable competition drives down supercompetitive price

Funding Mechanisms n USO Fund (USF) n Access Deficit Charge (ADC) n Government Funding: Grants and License fee waiver n Roll-Out Obligations: Access Providers to cover 50% of DHQs and NLDOs to set-up POPs in every LDCA

USO Fund Policy n Came into effect from April 1, 2002 n USF: statutory non-lapsable Indian Telegraph (Amendment) Act, 2004 n Expediting disbursements effectuating universal service policy n Administration: a separate administrative organization attached office of the Department of Telecom n Disbursement through least cost subsidy auction: subventions placing companies in competition through a system of inverse bids

Status of various USO Projects in India Project Operation and Maintenance of Village Public Telephones (VPTs) in the revenue villages identified as per Census 1991 Implementation Dates March, 2003 Approximately 520, 000 villages Replacement of Multi Access Radio Relay Technology VPTs installed before 1 st April 2002. 1, 80, 000 MARR VPTs September, 2003 Comments Includes support for 9171 VPTs installed by the Six Private BSO and remaining 5. 09 million VPTs installed by BSNL. This provides coverage of more than 90% of the villages where VPTs. are to be provided. Firms participating in this auction bid exactly the benchmark Since VPTs were mainly BSNL’s, subsidy went to BSNL with a zero cost reduction: bid exactly the benchmark

. . . contd. Provision of additional rural community phones (RCPs) in areas after achieving the target of one VPT in every revenue village (2 nd VPT). September 30, 2004 Out of 300 Secondary. Switching Areas (SSAs), BSNL was the successful bidder in 184, Reliance Infocom won 97 Competition between two service providers in only 115 SSAs. The competitive bidding has resulted in bringing down the cost of the project by about 17% from the reserve price November, 10, 2004 BSNL emerged successful for 12 service areas where six companies participated. BSNL oneto-one competition with Bharti Cellular Ltd. in three service areas. A reduction of 15 -20 percent In nine service areas BSNL was the sole bidder BSNL emerged the winner in all Service Areas 46, 253 RCPs. Provision of VPTs in revenue villages as per Census 1991 without any public telephone facility. No. of villages covered : 66, 822

…contd. Provision of rural household direct exchange lines (rdels) in specified short distance charging areas March 15, 2005 The project covers 274 (SSAs), competitive bidding in 215 SSAs, BSNL emerged the most successful bidder winning in 171 SSAs across 19 States, Reliance Infocomm emerged the winner in 61 SSAs spread across 15 States while Tata Teleservices got the project in 42 SSAs across 9 States, competitive bids have brought down the cost of the project by 60 -75 per cent

Disbursement Schedule Year Amount Disbursed/ Provision 2002 -2003 Rs. 300 crores (66 million USD) of the Rs. 1653 crores (USD 367 million) collected 2003 -2004 Rs. 200 (44 million USD) crores of the estimated Rs. 2, 143 crores (476 million USD) collected 2004 -2005 Rs. 1200 crores (266 million USD) 2005 -2006* Rs. 1200 crores (266 million USD)

Costing Model: Determination of Benchmark n Benchmark: Reserve Price for invitation of bids n Evolving of benchmark for each activity for different areas n Fully allocated current costs: costs for bulk procurement of latest technology-based equipment Determination of Net Cost (NC) for new facilities n Net Cost = [ {Annualized Capital Recovery + Annual Operating Cost} - {Annual Revenue}] (Where Annualized Capital Recovery = Aggregate of depreciation + return on equity plus interest on Debt) n Different Approach: 8. 6 million rural DELs installed prior to 1. 4. 2002 n Alternative Proxy cost model

Issues n Universal Access Vs. Universal Service: Payphones, broadband kiosks n Broaden the mandate: voice and low speed data to broadband connectivity n Technology “Neutrality” n Eligibility Criteria: Impact on the success of auctions, left huge rents for the incumbent n Costing Models and Auction Procedure n Market “Efficiency Gaps”: Regulatory levies n Spectrum Availability and Pricing n Sharing of Backbone

Conclusion of USF n Tend to be used by market players to extract too many concessions n Important strategic implications: effect the way firms compete against each other n Benefits from using auctions to assign USOs: difficult to have sufficient participants bidding against the incumbent n Asymmetry of information between the incumbents and new entrants n Financing these costs imposes distortions: try to minimize losses of allocative inefficiency

Background to ADC n Pre Reform n Cross subsidy from national and international LD tariffs n Reform n Falling prices in NLD, ILD, FL, WLL [M] and cellular n FL cannot sustain “social pricing” in rural areas; others have forbearance n Enter ADC n Normally [several other countries] a charge imposed on long-distance services and passed on to fixed-line access providers who are mandated to provide services below cost [but many are withdrawing ADC: US, UK, France, Canada, EU…] n Original implement date: 1 April 2003 n Implemented 1 May 2003

ADC ILD: Origination/Termination on FL: Rs 5. 00/minute + 0. 50 termination charge. None for WLL[M], Cellular

ADC n Objective is rapid growth in teledensity [affordable access to basic service NTPL 1999]; so cannot increase tariffs ADC until market is large [stable] enough to do without. n Total access deficit in FL INR 130 b [USD 3 b] n n n Applicable rental < cost based rental Free calls Below cost LD [0 – 50 km] n Calculated using a return of 14% ROCE n BSNL 2001/2 ROCE 7. 5% n 2002/3 1. 1% n ADC as a share of TR of Telco. Sector: Chile 2. 0%, France 2. 5%, US 6%, SA 0. 3%, India 30%

ADC: original thinking n Together with IUC [carrier, termination] n Connecting fixed and all else n [1] Uniform charge and [2] escalating with distance n Assumed cost per FL INR 424/mo [BSNL ADC INR 296] n Wide variation of call charges: particularly if FL-FL n Advantage to WLL [M] and Cellular n ADC only if FL; favored cellular-cellular the most n Could not apply IUC+ADC charges, TRAI authorized below cost tariffs to keep FL [incumbent] in competition. [Not predatory pricing]. Other FL BSO also followed suit n ADC questioned

Problems n TRAI had created a unequal playing field by bringing in complex and confusing arguments to determine ADC n n n Technology matters; distance matters Choice of regime [Distance does not matter] Unsubstantiated costs etc n BSNL complained that while they were the largest service provider in rural [>30%] they had the highest AD, but TRAI in its calculations did not consider this fact and specified equal ADC [based on BSNL costs]. n Bias built in against FL n n Cellular and WLL[M] was becoming much more competitive than FL WLL[F] was considered FL

Problems continued n Consistency of IUC under various schedules n Who should get ADC n BSNL n Others? n Below cost FL tariffs [to compete with Cellular and WLL] and its sustainability n Vicious circle cost is high; but keep tariff below cost to compete; deficit; apply ADC; high ADC makes FL less competitive; higher deficit… n Cost of NLD carriage > than TRAI specified cost n IUC of INR 5. 50/min termination of IT grey market n Led to May 2003 Consultation

Led to 2003 May consultation n Reassess ADC regime n Should BSNL and other BSO be given ADC? n given their urban presence and n unmet roll out n Should ADC be linked to roll-out? n Should ADC on ILD be reduced to discourage grey traffic? n Should ADC have a cut off date and/or merged with the USO regime?

2003 May Consultation n Was the calculation method correct [BSNL hist. avg]? Why not FLLRIC to account for technology change? n n n FLLRIC is necessary; but a single year shift would impact heavily on BSNL. So stick to historical [but 2002/3] audited BSNL a/c However, BSNL shifting to lower cost wireless technology Over a few years ADC to be merged with USO n GOI grants to BSNL for rural telecom need to be factored in the calculations of ADC n n n Reimbursement of license fees Moratorium on capital and interest payments Maximum 10% dividend etc. lower WACC lower ROCE

2003 May Consultation n Use of cost estimates and minutes of others [not BSNL] not yet possible n n n Un-audited Inconsistent, but higher cost compared to BSNL [even MTNL] In some cases “extreme” and “absurd” n Net AD for BSNL INR 53 b [including GOI comp. ] n ADC for others higher with their own data, but lower with normalized for BSNL n Consider linking ADC to roll-out n For BSNL and MTNL cover costs from high growth cellular [zero entry fee]

Revised ADC mechanism n Paid to all BSO on a per minute basis n Paid by Basic, Cellular, National LD, International LD service providers n ADC for fixed line operator or BSNL

Revised ADC n Implement date 1 Dec 2003; delayed 15 Dec 2003; delayed 1 Feb 2004 n ADC is lower [include GOI support to BSNL] n Shall fund INR 53. 4 b n Scrap 2 ADC regimes; stick to escalating ADC n Applicable to all calls except FL-FL, 0 -50 km intra circle, intra circle Cellular/WLL[M] to C/W[M] n Non BSNL to keep ADC, but less than BSNL [limited for of IUC] n n n Originating: keep ADC Terminating: keep ADC + Termination charge No WLL[M]/Cellular to-from WLL[M]/Cellular

Cont… Revised ADC n ADC to be merged with USO in 3 -5 years n All intra-circle INR 0. 30 per minute; inter- circle INR 0. 30, 0. 50 or 0. 80 n Earlier 92% of ADC funded by BSO [BSNL]; 40% as proportion of revenue. Now down to 12% of revenue for FL, 9% for Cellular and 16% for WLL. n In the future possibilities of ADC as a percentage of revenue?

Consultation June 2004 n Serious implementation problems n Payments not made n Data questionable [INR 0. 30 – 0. 80, 4. 25 for ILD] n Technical problems due to distance measures n BSNL billing system delays have made problem worse n Bypass [cannot identify calls from other networks] n Consider a simpler approach n Not distance based n Not call based

Proposed new ADC n ADC period n 10/2004 to 9/2005 n Revenue share n Less complex and easier to implement n Revenues for relevant period n [avg. subscriber base march 2005] X [monthly ARPU] n At INR 200/mo rental 2. 2% n At INR 156/mo rental 5. 3%

Revenue share ADC shot down n Amendments to Rev Share ADC calculation rejected n Currently main ADC contributor ILD, if Rev. Share, tariff on local calls will increase; drop in ILD illogical n ADC rev share would be on top of already rev share license fee n Later possible with increasing minutes and lower ILD share n New ADC from 1 February 2005 [previous method] n Given exceptionally high growth in minutes ADC per minute reduced, but total ADC unchanged n Only BSNL will receive ADC on incoming ILD and outgoing Cellular/WLL[M]. Others can on outgoing. n Over time USO will increase and ADC will decrease merge

New ADC of 1 February 2005

Expectation n Huge increase in traffic, so can bring down ADC per minute and still provide BSNL annually INR 50 b in ADC. n Largest drop is in ILD 60% [attempt to check the grey market; private ISD call cost to drop by 11%, BSNL by 24%]. NLD 40% n n Example Air. Tel to US: INR 16/min 14. 24/min BSNL: INR 7. 20/min 5. 45/min n Migrating to a Revenue Share and merged USO regime

Again, consultation March 2005 n Yet another consultation in March 2005 n Should ADC be restricted to rural FL? n Tariff ceiling only on Rural FL, AD very high in R-FL

Consultation March 2005 n Should ADC be available to non-BSNL? n Actually no. No deficit once local call surplus is considered. But given part [outgoing ADC] now. n Why ADC for wireless access? [WLL[F] can be moved around just like WLL[M] or Cellular] n n n Private operators 80% fixed access through WLL[F] Lower last mile cost, higher equipment cost [? ] But cannot distinguish b/w FL and WLL[F] so kept WLL[F] in ADC n Moving to Rev. Share n With reduced ILD ADC, and increasing overall minutes along with uniform ADC for domestic calls can TRAI shift to rev share?

Current status n 183 pages of responses [posted 17 May 2005? ]. Summary of main responses… n BSNL opposes the ADC reduction n n n “telecom provider of last resort!” Annual revenue loss of INR 12. 5 b [TRAI calculations] INR 79 b [BSNL calculations]; arrears INR 110 b WL service totally unviable TRAI too many consultations; confusion TRAI not submitted calculations non-transparent ADC includes self-funding [calls w/in network 80%] but should be Net ADC [from external networks] n Also oppose revenue share n Higher local call costs

Current status n MTNL argues for urban ADC Delhi, Mumbai 92% basic service [large legacy network]; annual loss INR 10. 8 b serving the urban poor [at below cost rental] without full ADC [INR 4. 5 b annually elsewhere] n But, no revenue share n Lower ILD loss of forex. , foreign carriers benefit n Tata/VSNL n In principle “market forces” but given social obligation need ADC support n Combine USO+ADC and subsidize all “below cost” service by everyone. n BSNL/MTNL got free entry to cellular; license fees reimbursed by budget grant etc. n WLL[F] is the way forward in rural in the future; need ADC support n TRAI does not create competition in FL n

Current status n Reliance n No justification for ADC in India; if ADC then should be uniform across services and operators n No economic rationale’ [only notional] Can apply only for FL in “rural area” n But BSNL earns revenue from various services, not stand alone [unfair advantage for BSNL] n Tariffs are based on forbearance except for rural FL n Define AD [rural access or affordable access] n Define “rural area” n Phase out ADC; USO is sufficient to meet social, economic and national objectives n

Issues n BSNL network is unviable, they did spend enormous amount then, but n Should it be sustained at such a cost n Why cannot it be funded through [simpler] USO n All non-rural users [via operators] pay for rural roll out. n Why bias towards FL? n Is ADC paying for “technology mistakes of BSNL”? n Why not technology neutrality n Open doors for options such as Wi-Fi and Wi-Max also n ADC “grey market”

Issues n Regulation should not hinder development through technological advancement and market forces n “Whenever there is a conflict between dumb regulation and consumer benefit, it is regulation that should yield space, not the consumer” [ET editorial 24 March 2005]

ADC in sum n Conceptually complicated n Objective not clear [definition; basic? ] n Technology bias that defeats the purpose n Encourages parallel markets [by-pass] n Design flawed n Need detailed information from commercial entities n Junk in junk out n Nightmare to implement n Keep changing rules of the game [2003 May, 2004 Feb, 2005 when again…] ~ not conducive for business n Should be merged with USO on a simple, technology neutral, revenue share model

Discussion

Contacts n www. lirneasia. net n Harsha de Silva n hdes 1@yahoo. com n Payal Malik n payal. malik@gmail. com