1. Portrait of the Village Phone Programme¹

1.1 Grameen Telecom's Village Phone (VP) programme
1.2 Origin of GrameenPhone and Grameen Telecom
1.3 How the VP programme works
1.4 Why GTC and the VP are unique
1.5 The cellular phone market world-wide
1.6 The cellular phone context in Bangladesh
1.7 Context of telephone services in Bangladesh and
the Emergence of the Village Phone

1.1 Grameen Telecom's Village Phone (VP) programme

Personal voice communication beyond hearing distance is simply not possible today from 90 per cent of the 68,000 villages of Bangladesh. The Village Phone (VP) programme established by Grameen Telecom provides an opportunity for universal access: according to Grameen Telecom, a person may not own a telephone but he/she should have access to a telephone within a ten minute walk.

The Village Phone initiative was developed by combining the Grameen Bank's expertise in village-based micro-enterprise and micro-credit with the latest digital wireless technology. Grameen Telecom (GTC) is a non-profit organization that forms part of the Grameen Bank family of organizations, and it focuses exclusively on the deployment of the Village Phone programme in rural parts of Bangladesh. GTC buys airtime in bulk from GrameenPhone Ltd. (GP), a for-profit corporation holding a nation-wide license to operate and maintain a mobile cellular network throughout the country. Bulk airtime purchasing from GrameenPhone Ltd. enables GTC to pass on savings to the Village Phone (VP) operator. GTC owns 35% of the shares of GrameenPhone Ltd., and plans to have controlling interest in GrameenPhone Ltd. in the near future.

The Village Phone is a unique undertaking that provides modern digital wireless telecommunication services to some of the poorest people in the world. A Grameen Bank member (most often female) purchases a phone under the lease-financing programme of the Bank and provides telephone service to people in her village. Each Village Phone operator is responsible for extending the services to customers for both incoming and outgoing calls, collecting call charges, remitting payments to Grameen Telecom, and ensuring proper maintenance of the telephone set. The operator's income consists of the difference between charges paid by customers and the airtime charges billed to the operator by Grameen Telecom. Repayment of the loan for the phone set is processed through the existing loan granting and collection procedures of the Grameen Bank. GTC prepares monthly bills for the airtime charges for each phone, and Grameen Bank managers at the community level collect monthly payments from operators, in person, at the village level. [Click here for a related video clip - 482K]

As of November 1999, Grameen Telecom reports that there were 950 village phones in service providing telephone access to more than 65,000 people. Grameen Telecom expects that when GrameenPhone has completed its network in December 2002, 40,000 VP operators will be employed for a combined net income of $24 million USD per annum.

The objectives of GTC are:

• to provide easy access to telephones when needed all over rural Bangladesh;
• to introduce a new income generating source for villagers;
• to bring the potential of the Information Revolution to the doorsteps of villagers;
• to introduce telecommunications as a new weapon against poverty.

1.2 Origin of GrameenPhone and Grameen Telecom

In 1994, a Mr. Iqbal Quadir from a U.S.-based company called Gonofone contacted the Grameen Bank about the possibility of establishing a socially beneficial and profitable telecommunications venture in Bangladesh². Quadir proposed extending telephone service to the poor in rural Bangladesh using the established principles and financial systems of the Grameen Bank (click here for Mr. Quadir's first-hand description of the development of the Village Phone concept). This proposed telephone system would be implemented in such a way as to promote self-employment while furthering economic development goals. The International Telecommunication Union provided a telecommunications expert to assist the Grameen Bank in assessing the prospects of this venture.

In November 1995, Grameen Bank and Gonofone formed a consortium with Norway's Telenor (acting as the technical partner and investor) to develop their project. A few months later, Marubeni also joined the consortium as an investor. In mid-1995, the Government of Bangladesh formally announced the bidding process for the licensing of three additional cellular operators. This provided the sponsors with the opportunity to enter the profitable Bangladesh telecommunications market and thereby pursue their goal of providing telephone service throughout Bangladesh. To represent its interest in the venture, Grameen Bank formed a wholly owned subsidiary called Grameen Telecom. In November 1995, with technical support from Telenor, the sponsors submitted a bid for the license under a competitive bidding process established by the Government.

Throughout 1996, the sponsors proceeded to plan their project. They set up a project company, incorporated in Bangladesh in October 1996 as GrameenPhone Limited. In November 1996, after a protracted evaluation, three cellular licenses were awarded. GrameenPhone was one of the successful bidders. It signed its license agreement with the Government on 11 November 1996 and began operations on 26 March 1997.

1.3 How the VP programme works

Selection, Subscription and Training. GTC first gathers information on villages that have cellular coverage emanating from GrameenPhone's existing network of cellular towers. (The current coverage expands from base stations connected to the fibre optic cable along the Bangladesh Railway network, from base stations installed in major urban centres, and from GP's wireless tower microwave trunk spanning from Khulna to Chittagong, southeast to southwest; see Figure A). GTC Unit Officers then visit the Grameen Bank branches in that area and prepare a list of villages where network coverage is satisfactory to provide Village Phone service. The GB branch manager selects women from among Grameen Bank members in those villages to act as VP operators.

The Grameen Bank has a special set of criteria for the selection of VP operators:

• She must have a very good record of repayment of Grameen Bank loans;
• She should have a good business, preferably a village grocery store, and have the spare time to function as the VP operator. Initially, this may be a side business and eventually switch over to telecom business on a full-time basis after services and revenue justify a full-time commitment (note that analysis of data for this report suggests that the focus on village grocery store ownership may inhibit phone use by village women: see Section 2.5);
• She should be literate or at least she must have children who can read and write;
• Her residence should be near the centre of the village.

After the initial selection of a suitable VP operator by the GB branch manager, the respective Unit Officer of GTC ensures signal availability in the house or shop where the VP operator intends to use the phone. GTC experience has demonstrated that a weak signal can be augmented by use of a high gain antenna mounted on a four- to five-metre bamboo pole affixed to a dwelling or shop, with a coaxial cable running between the antenna and the phone. The high gain antenna effectively makes the GSM cellular phone a "fixed-mobile" phone for many VP operators, although the high gain antenna and pole can be moved within a village, if needed.

The final selection of the VP operator is made by the zonal manager of the Grameen Bank. GTC then buys a cellular phone user subscription on her behalf from GrameenPhone and provides the connection to the member. GTC supplies the necessary hardware and the GTC Unit Officer provides training to operate the phone. The price of the phone and the connection fee are paid by GB to GTC and the member pays these costs back to GB within the stipulated 2 or 3 year period through the existing weekly repayment system for the loan (whereas airtime bills are paid on a monthly basis). The telephone is provided to members under a lease finance programme of GB.

Billing. GrameenPhone prepares a monthly bill at the bulk airtime rate for Village Phones for total airtime used by all VPs. The bill includes the monthly fee for the rental of the line. The bill showing the net amount payable to GrameenPhone is sent to GTC's head office for processing by the end of each month. GTC remakes the individual bills and sends them to the corresponding GB branches with a summary bill due from the branch. Grameen Bank branch managers are responsible for collecting the bills from VP operators and the branch pays the bill to Grameen Telecom six weeks later. This process provides an interesting solution to the problem of revenue collection that sometimes plagues the operators of rural telephone systems.

After Sales Support. GTC has established Unit Offices to cover the VP operation in the field. The duties of the Unit Offices include: finding new areas with coverage, helping the GB branch managers select new members for becoming VP operators, training VP operators, and backstopping any problems faced by VP operators regarding handsets, or billing. GTC receives monthly computerized statements from GrameenPhone and uses these to monitors phone usage among all Village Phone operators. The statements enable GTC to track trends in phone use and spot anomalies in individual bills that could indicate billing errors or indicate that a Village Phone operator is experiencing some difficulty. Unit Office staff are then able to review individual cases with personal visits to the operators.

Thus far, GTC has 5 Unit Offices located in Dhamrai (Dhaka Zone), Norsingdee (Narayanganj zone), Comilla (Comilla Zone), Feni (Noakhjali zone) and Chittagong (Chittagong zone). The number of Unit Offices will increase as the coverage area grows. According to the present plan, GTC will have one Unit Office for two Area Offices of the GB and each Unit Office will be able to take care of approximately 300 to 400 VP operators.

Purchase & Payment. The basic Village Phone package (as of Oct.1999³) contains:

• Nokia 1610 transceiver
• 1200 mAh battery
• fast charger
• sign board
• calculator
• stopwatch
• user guide in Bangla
• price list for calling different locations

The basic Village Phone package costs 15,000 Taka, or approximately $310 USD⁴. The VP operator pays for the phone through weekly loan payment instalments of 220 Taka or approximately $4.50 USD. These payments are made through the local Grameen Bank branch, which is responsible for collecting on behalf of GTC. For the usage charge, the VP operators pay a minimum monthly bill of 154 Taka or approximately $3.20 USD that includes the monthly fee for the line, Value Added Tax (VAT), a GTC service charge, and 100 Taka for the annual government license and royalty fee. Actual airtime charges are added on top of all this.

Electricity. Widespread access to electricity in rural Bangladesh enables the phone operator to recharge batteries or power the phone directly from an electrical outlet. Solar power sources are provided for two Village Phones in non-electrified villages. As Grameen Telecom expands its service, it will be providing phones to many more non-electrified villages. It thus provides a solar-powered solution for powering phones in these villages.

Each solar power package supplied for a VP is made up of a sunlight receptor panel, a battery, a charge controller, associated circuitry, wires and a 12-volt 6-watt tube light. The panel has two sections, each of which have a power rating of 6.5 watts. So the whole panel can provide a total power of 13 watts. The voltage rating is 12V. Its capacity is 31 ampere-hours. The controller automatically controls the level of charging and discharging of the battery. With normal sunshine, this installation can support the power requirement of (i) a VP continuously and (ii) the tube light for over 4 hours in each 24 hour period. The VP itself consumes 6 to 7 watts at 9 volts. Consumption of power can be kept up at this rate - even if sunlight is blocked by clouds - for a continuous period of up to 3 full days, and only if the cloudy situation persists for longer, consumption must be cut down.

Each solar power package supplied for a VP is made up of a sunlight receptor panel, a battery, a charge controller, associated circuitry, wires and a 12-volt 6-watt tube light. The panel has two sections, each of which have a power rating of 6.5 watts. So the whole panel can provide a total power of 13 watts. The voltage rating is 12V. Its capacity is 31 ampere-hours. The controller automatically controls the level of charging and discharging of the battery. With normal sunshine, this installation can support the power requirement of (I) a VP continuously and (ii) the tube light for over 4 hours in each 24 hour period. The VP itself consumes 6 to 7 watts at 9 volts. Consumption of power can be kept up at this rate - even if sunlight is blocked by clouds - for a continuous period of up to 3 full days, and only if the cloudy situation persists for longer, consumption must be cut down.

The whole package and its installation costs Tk 8,500 ($175 USD approx.) and is supplied by Grameen Shakti, which is a subsidiary organization of Grameen Bank, just like Grameen Telecom (see Appendix 9 for a description of the Grameen Bank Family of Organizations. They provide a warranty of 20 years for the panel (imported from Japan), 5 years for the battery (acquired from a local supplier) and 1 year for the circuitry including the controller. Grameen Shakti reports that the cost of the whole unit is likely to go up in future because the solar panels will, in course of time, probably cost more on the international market.

The solar panels have stood the test of time for over 6 months in one of the villages and nearly 1 year in the other. So far, as we observed, those are functioning normally. The VP operators say that there has not been any problem with the units as yet and that they are satisfied with the performance.

Call Charges & Profit. GrameenPhone charges Grameen Telecom Tk.2 ($.043 USD) per minute for a local call at peak hours, whereas urban GP subscribers pay Tk.4 ($.086) per minute. For NWD (national) and ISD (international) calls, GP charges the BTTB (Bangladesh Telegraph and Telephone Board state monopoly) long distance rate plus the VP airtime charges. A 15% Value Added Tax (VAT) is added to call charges. To cover the administrative cost of GTC and GB, a 13% service charge is added to the total GP bill. VP operators are supplied with a price list that includes all kinds of charges and a margin of profit for them.

For example, for a local VP call to another mobile phone, the retail rate is Tk.5 per minute, out of which the GP rate is Tk.2, the VAT is Tk.0.30, and the service charge is Tk.0.30. The remaining Tk.2.40 ($.051 USD) is the profit for the VP operator from which she has to pay the monthly line fee, the royalty fee and loan instalment for the handset.

A different rate applies for a local call to a fixed BTTB line as an additional Tk.1.70 is charged by BTTB; this charge is applied per call and is therefore charged for the first minute only. While GTC provides a set of recommended call rates, our site visits revealed that VP operators charged variable rates depending on their relationship to the user. Moreover, we also came across a non-VP Public Calling Office (PCO) operator in a small town who was using a GrameenPhone telephone only for calls to Dhaka, at a cost to the client of Tk.7 per minute, whereas his real cost is Tk.7.52/first minute, and each additional minute is Tk.5.24. Village Phone use appears to be very price sensitive, and where Village Phones are located near to towns with functioning PCOs, Village Phone operators are very aware of competitive pricing.

The following table shows the breakdown costs for a one-minute peak-time call (in Taka). The first column shows a call from a Village Phone to a fixed BTTB line (VP to BTTB); the second shows a call from a Village Phone to a GrameenPhone mobile line (VP to GP - this amount also applies for each extra minute of a VP to BTTB); the third column shows a call from a GP to a fixed BTTB line (GP to BTTB), and the last column shows the cost of each extra minute of a GP to BTTB call.

Table 1.A Comparison costs for a one-minute peak-time telephone call (in Taka)

The table above begins to explain why there is a wide range of fees charged for telephone calls. Our observations and survey data suggest that telephone users are very price sensitive and are generally reluctant to pay more for a telephone call unless the call involves important financial matters.

According to Grameen Telecom data for the month of September 1999, the average total bill was 6,110.64 Taka (approx. $125 USD) out a total of 652 Village Phones; the component costs are as follows:

Table 1.B Component costs for the average total phone bill (September 1999) of a VP operator

On average, the VP operator earns a net profit of 277 Taka per week (Bayes et.al., 1999); however, a much higher average net revenue is reported by GTC at 2,000 Taka/month (Grameen Telecom, 1999).

Cost of Phone Use and Demand. In Bangladesh, rural people's occupations are becoming increasingly diversified and many people live outside their villages. Fifty percent of rural households do not own any land; they seek off-farm and non-farm income earning opportunities. Less than 8% of the Grameen Bank members surveyed for this report indicated that their main occupation is agriculture. Labour mobility has increased enormously in the recent past and millions now work outside their own village and abroad. Population mobility is a key indicator of demand for telephone services because when people move and relocate, they have a greater need to communicate with family and friends, particularly when labour mobility involves significant remittances to family, as is the case in Bangladesh. Furthermore, the volume of rural-urban trade has increased at a rate much higher than the growth of the economy. Thus, two important factors - enhanced labour mobility and marketization of agricultural production - should result in a growing demand for telephone services.

The Government's stated plan to see card phones installed in towns on an extensive scale may facilitate outgoing calls from thanas (sub-districts), but not from the villages themselves. Data from Grameen Bank members surveyed for this report suggest that there is a demand for increased numbers of card phones and better access to pre-paid calling cards. Because tariffs for card phone use are lower than tariffs for Village Phone use, and because rural phone users are very price sensitive and express a willingness to travel to make phone calls, the Village Phone initiative may see a drop in demand for outgoing calls if card phones and pre-paid calling cards are more accessible in towns. Grameen Telecom will have to monitor this situation and may have to adjust its tariffs in order to remain competitive if alternative phones with lower rates are extensively deployed.

Village Phones enable rural people to make calls from their villages and also receive calls from outside. A VP operator has a financial incentive to ensure that incoming calls are completed and she is therefore willing to make the extra effort to find the person for whom the incoming call is destined. According to Grameen Bank member phone users surveyed for this report, 61% of the last phone calls completed were incoming calls received by villagers, and 58% of the last phone calls completed were connections with people (primarily family members) living outside of Bangladesh. Thus, the Village Phone provides an important link that enables relatives who have left the village to stay in touch with family at home. And almost 42% of Grameen Bank member phone users indicate that their main reasons for using the phone involve discussions of financial matters (primarily discussions about remittances) with family members. Only 7% of calls were made for business and trading reasons.

Our data reveals that the Village Phone plays a key role in facilitating family relationships and the flow of remittances to family members in the village when a family member has left the village for work in another country or Dhaka City. For example, 54% of Grameen Bank member phone users indicated that they were willing to spend between 100 to 300 Taka ($2 to $6 USD) for a three minute phone call involving a financial matter with a family member overseas, and 27% said they were willing to spend between 300 to 600 Taka ($6 to $12.25 USD) for this kind of call. Given an average reported monthly income of 5,000 Taka ($102 USD) for respondents' households, these figures represent significant proportions of monthly household income ranging from 2% to 12%.

1.4 Why GTC and the VP are unique

The Village Phone programme contains many rural development "firsts":

• First rural development micro-credit facility in a developing country to target the creation of micro-enterprises based on information and communication technology (ICT) services
• First rural development micro-credit facility in a developing country to assist in the creation of village telephone service businesses using digital, wireless telephony
• First private sector rural telecom initiative that specifically targets poor village women for establishing micro-enterprise (targeted, micro-level program)
• First private sector telecom initiative with an explicit purpose of rural poverty reduction

As a result of these factors, the Village Phone programme should be examined in-depth by both international donor and investment agencies and by technical providers as an innovative model. However, the Village Phone programme must be examined in context because it has evolved within a unique set of regulatory, commercial and technical circumstances. We should not assume that the Village Phone programme can easily be replicated elsewhere as a complete package.

1.5 The cellular phone market world-wide

Mobile cellular phones using the GSM (Global System Mobile) standard made their debut in Scandinavia in 1992. GSM telephony has become the cellular phone standard for Europe and many countries in Asia and Africa. In contrast, North American cell phones tend to use the TDMA (Time Division Multiple Access) and CDMA (Code Division Multiple Access) standards. Dublin-based GSM Association estimates that there are currently 200 million GSM subscribers world-wide versus 35 million CDMA subscribers. The GSM Association's 5-year growth forecast is for more than 700 million subscribers. Other projections describe this growth with regards to the number of landline subscribers: by 2002 there will be 2 cell phones for every 5 landline phone subscribers world-wide (The Economist, 1997).

In developing countries, the cell phone market is almost exclusively urban. Cellular technology provides significant return on investment in urban areas with higher than average densities of upwardly mobile subscribers.

While the technology is primarily targeted to higher-income, mobile business people, many other middle-income urban users opt for cellular phones because of the reliability of the service and the ease in obtaining a line. In countries where the landline telephone system remains dominated by a state-owned PSTN (Publicly Switched Telephone Network) monopoly, such as in the Philippines and Bangladesh, it is common for people to wait a year to be connected to the main landline grid. As a result, in developing countries cellular phones are a substitute for landlines, whereas in industrialized countries they are a complement.

1.6 The cellular phone context in Bangladesh

The GSM cellular phone market in Bangladesh has emerged during the last three years. The first licenses were granted in November 1996. Today, there are four national cellular operators with 100,000 subscribers centered in urban areas, and 80,000 additional subscribers are expected during 2000. (For a list of all national telephone operators in Bangladesh, please see Section 3.3) To put this growth in context, Bangladesh has approximately 500,000 fixed line telephones (350,000 considered operational), and 2/3 of these are in the Dhaka area. The country's teledensity is somewhere between 0.26 and 0.3/100 people - in other words, there are 2-3 fixed line phones per 1000 people. This teledensity is one of the lowest in the world. Teledensity is closely related to the development status of a country, as estimated by UNDP's Human Development Index (see Figure B). In the context of Figure B, this means that Bangladesh, with 28% of phone subscribers using mobile cellular technology, is poised to be off the Economist chart (see above) in a matter of months.

Unfortunately, the current growth rate of cellular subscribers is limited by a restricted number of interconnections to the national backbone owned and operated by the Bangladesh Telegraph and Telephone Board (BTTB). This has implications for both service providers and service purchasers. In order to avoid this constraint, several of the cellular operators have successfully opted to sell phones that only connect with other cell phones.

GrameenPhone Ltd. (GP) is one of the dominant national cellular phone operators with 50,000 subscribers as of October 1999. GrameenPhone leases and operates a 1,800km-long optical fibre cable from Bangladesh Railroad, effectively providing a parallel nation-wide network to the one operated by the BTTB state monopoly. Figure A shows the current and forecasted national cellular coverage offered by GP.

The rural telephone market is licensed to two operators with exclusive rights for 25 years: Sheba Telecom covers the south (195 thanas) and the Bangladesh Rural Telecom Authority (BRTA) covers the north (199 thanas) (please see maps published by BRTA in Appendix A.1). The licenses are for PSTN and allow for the use of any technology. Both Sheba Telecom and BRTA have opted for a combination of wireless local loop (WLL) technology systems and point-to-multi-point microwave technologies. BTTB remains the operator in the remaining 70 urban thanas (out of a total of 464). Sheba Telecom is the only operator with both a national cellular GSM license and a rural PSTN license.

We understand⁵ that there are 25-year exclusive territory license arrangements for Sheba's WLL line deployment in the south and BRTA's line deployment in the north, and we understand that these licenses were introduced with the assumption that rural telecom is not profitable – an obviously false assumption as shown in Sections 1 and 3 of this report. Grameen Telecom (GTC) can examine licensing options for becoming an alternative, national, rural telecom operator. Through its relationship with GrameenPhone, GTC has access to substantial infrastructure in the form of wireless towers, backbone service and operating experience.

1.7 Context of telephone services in Bangladesh and the Emergence of the Village Phone

Grameen Telecom (GTC) is a non-profit organization that owns 35% of the shares of GrameenPhone Ltd. (GP), a private sector, urban cellular telephone operation. GTC buys airtime in bulk from GrameenPhone and passes on most of the savings to the Village Phone (VP) operator, making use of Grameen Bank's extensive network (1,140 branches spread over 39,346 villages) and its loan collection system to collect revenue from the VP operators. GTC is using GSM (Global System for Mobile Communication) cellular telephone technology at the village level, taking advantage of the GP-installed urban capacity. GP leases and operates a 1,800km-long optical fibre cable from Bangladesh Railroad, effectively providing a parallel nation-wide network to the one operated by the state monopoly BTTB (Bangladesh Telegraph and Telephone Board).

A conservative estimate of the percent of cellular vs. landline telephone subscribers in Bangladesh is over 20%: 500,000 landlines vs. 100,000. A more realistic estimate that acknowledges the number of fixed lines that are not actually functioning comes closer to 30%. GrameenPhone alone plans to double the number of subscribers each year (in October 1999, GP had 50,000 subscribers). This is an unprecedented situation where the cellular network is poised to dominate the national telecom industry (the total number of cell phones is estimated to rise from 100,000 to 180,000 during 2000).

While there are several private sector telecommunication operators in Bangladesh, the Bangladesh Telegraph and Telephone Board (BTTB) currently operates a state monopoly on interconnection services to all operators. Telephone services are dependent upon smooth interconnection among operators so that phone users can place calls to any other phone user, regardless of who operates the lines. BTTB has not provided sufficient interconnection to meet the service demands of its competitors, and this has distorted the telecommunication market in Bangladesh.

BTTB has granted rural licenses for 25-year monopoly "fixed-line" service to rural districts in northern and southern Bangladesh. These were awarded to the Bangladesh Rural Telephone Authority (BRTA), a private sector operator despite the official- sounding name, and Sheba Telecom respectively. Both operators are utilizing wireless local loop (WLL) technology that is generally recognized as a cost-effective, technical solution for universal telephone access in rural areas. GrameenPhone's GSM cellular services are cost-effective solutions for dense user bases in urban areas, but GSM cellular is not the most cost-effective technology for rural service targets. Per line installation costs for BRTA are ranked as being among the lowest in the world for rural service, at about $1,000 USD per line (Kayani and Dymond, 1997). This amount is far less than the installation costs for GSM cellular service targeted to a small number of rural PCOs or Village Phone operators.

GrameenPhone is able to provide rural service, despite the monopoly licenses held by BRTA and Sheba, because it holds a license for "mobile" service. Thus, the mobile character of GSM cellular phones enables GrameenPhone to work with Grameen Telecom to provide PCO-style operations in rural areas where its towers and signals can reach. Grameen Telecom is limited to providing Village Phones within the thin corridors covered by its network of cellular towers (see Figure A). There are large gaps in its rural coverage that must be covered if Grameen Telecom is to meet its goal of universal access.

BRTA and Sheba have been prevented from deploying extensive rural telephone coverage because BTTB has not provided them with sufficient interconnection to meet demand. They could deploy more phones, but users would experience problems when they try to connect to the phone lines of other operators. Even with the limited number of phones each operator has deployed today, users experience significant frustration with interconnection, often trying for over an hour to place a single call from a rural area to a number in Dhaka. BRTA currently operates approximately 20,000 lines, the majority in the form of official or unofficial Public Calling Offices, with plans to double this number if the interconnection problem is solved. Indeed, when we visited with BRTA, we were informed that it had a warehouse full of new digital wireless loop equipment that it could not deploy because of the interconnection problem. One of the competing operators to GrameenPhone currently operates 1,500 rural lines as Public Calling Offices and will likely also expand its coverage if the interconnection problem is solved.

It is probably not justified to conclude that the Village Phone programme exists only because BTTB does not provide sufficient interconnection to competing rural operators in Bangladesh. However, this must certainly be a significant factor that led to the emergence of Grameen Telecom's service. It would be financial suicide for a GSM cellular operator to only target rural subscribers, but when a service targeted to urban customers also reaches into un-served rural areas with high demand, the additional cost of providing rural service is small. Combine that with a proven financial and revenue collection process in the form of the Grameen Bank and its extensive rural coverage, and you have a profitable scenario.

Demand in rural areas is high and cannot be met because so many rural areas in Bangladesh have remained un-served. This, in turn, is due to the interconnection obstacles faced by BRTA and Sheba that have prevented them from expanding profitably in parallel with demand. BRTA and Sheba both use WLL infrastructure which is generally much more cost-effective than GSM cellular for providing rural telecom services. In our interviews with representatives of BRTA and Sheba, we found there was strong desire for rural expansion because rural lines are very profitable; one source told us that their revenue from 12,000 urban cellular lines equaled the revenue from 1,500 rural PCO lines! Only the interconnection problems with BTTB prevent expansion of BRTA and Sheba networks.

Based on the International Telecommunications Union's (ITU) typical model for rural service, we can expect that people in rural Bangladesh will spend not less than 1.5% of GDP per capita on telecommunication services (ITU, 1994) "if they are appropriately deployed" (Kayani & Dymond, 1997: 9). Kayani and Dymond (1997) estimate rural income at $171 USD per person per year (based on an estimate of an overall GDP per capita of $220 USD⁶). This suggests $2.57 USD per capita expenditure. Given a conservative estimate of rural population of 80 million people, this translates into a potential rural telecommunication revenue of $205 million USD per year. And, as Kayani and Dymond (1997) note, rural expenditures can often exceed expectations because of the high cost of alternative forms of communication, such as travel by vehicle.

With this level of expected demand/revenue, it is not difficult to see how a technology designed as a personal communication tool for urban users in the north found an appropriate and profitable configuration for villages with hundreds of users in the south. That technology has been combined with a proven rural financial services organization, a commitment to rural poverty alleviation, experienced rural development professionals, and good management that is capable of exploiting narrow market opportunities in order to create the Village Phone programme.

One final consideration that is of considerable importance for future Internet and data services is that GSM cellular phone technology is currently not a viable option for inexpensive email/Internet/data connectivity. WLL and other options can provide much better bandwidth and cost of service. Indeed, Grameen Communications' pilot rural Internet service centre uses BRTA's WLL service for dial-up Internet access to Dhaka because their experiments with GSM cellular phones for Internet access proved unsuccessful. If Grameen Telecom is to move forward with Grameen Communications in providing Internet and data services, it will find that only the cellular phone towers will be of much value: the GSM system is currently not upwardly compatible with high bandwidth Internet service. BRTA's WLL systems are much more likely to be an infrastructure capable of handling Internet traffic. This is one of the reasons that we later recommend that Grameen Telecom consider entering into partnerships with operators such as BRTA and Sheba.

FOOTNOTES

Return to Full Report