REGULATORY EXPERIENCE IN DEVELOPING NATIONS

Telecommunications Re-regulation in Bangladesh: A Broadband Future Through a Development Initiative?

ABSTRACT

The latter half of this decade was witness to significant changes in the regulatory environment in Bangladesh. Various forms of competition have been allowed into the telecommunications landscape, most notably involving wireless technology. Bangladesh has taken a unique road in deregulation of telecommunications. Bangladesh has competition only in the local loop: internal and international long distance remains a monopoly. Using the lens of GrameenPhone we view the regulatory innovations in Bangladesh and conclude that such regulatory experiments are certainly worthwhile if not always optimal, and thereby implicitly oppose the economic orthodoxy of complete deregulation and immediate privatization as supported by the WTO and ITU.

In 1996, the wireline network in Bangladesh was in dismal shape. The same year, wireless licensing was expanded. One of the wireless providers to surface soon thereafter was a subsidiary of Grameen Bank, the premier development organization in the region. This paper will discuss how regulatory changes in Bangladesh enabled GrameenPhone (the cellular provider under Grameen Bank) to offer a development program for the explicit purpose of expanding access to POTS (Plain Old Telephone Service).

The GrameenPhone Village Program, Grameen's telecommunications for development initiative, offers a potentially sustainable solution to poor connectivity in rural Bangladesh. The program achieves this end by combining the use of new technology with a unique entrepreneurial approach to marketing and financing their service. Is that because of or despite the regulation? This is the question we explore.

INTRODUCTION

Much of the world has, at best, limited access to basic telephone service. Bangladesh is no exception to this rule. A brief survey of the telecommunications sector in Bangladesh (PBC) shows that in Bangladesh, which ranks among the most densely populated countries on the globe, one telephone serves 275 people, equating to one of the lowest teledensity measures worldwide. Of the total telephone count in Bangladesh, approximately twenty percent serve the rural areas - areas which comprise eighty percent of the total population. The need for telecommunications sector investment and growth is obvious. This is especially so if Bangladesh intends to participate and coexist economically and socially in a larger information intensive world.

This paper discusses, through the lens of Bangladesh and GrameenPhone, the possibility of implementing a wireless infrastructure technology (GSM) to simultaneously offer expanded access to affordable basic telephone service and broadband. The importance of the latter consideration cannot be overstated, as the gap separating information technology have's and have-not's displays every appearance of continued widening.

TELECOMMUNICATIONS FOR DEVELOPMENT

Telecommunications for development is a neo-liberal development strategy based on deregulation and the potential of competition to expand access to communications facilities. Telecommunications expansion is not a goal in and of itself but rather is meant to support economic, political, and social goals.

Telecommunications for economic development is not generic, but rather may be focused on any of several goals. First, nations may seek to connect literate populations with employment abroad. This human-resource focussed strategy is being pursued by India, which has a large number of English literate professionals is in a time zone complementary to the American West Coast. In effect this is making distant jobs virtually local. This connects entire industries with the global market (e.g. software development), by allowing employment of transcribers, translators, and data entry clerks and the like to serve industries of other nations.

Secondly, telecommunications development may seek to reduce the cost of locating in a country by providing connectivity to all world markets. Connectivity makes is possible for nations to leverage low wage rates and favorable tax environments in order to compete as a location for manufacturing facilities. This was the goal of Malaysia's Multimedia Super Corridor. This is a tactic to move jobs to local people.

Alternatively, telecommunications for economic development may focus on connecting local markets for handmade goods to distant locations. A classic example of this is People Link (www.peoplink.org) which uses its Internet-based catalogue to connect pre-industrial producers with post-industrial consumers. People Link provides the scanners, cameras, connectivity and training to enable producers to sell their goods over the network.

Finally, telecommunications for development has the potential to connect internal markets more effectively with each other. Korea (although hardly a developing country) exhibits this focus for telecommunications. By connecting the country and the industries effectively, national industries are able to move to less desirable areas in Korea. Consequently, Korean industry can more effectively compete overseas. This approach to development concerns improving the connectivity of internal markets, ultimately improving waste reduction, and overall government and private sector efficiency within the country. This dimension focuses on developing assets (specifically human assets) within the country rather than connecting to the global market per se. As with Korea, telecommunications in Bangladesh focuses on connecting the country to itself more effectively. Telecommunications reform in Bangladesh has followed the neo-liberal paradigm to increase internal connectivity, while the World Bank has invested in satellite links to increase Bangladesh's international connectivity.

There are several reasons to consider telecommunications development as equivalent to development in general. Rapidly advancing telecommunications technology is a prime means to achieve a wide range of social and economic goals in numerous socially oriented sectors. Sector improvements can include the delivery of education and health services, and the facilitation of poverty alleviation programs (Schmandt, et al 1989).

Abstracting beyond Korea and Bangladesh, the major goals of telecommunications for development are the construction of an advanced and efficient telecommunications infrastructure and the liberalization of national telecommunication industries. This development is to be enabled through privatization and deregulation, without emphasizing the use of mass or interpersonal media to bring about attitude change or to diffuse the ideas and innovations founding the approach. In this light, the value of telecommunications is considered the sum total of increases in access to information and opportunities for entrepreneurs (McDowell, 1997). The choice of technologies used to modernize and expand the main telecommunications network influence the development of the sector and the applications of information technology in the future (Saunders, et al, 1994).

EXPANSION MODELS

There are two fundamental mechanisms for providing universal service. The first is a state-owned PTT or regulated monopoly which uses cross-subsidization to pay for reduced rate services to targeted populations. The benefits of rate averaging, economies of scale, interconnectivity, and reductions in infrastructure redundancy and duplication, all contributed to the global embrace of telecommunications as a natural monopoly. Natural monopoly, as warranted by arguments of national security and public benefit, underlies the post, telegraph and telephone (PTT) model. Through this government monopoly model, infrastructure inter-operability is virtually guaranteed. However, decades of experimentation have shown that with the absence of competition, affordability and an optimized rate of expansion are sometimes sacrificed (Mueller, 1993).

The second mechanism for providing universal access is facilities competition. Current attempts to produce facilities-based competition across the globe, including the GrameenPhone Village Program detailed above, offer a range of market-based models for providing service to low-end consumers. Technological changes have been the dominant driver of deregulation in telecommunications and have marked the advent of market-based regulatory approaches. Technical change alters the expense of access expansion, potential for seamless interconnection, and capital requirements for infrastructure investment. As a result, privatization and competition are becoming increasingly adopted solutions to the problem of universal access. Several projects of infrastructure expansion by private, competing bodies are currently in operation and include the use of prepaid phone cards, and a myriad of sub-leasing schemes such as privately owned wireless public phone services. Market-driven expansion is arguably particularly well suited to developing nations where state subsidization does not necessarily afford reasonable rates of expansion. Deregulation leading to competition can also enable access to private capital markets and free the state from making initial expenditures or taking up-front market risks.

There is a historical argument for competition as an effective means for expanding universal access. In the United States, after the expiration of the original Bell patents and prior to the initial AT&T monopoly, access expansion was fiercely competitive (Mueller, 1993). Expansion of service and access through competition was extremely successful but produced connectivity without adequate inter-operability and with a wide range of quality of service (rural farmer-owned cooperatives did not offer high quality or 24/7 service).

Bangladesh adopted a national PTT as a result of the natural monopoly of telecommunications infrastructure. Today, in order to upgrade the telecommunications infrastructure, attract capital, and expand the diversity of service options, Bangladesh is opening its national telecommunications market to competition.

BANGLADESH: THE TELECOMMUNICATIONS INFRASTRUCTURE

Bangladesh, a country slightly smaller in size than the state of Wisconsin, shares over ninety percent of its border with India and the small remaining portion, to the Southeast, with Burma. The country encompasses an area of approximately 144,000 square kilometers and a total population of 131 million, yielding a population density of over eight hundred inhabitants per square kilometer. The geography of Bangladesh is hilly and somewhat mountainous in several regions but as a whole is primarily alluvial plain, the result of the confluence of the Ganges, Meghna, and Brahmaputra rivers. The region experiences a monsoon season from June to October during which time much of the country can be flooded (CIA, 1999; World Bank, 1999).

Bangladesh remains one of the poorest countries in the world despite sustained domestic and international efforts to improve the economy. Major impediments to economic growth include frequent cyclones and floods, the inefficiency of state-owned enterprises, a rapidly growing underemployed labor force, a predominantly agricultural economy, inadequate power supplies, a continually increasing birth rate in combination with natural and economic limitations The transportation infrastructure in the country is not highly developed. In the mid to late 1990's, railways covered approximately 1,800 total track miles and paved roadways covered approximately 8,500 total miles with a majority (approximately 5,000 miles) being paved in 1992 (CIA, 1999; World Bank, 1999; ITU, 1999). In comparison Wisconsin had 110,000 miles of highways, roads, and streets. Thus the telecommunications infrastructure does not stand out as the poorest of the national infrastructures. Bangladesh has not been able to secure significant capital to modernize or significantly increase the rate of expansion of the telecommunications infrastructure. (Shetty et al, 1997).

Bangladesh, like most developing nations, balances a recognition of need for affordable communications technology with reasonable fears about job security for the state workers in the event of privatization. The most complete picture of the state of wireline telecommunications in Bangladesh is offered by the ITU. Given the historical trends in telecommunications and recent flooding (1998) in Bangladesh, it is unlikely that fundamental change within the state firm (the dominant wireline provider) has occurred.

Teledensity is perhaps the most recognized measure as an expansion indicator in the telecommunications industry. Teledensity measures relate most directly to homogeneous population distributions in which sharing of resources is not explicitly accounted. Total access to the telecommunication infrastructure in an isolated, homogeneous population is equated with a teledensity of 100 (in essence a one-to-one ratio of telephones to population). Teledensity measures in Bangladesh hint at lagging expansion efforts.

It is safe to conclude that the telecommunications sector in Bangladesh rests on a very weak infrastructure. In 1996, the industry showed an extremely low teledensity of 0.26 telephones per 100 inhabitants. In the same year there were 155,000 people wait-listed for telephone service. The wait-list duration averaged 6.6 years. Thus, assuming complete recovery from the floods and steady connection in the intervening three years 77,500 of those people have been connected. Given the population 131 million this is not a significant change in teledensity. During the same period there were only 3,410 public telephones in the country.

An overwhelming majority of the country's telephone lines (91 percent) were concentrated in urban areas (36.6 percent are in Dhaka alone) constituting an urban teledensity of 1.09 telephones per 100 inhabitants. In stark contrast, only 9 percent of the total main lines existed in rural areas, a teledensity of 0.02.

In 1995 there were 19,300 government employees working in the telecommunications sector, effectively constant from 1992 (ITU, 1999). The revenues obtained form the sector amounted to US$ 200.4 million.

In contrast to the homogeneity ideal implicit with teledensity measures, Bangladesh has a heterogeneous population distribution with the rural population more accurately modeled as a distribution of villages. The rural population is nucleated in villages where devices are easier shared (from economic and geographic perspectives). In villages, total access to the telecommunication infrastructure can exist with a teledensity considerably less than 100. Though, as argued before, low teledensity measures are not necessarily in direct proportion to telecommunications accessibility, the teledensity measures listed above concerning Bangladesh do indicate the relatively limited scope of the state supported wireline infrastructure.

The reliability of the BTT wireline infrastructure experiences limitations of its own. The reliability of the infrastructure is illustrated through an annual main line fault rate of approximately 600 faults per one hundred mainlines.

The PTT infrastructure is old, in disrepair, and that given over ninety percent of the total mainlines were urban the infrastructure has been, and continues to be, predominately localized away from rural society. Rural Bangladesh is not served well. Alternative infrastructure projects do not yet compensate for this. As of 1996, cellular was not a competitive access technology (only 0.9 percent of the total phone subscriptions were cellular) (ITU, 1999), and there have been only 2 rural landline competitors (operating in separate regions). The low number of cellular subscriptions reflects less a non-viable technology and more entry and artificial cost barriers.

Until recently, the BTTB was both a competitor and a regulator in the local loop, now the BTTB functions only as a regulatory agency. Main telephone line growth has failed to keep pace with demand for basic service and the waiting list for telephone services is growing at about 9% per annum. With a call completion rate of about 40%, there is widespread dissatisfaction among key business customers. The government expects private investments and operations to play an increasing important role in the provision of telecommunication services. Even the most ambitious plans privatize BTTB do not place the change of corporate structure until 2010 (ITU, 1999).

Early in 1997 competition to supplement the state-run national operator appeared in the form of two of the three newly licensed digital cellular services became operational. However the government did not permit cellular operators to install their own long distance infrastructure, and thus the new operators have no choice but to turn to the BTTB for interconnection to the public network. As additional private operators are licensed and the traffic increases, interconnection bottlenecks and traffic congestion are likely to worsen as all competitors which require long distance depend on BTTB to make adequate network investments (Bristow, 1999).

At present the telecom market in Bangladesh is comprised of a half million phone connections, of which four fifths are with the fixed line network. The remaining one-fifth, approximately 90,000 connections, is divided among several private service providers, some wireline, some wireless: Pacific Bangladesh Telecom Ltd., GrameenPhone (since 1997), Bangladesh Rural Telecommunications Authority (BRTA), TMIB (since 1997), International Communication Technologies, and Sheba Telecom (since 1998). Just recently, Telecom Malaysia and the AK Khan Group of Bangladesh launched the AKTel mobile phone service in Dhaka (ITU, 1999).

THE REGULATORY REGIME

Bangladesh in 1990 had a single state-owned telecommunications provider. This single provider is still exists as government-run Bangladesh Telegraph and Telephone (BTT). The Bangladesh Telegraph and Telephone Board (BTTB) was created out of the Telecommunications Department in 1979 to oversee the BTT. There is considerable government concern over the privatization of the BTT reflecting fears of immediate and unavoidable employee layoffs. The BTT controls three long distance monopolies: domestic, international (terrestrial) and fixed satellite. (The majority of the overseas bandwidth is provided by fixed satellite - this monopoly is of unusual interest in the Internet context.) The BTT does pay into a universal service fund; however, there is no legal definition of universal service so the efficacy of the fund is impossible to evaluate against any stated goals.

Given the systematic incompetence of the BTT in terms of network upgrades, expansion and reliable interoperability, and the regulatory changes in partial compensation, competition does exist in Bangladesh's wireless and fixed local loops. The convergence of long distance monopoly and local loop competition can be considered an experiment in the provision of universal service, contrasting a classical universal service funds approach with a competitive mechanism.

It is important to note that universal service has vastly different connotations in the developed and developing worlds. In the developed world, the issues of universal service are primarily focussed on connecting those who are unconnected and therefore may be more appropriately considered in terms of access as opposed to service. Consequently, infrastructure development is the critical element of universal service drives in the developing world.

While telecommunications deregulation is much heralded, it is quite unusual to provide competition only in the local loop as this leaves competitors dependent on the long distance monopoly incumbent for long-line connections. In the United States, it was the dominance of the long distance market (as well as mutually profitable association with JP Morgan) that allowed AT&T to become an effective end-to-end telecommunications monopoly. Despite the existence of long distance monopoly in Bangladesh, the telecommunications sector may not display the same fate due to the state of the BTT network.

UNIVERSAL SERVICE & TECHNOLOGICAL CHANGE

Though universal service, in its broadest sense, is an economic goal for developing nations, in predominantly rural developing nations with low telecommunication penetration rates, basic considerations of accessibility to POTS infrastructure is the dominant universal service concern. Tariffs must be affordable and on the supply-side. In short, services must be available and affordable to the end user. The ITU names two primary supply-side strategies. One is to increase the number of carrier licenses and the other is to exploit advantages in telecommunication technology. Ideally the former introduces competition and expands resources, the latter can afford non-traditional, cost effective solutions to telecommunications infrastructure development.

Alternate infrastructure technologies are often well suited for expanding access in developing nations. Specifically, wireless local loop technologies have proven cost advantages in covering large areas with phone service. There are some fears that any direct competition enables regulatory arbitrage, and thus undercuts the cross-subsidization historically used to afford infrastructure development, maintenance, and access for universal service. GrameenPhone offers the possibility of self-sustaining expansion and access in the absence of the long distance fees used to subsidize universal service. In this way, GrameenPhone is the ideal for considering the possibility of the self-supporting wireless local loop as a mechanism for expanding universal service.

There are several recognized initiatives employed for the achievement of telecommunications affordability and availability: privatization, infrastructure franchising, encouraging and fostering wireline local service competition, and licensing to serve underserved areas. In Bangladesh, the state run wireline infrastructure has not been privatized, nor should the Village Program be considered infrastructure franchising. We view the GrameenPhone Village Program, a wireless solution to access expansion, as representing the licensing initiative, noting that there is no meaningful competition with incumbent local wireline providers. In fact, Grameen Phone depends upon the cooperation of the wireline incumbent for long term efficacy as well as near term long-distance connectivity.

GRAMEEN

The Grameen Bank is an organization of micro-enterprises and a human development organization with integrated financial, social, and communication services. The goal of the bank is to empower poor people to improve their socioeconomic conditions in an environmentally sound and sustainable manner. Toward this end Grameen organizes the poor, landless, and mostly illiterate people of rural Bangladesh. Grameen's target is the bottom 25 percent of the population, who are primarily women. Mohammad Yunus began the Grameen Bank as an action research project in 1976 and it became a full-fledged bank under his control in 1983. So far the Bank has inducted 2.3 million members (94% of these are women), reaching out to 37,000 of Bangladesh's 68,000 villages. In June 1997 the bank crossed the US$2 billion mark in cumulative loans averaging less than US$175.2 and the repayment rate for the bank is 95%. Apart from providing credit to the poor, Grameen also offers training in adult literacy, family planning services, environmental protection, health care, and nutrition. Yunus's goal, as voiced through the Grameen Bank mission, is for the bank to operate as a successful capitalist enterprise enabling both financial and social returns.

GrameenPhone, the brainchild of Iqbal Z. Quadir, was established to provide Global System for Mobile (GSM) cellular service at affordable prices to all people irrespective of their locations in Bangladesh. According Yunus, GrameenPhone is an example of a successful business enterprise driven by both profit and a social consciousness motives.

GrameenPhone is focusing on rural villages where there is rarely wireline competition. The services of GrameenPhone are targeted to help Bangladesh keep pace with other countries and reduce its existing disparity in telecommunication services in urban and rural areas. GrameenPhone strategic goals are to earn healthy returns for its shareholders and at the same time, contribute to genuine development of the country.

Grameen Bank established Grameen Telecom to manage the bank’s interests in telecommunications, as the bank believes that a lack of communication facilities in the rural areas is one of the major obstacles to rapid economic development in Bangladesh. Though GrameenPhone and Grameen Telecom operate in close associaion, Telenor is responsible for the management of GrameenPhone as the guiding economic and administrative force behind the GSM provider's goal of building convenient and cost effective communication facilities throughout Bangladesh. Grameen Telecom owns 35% of GrameenPhone and shares ownership with Telnor (51%), Marubeni (9.5%), and Gonofone Development Corporation (4.5%). (ITU, 1999)

GrameenPhone is seeking rapid growth in the subscriber base by marketing its service as a wireless local loop service than a straightforward mobile phone network. GrameenPhone is primarily a commercial operator providing cellular services in both urban and rural areas.

GrameenPhone is going for rapid growth in the subscriber base by marketing its service as more of a wireless local loop service than a straightforward mobile phone network. Today, GrameenPhone is a commercial operator providing cellular services in both urban and rural areas and with 25,000 customers. The goal is to expand this an order of magnitude in the next year, with at least 300,000 subscribers signed up nation wide by 2001.

To finance this infrastructure expansion GrameenPhone has an initial funding of $125 million, including a $50 million loan from International Finance Corporation, Asian Development Bank, and Commonwealth Development Corporation in Britain. The Soros Economic Development Fund loaned Grameen Telecom $10.6 million to help it bring reliable and affordable telephone service to virtually all of Bangladesh (Financial News, May 5, 1999). This income makes GrameenPhone more likely to survive in the long term, however the terms of the loan make Grameen a less than perfect case for universal service expansion through local-loop only competition.

GrameenPhone was established to provide Global System for Mobile (GSM) cellular service at affordable prices to all people irrespective of their locations in Bangladesh. There is one additional GSM competitor and a CDMA competitor. GrameenPhone is focusing on rural villages where there is rarely wireline competition. The services of GrameenPhone are targeted to help Bangladesh keep pace with other countries and reduce its existing disparity in telecommunication services in urban and rural areas. GrameenPhone strategic goals are to earn healthy returns for its shareholders and simultaneously contribute to genuine development of the country.

The standard GrameenPhone product connects to the local fixed network maintained by BTTB as well as to all GP mobiles (Financial News, 1999). GrameenPhone's Village Program was started to bring the information revolution to the doorsteps of the villages in Bangladesh and also to use telecommunications as a weapon against poverty. The program is financed by the communications businesses developed by those who own the phones. The village program provides the capital to purchase a phone and begin selling calls. GrameenPhone is also a reseller, and sells air-time at cost to its village merchants. The initial loan is approximately $500, usually borrowed by a poor woman in a village. This covers the purchase of a GrameenPhone GSM phone, and she can immediately begin to sell communication services to fellow villagers.

A pilot program involving 150 villages suggests that the village phone concept is economically viable. Each of the village operators made an average of $2 a day or $700 a year after covering all her costs. This earning is more than twice the country's annual per-capita income and is a proof that there is a demand for communications in the villages. The initial expectation was that there was little or no demand for communications in the villages -- after all, whom would the potential customers call? In fact there is a demand for personal communications between villages, between villages and cities, as well as business communications, for example, information about prices in the centers of commerce. There are currently more than 500 village pay phones in operation.

GrameenPhone selection depends on three factors: the applicant's history, location, and facility in English. GrameenPhone selects women for this program based on their past borrowing records with Grameen Bank. Residence location is important in considering loan requests because for many villages the Village Phone will be the only link to the telecommunications infrastructure and therefore a central location will prove beneficial, both to business and to the community. GrameenPhone also ensures that at least one member of the family knows the English letters and numbers. According to GrameenPhone, women choosing to retail telephone services need only a one day training session. Their success in rapid training and in their business has partly been attributed to their general entrepreneurial skills and the confidence they have built through their past income generating businesses with the help of the Grameen Bank.

GrameenPhone's aim is to connect a wider range of the rural population and promote general economic development within Bangladesh. GrameenPhone emphasizes rural coverage, as these areas have traditionally been underserved. Service to the villages helps them develop economically, and allows GrameenPhone to develop a rural market for its services in regions with minimal landline and wireless competition. From a development perspective, this strategy also provides self-employment opportunities for the rural poor. An implicit goal in this framework is the ideal of sustainability. The GrameenPhone Village Program offers a potentially viable approach to this end.

REGULATORY OBSTACLES AND COMPETITORS

In 1997 the government of Bangladesh opened telecommunications services to competition by removing the telecommunications sector from the "Reserve List" but did not develop an adequate regulatory framework (Bristow, 1999). The Telecommunications Regulatory Board (TRB) is still being formed, thus it is not currently equipped to adequately handle all regulatory functions. TRB has the responsibility to monitor the interconnection facilities that BTTB makes available to private operators. Over the long run TRB's revenues from operator license fees are likely to increase substantially, as existing service providers expand their operations and new operators are licensed. A National Telecommunication Policy is under preparation and a new telecommunications law is being drafted.

The government's strategy is to allow both the Bangladesh Rural Telecommunications Authority (BRTA), a wholly indigenous company, and Sheba Telecom, under a separate license, to provide rural services in competition with BTTB. BRTA, in cooperation with International Communication Technologies (Bangladesh) Ltd., is operating a rural telephone network in Northern Bangladesh. The aim of this project, partially funded by the World Bank, is to increase telephone lines to over 120,000 over the next three years through a combination of microwave and wireline (potentially fiber lines) systems (World Bank Group, 1999).

Sheba Telecom uses a wireless local loop system to provide services to 1,000 subscribers spread across five towns and cities in Southern Bangladesh, while BRTA serves about 16,000 in the North of the country. Sheba has acted unilaterally to ensure sufficient interconnection was available before investing in its planned mobile phone system by offering to subsidize BTTB for the low cost of installing additional microwave transmission facilities.

BTTB has been unsuccessful in attracting foreign investment into network upgrades and thus has had to finance the build-out of 67,000 digital fixed lines itself, through the sale of telecommunication bonds. The organization also hopes the similar bonds will raise enough capital to replace old analog exchanges and expand the long-distance links (Shetty et al, 1997). This may be, in part, due to the lack of efficacy of BTTB investment. In 1999 US$110.7 million was spent in telecommunications investment and 29,500 new telephone lines were added (ITU, 1999). In contrast, approximately one hundred and ten cellular sites could have been placed for the same cost, covering more territory and serving an order of magnitude more people. Consider teledensity in the perspective of Bangladesh. Between 1998 and 1999 competitive cellular providers went from 57,000 to 105,000 subscribers. (Budde, 1999)

Pacific Bangladesh Telecom Ltd (PBTL) is a joint venture partnership company, owned 50% by a local conglomerate, the Pacific group. The company was established in 1990 as the first and only fixed and mobile cellular telephone operator in Bangladesh. The development of Bangladesh's cellular phone market, until recently, was hampered by the failure of BTTB to provide enough connections due to the upgrading of its own network. PBTL had 7,000 subscribers in 1997 and was growing at the rate of 600-700 subscribers per month. On the analogue front PBTL also expected the number of registrations to increase (Asiamoney, 1997). The company plans to digitize and expand its system to take on about 100,000 subscribers by the turn of the century. As of March 31, 1999 PBTL launched an 800 MHz cellular system supplied by Motorola Inc. With a capacity for 50,000 subscribers, the new service in Dhaka will be operated alongside the Motorola-supplied analog network that serves an additional 25,000 subscribers. PBTL has also commenced a second service in Chittagong, which has the ability to serve 50,000 customers (Wireless Today, 1999).

PROMISES AND PROJECTIONS

The market mechanism for expanding telecommunications infrastructure, as exemplified by GrameenPhone, is currently viable for several compelling reasons. Given the population, poverty and lagging infrastructure of rural Bangladesh, it is clear the rural shared cellular market is far from saturated. Before the success of the GrameenPhone pilot project, the existence of demand was questioned. Now the existence of demand has been proven. Finally, due to mature cellular technology, prices in cellular equipment are affordable. These factors highlight the potentials in the neglected Bangladeshi rural market, which commands 80% of the population, and purportedly 50% of the economy.

What, in effect, GrameenPhone has been able to develop is a niche market where:

• it does not competes for connectivity with the wireline incumbent (there is really no competition as the BTTB does not offer services in many of the villages)
• there are enough opportunities to establish local markets without direct, local competition
• and where (as an aside) consumer price elasticity keeps the prices reasonable (though it is interesting that there are documented cases — at least one- where the phone lady undercut the local competition)

The price of performance continues to fall in communications as well as computing. While Moore's law predicts a doubling of processing power per chip area every eighteen months, no popular law exist for the equivalent result in communications in optical, mobile wireless, and fixed wireless systems observed the past decade. An implicit corollary is that the cost of processing power and connectivity should be decreasing. Universal communications requires that both trends continue. This is most obvious in the developing world today through proposals of free hardware for attention span and service contracts (e.g. www.peoplepc.com gives computers away for a service contract for internet access). In developed nations, expected subscriber revenue is adequate to pay for both connectivity and end-point equipment. (We do not mean to imply that individual consumers at the low end of the income spectrum in the developing world will have the purchasing power to support such business plans in the near term.)

Wireless costs are a result of the fact that an exponential increase in performance will not correspond to an increase in demand for processing power (supply will surpass demand). Soon we can expect wireless to be among the increasing pool of commodity technologies.

The combination of technological affordances, and a nascent long-distance market illustrate the economic feasibility of the GrameenPhone development project. The dominant costs (technology) of GrameenPhone will drop. This means that the continuing driving costs will be the administration of loan management. The dominant costs of the PTT, however, are from the costs of buying the physical plant. These costs are dominated by labor rather than technological costs. Thus there is no expectation of subsequent reduction in subscriber costs.

It has been widely documented that wireless networks are currently less expensive in terms of area coverage. In addition, the personal costs of GrameenPhone, services, subscriber management and loan management, are exactly those services that are expected to drop in price with technological advances and automation. In contrast, there are no expectations in reduction of cost of digging trenches and burying cables, or hoisting poles and stringing lines.

BROADBAND

Internet access is the long-term goal of telecommunications infrastructure investment. The wireless infrastructure provides lower bandwidth than a wireline infrastructure. Thus the question, can a GSM wireless infrastructure provide Internet access? Bandwidth comparisons for different technologies is shown in the following table:

GSM networks in theory can provide 14.4K service. It is worth noting Internet growth before 1997 when home and dial-up access were dominated by connections at or less than 14.4k.

The Internet has experienced exponential growth on every continent, as shown in the following table. This growth did not depend upon broadband speeds to the home. For much of this time those speeds now considered narrowband dominated.

Wireline networks using DSL can provide speeds of 8.2Mbps (up, of course considering asymmetry) The ability of GSM provision of Internet access is feasible under two conditions. First, that 14.4K is an acceptable speed for access. This necessitates there be no assumed minimum bandwidth in the design of web services - 144 k in 2000 is not the equivalent of 14.4k in 1997. Wireless technologies per se are not limited to 14.4k. In fact fixed point-to-point microwave promises speeds of up to OC3. The existence of the towers may provide a lower price for upgrading the technology, since, while the switches cannot be used at speeds higher than 14.4 the towers are an expensive, critical element of the wireless infrastructure useful for any but satellite wireless hardware.

The hypothesis that 14.4k is tolerable seems at least feasible. The second condition is that ISDN sees greater success internationally than domestically. This may be possible should international providers adapt to the lessons (predominately pricing) learned from ISDN. The technology itself is applicable. In terms of long-term Internet access there is a bandwidth advantage to wireline, yet there are reasonable scenarios under which GSM can provide Internet access for how ever many years it will take to implement higher bandwidth technologies. In developing nations the dominant ISP market is controlled by the State. The cost of access to the Internet has a lower bound controlled by the state PTT. GrameenPhone does not contract with the state PTT and therefor offers only narrow band connectivity to the Internet. It may be of interst that the PTT neglected the Internet market for an extended term by any standards. Only North Korea lagged Bangladesh in requesting and obtaining a ccTLD.

Bangladesh is subject to flooding, which will no doubt increase in intensity as the exacerbating environmental factors continue. Thus any infrastructure must be easy to protect or easy to replace. GSM systems require protections at the base station, base-station controllers, and hand units. Hand units stay with individuals. The end unit themselves travel with environmental refugees, and therefore have potential to maintain communication, whereas landline phones are inherently linked to the home or location. The implementation of GSM should include flood-hardened base stations and base-station controllers. The GrameenPhone cellular system is inherently better suited for geographical and meteorological conditions in Bangladesh than a wireline system.

The GrameenPhone approach is based on the costs of management of accounts and the costs of hardware, which are both likely to decline. Wireline systems are based on the cost of manual labor, which is already remarkably low in Bangladesh by global standards. The GrameenPhone approach will have long-term success if the base station and base-station controllers are protected against floods.

Looking ahead to data communications, the ability of GrameenPhone to enable widespread wireless depends, in the near term, on the willingness of the community to accept 14.4K connections. In the long term, Internet access through wireless will depend on the costs of upgrading the hardware once towers themselves are in place.

SUMMARY AND CONCLUSIONS

Bangladesh has more competition in the local loop than in domestic long distance. While it is not unusual for states to maintain the highly profitable international long distance market the domestic long distance market should be privatized first or simultaneously with the local loop according to neo-liberal development orthodoxy. Bangladesh is one of three non-communist nations (Bangladesh, Sri Lanka, and India) where the local loop is more competitive than domestic long distance. The results have been mixed. The regulatory changes include the opening of the wireless arena resulting in new possibilities for relatively affordable infrastructure expansion.

This bold regulatory strategy is helping to offset the challenge of providing universal access to the telecommunications infrastructure. The BTT network is severely antiquated, rife with reliability problems and poor interconnectivity.

Options for increasing teledensity include privatizing monopoly operators, opening the sector to competition (facilities competition), and allowing for entrepreneurial approaches to capitalize on the niche local loop markets. GrameenPhone operates through the latter model. The universal service perspective on niche local loop markets reads that in the place of universal service funding (which might be difficult to generate in a developing nation) entrepreneurial, locally competitive mechanisms can operate to the same end. This is especially convincing the higher the chances that such a solution (in the case of GrameenPhone, the development initiative) is sustainable.

Beyond the provision of POTS, it is important to recognize the need to consider broadband communications. GSM may be able to implement broadband communications but its potential is uncertain, and depends, among other things, on the definition of broadband.

Finally, there are clear arguments for opening up the long distance market. Arguably Grameen Phone would exist in any competitive regulatory regime. According to GrameenPhone, their Village Program derives directly from the situation of poverty and general infrastructure weakness in Bangladesh detailed above. Such conditions would clearly exist regardless of the state of domestic long distance. Grameen Phone is a development organization, a near term market opportunity, and as a means towards developing a national long distance market in the long term. Thus the openness of long distance would free Grameen Phone to compete in the long distance market.

The lack of competition in the long distance market has prevented the presumable major influx of capital resulting from the sale of the national phone company. Though not stated explicitly, it is clear that a more lucrative urban cellular market could directly subsidize a low yielding though presumably self-sustaining rural program.

Conversely, the lack of an open long distance market has pushed investment into the local loop, which is in the most dire need of investment. Investment that is targeted at the purchase of a company is certainly a movement of funds. However, the end result of such movement of funds need have no effect on the ground, especially in isolated areas with markets traditionally considered of dubious value. Exchange of ownership does not in itself create local plant. If competition is the best solution to the problem of universal service, the ideals would be to concentrate competition in the areas which are in most dire need of POTS service. The liberalization of the local loop and constraints on long distance have had this effect.

Grameen Phone may have been unable to compete with global providers were the sale of BTT to have coincided with the opening of the marketplace. The fewer, more capital-intensive players would have been unlikely to concentrate on the landless poor as a cellular market.

Thus Bangladesh is offering a clearly different path for universal service by concentrating competition in wireless local loop. The success of this effort is in no small part a function of the success of Grameen Phone. What early data are available suggest that the statategy taken in Bangladesh is functioning well for that nation. This argues against the wholesale deregulation according to a single model advocated by the World Trade Organization and for a more and locally sensitive detailed understanding of the effects of telecommunications as a tool for development.