The Role of Competition in Institutional Reform of Telecommunications: Lessons from Sri Lanka 

Rohan Samarajiva*

*Economics of Infrastructure Program, Delft University of Technology, The Netherlands & School of Journalism & Communication, The Ohio State University, Columbus OH, USA. E-mail: samarajiva.1 @  The author is grateful to the members and committed staff of the Telecommunications Regulatory Commission of Sri Lanka who made the activities described in the article possible during the time he served as Director General of Telecommunications in Sri Lanka. This is the author's prepublication version of an article that will appear in Telecommunications Policy Vol. 24 Nos. 8/9
Abstract Consolidation of the recent wave of institutional reforms in telecommunications requires assessment of the position of competition in relation to organizational reform of incumbents and the establishment of regulation. Drawing lessons from the experience of Sri Lanka, a country which has achieved good results from comprehensive reforms over a nine-year period, the article concludes that competition plays the most important role. In addition to the direct benefits of increased investment, lower prices and higher levels of connectivity, quality of service and innovation, competition reinforces organizational reform of the incumbent, contributes to the consolidation and legitimation of regulation and prevents the rollback of advances in organizational reform of the incumbent and in regulation. The article also draws general conclusions from the policy choices made by Sri Lanka regarding the strategy of introducing competition; the sequence of privatization and competition; the balance between telecommunications sector objectives and yield maximization from privatization/ownership; universal access/service; managing competition within technology-defined subsectors; and ongoing regulation for competition. 


Competition, regulation, privatization, institutional reform, Sri Lanka


A wave of institutional reform in telecommunications is sweeping the world. A majority of the world’s nations have enacted or are in the process of enacting new legislation governing their telecommunications sectors. Sector-specific regulation has played an important role in the reforms. There are now over 84 "independent" telecommunications regulatory agencies in the world as against only ten before 1990. Most of these new agencies are in the developing world (International Telecommunication Union, 1999a, p. 5).

Consolidation of the new reforms is not proving to be easy. The difficulties are exemplified by the crisis of telecommunications regulation in India. The highly professional and active Telecommunications Regulatory Authority of India (TRAI) that appeared destined to become one of the leaders of telecommunications regulation in the world found itself in a stalemate with the government-owned incumbent and was reconstituted as two separate bodies in early 2000, after less than three years in existence (ENS News Bureau, 2000). If difficulties of this type are to be minimized it is important that intelligent policy choices be made in the course of telecommunications reform. This article assesses the institutional reform process in Sri Lanka with the view of clarifying the policy choices necessary to consolidate the inchoate reforms in all developing countries, especially the relative importance of competition.

Sri Lanka is a country with a population of approximately 19 million, a population density of almost 300 per km2, and a per capita GDP of around $840. Economic liberalization policies initiated in 1977 by the current opposition party have been continued (and, in some cases such as privatization, accelerated) by the center-left coalition in power since 1994. Sri Lanka's development prospects have been affected by an ethnic conflict that has continued since 1983. 

Institutional reforms in telecommunications began in 1980, driven in part by the inability of the sluggish government Department of Posts and Telecommunications to meet the demands of the newly liberalized, export-oriented economy, and were advanced in two stages. New legislation separating policy, regulatory and operational functions was enacted in 1991. In 1996-97 the legislation was amended to strengthen the regulatory agency, fixed-access competition was introduced and the incumbent operator was partially privatized (Jayasuriya & Knight-John, 1998; Samarajiva, 1997). The range of reform actions undertaken over a long period with relatively good results in a country of manageable size makes Sri Lanka a good case for analysis and for drawing out generalizable policy conclusions (Gunaratne & Wattegama, 1999; Shetty, 1999). The unique association that the author had with the institutional reform process, as an academic researcher in the first phase of reform (Samarajiva, 1993; 1997) and as the Director General of Telecommunications in the second phase may also add value to the analysis. 

Institutional Reform In the classical PTT form of telecommunications supply found in most countries outside North America until recently, a government department or an agency such as the Post Office combines in one entity the three functions of policymaking, regulation and operation. Generally, experience has shown that governments have failed in all these functions. For example, with the exception of Scandinavia and Australia, all PTTs failed to provide universal service until the current wave of institutional reform. As a result the separation of the three functions is seen as a necessary, but not sufficient, condition for improved sector performance. 

Melody (1997, p. 22) sets out the decentralized nature of telecommunications reform, showing that reforms must occur at the government/Ministry level in the form of setting the overall context, and at the regulatory agency and within the operator. It is now commonplace to say that modern management techniques must be applied to telecommunication operators and that they cannot be administered as government departments subject to political and bureaucratic interference. This necessitates the separation of this function. In the same way, there must be specialization in policy and regulation which requires separation of these functions as well. Government must specialize in the broad policy functions of setting out the objectives and the parameters of activity within the sector. The regulatory agency must specialize in the implementation of the broad policy directives and, where such policy emphasizes private investment, exercise the discretionary and quasi-judicial powers essential for a stable environment for private investment. The operator must operate as efficiently as possible. 

The policy must set out the sector objectives and the ways in which they are to be achieved. In many countries, governments have included certain industrial-policy objectives such as strengthening of manufacturing or exports within their telecommunication policies. In a few countries, government has chosen monopoly or duopoly provision as the means of achieving the policy objectives. But today, the overwhelming tendency is for governments to adopt competition as the mode of achieving sector objectives. Because many developing countries seek to mobilize foreign private investment to build out their networks, it is common to see competition introduced on the edges of the system for mobile-access or for Internet services. Attracting private investment requires a stable environment and credible assurances against arbitrary takings. This can be done through contracts, but generally some form of regulation is also required (Levy & Spiller, 1994; Wellenius et al., 1993). In addition, there is the natural pressure to reform the incumbent, the recipient of much public investment, the source of revenues for both government and politicians, and also a major source of irritation in a government’s relations with entrepreneurs and the middle class. Therefore, institutional reform in telecommunications usually takes three forms: organizational reform of the incumbent operator, the introduction of competition, and the establishment of regulation. The three components are irreducible to each other, but are intimately connected. It is technically possible to engage in institutional reform that does not include all three elements, but generally all three are found in some form in current reform processes (International Telecommunication Union, 1997; Wellenius et al. , 1993)

At different stages of the Sri Lankan process, different elements were emphasized. Sri Lanka, which failed to achieve higher rates of network growth by reliance on the incumbent operator in the 1980s and the early 1990s, adopted competition as a central plank of its policy in the middle 1990s (Sri Lanka, 1994). This article seeks to assist in identifying the relative weights that must be assigned to the three elements of institutional reform in the current environment. 

Organizational reform of the incumbent operator 

In 1980, shortly after the closed economy was opened up to export-oriented activities and greater trade, the government split telecommunications operations from the postal service, while keeping them in the same ministry. Subsequently, international and domestic telecommunications operations were merged. Major investments were made in telecommunications through aid and government-guaranteed credits. The government department supplying telecommunications services was converted to a government-owned corporation in 1991, following an abortive attempt to privatize it in 1988-89. In order to accelerate network rollout and break from the old departmental culture, a fully owned subsidiary company was created in 1993 to supervise private contractors installing new lines. The government initiated, but did not complete, build-operate-transfer contracts. Both actions were interpreted by some within the incumbent as threats of competition. However, growth was slow (the compound annual growth rate (CAGR) was 14.68 percent for 1992-95) and little progress was made in reducing the waiting list and in improving service to consumers. 

The introduction of direct competition in 1996 saw a qualitative increase in the rate of network rollout. The number of new connections went up in 1996 itself in response to the introduction of credible competition even though the new operators were only able to install a miniscule number of lines that year because of litigation. The incumbent was converted to a fully government owned company in 1996, preparatory to partial privatization. In 1997 August, the government divested itself of 38.5 percent of the shares of the operator and signed a management agreement with NTT Corporation, the government-controlled incumbent operator in Japan. NTT had purchased 35 percent of the shares for US$225 million in the largest privatization transaction in Sri Lanka. The employees of the company were given 3.5 percent of the shares. In 1999, the government initiated action to divest itself of an additional tranche of shares, which is expected to reduce the government holding below 50 percent (Reuters Staff, 2000). Currently, the incumbent is managed by NTT, subject to limited oversight by a Board of Directors, which has a majority of government appointees. The internal organization of the company has been radically changed. The CAGR for 1995-99 was 31.97, for the incumbent alone. Chronic under-utilization of capacity has been eliminated. Serious efforts are being made to improve customer relations, especially through Teleshops staffed by newly hired and trained employees. Opportunities for petty corruption that were endemic in the old order have been reduced, wherever competition exists.

It was not expected that the center-left coalition that came to power in 1994 would accelerate the privatization process. Indeed, there were fears that the new government would halt, or even roll back, the liberalization policies initiated in 1977. However, the shortage of resources caused by the war, the need to reassure investors of the government’s commitment to liberalization, and the shift in economic thinking since the center-left coalition was last in power created the conditions for an aggressive privatization process led by an agency under the Ministry of Finance. Better relations with unions enabled implementation of privatization decisions. A young, active, and committed Minister assisted by an able and intelligent secretary created the conditions for the privatization agency to complete the incumbent’s privatization faster than anyone had expected.

Introduction of competition

Starting with the grant of the first mobile-access license in 1989 under the Telecommunications Ordinance, competition was gradually introduced on the edges of the system. The Sri Lanka Telecommunications Act, No. 25 of 1991 (Sri Lanka, 1991), which made provision for the issuance of licenses for the operation of telecommunications systems was enacted and two licenses for the provision of data and other communications services as well as the license of the incumbent operator (at that point, a government corporation) were issued in 1991. Additional mobile-access licenses were issued subsequently and by 1995 four mobile-access operators, using three different standards, were engaged in vigorous competition. Eight operators, including the incumbent, are licensed to provide data communication services using their own facilities. Additional operators continue to be licensed to provide services (including Internet), on the basis of leased facilities. Five paging operators, a trunk-radio operator, a specialized infrastructure provider and two payphone operators have also been licensed.

The first phase of introducing competition was marked by caution. Competition was introduced at the margins of the sector, even though no formal guarantees of monopoly were given for any service. The caution was a result of the traumatic experience in 1988-89, when the government was forced to retreat from a long-planned partial privatization involving a foreign strategic partner in the face of violent opposition by unions at a time of major civil unrest. It must also be seen in the context of the prevalent thinking on competition, which was quite cautious at that time. Nevertheless, the foundation of a competitive market was laid during this cautious first phase with the grant of around twenty system licenses. While various policy pronouncements were made (and changed) about the number of licenses that would be issued, no formal market-position guarantees were given to any operator during the first phase (Samarajiva, 1997). 

A qualitatively higher stage of competition was introduced in 1996 with the licensing of two fixed-access operators using wireless local loop (WLL) technology to compete directly with the incumbent. In contrast to the previous licenses that were issued on an ad hoc basis, the two fixed-access licenses were granted on the basis of open solicitation and bids. However, this also marked the first time that formal assurances of market position, a duopoly, were included in licenses and related commitments. The new entrants were promised a continuation of the duopoly and exemption from price regulation, subject to specified good-performance targets being met. However, a license provision that allowed the incumbent to use a specified range of frequencies for WLL services somewhat diluted the duopoly. 

In 1997 August, the government of Sri Lanka entered into several agreements with the NTT Corporation that transferred partial ownership and management of the incumbent to NTT. These agreements, and the consequent license amendments, committed the government, inter alia, to

  • issue no further wireline licenses until 5th August 2002, 
  • issue "no other license . . . for the provision of international telephonic services in and from Sri Lanka prior to 5th August 2002," and 
  • approve adjustments to rentals, call charges, and connection charges in a manner that would yield a minimum 148 percent increase in domestic revenue (not adjusted for inflation) over a five-year period. 
The second of the above commitments has given rise to considerable friction as a result of the incumbent and its strategic investor interpreting it to constitute an all-encompassing monopoly over international communication services. Several other licensees, who claim that licenses predating the privatization allow them to provide forms of international voice services, have challenged this view. The incumbent was denied an injunction to prevent one operator offering such services in 1999 and the matter is under litigation in a series of cases. The third commitment has formed the basis of the rate rebalancing process undertaken since 1997. The second and third commitments, taken together, modified the commitment made by the government of Sri Lanka in relation to the Fourth Protocol of the General Agreement on Trade in Services to end the monopoly on international services subject to the satisfactory completion of rate rebalancing by 1999 (Samarajiva, 1999; WTO, 1997).

The decision by the People’s Alliance government to issue two fixed-access licenses shortly before privatizing the incumbent was courageous. Despite the fact that prevailing wisdom on competition had become less cautious, licensing two operators to compete directly with the incumbent in all services except international evidences uncommon leadership by the "line" ministry and a victory of sector objectives over the Finance Ministry’s desire to maximize privatization yields. The clarity of the policy solution was, however, sullied by a last-minute compromise that postponed the liberalization of the international market by 32 months. 

Establishment of regulation

The government separated regulatory functions from operations in 1991. The 1991 Act created a single-person authority, modeled on Britain’s OFTEL, headed by a Director General of Telecommunications to perform the regulatory functions and advise the Minister on policy matters. Being constituted as a government department, the Office of the Director General of Telecommunications (ODGT) lacked the resources to recruit qualified personnel. As a result, many provisions of the 1991 Act lay dormant and most conditions of system licenses were not enforced. However, major progress was made in the issue of licenses, formulation of proposals for amendment of the enabling legislation, and the preparation of the infrastructure for an effective regulatory agency. Key individuals associated with the formation of the ODGT reported that giving it greater autonomy and financial resources was seriously considered but that forces at the Ministry and Cabinet levels opposed it. This may have been partially caused by fear of the unknown (independent regulation being a novel and foreign concept at that time), a desire to maintain Ministry control over the new agency, and bureaucratic jealousy about the prospect of higher salaries for ODGT staff. 

By the amending legislation of 1996, the regulatory authority was strengthened by the creation of a Commission with the resources to build up the necessary regulatory expertise and in-built checks and balances to assure a degree of independence. The powers and duties of the Director General of Telecommunications under the 1991 Act were transferred to the Commission by the 1996 amendment. The Commission was also given the powers to collect license fees and to utilize them for the performance of its functions and to generally organize itself in a business-like manner, subject to government financial and administrative regulations. Three part-time members with security of tenure and two ex-officio members constitute the Commission. The Chairman is the Secretary of the Ministry in charge of the subject of telecommunications and represents the interests of government. The Director General of Telecommunications who is designated as the Chief Executive Officer of the Commission serves as the other ex-officio member. The Act specifies that the Minister shall appoint qualified persons who have distinguished themselves in the fields of law, finance and management to serve as part-time members. The coincidence of the strengthening of the regulatory agency with the licensing of the fixed-access competitors suggests that the primary purpose was to give a greater assurance of stability and fair play to those investing in the sector. A higher level of familiarity with regulation after five years of experience with the ODGT may have also contributed, as would awareness of prevalent thinking on the value of independent regulation.

Also at this time, the government chose to commit to the Regulatory Reference Paper that is part of the Fourth Protocol to the General Agreement on Trade In Services of the WTO (World Trade Organization, 1997). Sri Lanka is the only South Asian country to make this commitment. The inclusion of exceptions has postponed the creation of "single-market" conditions for a few years, but the general effect of the six principles of the Reference Paper (anti-competitive practices, interconnection, universal service, transparency of licensing, independent regulation and fair allocation of scarce resources) has already been felt in the Commission’s work (Samarajiva, 1999) and may prove pivotal in the various legal actions that are underway. 

In 1998-99, the Commission recruited new personnel and launched a series of initiatives including the fixed and mobile interconnection proceedings, the first public hearing on the improvement of billing, the first public-notice proceeding that yielded an innovative policy on GMPCS [Global Mobile Personal Communication by Satellite] licensing, the first successful order against an operator for the violation of license conditions, and a number of competition-related initiatives including the authorization of direct satellite access for data operators and reporting requirements for payphone installation to prevent unfair competition by providers of loops. The Commission also launched major initiatives to improve emergency telecommunications and Y2K preparedness. These actions were facilitated by a credit from the World Bank, approved in 1996, and a partnership with the Government of Canada, established in 1998. Expenditures on consultancy and training under the World Bank credit were increased ten-fold in 1998, reflecting the higher absorptive capacity of the Commission and the activation of the Canadian partnership. 

The enforcement powers of the Commission are being defined in legal proceedings initiated by the incumbent who objected to the fixed-access interconnection determination. Significant, but far from perfect, forward movement can be seen in the areas of frequency management and numbering administration. In 1999, the Commission finalized the procurement of a frequency monitoring and management system that will hopefully be the basis of efficient management of the spectrum, a foundational element of a competitive telecommunications sector. A closed numbering plan was adopted but has been stalled by the incumbent. In its policy-advisory role, the Commission has generated several innovative and competition-compatible proposals to promote access to the network from rural areas. A Commission-proposed scheme to provide targeted subsidies for rural payphones using license-fee revenues is being implemented.


Sri Lanka’s institutional reforms have yielded relatively good results on the dimensions of connectivity, quality of service, affordability and innovation. Connectivity, particularly in the form of teledensity, is the easiest to measure. It is therefore given the most weight in assessing telecommunications sector performance. However, sustainable growth with the desired effects on the economy as a whole and on living standards requires progress on all four dimensions. Multifaceted improvement of performance is critical for the legitimacy essential for the sustenance of the institutional reform process. 


Sri Lanka shows an exceptional growth rate in fixed-access connectivity over 1995-98, compared to ITU averages for low, lower-middle, upper-middle and high income countries (ITU, 1999b). While higher than the fixed rate, the mobile growth rate was lower than that of Sri Lanka’s peers. Fixed-access teledensity increased from 0.73 in 1991, at the start of the serious reforms, to 3.55 by end of 1999. Mobile teledensity increased from 0.01 in 1991 to 1.35 by end of 1999. The combined fixed-mobile teledensity was 4.9 by end 1999 and the fixed/mobile ratio was 2.63. Most of the connections (fixed and mobile) are still in urban areas, but growth is accelerating in the rural areas. Payphones per 1000 population grew from 0.03 in 1992 to 0.3 in 1999. Despite this ten-fold increase, Sri Lanka still lags behind its peers in availability of payphones. Non-incumbent operators have built stand-alone data networks that hold the promise of rapid growth with new technologies and innovative regulation.

Quality of service

Adequate progress has not yet been made in this regard, but operators are moving in the right direction and the Telecommunications Regulatory Commission has signaled its strong commitment to improvement of quality of service. A multi-million rupee decision to compensate customers for delayed connections was issued and implemented in 1999. Measures to compensate subscribers for delayed fault repairs that create incentives for improved performance by the operator have been in place since May 1998. The incumbent operator completely revamped its fault-recording procedures in 1998 and is expected to provide itemized billing in 2000 as a result of a Public Hearing. Quality of Service rules have been developed following a process of consultation that included the country's first sample survey of fixed-access subscribers. The networks of most operators continue to be plagued by low call-completion, partly because of high rates of growth and partly because of periodic constrictions of interconnection by the incumbent. The Commission has sought to alleviate the former problem through encouragement of off-peak use and the latter through the establishment of stable interconnection arrangements, including creation of multiple interconnection points. Complaint rates increased dramatically in 1998-99 (the average of complaints received in March-May 1999 was 287 per month), but this is partly an indication of the higher profile and credibility of the Commission and higher expectations of the privatized incumbent. 


The affordability of fixed-access telecommunications services is affected by a tariff-rebalancing process intended to increase domestic revenues of the incumbent operator by a minimum of 148 percent over five years. Within the constraints of the rebalancing process, the Regulatory Commission created low- and medium-user tariffs to ease the transition for a majority of users in 1998 and 1999 (Telecommunications Regulatory Commission, 1999a). The lowest one-third of users paid less than US$6 per month for monthly rental and call charges (local and national) and the middle one-third of users paid between $6 and $15 in monthly domestic telephone charges in 1998-2000. The 2000 tariff decision continued the policy of moderating the impact on low- and medium-users while giving the incumbent a domestic revenue increase of 20 percent (Anon., 2000). 

Connection charges remain at high levels (around $200 for Greater Colombo and between $200 and $1000 for the rest of the country). However, the Commission introduced an installment plan for paying these charges and prevented their collection without prompt connection. In addition, the increased supply of connections has reduced the opportunities for bribery, resulting in a de facto reduction of connection charges. The incumbent is reducing international rates by 8-10 percent every year. A significant amount of residential traffic has been moved to off-peak periods using economic incentives and advertising. Special off-peak rates for Internet users (as low as 1/18th of the peak rate for users outside Colombo) have been introduced. Intense competition among mobile-access operators has resulted in low mobile rates. The ITU (1999b) reports that a 100-minute basket cost US$ 17.80 in Sri Lanka, well below the averages for low-income countries ($ 35.99), lower-middle income countries ($ 39.69) and the world ($ 38.15). With the proliferation of home-zone pricing for mobile calls and the fixed-access rate rebalancing, it is likely that mobile rates will become competitive with fixed rates within the next few years.


Telecommunications operators are introducing new services and new service configurations at a rapid pace. Fixed-access operators have introduced ISDN services. Sri Lanka was early to introduce pre-paid mobile services, which have come down to the same levels as subscription rates under competitive pressures. Pre-paid fixed-access services are also available. Similarly, competition has caused consumer-friendly practices such as per-second billing (as opposed to billing in increments of minutes or longer) to be introduced without regulatory intervention. The mobile operators have introduced many hybrid Internet-mobile services, including Wireless Access Protocol (WAP) services. The data communication operators offer services on par with their counterparts abroad, subject to bandwidth and related limitations. The telecommunications sector was the largest advertiser in the country in 1998.

Key Policy Choices Competition at the margins or at the center?

Until 1996, Sri Lankan policies with regard to the introduction of competition were cautious. Competition was introduced only in services portrayed as marginal to the interests of the incumbent operator, moderating the reactions of the vociferous telecommunications trade unions and allowing the new operators to get established. However, the sector may have suffered as a result of the continuation of the inefficient administrative culture of the incumbent, including slow network rollout, inefficient utilization of capacity, corruption, poor quality of service and lack of innovation, as well as the stunting of the growth of mobile telephony and data communications by the arguably anti-competitive practices of the unreformed incumbent. 

The significant improvement of network rollout performance of the incumbent starting from 1996 illustrates the shortcomings of the cautious approach. In 1996, well before the entry of the strategic investor, the incumbent doubled the number of new connections per year. The incumbent's compound annual growth rate for the four years after 1995 was 31.97 percent, more than double the 14.68 percent rate for the preceding four years. What galvanized the incumbent was the credible and serious threat of competition, more than the actions of the competitors themselves, because the competitors installed only 527 lines in 1996. Where previous efforts to awaken the incumbent from its stupor by the creation of a company to manage the network rollout projects (interpreted by some union leaders as a stalking horse for privatization) and attempts to introduce Build-Operate-Transfer operators failed, the licensing of direct competitors succeeded. 

Competition at the margins can be justified in some cases by the logic of incubation: firms can be allowed to establish themselves at the margins and then move into the center for the big fight. However, no niche players have developed into major competitors in Sri Lanka yet. Ongoing litigation initiated by the incumbent against two operators, who were previously active only in data, seeking to prevent the termination of international telecommunications traffic in Sri Lanka may perhaps, unintentionally, give rise to that outcome.

Despite respectable growth, it is possible to argue that the mobile-access and data sectors would be in healthier condition today had their formative period been marked by a greater degree of competition at the center of the system. It is possible to speculate that without the highly restrictive interconnection arrangements imposed by the incumbent on the mobile operators and the problems experienced by data operators regarding access to leased lines and international capacity, both sectors would today show better performance. But because effective regulatory action could also have reined in the incumbent, blame cannot be laid solely at the door of the policy of introducing competition at the margins.

Of course, the very formulation of the question can be challenged. Technological change makes fluid the definitions of what is center and what is margin. For example, there is a possibility that circuit-switched networks that constituted the "center" in 1991-96 may lose their centrality because of rapid growth in IP telephony. In such an event, the so-called data operators who were licensed to provide services at the margin could move to the center and provide real competition to the incumbent. 

Privatization before competition or vice versa?

In the original planning for institutional reform, the Sri Lankan government intended to privatize the incumbent operator at the same time as the new Act creating licensing procedures and a regulatory authority came into effect (Samarajiva, 1997). Many countries such as Mexico, Venezuela, and even South Africa have given primacy to organizational reform of the incumbent through some form of privatization over sector-wide liberalization, including the introduction of competition and strong regulation. Because of union opposition, the Sri Lankan government was forced to give primacy to liberalization and regulation, over the reform of the incumbent, which though unintended, appears to have had beneficial effects. 

The Peoples’ Alliance government that assumed power in late 1994 had a clear choice between the two options and chose to introduce direct competition and strengthen the regulatory agency prior to the reform of the incumbent through partial privatization and management by strategic investor. This was especially noteworthy because the privatization managers in the Ministry of Finance may have felt that the introduction of competition would reduce the value of the shares of incumbent. The difficulties experienced by the Commission in gaining compliance from the rejuvenated incumbent backed up by a powerful strategic investor with a monopoly culture suggest that it would have been exceedingly difficult to introduce competition and strengthen regulation after 1997. Sri Lanka’s lesson is that competition should precede privatization. 

Incumbent or competitors as policy instrument?

Effective institutional reform must address the question of whether telecommunications sector performance objectives (e.g., rapid network rollout) are better achieved through multiple competitive providers or though a reformed incumbent. The experience in most countries, including Sri Lanka (emphatically), supports the competition-based strategy. China and South Korea exemplify success in achieving rapid network rollout through monopoly incumbents, though closer examination may show shortcomings on the other dimensions of quality of service, affordability, and innovation.

Another related question is whether competitors are seen as instruments for improving incumbent performance or whether they are seen as direct instruments for the achievement of sector performance objectives. One interpretation of the UK experience with competition is that the financial interests in the City of London promoted the original competitor, Mercury, primarily in order to reform the incumbent, which may explain Mercury's decline after that outcome was achieved. The fate of the competitors who rejuvenated the Sri Lankan incumbent is still not decided, but the experience so far shows that competitors have made decisive contributions to the achievement of sector objectives directly, as well as indirectly through the reform of the incumbent.

Telecommunications sector objectives vs. yield maximization from privatization/ownership

Not so much a policy choice, this is a perennial tension that bedevils the institutional-reform process in every country. The pre-privatization actions of the Mexican government in selling a countrywide network that could have provided a platform for effective competition to the incumbent (Pérez-Chavolla & Samarajiva, 1997) is an example of this widespread phenomenon.

In the case of Sri Lanka, the balance was struck at different points at different times. In 1996, when the fixed-access operators were licensed, the balance favored telecommunication-sector objectives. Less than one year later, in the last stages of the privatization negotiations, the balance tilted toward the Finance Ministry’s objective of maximizing privatization revenue with the inclusion of the "no other licenses" provision that modified the WTO commitments of the government. The balance may move even further toward Finance-Ministry objectives in 2000 in the context of the government’s heavy reliance on the receipts of the sale of additional government shares of the incumbent (Reuters Staff, 2000).

It is natural for a privatization agency to seek to maximize privatization yields. The Telecommunications Ministry may stand a slightly better chance if its opponent is the Finance Ministry, and not the privatization agency, and if it succeeds in framing the choice as broader than privatization yields, but inclusive of foreign investment inflows, license fees, and taxation revenues from higher levels of output caused by competition. For example, the US$ 225 million realized by the 1997 privatization of the Sri Lankan incumbent was less than the total foreign investment flows attracted by the telecommunication sector. The approximately 18 percent tax levied on all telecommunications bills yielded a major proportion of the government’s sales-related taxation revenues, a proportion driven up by the higher levels of activity generated by vigorous competition. Yet the tight integration of the Finance Ministry and the privatization agency resulted in the privatization yield being given undue prominence. The appropriate policy response in this case appears to be an effective campaign to persuade senior decision-makers of the net revenue and other benefits of a competitive telecommunications sector. 

Is universal access/service compatible with competition? 

It is common for incumbents to bemoan the negative effects of competition on service provision to rural, high-cost, or low-revenue areas and to special groups (Melody, 1998). These claims must be measured against the actual performance of these companies in extending service to such areas and groups in the monopoly years. The Sri Lankan incumbent had more switching capacity and lines in service in Metropolitan Colombo than in the entirety of the rest of the country at the end of 1998, even after two years of competition, which suggests a negligible commitment to rural connectivity.

Competition since 1995-96 has done more to improve connectivity in Sri Lanka than a century of so-called public service. Twice as many connections have been provided in the past five years than since the introduction of telephony to Sri Lanka in the 19th Century. Persons outside the networks of power and affluence are no longer shut out from the possibilities of obtaining services. Unlike in the past, service can, in most cases, be obtained without bribing the incumbent’s employees. In the face of satisfied demand in metropolitan Colombo, the incumbent and the fixed-access competitors are rolling out the network in rural areas that were hitherto considered unprofitable. Mobile-access operators who built out their networks to accommodate the communication needs of their vacationing Colombo customers have started to see the potential of rural dwellers who live within the extended coverage areas and are offering them lower home-zone rates. Prepaid mobile service that was introduced to reduce bad debts has opened up access to urban dwellers without fixed addresses. Serious efforts are being made by non-incumbent operators, including the establishment of a prototype VSAT connected Internet business center by one non-incumbent operator, to address the lack of Internet connectivity in areas outside Colombo. 

However, policymakers and regulators cannot rest satisfied with the substantial progress on universal access yielded by competition. What the extension of the market by competitive forces does is to allow policymakers and regulators to focus their energies on providing connectivity to specific populations that are for various reasons ignored by market forces. Most regulatory mandates include promoting connectivity for those who may be ignored by the market, such as rural citizens and the disabled, and it is both a legal requirement and a strategic necessity for regulatory agencies to implement universal access harnessing the forces of competition. For optimal effect, it is necessary to mobilize competition in the design of universal service policy (Wellenius, 2000). For countries such as Sri Lanka which have made commitments to the WTO Regulatory Reference Paper it is also necessary to ensure that the universal service mechanisms are competition-neutral (Samarajiva, 1999).

For many years the Sri Lankan government had allocated "universal-service" funds from the general treasury to connect remote post offices to the network. Finding the practice of obtaining these connections solely from the incumbent to be inefficient to say the least (connection charges per post office were around $4000 in 1999), the Commission recommended that the government allow all fixed- and mobile-access operators to bid for the provision of this service. The Ministry has accepted the recommendation but appears to have difficulty in implementing it. 

Finding the payphone density in the country to be significantly below those of its peer economies (0.2 per 1000 inhabitants versus 1.57 per 1000 inhabitants for lower-middle income countries in 1998), the Commission proposed a subsidy scheme to encourage installation of rural payphones, to be funded by license-fee revenues. Each operator is entitled to a subsidy of approximately $750 per payphone (a portion of the installation costs) for up to 25 new installations per district. The scheme will sunset when 100 payphones are installed in each of Sri Lanka’s 25 districts (International Telecommunication Union, 1999a, p. 67; Telecommunications Regulatory Commission, 1999b). The first disbursements were made in late 1999. 

Managed competition in technology-defined subsectors vs. technology-neutral competition

The Sri Lankan strategy has been to introduce managed competition in technology-defined subsectors. The national telecommunications policy and the licenses show evidence of efforts to define subsectors such asfixed- access, mobile-access and data and to license limited numbers of operators in each subsector. This strategy appears to have been driven in part by the government’s desire to prevent the leveraging of market power across markets and in part by investors’ desires for certainty and stability. In higher-risk environments such as Sri Lanka, it is natural that weight is given to investor demands.

Sri Lanka has done relatively well with its policy of technology-defined managed competition, but it has created problems that are becoming increasingly difficult to manage with the advance of convergence. When the first licenses under the 1991 Act were issued, the distinction between circuit-switched voice telephony and data communication may have been relatively clear. This distinction may have been assumed as late as August 1997 when NTT and the government of Sri Lanka included language in their agreements that has been interpreted by the incumbent as a broad form of exclusivity for international telephone services for five years. However, rapid technological and market changes in international telecommunications in the past few years have erased the distinction between data and voice for all practical purposes. The legality of the termination of international voice traffic in Sri Lanka by operators licensed to provide enhanced voice services is currently under adjudication.

In common with many countries, Sri Lanka’s initial mobile-telephony policies assumed a small niche market of wealthy individuals. However, as competition drove down prices and speeded up rollout after the entry of the fourth operator in 1995, the assumption no longer holds. The fixed to mobile ratio at the end of 1999 was 2.63, the lowest it has been since mobile telephony was introduced. Individuals and enterprises in remote locations have connected cellular phones to antenna and use them as fixed-access phones. Innovative tariffs that give lower rates to customers who stay within their "home zones" have potential of making mobile rates comparable to fixed-access, especially because of rate-rebalancing on the fixed side. It is evident that mobile is increasingly becoming a substitute for fixed services, rather than a separate service. 

The national telecommunications policy and provisions in licenses and tariffs require separation of mobile and fixed access. For example, the fixed-access WLL licenses require affixation of instruments. Neither the Commission nor the government has formally opened up the question of removing the fixed-mobile distinction, but the Commission’s 1999 fixed-mobile interconnection ruling suggests that cellular telephony is no longer perceived as a luxury (Telecommunications Regulatory Commission, 1999c). However, if the government were to consider allowing greater interpenetration between fixed and mobile markets, the limitations written into the fixed-access WLL licenses could pose problems. 

In contrast with many Asian countries, Sri Lanka has steered clear of specifying technologies or standards in licenses. For example, cellular operators were not mandated to follow any particular standard and in fact use three different standards (AMPS, E-TACS and GSM; now in transition to more advanced standards) in a relatively small market that has just reached 250,000; the data licensees were not limited to specific technologies such as VSATs but were free to use whatever technologies they considered appropriate. While the overall results have been satisfactory, the question now is whether the specification of the service itself will prove to be too constrictive.

The remedy to the problem is to give greater discretionary authority to the regulator to manage convergence. Of course, discretion cannot be given without accountability and cannot be sustained without legitimacy (Samarajiva, 2000). And unless the regulatory agency is perceived to be truly independent and fair, it is unlikely that investors will cease to seek the certainty of detailed license commitments. 

Ongoing regulation for competition

In network industries such as telecommunications, it is not enough simply to introduce competition through licensing. Even at the point of introducing competition a number of related actions are necessary, especially with regard to interconnection and access to scarce resources such as frequencies, numbers and rights of way. The Sri Lankan government outlined some broad principles regarding interconnection in the course of licensing the fixed-access WLL operators but left the actual agreements to negotiation by the parties, failing which the Regulatory Commission was empowered to intervene. No guidance at all was provided when the mobile-access licenses were issued. This was in contrast with what was done in Australia, where the basic terms and conditions of interconnection were specified prior to the entry of the competitor.

In the case of the mobile-access operators, the incumbent imposed what appear in hindsight to be patently anti-competitive terms and conditions: mobile operators had to pay the higher "national-rate" calling charges for all calls terminated on the incumbent’s network, a position less favorable than a retail customer's; they had to pay the full retail charges for international calls originating in their networks; they were allowed to interconnect only in one location and had to pay for the entirety of the costs of interconnection including for use of the incumbent’s switch; and the incumbent paid nothing for terminating calls on the mobile networks. These arrangements were changed only in 1999 by the Commission’s fixed-mobile interconnection and related decisions (Telecommunications Regulatory Commission, 1999c).

The Commission attempted to incorporate the broad principles of interconnection set out in the WLL licensing documents in the fixed-access interconnection proceeding of 1998. However, the language allowed divergent interpretations by the incumbent and the competitors. Despite painstaking efforts to bring the parties to agreement through alternative dispute resolution and a subsequent formal proceeding, the disagreement persisted and resulted in the appeal of the Commission's determination to the courts in 1999. The conflict has involved several crises caused by disruption of technical interconnection by the incumbent and/or parties within the incumbent organization. 

The question then is whether Sri Lanka should have followed the Australian policy of setting out detailed interconnection rules prior to licensing without leaving it to subsequent negotiation and/or determination. The difficulties experienced in arriving at mutually acceptable interconnection decisions indicate that, at minimum, it is necessary to reduce the ambiguity of the language used in legal documents. 

The danger of preparing interconnection rules before competitors enter the scene is that the rules are likely to favor the incumbent without the active participation of other stakeholders. Even if such a result is not intended, the incumbent being the only stakeholder with the opportunity to influence the process is likely to yield such an outcome. Some aspects of interconnection agreements require detailed knowledge of network configurations and other details that would be simply unknown prior to the competitors commencing operations. Therefore the remedy is to lay out clear broad guidelines consistent with international best practices and require the use alternative dispute resolution and other techniques to bring the parties to consensus. Absent consensus, interconnection terms consistent with international best practice should be imposed on the parties. 

In addition to interconnection, which has been described as the cornerstone of competition, fair and transparent processes are necessary for the allocation of scarce resources such as frequencies, numbers and rights of way. These actions require the continuing intervention of a strong and pro-competitive regulatory agency. The Sri Lanka regulatory commission has made progress in each of these areas, but it is possible that the efficacy and timeliness of that progress may be questioned in light of Sri Lanka's WTO commitments.


Sri Lanka’s decade-long experience with telecommunications reform yields some lessons on the optimal design of effective institutional reform. It also enables the clarification of policy choices that must be made in context-specific ways within the overall design. 

The three components of institutional reform – organizational reform of the incumbent including corporatization and privatization, the introduction of competition and the establishment of regulation–are widely recognized and applied (e.g., International Telecommunication Union, 1997; Wellenius et al.,1993). What is lacking is consensus on the relative weight to be given to each component and the sequencing. While a certain degree of context-specific variation must be allowed, the Sri Lankan experience shows that competition not only yields good sector performance but, perhaps even more importantly, energizes organizational reform of the incumbent and contributes to consolidating and legitimating the regulatory process. Without the external impetus provided by competition, internal reform of incumbents and efforts to create modern regulatory agencies are likely to succumb to the inertial forces that have held back telecommunication developments in the first place. 

In low-teledensity countries such as Sri Lanka, attracting investment to build out the networks and making optimal use of that investment must be the primary objectives. With most concessionary lenders, including the World Bank, ceasing to lend to inefficient government monopolies, there is little choice but to allow private firms to enter the market. Competition is also the best choice under conditions of rapid change and high uncertainty. No single private firm, let alone government or a government-owned operator, can predict the volatile and complex technological and market developments of telecommunications in our time. The meteoric rise of Cisco as a telecommunications equipment supplier based on IP expertise (perhaps to be overshadowed soon by the rise of optical switching) illustrates the unpredictability of the environment. In such a setting the only effective strategy is to rely on the discovery processes embodied in the aggregation of the actions of competitive operators. 

But competition need not be seen as an unavoidable imposition. In addition to its most obvious benefits of attracting investment to a capital-hungry sector and allowing for flexible responses to rapid change, competition reinforces and accelerates organizational reform of the incumbent. Sri Lanka shows clearly that external pressures from competition are more effective in reducing inefficiency, waste and corruption, than internal efforts alone. The Sri Lankan incumbent’s pre- and post-competition average connectivity increases of 14.68 and 31.97 testify to the power of competition.

Regulation in low-teledensity countries like Sri Lanka is, at core, a safeguard against arbitrary expropriation of private capital invested in long-term projects. In addition, regulatory agencies are given mandates that include implementation of government policies of universal access, consumer protection, etc., and the government’s international commitments to international organizations such as the ITU (frequencies and standards) and the WTO. 

Irrespective of the legislative mandate, fledgling regulatory agencies must proactively take on high-profile public-interest tasks such as improving connectivity in rural areas and protecting consumers from operator abuse. If not, they will lack the legitimacy to perform any of their functions, including those of ensuring the continued inflow of investment and the prevention of disinvestment. The perception of an emphasis on pure competition-enhancing measures by the Telecommunications Regulatory Authority of India (TRAI), as against those with more popular appeal such as universal access and consumer protection, may have contributed to its loss of legitimacy, first in the eyes of the media, and then in the eyes of the government. The TRAI also suffered from the unfortunate integration of operational and policy functions in the incumbent Department of Telecommunications and the undeveloped state of competition in India at that point. The alignment of the policy and operational agencies created a powerful force that overwhelmed a regulatory agency that was buttressed solely by expertise and statute, and lacked broad legitimacy and the strategic support of competitive entrants. 

It is one thing to expect regulatory agencies to do the right thing; it is quite another to create the conditions for the right thing to be done. The expectation is that regulatory agencies will exercise discretion in ways that are fair, based on the best expertise, and in the public interest. But many regulatory agencies are staffed either by former employees of the privatized incumbent, or by government officials on secondment, or by raw new recruits. In most cases, the staff are, understandably, not very knowledgeable about regulation. In almost all cases, there is a lacuna of understanding about the basic principles of competition and natural justice (or due process). Salaries are generally inadequate, creating the necessary (but not sufficient) conditions for corruption. Training opportunities exist, but, in many cases, are not optimally utilized because many decision-makers lack the ability to identify what is needed, or lack the self-confidence to sift the wheat from the chaff of foreign-training offers, or because the subsidiary role of foreign training as supplementary compensation is given too much weight. 

Many things can and should be done to improve the functioning of regulatory agencies, including the introduction of modern management techniques, effective development of human resources, improvement of the transparency of regulatory activities, the establishment of practices unconducive to corruption, the inculcation of ethics and natural-justice principles and effective communication with stakeholders (Samarajiva, 2000). Yet experience shows that articulate and assertive participation by competitors in the regulatory process, in some cases even taking the form of a well-placed court case, can be as effective as these internal efforts. Such external discipline emerges from the competitive environment, especially from competitors who stand to lose from the predatory actions of the incumbent or from the arbitrary decisions of a weak regulator. Competition provides a powerful reinforcement to the consolidation of effective regulation. 

Competition is developed gradually, but it is difficult to roll back. Not only that, it prevents other institutional reforms being rolled back too. The existence of viable operators who have something to lose from the re-imposition of monopoly practices is the best guarantee of consolidating institutional reform. While a sector-specific regulatory agency does provide a locus for the coordination of institutional reforms, it is too much to ask that a fledgling agency with underpaid and mostly under-skilled staff with an ambiguous relationship to government take on the full responsibility of defending reform against the many beneficiaries of the status quo ante. For example, the greatest hope for the Sri Lankan institutional reforms in 2000 lies in the actions of competitors in the courts and elsewhere, not in the Regulatory Commission which is under intense pressure from a Finance Ministry intent on maximizing privatization yields. As numerous studies of regulatory agencies in the United States and Canada (the countries with the longest experience of independent regulation) have shown, rarely do these agencies, however well designed and well staffed, serve the public interest consistently (Melody, forthcoming). Good people within well-designed regulatory agencies can perform best when external forces engendered by competition exist. It is not that such people cannot achieve results on their own, but that such outcomes are exceptions.

Independent, proactive regulatory agencies that use competition as their primary instrument are vital for the success of institutional reforms that will not only yield better results in terms of connectivity, quality of service, affordability and innovation, but will enable the people of developing countries to participate in the information society. Vibrant competition provides the best environment for the success of such agencies, which in turn keep that environment vibrant. Of the three components of institutional reform, regulation in the form of an independent sector-specific or multi-sector regulatory agency provides the necessary cohesion; competition reinforces the other components and provides the essential momentum. 


1. The ITU does not provide a definition of an independent regulatory agency.  The broadest definition is found in the WTO Regulatory Reference Paper (World Trade 0rganization, 1997):  “The regulatory body is separate from, and not accountable to, any supplier of basic telecommunications services.”  Another definition hinges on insulation from direct interference by the government, especially its executive branch.  The actual number of independent agencies will be considerably smaller than the ITU figure if the latter definition is rigorously applied.

2. Fixed access is conventional voice telephony, provided over wire and wireless local loops.  It is distinguished from mobile access.  In light of the small geographical and financial size of the Sri Lankan market, no distinction is made between local and long-distance markets.  Each operator provides local and long-distance service in an integrated form.  The international telephony market is treated separately in policy, and may be seen as equivalent to the long-distance markets of larger countries. 

3. See news releases and other information at for more details.

4. Teledensity is the number of telephone main lines per 100 population.  The existence of a certain amount of fixed-mobile substitution is recognized by the separate calculation of fixed and mobile teledensities, along with a combined figure.  Except where otherwise stated, the Telecommunications Regulatory Commission is the source of data on sector performance.
5. See news releases and other information at for more details.
6. This does not exclude the possibility of discipline being imposed by incumbents.  Incumbents, especially in South Asia, have been quick to appeal to the courts against regulators exhibiting any signs of independence. However, emphasis is placed on competitors in this case because incumbents do not usually have an interest in strengthening regulatory institutions. 


Anon., (2000). "80 percent of subscribers sheltered." Sunday Times (Sri Lanka) June 4.

ENS Economic Bureau, (2000). "TRAI disbanded, stripped of judicial powers," The Indian Express, January 20. At

Gunaratne, S.A. & Wattegama, C. (2000), "Sri Lanka." In Gunaratne, S.A. (ed.). Handbook of the Media in Asia. Thousand Oaks: Sage.

International Telecommunication Union, (1997). Telecommunication Development Bureau, ITU-D Study Groups, Recommendations emerging from the study of Question 2/1, telecommunication policies and their repercussions at the level of institutional, regulatory and operational aspects of services. Document 1/193 (Rev.1)-E. September 23, Geneva: ITU.

International Telecommunication Union, (1999a). Trends in telecommunication reform: Convergence and regulation. Geneva: ITU.

International Telecommunication Union (1999b). World Telecommunication Development Report 1999: Mobile cellular. Geneva: ITU.

Jayasuriya, S. & Knight-John, M., (1998). "Liberalisation and privatisation: Telecommunications reforms in Sri Lanka." In S. Macdonald & G. Madden (eds.), Telecommunications and socio-economic development, pp. 353-74. Amsterdam: Elsevier.

Levy, B. & Spiller, P., (1994). "The institutional foundations of regulatory commitment: A comparative analysis of telecommunications regulation." Journal of Law, Economics & Organization, 10(2): 201-246.

Melody, W.H., (1997). "Policy objectives and models of regulation." In Telecom reform: Principles, policies and regulatory practices, pp. 13-28. Lyngby DK: Den Private Ingeniorfond.

Melody, W.H., (1998). "Universal service in an information society," Telecommunications and Space Journal, 5: 35-42.

Melody, W. H., (forthcoming). "Designing Utility Regulation for 21st Century Markets" Michigan State University Press, East Lansing.

Pérez-Chavolla, L. & Samarajiva, R. (1997). "Privatization, market liberalization and regulatory reform in the Mexican telecommunication system." In Telecommunication: Privatization and competition, ed. D. J. Ryan, pp. 149-65. Westport CT: Praeger.

Reuters staff (2000,). "Sri Lanka to offer over 20 percent of Telecom." February 17
Samarajiva, R., (1993). "Institutional reform in Sri Lanka's telecommunication system: Regulation, corporatization, and competition," Asian Journal of Communication, 3(1): 37-63.

Samarajiva, R., (1997). "Institutional reform of Sri Lankan telecommunications: The introduction of competition and regulation." In Telecommunication systems in Western Asia and the Middle East, ed. E. Noam, pp. 38-61. New York: Oxford University Press.

Samarajiva, R., (1999). "Sri Lanka's WTO implementation and universal access," Connect-World Asia, 1st Quarter: 11-14. At:

Samarajiva, R., (2000). "Establishing the legitimacy of new regulatory agencies," Telecommunications Policy, 24(3): 183-87. At

Shetty, V., (1999,). "As the wheel turns." Communications International, pp. 48-50. March 1

Sri Lanka, (1991). Sri Lanka Telecommunications Act, No. 25 of 1991, as amended by Sri Lanka Telecommunications (Amendment) Act, No. 27 of 1996.

Sri Lanka (1994). National policy on telecommunications. At

Telecommunications Regulatory Commission, (1999a). "Approved SLTL tariffs for 1999 released; No rate hike for low and moderate users: News release." April 8

Telecommunications Regulatory Commission (1999b). "Asia Pacific Telecommunity Workshop on Rural Telecommunications: News Release."

Telecommunications Regulatory Commission (1999c). "Telecom Commission releases mobile interconnect decision: Fairer interconnect terms now; public hearing to seek comments on calling-party-pays decision: News release." June 17

Wellenius, B., (2000). "Extending Telecommunications beyond the Market: Toward universal service in competitive environments," Public Policy for the Private Sector, World Bank, March.

Wellenius, B et al., (1993). Telecommunications: World Bank experience and strategy. World Bank Discussion Papers 192. Washington DC: World Bank.

World Trade Organization (1997). Fourth Protocol to the General Agreement on Trade in Services. Geneva: WTO. and