Black Rural	Black Urban	Coloured	Asian	White	Total
Population ('000)		12 095		1 136		37 859
Households ('000)		2 503	750	254		8 721
"universal service" % households with wireline telephone		29%		74%		32%
"universal access" % households without wireline phone, but with access to one within 5 km		57%		23%		43%
% households with no universal service or access		14%		3%		24%

The stark disparity in the allocation of resources between whites and blacks is illustrated by the fact that 55% of South Africa's telephones (1.5 million instruments) are in the hands of its whites, who constitute only 13% of the total population. ¹³¹ There are also wide disparities in network access from region to region: in the wealthy and predominantly urban provinces of the Western Cape and Gauteng, with their large white populations, access line penetration exceeded 20 lines per 100 people in 1996, but none of the remainder of South Africa's nine provinces had an access line penetration of more than 10 lines per 100 people; in one of the poorest and least urbanised provinces, the Northern Province, access line penetration was below 2 lines per 100 people. This pattern of insidious disparity was replicated inside the company: of Telkom's 60 000 employees in 1995, the vast majority of those in managerial and skilled technical positions were white males, reflecting the priorities of Apartheid social engineering. ¹³²

International trends towards liberalisation of telecommunications services which accompanied the transition from basic telephony to the new information-rich telecommunications era in the 1980s, were felt in South Africa: whilst the state retained its control over basic telephony and the provision of the telecommunications network, numerous private-sector undertakings were permitted to lease state-owned telecommunications lines in order to operate their own private data networks. Complementary to private networks, data communication systems with special service features - dubbed "Value Added Network Services" or "VANS" - began to emerge, where private operators provided services such as protocol conversion and access to databases, ¹³³ using the PTT's telecommunications grid for data transport and adding a margin of commercial value to it through the services offered. By the end of the decade, corporate customers of the PTT were able to lease lines on a trial basis, at special tariffs, for the operation of private voice networks. ¹³⁴

Notwithstanding these small concessions to the corporate market, Telkom was still their sole option for provision of the underlying network, and it enjoyed almost complete hegemony over the country's telecommunications. ¹³⁵ Its fixed assets at the 1993-1994 financial year-end were valued at R13,8 billion (then US$4 billion), with annual revenue of R8,35 billion (then US$2.4 billion), ¹³⁶ rising to R13.3 billion (then US$3.3 billion) in the 1995-1996 year-end. ¹³⁷ In that year, Telkom generated an operating profit of slightly more than R3 billion (then US$¾ billion), ¹³⁸ but like most PTT's it carries a significant debt burden, ¹³⁹ which ran to R8,7 billion (then US$2,2 billion) at the 1995-1996 year-end. ¹⁴⁰ At the 1993-1994 year-end the company was operating 3,66 million main telephone lines, of which 60% were digital, and 50 000 public pay-phones. ¹⁴¹ By the 1995-1996 year-end these figures had risen to 4 million main telephone lines, of which 70% were digital, and 70 000 public pay-phones, ¹⁴² these impressive increases reflecting new priorities for Telkom following the advent of democratic government in South Africa in April 1994. In particular, Telkom was expected to roll out universal access to basic telephony for the majority of people, thereby starting to redress historic discrimination by the state telephone company against black South Africans. Ranked internationally by size in 1996, Telkom was the 32 ^nd largest telephone company in the world.

4.2 The introduction of digital cellular telephony in 1994

As noted in section 3.1 above, plans for the privatisation of the state's telecommunication assets were first mooted in the late 1980's, and Telkom's creation in 1991, to hold these assets in a corporatised vehicle, was one of the results - though nothing further could be done until the new, democratically-elected government had taken power. But there were stirrings of market forces in the early 1990's, principally through the licensing in September 1993 of two GSM cellular networks, Vodacom and MTN. Vodacom began operations in the early part of 1994, and currently has approximately 900 000 subscribers. ¹⁴³ Telkom holds 50% of Vodacom's equity, with 31.5% held by British technology partner Vodafone Group plc, and the remaining 13.5% by Rembrandt, a South African-based conglomerate best known for its tobacco interests. MTN launched a few months after Vodacom, and has made up the backlog to achieve around 800 000 subscribers at present. MTN's equity is shared amongst a broad group of owners: the state-owned Transtel has 20%, ¹⁴⁴ the M-Net group has 25% via its subsidiary M-Cell, Cable & Wireless plc has 25%, SBC has 15.5% via a South African subsidiary, and the balance of 14.5% is directly held by empowerment groups, the largest single investor being the National Africa Telecommunication Company (Naftel) with 10%. Both the SBC and the Cable & Wireless stakes are about to be sold, with 20.5% to be divided into portions and sold to various black businesses and empowerment groups, ¹⁴⁵ the remaining 20% to be sold to a new international management partner to replace SBC. ¹⁴⁶ MTN has made some impressive forays into other African markets, most recently winning the second Ugandan cellular licence together with Sweden's Telia and local investors. ¹⁴⁷

Notwithstanding the dramatic growth rate of digital cellular telephony in South Africa, its relative cost makes it a premium service, and data collected thus far indicates that while 19% of white South African households own cellular phones, the penetration figure for blacks, coloureds and Asians is in the region of 3%. ¹⁴⁸ There is an important cross-subsidisation measure in the operators' licences, namely that Vodacom is required to install and maintain 22 000 GSM public pay-phones, and MTN 7 500, by September 1998. But, in terms of direct subscribers, while the existing GSM services fulfill the needs of some lower-income earners such as informal traders (approximately 60 000 of whom operate their cellular phones using pre-paid airtime sold by Vodacom), ¹⁴⁹ they do not significantly advance the key overall policy aim of providing the vast majority of historically disadvantaged South Africans with access to telephony.

4.3 The telecommunications policy process, 1994-1996

Future telecommunications policy, and in particular the liberalisation of the South African telecommunications market, had already become a contested issue during the CODESA negotiations process, with licensing of the cellular services in 1993 being particularly contentious. Following the April 1994 national election, a Technical Task Team began work on the basis for subsequent broad-based consultation, the Green Paper on Telecommunications Policy. ¹⁵⁰ Released by the Ministry of Posts, Telecommunications and Broadcasting in July 1995, the Green Paper identified a number of possible privatisation models, liberalisation approaches and regulatory options for discussion. Three basic policy premises framed the debate as to which options the country should choose: first, the Green Paper recognised that telecommunications "is not simply an aspect of development, but a precondition for its success." ¹⁵¹ Second, a primary objective of the new telecommunications policy had to be the achievement of a rapid increase in telephone penetration rates, particularly in black urban and rural communities. ¹⁵² However, the Ministry recognised that while the long-term goal was universal service - i.e. a telecommunications line in every household at affordable prices - the pragmatic, medium-term aim should be the achievement of universal access , meaning that shared community or public telephones would be placed within walking distance of people's houses. ¹⁵³ Third, the Ministry made it clear that it was well aware of the new global realities, namely that new technology was undermining the traditional PTT monopolies, that cross-subsidy mechanisms were not maintainable in the longer run and rates had to be rebalanced, and that the Uruguay Round trade agreements had acted as a catalyst for a new era of market liberalisation worldwide. ¹⁵⁴ Reflecting the deeply tensioned character of the Ministry's approach to liberalisation of the South African telecommunications market, and in sharp contrast to the enthusiasm for the market displayed by the IBA, the Green Paper states:

In the final analysis, competition is not an end in itself, but a means to an end, that is a tool that can be used to achieve certain objectives. Thus, determining the optimal market structure for the telecommunications sector, that is, striking the proper balance between competition (where and how it should be used) and monopoly (where it is still preferable), is of crucial importance. ¹⁵⁵

So, the market romance in telecommunications was to be much more cautious and measured. The debate over how far market liberalisation should go, and whether continued monopoly and cross-subsidisation in order to roll out universal access were a viable policy option, was canvassed in 131 submissions made by interested parties in response to the questions posed in the Green Paper. In order to attempt to reach some kind of industry and societal consensus, the major issues which emerged from the submissions were referred to a National Colloquium in November 1995. The Colloquium in turn nominated an Eminent Persons Group, which drafted the binding policy document, the White Paper, in conjunction with the Minister. This draft of the White Paper was tabled for discussion at the Plenary of the National Telecommunications Forum, the broad stakeholders' group which had been instrumental in initiating the policy process. Finally, after revision and consideration by both the Minister and the Cabinet, the White Paper on Telecommunications Policy was issued in mid-March 1996. ¹⁵⁶

The depth of consultation evident in this lengthy process was characteristic of the then-Government of National Unity, in which NP leader F.W. de Klerk was Second Deputy President despite the ANC's convincing election victory, and South Africa was still negotiating its path from the past into the more democratic present, and new policy initiatives had to be negotiated with careful regard for the realities of the continued economic power of white South Africans. The difficulty with the consultation process was that, of the 131 submissions made in response to the Green Paper, the large majority reflected the interests of well-financed and organised groups, rather than those of the most impoverished members of society, which the Green Paper suggested should be a focus of South Africa's new telecommunications policy. Ultimately, it was the ANC's responsibility to speak for those historically disadvantaged South Africans who had given it an overwhelming election mandate, but whose individual day-to-day battles with poverty made their engagement in each and every policy consultation process unlikely.

The White Paper carefully seeks to reconcile what it calls "two seeming opposites" of providing universal service to disadvantaged rural and urban communities, versus the delivery of high-level telecommunication services required by South Africa's sophisticated business, finance and industrial sectors: ¹⁵⁷

Liberalisation trends associated with the spread of the global information highway and the legitimate needs of South African business and urban areas for advanced services could easily combine to draw interest and resources away from the delivery of service to rural and disadvantaged areas. The potential development impact of telecommunications would be limited; the opportunity would be lost for South Africa to leapfrog traditional stages of development through the use of telecommunications to foster the application of new information technologies. ¹⁵⁸

The White Paper's solution to the dilemma of how to balance the sometimes conflicting goals of marketization and universal service, was a phased process of liberalisation, with Telkom to continue enjoy exclusivity in the provision of the public switched telecommunication network (PSTN) for a limited period of time, and liberalisation to be implemented, sector by sector, by an independent ¹⁵⁹ regulatory agency in line with the Ministry's policy directives. ¹⁶⁰

Telkom will function as the primary provider of the local loop, exchanges, transmission, and international services during the period of exclusivity. Because the central goal of the period is the building out of the basic network as quickly and as extensively as possible, Telkom should engage traditional cross-subsidy mechanisms to help facilitate network roll-out, including maintaining nationally uniform tariffs for standard [PSTN] services. Because the South African telecommunications market is already in part liberalised, this means that during the exclusivity period reasonable efforts must be put in place to safeguard against private service offerings and practices that unduly divert revenues from Telkom. Perhaps the central element here is the control of bypass and resale. ... On the other side of the coin, because Telkom has a number of advantages when it competes with private sector companies in value-added services, reasonable steps must be taken to ensure that all competitors operate on level playing fields. ... In brief, Telkom itself has two essentially conflicting requirements during the period of exclusivity: to roll out the network and keep usage affordable, and to rebalance its tariffs in order to prepare for competition. ¹⁶¹

Unusually for a broad policy document, the White Paper went so far as to suggest deadlines for liberalisation, which it stated were necessary "so that potential competitors are able to plan their entrance and investments accordingly." ¹⁶² The exclusivity period was set at five years, to begin following the establishment of the new Regulator and its licensing of Telkom. ¹⁶³ Telkom would be required to complete its tariff rebalancing by the end of this period, with the most significant rebalancing taking place in year 4. ¹⁶⁴ The use by businesses of private networks was legitimated, including switching capabilities within the private network, and the White Paper stated that no distinction would hence forth be made between voice and data carriage on these networks. ¹⁶⁵ Private networks would be permitted to resell leased capacity from year 4, ¹⁶⁶ though the same call could not be patched into a private network and then be routed back into the PSTN at its destination, as a means of by-pass. ¹⁶⁷ Competitors who entered the market would be expected to make contributions to a Universal Service Fund (USF), which would be used to compensate for the undermining of the ability to achieve universal service objectives due to the impact of competitive forces on Telkom. ¹⁶⁸ Similarly, private-sector VANS were legitimated, with the requirement that they use the Telkom infrastructure until the expiry of the exclusivity period. ¹⁶⁹ After the five-year exclusivity period had elapsed, i.e. at the beginning of year 6:

the national long-distance market would be opened to competition;

competitors would be allowed to enter the public pay-phone market, though using the Telkom local loop infrastructure;

the interlinking of private networks to create metropolitan area networks (MANs) would be permitted,

again, on the express proviso that entrants into these newly competitive sectors would be obliged to pay contributions to the USF. ¹⁷⁰ The final phase of the liberalisation programme would commence at the beginning of year 7, when:

international long-distance services could be provided on a competitive basis; and

a second full service network operator would be licensed to compete directly with Telkom in all areas of telecommunications services. ¹⁷¹

The balance of the White Paper dealt largely with the structure and powers of the new independent regulator, the South African Telecommunications Regulatory Authority (SATRA), a full discussion of which is, unfortunately, beyond the scope of this paper. (We will, however, be making some incidental comments on the extent of the new agency's independence as conferred by the subsequent legislation, in sections 4.4 and 4.5 below.) Despite questions raised in the Green Paper calling for submissions on the possible privatisation of state-owned telecommunications assets, ¹⁷² the White Paper does not make any policy statement on the issue. It simply records that no consensus could be reached in the National Colloquium, since organised labour (led by the ANC's alliance partner, the Congress of South African Trade Unions) wanted Telkom to remain in state hands, while the remainder of the Plenary favoured "strategic equity partnership with an international partner and/or private equity participation." ¹⁷³ In any event, noted the White Paper, "[t]his issue is subject, at present, to the outcome of the state assets restructuring process in the Government of National Unity." ¹⁷⁴ Decisions in that broader policy process required cooperation between several government offices, including the Department of Trade and Industry and the Public Enterprises Ministry. The government's policy decisions on privatisation were only released three months later, in its macroeconomic medium-term strategy document, Growth, Employment and Redistribution (GEAR). ¹⁷⁵ A key element of the medium-term strategy was set out to be,

the implementation of the public sector asset restructuring programme, including guidelines for the governance, regulation and financing of public corporations, and leading off with the sale of non-strategic assets and the creation of public-private partnerships in transport and telecommunications. ¹⁷⁶

In other words, the favoured mode of privatisation for Telkom would be the sale of a strategic equity share to a private sector partner, rather than its wholesale divestiture. By this time, the Minister who had steered the telecommunications policy process through to the White Paper, Z. Pallo Jordan, had been replaced by Jay Naidoo. The new Minister faced a tricky dilemma. Estimates put the cost of debt servicing by the South African government at more than 20% of its total budget, capital expenditure included, and a primary aim of the programme for restructuring public enterprises was to obtain funds, preferably from foreign investors, which would assist in lowering these servicing costs. Therefore, the government needed to obtain as much money from the sale of a strategic equity stake in Telkom as possible. Achieving good value, however, would require that potential foreign partners for Telkom be given cast-iron guarantees that their investment would be protected, and the longer Telkom's monopoly over basic services and its exclusive privilege could be maintained, the greater the value of a strategic stake would be to a foreign investor. After publication of the White Paper, and while these considerations were being weighed in government, the new telecommunications legislation went through 16 drafts, was publicly tabled before the Parliament Portfolio Committee and Senate Select Committee, and finally emerged in November as the new Telecommunications Act of 1996.

4.4 The Telecommunications Act of 1996

While Minister Naidoo steadfastly maintained that the White Paper would continue to be the policy framework for liberalisation of the telecommunication sector, there were two notable omissions from the new legislation: first, the White Paper's precise deadlines for liberalisation of each telecommunications sector were excluded, and replaced by a ubiquitous "dating" provision that,

No person other than Telkom shall be granted a licence ... until after a date to be fixed by the Minister by notice in the Gazette ; ¹⁷⁷

and second, the Minister would be issuing Telkom's new PSTN, radio and VANS licences, and not SATRA. All three of the latter licensing provisions use the same peremptory grandparenting language, stating,

... Telkom shall be deemed to have applied to the Minister for a [relevant licence type] and ... the Minister shall grant the application and the Minister shall issue such licence to Telkom ... ¹⁷⁸

The need for Telkom's future strategic equity partner to be assured of the security of its investment by such black-letter provisions is largely understandable, and the fact that the Minister was to issue Telkom's licences does not amount to a significant diminution of SATRA's independence. After all, in 1993 incumbent broadcasters were deemed to be licence-holders in a similar fashion. ¹⁷⁹ As for the omission of the precise liberalisation schedule from the legislation, the Communications Department's explanation that inclusion would have been undesirable and without international precedent has considerable merit. Nevertheless, the omission of precise deadlines thoroughly alarmed the business community, which had ardently championed liberalisation, and has dampened prospects for a clean cut to competition. Although the licence subsequently issued to Telkom does provide for expiry of its exclusivity period, largely as set out in the White Paper (see the fuller discussion in section 4.5 below), this does not mean that the Minister necessarily has to open the relevant market sectors to competition. Instead, in the case of international, national long-distance and local telecommunications competition in the South African market, potential competitors are left dangling, waiting anxiously by the 'phone for the Minister to call and give them a date.

Aside from these two controversial elements, the new legislation does faithfully reflect the policy intent of the White Paper. The Act has a list of diverse and tensioned objectives set out in section 2, that is similar to the IBA Act, and is based on a similar "wish-list" approach in the 1993 Canadian Telecommunications Act ¹⁸⁰ . Objectives such as universal service at affordable prices ¹⁸¹ are balanced against explicitly market-oriented goals such as ensuring "fair competition", ¹⁸² the encouragement of investment and innovation in the South African telecommunications industry, ¹⁸³ and the promotion of small, medium and micro-enterprises within the industry. ¹⁸⁴ As in the IBA Act, ownership and control of services by persons from historically disadvantaged groups is to be encouraged; ¹⁸⁵ and the need to "promote the empowerment and advancement of women in the telecommunications industry" is expressly realised. ¹⁸⁶ As has been said of similar provisions in the Canadian Act, the objects of the South African Telecommunications Act are "not dispositive conclusions, but merely the constraints within which argument must take place." ¹⁸⁷ In a similar fashion to the broadcast privatisation and liberalisation process, in the South African telecommunications sector this process has a multiple agenda: marketization; empowerment; and reconstruction and development.

Overseeing the achievement of these multiple aims is the new regulator, SATRA, which was established by the Telecommunications Act ¹⁸⁸ and has been operating since the appointment of its six-person Council in early1997. SATRA does not enjoy a constitutional guarantee of independence similar to that of the IBA, but South Africa has agreed to abide by the principles of independent regulation contained in the Reference Paper appended to the Fourth Protocol to the GATS. (South Africa's specific commitments in respect of basic telecommunications, and its regulatory commitments, are contained in Appendix 4 to this paper.) ¹⁸⁹ SATRA itself is bound by the Act to honour South Africa's commitments in terms of international agreements, ¹⁹⁰ and hence must abide by the regulatory principles to which South Africa has committed in its Schedule, which adopts the Reference Paper. A full discussion of the functions of SATRA and the degree of its independence in comparison to other regulators world-wide, is, unfortunately, beyond the scope of this paper. Essentially, while appointment to SATRA is a transparent, public process and there exists a substantial degree of functional independence, ¹⁹¹ the regulator's activities are circumscribed in three significant ways: first, in that SATRA is obliged to perform its functions in accordance with policy directives issued by the Minister of Communications ¹⁹² - though such Ministerial directions must, in turn, be consistent with the objectives clause in section 2 of the Act; ¹⁹³ second, as part of this general power to determine policy, the legislation has reserved to the Minister the right to set the liberalisation timetable, as discussed above; and third, in respect of the current areas of Telkom's "exclusive privilege" ( e.g. all public voice telephony services) and mobile cellular services, when the market is ultimately opened to competition, SATRA is not empowered to grant licences itself, but may merely "recommend that the Minister grant the application". ¹⁹⁴ Therefore, notwithstanding its express enthusiasm for market liberalisation, ¹⁹⁵ the new SATRA Council has little scope to emulate the IBA's past policy initiatives. Much of any assessment of SATRA's independence will, in any event, depend not so much on the letter of the legislation as on the extent to which the regulatory authority is willing to test the boundaries of its independence - and, as will be briefly discussed later, SATRA has already been aggressive in asserting its independence, in the recent dispute between Telkom and the private-sector Internet industry in South Africa over the provision of the Internet protocol.

The new Act vests SATRA with the powers to control, plan, administrate, manage and licence use of the radio frequency spectrum ¹⁹⁶ and the provision of telecommunication services using both wireless and wired means. ¹⁹⁷ Practically, this translates into immediate licensing authority for the provision of competitive private-sector VANS, ¹⁹⁸ private networks which interconnect to the Telkom network, ¹⁹⁹ and privately operated radio networks. ²⁰⁰ Potential authority to recommend licensing is established in respect of the remainder of services which currently fall within Telkom's exclusive privilege, i.e. the provision of a public switched telecommunication network, ²⁰¹ international and national long-distance telecommunication services, ²⁰² local access telecommunication services, ²⁰³ and public pay-telephone services. ²⁰⁴ SATRA has also been instructed to conduct an investigation into the feasibility of introducing a third mobile cellular service, ²⁰⁵ a process which is well advanced. ²⁰⁶ The outcome of this inquiry will almost certainly be positive: SATRA has already expressed strong opinions on the desirability of a third cellular operator being licensed, by the end of 1998 if possible. ²⁰⁷ International telecommunications operators said to be interested in bidding for the licence in conjunction with South African empowerment groups include Bell Canada International, Sweden's Telia, Telecom Portugal, ²⁰⁸ and Hong Kong's Hutchison Telecommunications. ²⁰⁹ One important difference between this greenfields digital cellular licence and the IBA's greenfields radio licences is that there is no foreign ownership cap in the South African Telecommunications Act; though, given the empowerment objects expressed in the Act and with SATRA's consultants recommending that historically disadvantaged South Africans obtain at least 25% of the equity in any new licensee, ²¹⁰ a foreign operator would be obliged to partner with appropriate South African companies if their application is to have any chance of success. From a service delivery perspective, the current position of the feasibility study is that the new licensee should be free to compete with the two incumbent operators for the affluent end of the market with high-mobility services, but that "the service obligation should be directed toward the low mobility market segment", ²¹¹ indicating that the new operator will be encouraged to compete with Telkom to provide basic telephony to currently under-served consumers.

The process of licensing itself, in a competitive application context, will first be tested in the licensing of the third cellular operator. The Telecommunications Act sets out a notice-and-comment process, with public hearings for applicants, similar to the IBA licensing process, ²¹² though adversarial oral representations against rivals have been dispensed with (the opportunity to lodge written representations remains). There is a more robust structure to the procedure by which decisions are made known, giving an opportunity to aggrieved rejected applicants to appeal to the regulator, which may serve to reduce the chances of successfully taking a regulatory decision on judicial review. This is a welcome development, since the IBA has found itself in court far too often in recent times, and the Telecommunications Act demonstrates that the relevant lessons in structuring regulatory authority have been learnt, by giving SATRA the ability to reverse itself on a decision if necessary. ²¹³ For private-sector companies applying for licences, the greater robustness of the licence-granting process (at least in comparison with the IBA process) will contribute to greater legal certainty in the liberalisation of the sector.

As noted above, the primary policy reason for retention of the Telkom monopoly over basic services during the exclusivity period, was to facilitate cross-subsidisation of the roll-out of universal access. However, the progressive introduction of competition and tariff rebalancing envisaged in the White Paper on Telecommunications Policy will mean that Telkom's monopoly revenues will decline over time, and with it ability to cross-subsidise network expansion. The solution proposed in the White Paper, a Universal Service Fund (USF) which would draw contributions from providers of competitive services, was put in place in the Act. ²¹⁴ A Universal Service Agency has been established by the Act to administer the fund, which is to be used exclusively for payment of subsidies:

to Telkom and other telecommunications providers who are obligated, in terms of their licences, to extend their public switched telecommunication services to areas or communities not served or inadequately served, in order to finance this extension of service; ²¹⁵ and

for the assistance of needy persons towards the cost of their being provided with or making use of telecommunications services, with direct subsidies becoming the sole mechanism of the USF once Telkom's rates have been rebalanced. ²¹⁶

While it might be argued that the USF and a companion human resources fund to which licensees will be obliged to contribute, ²¹⁷ constitute an additional burden on new entrants into the South African telecommunications market, the transparency of a publicly audited contribution fund is far preferable to leaving cross-subsidisation within the "black box" of the monopoly operator, and the substitution of USF payments for Telkom cross-subsidies will help to hasten rebalancing. What may appear at first to be an additional tax on new operators is, in fact, a necessary corollary to a competitive telecommunications market, particularly given the moral imperative to redress the inequities of Apartheid. This must be the price of any market romance in the new South Africa.

Other typical legislative supports for a competitive market also appear in the new Act: first, a licensee may not give itself "undue preference" or "cause undue discrimination" to any person or category of persons, ²¹⁸ and Telkom may not confer or propose to confer an undue advantage on itself vis-a-vis another licensee, current or future. ²¹⁹ In the event of SATRA determining that a violation of these fair competition provisions is taking place, it may issue a cease and desist order against the offender. Second, Telkom is obliged to fulfill any reasonable request by another licensee to allow it to interconnect to the Telkom telecommunication system; ²²⁰ and, similarly, Telkom is obliged to make its network available for use, by lease or otherwise, upon receipt of any reasonable request so to do. ²²¹ Therefore, while competition in basic telecommunications services may be a few years off, the structural elements to ensure that Telkom does not unfairly leverage its control of the existing public switched network into the future are at least in place, demonstrating that the new legislation is substantively committed to the eventual liberalisation of the market. This will be a long term, lasting romance.

4.5 Partial Privatisation: Sale of the 30% Strategic Equity Partnership Share in Telkom

As discussed in section 4.3 above, the South African government's macroeconomic policy document, GEAR, determined in mid-1996 that a private partner would be sought for Telkom. In the latter half of 1996, the Communications Ministry initiated a tender process for sale of up to 30% of Telkom. SBC Warburg were appointed to handle the sale on behalf of the South African government, and Goldman Sachs International to advise Telkom and prepare its due diligence programme. Since the Telecommunications Act (by then in its final draft) entitled the Minister to grant Telkom its licences, ²²² these could be prepared in advance and included in the Request for Proposal (RFP) that was issued to potential bidders. Thus Ministerial licensing, rather than being designed to impugn SATRA's independence, was in fact an important guarantor for potential buyers of the legal regime that would obtain after the purchase of the 30% Strategic Equity Partnership (SEP) stake. The simultaneous accomplishment by the Ministry in late 1996 of both the passage of a new and complex telecommunications regulatory scheme into law, in the form of the Telecommunications Act, and the sale of the SEP stake, was a very significant achievement. (The fact that the Ministry was so absorbed in telecommunications matters also explains why little attention was devoted to what the IBA was contemporaneously up to in its privatisation process, and created the policy vacuum which the IBA filled.)

The Ministry envisaged the following specific objectives for the SEP process -

to position Telkom to deliver basic telephony services on an economically viable basis to all South Africans by means of an aggressive network roll-out programme;

to achieve greater coverage of priority customers, e.g. schools, clinics, hospitals, governmental agencies, etc.;

to develop Telkom into a world-class telecommunications operator;

to obtain technology transfer and management expertise from the SEP, as well expose staff to world-class skills training;

to attract capital from the sale of the 30% stake, in order to finance network roll-out;

to facilitate transformation of Telkom into a company representative of the demographics of South Africa, through empowering historically disadvantaged South Africans;

to foster development in South Africa through Telkom creating infrastructure for a "knowledge society". ²²³

In exchange for enabling Telkom to achieve these goals and, of course, for the purchase price to be paid by the SEP for the 30% stake, it would obtain the five year exclusivity period and a pivotal role in management, as well the possibility of a further interest in Telkom down the line. On the other hand, the SEP would have to commit to Telkom achieving its network roll-out and specific service targets, with the incentive being the grant of an additional year of exclusivity should around 85% of the target figure be met, and sanctions in the form of penalties for failure to meet roll-out targets.

Finally, it was recognised that roll-out of basic telecommunications to under-served areas could not be on a wired basis alone, but instead South Africa would have to become a leader in deployment of wireless telecommunications; on the other hand, "leapfrogging" wired technology created opportunities to extend broadband services into disadvantaged areas. There are significant indications in both the Green and White Papers on Broadcasting that, at the time of the sale of the Telkom strategic equity stake, the SEP was offered a guarantee that it would be permitted entry into the broadcast/cable carriage market, and that this development in Telkom's activities can be expected within the next two years. ²²⁴

After the initial phase of the SEP tender, the number of interested bidders was reduced to four: Deutsche Telekom, France Telecom, Southwestern Bell Corp. (SBC), and Telekom Malaysia, the latter two eventually submitting a joint bid under the moniker of Thintana Communications (Proprietary) Limited, in which SBC Communications International (a wholly-owned subsidiary of SBC) held three-fifths of the equity and Telekom Malaysia two-fifths. Ultimately, though they saw the bidding process through to its conclusion in February 1997, neither the Germans nor the French submitted an offer for the SEP stake. Thirty per cent of Telkom was therefore sold to the joint SEP of SBC (18%) and Malaysia Telekom (12%), with the sale being concluded in May 1997 at a price of $1.2 billion (R5.6 billion, at the time).

Telkom's licences were duly issued by the Minister on 7 May 1997, ²²⁵ setting out Telkom's exclusive privilege period of five years ²²⁶ (with an option for a sixth, on condition that approximately 85% of the roll-out target is achieved by the end of Year 5), as well as the required roll-out, service targets and penalties. In the first five years of its licence, Telkom is required to instal 2,81 million new lines, including 120 000 pay-phones, and convert any remaining analogue portions of its network to digital. Though the network is only slated for full digitalisation by 2005, Telkom has moved this target forward to 2000. Priority customers such as schools and hospitals will get 24 000 lines, and Internet access must be provided to 2 000 institutions serving historically disadvantaged South Africans. An amount of R2,315 billion (US$450 million) is required to be committed to training for Telkom staff, of which at least 60% would be devoted to literacy, sales and service education, with historically disadvantaged South Africans being the primary beneficiaries. At least 35% of Telkom's management team will have to consist of South Africans from historically disadvantaged backgrounds. In total, R53 billion (US$10,3 billion) will be spent on the roll-out. ²²⁷

The penalties imposed for missing targets are R450 (US$87) per line for the first 100 000 lines missed, and thereafter R900 (US$174) per each additional line, in ordinary areas, and half of this amount in under-served areas. Penalties in respect of missing the target for installation of lines at priority customers are higher: for example, R4 500 (US$870) for each school line missed.

In addition to its roll-out targets, Telkom is also required to improve its service. For example, the residential customer fault indicator target per annum which was 600 faults per thousand lines in the 1997-1998 year, is to fall to 290 faults per thousand lines in 2001-2002. In the 1996-1997 year, to provide another example, 83% of business customer faults were required to have been cleared within 48 hours; by 1998-1999, the target figure rises to 90%, and by 2001-2002, it is 97%. Current waiting lists for overdue residential telephone installations (which, as an illustration, had over 120 000 installations overdue at any given time during 1996-1997) must be cleared by the end of October 1999. In 1996-1997, only 60% of residential customers had their orders for the installation of new instruments fulfilled within 28 days; by 2001-2002, this figure must be 80%, and 98% of orders must be met within 120 days. As with the case of lines missed, the failure to reach service targets is sanctioned with the imposition of progressively higher financial penalties on Telkom. ²²⁸

4.6 Telkom's Performance to Date

In the first few months following the SEP sale, Telkom's management was transformed. A new CEO, R. "Mac" Geschwind, was seconded from SBC in the U.S, and at the Board level three SBC and two Telekom Malaysia directors have been appointed. The SEP is, in general, now responsible for the day-to-day running of the company, though within a reporting structure in which non-executive directors representing the South African government have substantial involvement. A New York Times piece on the transformation of Telkom's management neatly captured the initial reaction to the SEP, as follows:

Suddenly, everything is changing. In the eyes of some South Africans, it's a miracle: the company is now being run by Americans, people famous for technical skill. In the eyes of others, it's a disaster: the company is now being run by Americans, people famous for arrogance. ²²⁹

Insofar as the achievement of line and service targets is concerned, the article's subtitle summed up the question everyone in the South African telecommunications industry was asking: "Can Phone Service Improve In Absence of Competition?" ²³⁰ While the first substantive answer will come later this month, when Telkom reports to SATRA on delivery in its first licence year, ²³¹ there have been some initial indications. Line delivery appears to be on target: with 340 000 new lines to be installed by March 1998, in January 253 713 new services had already been installed, including 177 000 in underserviced areas, and 22 400 pay-phones, a little ahead of the 20 000 required of Telkom by that time. ²³² Roll-out of digital wireless backbone also appears to be on target, with 520 projects using the Digital Enhanced Cordless Telecommunications (DECT) system to be completed in underserviced areas by the end of 1998, reaching more than 600 villages which previously did not have any access to telephony. ²³³ The major bugbear for Telkom at present appears to be its difficulty in achieving its service targets: complaints are increasing, ²³⁴ and customer surveys suggest that the South African public has seen no improvement in Telkom's service. ²³⁵ A recent South African Financial Mail article described Telkom's service woes as follows:

Telkom's Texas-based equity partner SBC is hell-bent on getting the monopoly-operator into the competitive saddle. Compelling self-interest lies behind SBC's determination. SA's fixed telephone market will be opened to competition within five years. In the interim, Telkom's monopoly is being eroded by de facto competition from the Internet, virtual private networks and other new technologies. High on SBC's priority list is the need to improve Telkom's service so that today's frustrated customers won't defect to competitors as soon as the market is liberalised. SBC's problem is that while Telkom's management has woken up to the fact that it will soon face real-world competition, most of its 56 500 staff are asleep. ²³⁶

Ahead of the first annual report to SATRA, executives from both SBC International and Telkom have been foreshadowing bad news, saying that it is going to take time to change Telkom's corporate culture in the absence of direct market competition. SBC attributes the problem in part to service lag in the wake of rapid and unprecedented network expansion, and argue that they have "read this book before" in turning around former Mexican PTT Telmex, which was granted a six-year exclusivity period that concluded in 1996. ²³⁷

But Telkom has already upset many corporate clients through their unfortunate handling of an ongoing dispute over provision of the Internet protocol (IP) in South Africa. While full discussion of this dispute is beyond the scope of this paper, in essence it began with an antitrust action brought against Telkom prior to the arrival of the SEP on the scene, in early 1996. Private Internet Service Providers (ISPs) in South Africa had by that time propelled South Africa to one of the 20 most Internet-connected countries in the world, ²³⁸ using network infrastructure leased from Telkom. At that point, Telkom entered the market with its own service provider, SAIX, and was accused by private-sector ISPs of unfairly leveraging its control of basic voice and data networks to engage in predatory pricing, and run the private-sector ISPs out of business. ²³⁹ Following numerous Competition Board and regulatory twists and turns, the Telkom position evolved into a claim in March 1997 that the provision of IP was part of the exclusive privilege in respect of basic telecommunications accorded to Telkom by the Telecommunications Act of 1996. ²⁴⁰ The claim was regarded by the private-sector industry as preposterous and indefensible in terms of any reasonable interpretation of the Act, which expressly identifies the provision of e-mail as a VANS in the competitive arena. The strong suspicion was that Telkom was moving the goalposts in order to preclude any close scrutiny of the affairs of SAIX and another Telkom creation, Intekom. ²⁴¹ When SATRA issued a "Pronouncement" in the dispute, ruling against Telkom in October this year, ²⁴² Telkom promptly took SATRA's decision on judicial review - reinforcing the predominant opinion in South Africa's IT and telecommunications industries that Telkom was using litigation in order to steamroller the private sector ISPs into submission. Troubling for SATRA, the Supreme Court judge who heard Telkom's review application found that SATRA had not followed proper procedures in coming to its decision, so that its "Pronouncement" was of no effect. ²⁴³ (SATRA has had similar difficulties with its first ruling, declaring call-back operators illegal; the ruling was made without sufficient due process, and almost a year later the call-back industry in South Africa continues to operate.) ²⁴⁴ The judge referred the dispute to oral evidence, but Telkom have postponed proceeding with the matter for now, suggesting that a settlement may be negotiated. The private sector Internet industry in South Africa has continued to flourish all the while: the major South African ISPs have developed ever-closer links with major information technology companies, and are beginning to develop private data networks which may provide significant competition for Telkom in the future. ²⁴⁵ Given these real-world developments, Telkom's exclusivity period may not turn out to be as exclusive as the SEP partners were promised.

5. ab

Conclusions: some initial assessments of results

In June 1988, just a decade ago, South Africa was about to enter its third consecutive year of emergency rule by P.W. Botha's security commissars; state hit-squads were on the loose, murdering anti-apartheid activists; the South African Defence Force was still engaged in a brutish war on Namibia's border with Angola; and the economy was reeling under the impact of ever-increasing international sanctions. Two colossal state-owned entities were amongst the pillars perpetuating National Party government: the Department of Posts and Telecommunications, which controlled the vast majority of telecommunications assets in the country, and the South African Broadcasting Corporation, whose management saw its purpose to be the galvanising of the country's white population against the "total onslaught" of communism and the ANC-led liberation movement. An appreciation of the abyss which South Africa faced - indeed, by then had begun to tumble into - is the essential starting point for any analysis of developments in that country.

But, we have now passed beyond the moving images of patient queues of voters in the 1994 national election, and beyond the astonishing drama of Nelson Mandela's inauguration as South Africa's first democratically-elected President. South Africa is back in the world economy, with Finance Minister Trevor Manuel and South African Reserve Bank Governor Chris Stals implementing conservative fiscal and monetary policy, and committing the country to an extensive programme of trade liberalisation. South Africa's story has moved beyond mythical television images, to the difficult process of creating a functional economy that delivers growth, development, education, and opportunity to all its citizens. An important part of that story is being told in South Africa's communications industry, in the processes and outcomes of the partial privatisations of Telkom and the SABC, and the liberalisation of their markets. Because their starting points as market-hogging state monopolies are so similar, Telkom and the SABC offer fascinating material for comparison. Yet, we would caution that our assessment is an early one, since privatisation and liberalisation of the South African communications industry are still ongoing: the privatisation of the SABC's commercial broadcasting portfolio, consisting of at least one of its national terrestrial television channels and several of its radio services, is in the offing, as is the issue of new regional television licences and city and town-based radio licences; Telkom still has to face the "big bang" of full market competition in basic services, in either May 2002 or 2003. ²⁴⁶ With that caveat in mind, we would offer the following tentative conclusions.

First, the outcomes of the two privatisation and market liberalisation processes, in broadcasting and telecommunications, need to be measured against their particular objectives. As we have continually stressed, privatisation and liberalisation in South Africa have had a multiple agenda: marketization providing competition, efficiency and choice; empowerment extending ownership and control to historically disadvantaged South Africans; and reconstruction and development , creating an infrastructure for and channels to direct economic upliftment of disadvantaged people. Table 2 below summarises our assessment:

Table 2: Assessing Privatisation and Market Liberalisation, 1994-1998

Our opinion, therefore, is that while the 1994 point of departure for the telecommunications and broadcasting industry were very similar, the outcomes of their liberalisation processes have been quite different. In broadcasting, successful marketization and empowerment has come at the expense of delivery of services to, predominantly, the affluent sector of society. A further trade-off was that the IBA's desire to ensure that the most suitable bidders were awarded the SABC stations resulted in the public fisc not receiving full value for the state assets sold - though there may be some doubt as to whether optimum empowerment need come at the expense of the fisc; arguably, the under-valuing of the stations was more a consequence of a confused sale process. In telecommunications, on the other hand, the retention of Telkom's exclusive privilege in respect of basic services has either entirely prohibited or chilled competition, depending on the relevant market sector, in the short term. Obviously, the failure to develop competitive basic service providers at a time of great development in international telecommunications markets will also have long-term implications for South African telecommunications companies' competitiveness. On the other hand, the government's rationale, that the Telkom monopoly should be maintained in the short term so as to facilitate the SEP sale and roll-out of basic telephony to historically disadvantaged South Africans, has thus far been vindicated by the impressive results in network expansion in the past year.

One important variable which differentiates the Telkom and SABC privatisation processes is capital intensity , and it may go some way to explain why their outcomes have differed. Some of the SABC stations were picked up for as little as US$2 million, and a new station can be launched at an even smaller cost - a very different proposition to the US$1.2 billion purchase price made by SBC and Telekom Malaysia for their stake in Telkom. It may be argued that the cost of keeping Telkom whole was that no financing structure could be created to enable historically disadvantaged groups to take a significant share in it; instead, the government would remain a 70% shareholder, as a proxy for the majority of South Africans who will benefit from the network roll-out. The cost of maintaining Telkom intact, as a monopoly, goes further than that, of course: monopoly has its own odious characteristics, such as the poor service South Africans still experience from Telkom. Compare this to the abundant diversity of sounds and information available when turning the radio dial in Johannesburg, or the consumer's choice of more than 30 competing Internet service providers operating points of presence in that city. Competition in primary service delivery also spawns efficient secondary markets: for example, in the radio airtime sales market there has historically been only one national player, the SABC's Radioactive division. Now, with 13 private stations across the country, there are two national competitors, Radmark (representing the stations East Coast, Algoa, YFM, Jacaranda, and OFM-Oranje) and Radioworks (representing the Primedia-managed stations Radio 702, Highveld and Cape Talk), and seven stations are representing themselves in the sales market. In addition, an impressive secondary market in radio broadcasting talent has now developed, where previously being blackballed by the SABC meant the end of one's radio career. Leading radio show hosts in South Africa can now expect to sell their services for three to four times the amount of the salaries offered in the days of the SABC's monopoly.

For the fruits of competition, we can also look to the digital cellular network operators: Vodacom and MTN have both innovated pre-paid airtime cards which preclude the need for subscribers to enter into long-term contracts - important in the South African context because many people from disadvantaged communities have never held an account at a bank, and hence have no formal credit histories, usually required as condition of the standard 24-month airtime contract. Vodacom's "Vodago" technology, in particular, is so successful that it won the recent international GSM network operators' award. Perhaps it is fruitless to speculate whether there was any other option for Telkom than the SEP route: unbundling the company into numerous competing units was suggested as an option by market romancers, during the policy process. Whether or not that option was ever realistic, it is important to recognise the costs of monopoly and to continuously remind South Africans of them, lest the Minister's market ardour cool during the remainder of Telkom's exclusivity period.

Second, the very different roles of regulatory agencies in the two privatisation processes need to be carefully kept in mind. Whereas the IBA initiated and played the leading role in the sale of the SABC regional stations, SATRA was entirely excluded from the Telkom SEP sale, and the Minister still retains authority to both open and licence future basic services competitors. The Telecommunications Act's limits on SATRA's jurisdiction have confused several commentators. As August a journal as The Economist has got it completely wrong: in their May 16 ^th , 1998, edition, they claimed that SATRA "falls under the Department of Communications" (not true - it is an entirely separate entity, in legal form, staffing and physical location), that it is overly "understanding" of Telkom (since Telkom currently has SATRA in court over the adverse Internet protocol decision, it would disagree), and that "the regulator recently ruled that several Internet services fall under [Telkom's] monopoly" (a horrendous error of fact, since, as discussed in the previous section, SATRA ruled against Telkom and were taken to court for their trouble). ²⁴⁷ Minister Naidoo has understandably been annoyed by accusations that SATRA is a lap-dog for the Ministry:

There are those who have naively questioned this body's independence. The reality internationally is that there is no regulator of telecommunications in the world - including the oft-quoted FCC of the U.S. and Oftel in the UK - that operates independently of industrial policy and the framework of the sovereign nation state. ²⁴⁸

Clearly there has been insufficient understanding of the degree to which the SEP sale and the simultaneous determination of the ambit of the regulator's jurisdiction were linked. Furthermore, the pivotal role of the IBA in the SABC radio privatisation demonstrates how crucial the interplay between regulatory agencies and state privatisation can be - particularly when both the regulator and the privatisation policy are created simultaneously. Regulatory oversight of privatisation is not necessarily demonstrative of the regulator's independence; there is cause to have the regulator involved, particularly in broadcasting where important democratic values such as diversity of voice are involved, but management of the process is not a typical regulatory task. One important outcome of the broadcast privatisation process in South Africa is that the IBA's wings have been pinned back unceremoniously: the next time an SABC asset is privatised, says the White Paper, the Minister will be in charge. So long as the IBA (or it successor within the merged regulatory agency) retains ultimate licensing power, this does not compromise its independence. But this controversy over both the IBA and SATRA's powers and independence demonstrates that the relationship between regulation and privatisation deserves more study.

Third, there have been two very different approaches to privatisation: the IBA-led "big bang" approach to the SABC's commercial radio assets and liberalisation of the regional markets; versus the partial privatisation process adopted by the Communications Ministry in respect of Telkom. Stepping back from the peculiar circumstances of South Africa, we might have intuitively guessed the respective outcomes of the processes, just from their characteristics: the "big bang" approach would result in a competitive market, with high service levels and the creation of competitive secondary markets, but with the focus being on affluent audiences who can profitably be sold to advertisers. The partial privatisation approach, on the other hand, would enable the state to direct resources to disadvantaged consumers by way of cross-subsidisation in the short term, and phased-in competition would be helpful in giving a state-owned corporation time to adapt and reform its culture, though at the cost of poor service, artificially high costs, and lack of competitive choice while the titanic Telkom edifice was being turned in a new direction. The results which we have outlined in Table 2 would seem to conform to these intuitive expectations.

The decision to defer competition in South African telecommunications remains a controversial one; many believe that a more robust pro-competitive approach could have achieved the same goals as the SEP process. We will give Minister Naidoo the last word on this issue:

[The South African] Government has not chosen to go down the road of unbridled deregulation in the telecommunications sector. To have done this would have been an unmitigated disaster and would have compromised all the socioeconomic goals we have set. What we have done - contrary to the view of many analysts - is to set in place a world-class environment for developing dynamic growth in the sector. ²⁴⁹

Only hindsight will determine whether the Minister is right. What is interesting, though, is the suggestion that once begun, the deregulation of a sector can be controlled, and that it is possible to introduce competition in carefully planned phases. Advances in telecommunications technology have a nasty habit of blurring carefully pre-determined regulatory borders. We are reminded that Johnny Mercer continued his song, the first verse of which we quoted at the outset of this paper, as follows:

Fools rush in, where wise men never go

But wise men never fall in love, so how are they to know?

The illusion here, of course, is that one can take a fling and try out love, see if it is to one's liking, and reap all the benefits if the gamble is successful. Like Minister Naidoo's view of phased-in competition, this is all a little too controlled and pre-determined for true market romantics. The market is more seductive than that, particularly given an entrepreneurially-minded population such as that of South Africa. We believe that, in the current telecommunications environment, a partial monopoly is bound to be swept away sooner rather than later by the competitive and technological forces surrounding it. Any market romance is likely to prove to be very much more impetuous and less controlled and calculating than suggested by a series of dates to be fixed by the Minister. Caution will likely have to be thrown to the wind if the romance is to blossom.

Appendix 1: Bidders for Privatised SA Broadcasting Corporation Regional Radio Stations