Regulation by litigation
Integration is not a benefit of the network but has to do with what you attach to the network. The user is the integrator. The network doesnt integrate it is only a carrier… Telephone companies cant seem to get it straight in their heads they believe they are the integrators. The network doesnt deliver value. Speed, protocols, extra bandwidth and voice capability are cost issues. Adding value is a highly personal interaction If telephone companies begin generating information and playing with it this could lead to the ultimate monopoly. Canadian futurist J. James Mackie, addressing the TUANZ conference, August 1993.
In 1990 when New Zealand became a leader in deregulating telecommunications there was a sense that the country was breaking new ground and would soon be in a position, with affordable bandwidth, to develop and use smart new applications so we could play at the leading edge of a brave new world.
We were told this new openness would give our entrepreneurs and business pioneers an edge that would quickly reflect on the bottom line of our economy and lead the way to export growth. What we got was a series of costly experiments that turned New Zealand into a telecommunications laboratory for others to learn from and gave little back to the consumer other than cheaper toll calls and a fiercely protective incumbent carrier that hogged the data market.
While the universities were battling to get international access and trying to cope with the growing demand for their campus-based number 8 wire access arrangements, the telecommunications carriers seemed at first oblivious to the Internet. They stayed firmly focused on their own tactical war based on phone calls and high-end data services, which proved such a distraction the window of opportunity onto the brave new digital world began to close.
There were no clearly defined rules and regulations for network interconnection, and because Telecom allegedly owned all the overhead and underground copper lines into every home and business and claimed ownership of the phone numbers as well, there were prolonged battles for new players to strike agreeable local access terms. Telecom was doing all it could to invest in new infrastructure while keeping competitors at bay with every ploy possible. Millions of dollars were spent on litigation as competitors sought to connect into the last mile. Telecom held tightly to its tariffs so benefits at best trickled down. Data services, the key area for real business competition, were the last bastion; if you werent in the central business districts of the main centres you were in for a long wait.
Communications Minister Maurice Williamson found himself in a running battle, attempting to drag Telecom kicking and screaming into an arrangement where Clear and others could interconnect with its network at reasonable rates. What we ended up with was this Baumol-Willig junk which says if someone takes part of your business off you, then they have to pay you a full equivalent of what you were earning anyway. In the end Telecom could have lost its entire business and still retain the same revenue strength. So what was this Baumol-Willig business? After a number of legal disputes over local access interconnection terms between Clear and Telecom from 1991 onwards, Clear sought interconnection at incremental cost, with payments between the two companies on a reciprocal basis.
Telecom responded offering terms based on the Efficient Component Pricing Rule (ECPR), also known as the Baumol-Willig rule, which would have required Clear to pay it the opportunity cost of providing interconnection together with a contribution to common costs and profits, including any monopoly profit foregone by Telecom from loss of business. The case went to the Privy Council, which held that the Baumol-Willig rule did not breach the Section 36 of the Commerce Act which dealt with anti-competitive behaviour. The government was opposed to the use of this pricing principle because it had the potential to lessen competition. The companies finally agreed to set access prices at levels below those implied in the rule. With some form of interconnection finally agreed to, the formative competitive market could at least begin to make some headway.
The hype about the information superhighway hadnt slowed down, only the ability to deliver it. Demand was being driven by the growing power of PCs, the drop in price of computer memory and hard disk space, and the need to send more complex documents and images. The ability to move documents between PCs and across an office network at 10Mbit/sec over Ethernet cabling had existed since the late 1980s and many internal networks were now capable of 100Mbit/sec over copper and fibre. The desktop revolution was showing no sign of slowing but moving data at speeds beyond 2Mbit/sec over the Telecom network was impossible, unless you had your own WAN between locations, or leased a dedicated digital data line from Telecom.
Visiting futurist J. James Mackie warned users to start taking the lead in provision of new services and stop leaving it up to the telephone companies which, he said, shouldnt be in a position to dictate. Telephone companies dont know what to do with bandwidth; their mandate is to collect tolls. More should be happening here in this completely deregulated market. He said the fortunes of the future were not to be made in hardware or networking but in the provision of new services which should be one of the greatest business adventures of our time.
In its communications policy in July 1993, the Labour opposition party stated its deep concerns about the state of play in deregulated New Zealand, suggesting we were seriously slipping behind. The solution was to extend fibre-optic infrastructure to every home within the next ten years. Labour leader Mike Moore warned we needed to speed up our implementation of telecommunications or be left behind by what was now being taken for granted in the first world. He said the market could not be relied on to produce a national fibre-optic network, and it would fail unless the government got behind it.
Market forces be with you
Maurice Williamson, however, stuck to his hands-off stance, continuing to wax eloquent about New Zealand being a world leader in telecommunications. He claimed fibre-optic penetration and the level of digital exchanges in New Zealand was higher than in any other country in the world, despite that technology being largely reserved for Telecoms own services.
Telecom claimed to have invested $3.3 billion in its infrastructure, having introduced the concept of the intelligent network and No 7 signalling to add value to 0800 and 0900 calling, automated calling cards, Centrex virtual PBX, and enhanced ISDN. However nationwide ISDN and broadband (2Mbit/sec-plus) infrastructures remained tethered by interconnect arguments and endless court battles. By mid-1993 there were still only interim interconnection agreements between Telecom, Clear, new GSM cellular provider BellSouth, and third cellular player Telstra, which planned a service by the end of the year.
It took several years but in 1992 Telecom finally reached an agreement with Clear on local service interconnection and began to engage in a price war over toll calls. Telecom chairman Peter Shirtcliffe claimed the cost of long-distance calls had dropped 45 percent over five years. In the absence of strong guidelines for competition Telecom remained thede facto regulator and the courts the de facto policy makers, leaving the former government department free to glean the maximum from captive markets. The courts even ruled to disallow the Commerce Commission inquiry on telecommunications, which had identified eight obstacles to competition, reflecting Telecoms role as supplier and controller of the market. The report singled out lack of competition in network and value-added services as being of particular concern.
New Zealand remained the only country in the developed world with no telecommunications regulatory agency. It relied instead on the Commerce Act, the courts, and the Commerce Commission. As Sir Geoffrey Palmer pointed out, this situation reflected New Zealands inexperience in dealing with competition. For decades we had state monopolies. Once those were sold off, we were left with little awareness let alone adequate coping mechanisms of the potential dangers that private monopolies can pose to competition. The Commerce Commission obviously felt powerless to pursue the matter further when it had the courts to contend with and a government determined to keep its distance:
Don Wallace, chairman of TUANZ, likened the interconnection debacle to a car stuck in the mud with the occupants so busy arguing about whose fault it was they had no energy to push the vehicle back on the road. They need a tow truck, he said, but the only tow truck operator in town didnt want to get involved. Deregulation of the New Zealand market was supposed to have opened the way for cheaper, faster, more sophisticated services, but the delays infuriated Clear and BellSouth. Clear manager Neil Tuckwell claimed two years of opportunity had been wasted. BellSouth managing director Keith Davis doubted whether the company would have gone ahead with its plans for New Zealand had it known the difficulties that lay ahead. If it wasnt for the deep pockets of both competitors they may not have survived the wait and it was still far from over.
In 1993 BellSouth had launched the first mobile phone network to compete with Telecom, and Telstra purchased 20-year rights to operate a second mobile network, then sold its spectrum to BellSouth in return for mobile services. Meanwhile Ameritech and Bell Atlantic reduced their Telecom shareholding to a combined 49.6 percent. Massive investment was being made in increasing telecommunications capacity between New Zealand and the rest of the world. In March 1992 the $200 million Tasman 2 submarine fibre-optic cable between Australia and New Zealand was switched on. This was the first stage in the massive Pacific cable network due to ring the Pacific basin by 1994. In June 1993 the $500 million PacRim East undersea fibre-optic cable between New Zealand and Hawaii, a partnership between Telecom, AT&T, KDD, and Telstra Australia, became fully operational, offering voice, video, and data transmission. A year later the third stage of the Pacific cable network connecting New Zealand and Australia with Asia via Guam was live.
Who needs broadband anyway?
In July 1993 Eddie Paterson, Telecoms group product manager for data solutions, articulating his blinkered view of market demand, said he couldnt see a business case for anything beyond 2Mbit/sec, even among larger corporations. About 90 percent of users didnt need anything more than 48kbit/sec128kbit/sec for LAN to LAN traffic although there would be an increasing need to mix DDSs with ISDN, when usage reached a threshold. However he did admit New Zealand was way behind the United States where 30Mbit/sec was now common and many organisations were migrating to the new ATM standard for 155Mbit/sec speeds.
Clear had a fibre-optic backbone running down the centre of the country and a second around the East Coast, costing around $50 million, was under consideration. The competing carrier had invested $160 million and achieved about 16 percent market share since it kicked off in February 1991. That had translated to about 150,000 customers including more than 100,000 in residential areas. Clear had chosen not to develop the high-end data market because of perceived barriers to competition, including lack of access to the local loop.
Manager Neil Tuckwell said Clear wanted to offer a full range of services based on an intelligent network capability, including ISDN, and was investigating options. By 1993 it finally had its own version of 0800 (0508) and users could drop the 050 prefix in front of every Clear number they called. It had access to 15 centres and attempts to negotiate an additional 27 points of connections had been going on for a year. A small mountain of technical and political issues needed to be scaled before it had full access.
The main sticking point was Telecoms insistence on charging Clear as if it were a retail customer. Clear complained under section 36 of the Commerce Act, claiming Telecom had acted in an anti-competitive way by purposefully restricting access to a full range of services. The biggest concern was bundling a discounted grab-bag of telecommunications products often presented to customers on the condition they dont switch carriers. Clear couldnt reciprocate because it didnt have the breadth of products to bargain with, such as local loop access and Centrex. The long-standing legal battles between Clear and Telecom over competition law eventually ended up at the Privy Council in London. The only winners were the lawyers. The decision backed Telecoms right to maintain its monopolistic activities.
In the absence of true competition the focus remained on proprietary services or enhancements to Telecoms own network. There was little sign of lower data-networking costs or for that matter any discussions about the Internet at a level that businesses might take heart from.
Netway Communications, a partnership between Telecom and Freightways, had become something of a renegade outfit operating just outside of Telecoms corporate arm. It had done a deal with Tuianet to provide frame-relay circuits when all other negotiations had failed. Managing director Malcolm Dicks serious faith that competition was good and healthy and then backing it up by using Aussat for trans-Tasman communication and leasing space from Clear must have raised some eyebrows. In September 1994 he stepped out of the box again and announced a retail ISP service for corporate customers would launch within six months. It didnt. Netway had been pulled back into line, and Dick was soon off to Australia to seek his fortune.
By June 1995 Telecom had enjoyed seven consecutive quarters of revenue growth and an earnings growth rate that ranked it with the top 15 percent of telcos worldwide. Clear finally reached an agreement on local service interconnection after spending $8 million trying to resolve ongoing disputes. Remarkably, it had managed to achieve an estimated market share of 23 percent during its four years battling the incumbent.
It wasnt only the cost of faster services like ISDN or leased lines and the ongoing carrier squabbles that were concerning users and curbing the growth of businesses; the charging regime for bandwidth at the Waikato Internet gateway (NZGate) was also an issue. The speed of the international connection had doubled to 256kbit/sec in July 1995 but ISPs were feeling the pinch with per megabit charging.
According to Net Wizards and a report by Colin Jackson at the Ministry of Commerce, New Zealands Internet use had been through a rapid growth spurt, leading the world in uptake. However browsing the Web had become something of a luxury, particularly for those visiting offshore sites or for US users visiting local sites. While US users got to surf free-of-charge, New Zealanders were paying up to $8 a megabyte, regardless of which way the traffic was flowing. Several proposals were put forward to ease concerns, including set fee charges, home-page mirroring, and fee reductions at NZGate.
The problem was that NZGate hadnt shifted from its original billing approach, which enabled it to share the cost of traffic between universities and research institutions to cover international bandwidth costs. The subsidy from NASA had disappeared in 1994, and charges were now being passed on to the growing number of ISPs who linked directly into the Waikato exchange. Waikato was charging providers on monthly volume (around $3.50 to $4.50 a megabyte). Iconz, the PlaNet network, Actrix, Cybernet, and others had to pass on this charge to users, typically at $2.50/Mb$4/Mb for national traffic and $4/Mb$8/Mb for international traffic.
CompuServe, IBM, and Microsoft were currently the only providers offering access independently of Waikato, although David Dixs KC Computer Services had its own international gateway for select customers. Waikato gateway manager John Houlker had to justify the charging by reminding users that New Zealand was one of the most expensive ends of the Internet and suggesting larger providers look at their NZGate fees; if they were fairly stable he suggested they should consider charging a flat rate. NZGate was doing what it could to bring charges down and speed up the flow of traffic, including creating a mirror site for New Zealand content in the United States to cut down on costs. Houlker suggested web site owners keep rich content such as videos off their sites. However Alan Marston, founder of the PlaNet Network, was refusing to put up any web sites until the charging issues were resolved. Even without Web page access, PlaNet was paying Internet umbrella group Tuianet about $1000 to $1500 a month. It is inconceivable that well put up a web site with our stuff on it when we have to pay every time someone wants to look at it, he said. New Zealand was becoming a very sought-after location for international people to explore. Anything about the environment or alternative lifestyles is very popular. You should be able to make a Web page with valuable information that would benefit New Zealand without having to meet the complete cost or pass it on to subscribers.
Meanwhile ISPs were considering whether to charge per kilobyte stored on their Web servers or flat or monthly rentals per page to get around data volumes. Time-based charging was another option being considered. David Dix had hedged his bets with links to NZGate for his main users and dedicated international link to Hawaii for commercial customers who wanted more bandwidth. At the time he was hosting about 30 pages with different arrangements for each. One company has had four pages up for three months and it cost them less than $100. However he warned that New Zealands growing attraction for tourism and travel opportunities presented a problem. Youd be swamped if you put up that kind of home page, but you have to weigh up the cost and what the value is. Providers have to be flexible and help their customers work out what is best for them.
If the competition model wasnt seen to be working among the big carriers, the principles of letting the market decide were certainly starting to have an impact among small- to medium-sized ISPs, who had generally taken a relatively laid-back approach to the market theyd had to themselves since 1992. Early in 1996 there were about 40 ISPs servicing around 160,000 users. Competition was heating up, and everyone was looking to cut costs while still retaining and growing their customer base.
Most ISPs were still taken back by Ihugs bold move into flat-rate charging, and it was clear that telco and ISP tariff models needed urgent reworking. Increasingly the answer to the dilemma was being seen as flat-rate charges to a megabyte limit and then time, or per-Mb-based charging, thereafter.10 Full interconnection had finally been signed off between Telecom and Clear, and international carriers such as SITA, Telstra, and MCI were now vying for business. There was great hope the promised rewards of deregulation would at last be delivered to long-haul data customers, including ISPs.
Who’s the gatekeeper?
Having only peeped over the commercial horizon in 1992 the Internet was consuming ever-increasing bandwidth and in turn costing the Waikato University gateway and subscribing universities a small fortune.
Traffic through NZGate had quickly outgrown the original ANZCAN undersea copper network and maxed out the capabilities of a satellite connection. From 1993 backbone traffic had been delivered between New Zealand and the United States via frame relay over the $500 million PacRim East undersea fibre-optic cable owned by Telecom in partnership with AT&T, KDD, and Telstra Australia. Gateway manager John Houlker made frequent visits to Telecom asking for circuit upgrades on the satellite and then on PacRim. We were demanding more capacity than any of its other international customers. Id been trying to tell them this Internet thing was really important but I dont think this had much impact until they started adding up what we were doing with all this bandwidth. I wanted them to be aware of the demand so they could plan better on their side but also to keep in the loop because it was logical to me that they would ultimately be involved in this business.
Eventually the main carriers began to see this was indeed their territory and discussions about their involvement began in 1994 when the NASA funding finally faded out. Initial negotiations to operate wholesale services through Waikato involved Telecom, Clear, and overseas operators including MCI. Waikato University was now sharing the full cost of international links with all the other universities and CRIs on a volume basis.
NSFnet had moved across to three private backbone providers but NASA was still managing the US end of the connection at the Federal Internet Exchange (FIX West) gateway, with a link into the next-generation supercomputer networks. The arrangement with the PACCOM community, which Waikato was party to, was expected to come to an end with traffic moving across from NASA Ames to the proposed FIX East gateway. Houlker recalls being New Zealands representative at the Tokyo meeting of the International Engineering Planning group for the Internet in 1992, where half the agenda was taken up trying to decide on a name for proposed new commercial exchanges to attach to FIX East and FIX West.
There was eventually a consensus that they be called MAE East and MAE West, short for long haul metropolitan area network. However various political and technical obstacles meant the proposed shift to commercial routers at the MAE West exchange failed to eventuate. In 1995 when NSFnet reverted to its initial role as a Federal research and education network and NASA, under its public non-profit charter was no longer supposed to be in the business of carrying commercial traffic, New Zealand remained an exception, continuing to operate through NASA equipment.
It was a time of transition as the neutral NZGate and Tuianet, the non-profit body that administered the science and research network, prepared to hand over the commercial side of their businesses. Waikato was hosting two high-speed circuits to the United States, effectively serving the entire Internet community as well as the science and research networks. From May 1995 the amount of commercial traffic moving across NZGate and Tuianet grew significantly, because most universities were still acting as providers for their own campuses and delivering links into commercial ISPs. A few ISPs operated their own international links and while some ran long-distance links between particular cities, there was no apparent advantage in constructing a separate commercial backbone to mirror the Tuia backbone.
That all began to change rapidly and by the end of the year, research and educational institutions accounted for only half of New Zealands international Internet traffic. Commercial international traffic was divided roughly equally between the Tuia network and other links to the Hamilton gateway.
The highest-capacity connection within the Tuia network was the 512kbit/sec link between the University of Waikato and Victoria University. Most links were at 64kbit/sec or 128kbit/sec.
Both Telecom and Clear had signalled their interest in taking over the responsibilities of NZGate. The ideal time for the handover would be before the end of January 1996, when Waikatos international leases were due for renewal; it had committed to 12-month contracts to get the best possible deal on bandwidth. The complicated part for Waikato would be how to cleanly manage itself out of its existing technical and customer relationships. It had choreographed everything for a seamless handover at the end of 1995 but then Clear seemed to back off, and it appeared for a time that Waikato had acquiesced and simply handed over everything to Telecom as sole wholesale provider. In fact Waikato didnt sell anything to anyone. Clear had planned a concurrent entry but got the jitters. It was only through strong persuasion from Houlker and his associates that it was back in the game, although a month behind Telecom.
The shift to Telecom and ultimately Clear was anticipated as a natural evolution. The complicated bit was how Waikato would phase out the research and education network and the provision of ISPs without being left with a debt. Houlker admits he should have kept a closer eye on that commercial risk. In the end a handful of ISPs decided not to pay their last bills. They never settled their debts with the universities when they moved across to the new commercial services, and some of those debts were quite significant. Fortunately we had enough of a buffer to cover it but we were a little surprised.
The other unfortunate challenge was exiting responsibility for the third circuit which Waikato had established with Australia. This had initially gone in to the University of Melbourne in 1993 and then moved to a university exchange in Sydney. While the trans-Tasman frame-relay connection had hugely improved the flow of Internet traffic Australia had never paid for its share. Attempts to recoup the cost proved futile, and further complexity was added when the US backbone was privatised. Houlker said the Australian gateway router didnt have sufficient memory to view and process the complete routing table of New Zealand traffic, and it was never upgraded even when this clearly presented a problem.
Traffic from some New Zealand ISPs continued to route there even after Telecom and Clear took over the gateway. Australia had applied illogical decisions about routing which meant they couldnt see the detailed routing decisions being made in New Zealand which caused some traffic to flow over that line when it shouldnt have. Clear didnt have an arrangement with Australia but some of its ISPs continued to go direct through Australia. We attempted to bill those companies but only a few paid up. We had no way to stop them even after we were no longer operating the circuits and had vacated the scene. We got caught in the middle and took a hit on some of that. We could have blocked the routes but then a lot of companies would have lost their connection to Australia which would have been too draconian, said Houlker.
When Telecom and Clear took over, the AT&T and Sprint circuits were still going directly into NASAs FIX West gateway. Soon after, there was an exponential growth in private and commercial sector traffic. Telecoms subsidiary Netway Communications began to take a more active role in managing the provision of bandwidth, adding a second 512kbit/sec circuit through US carrier Sprint to complement the existing frame-relay service provided by World Partners, an AT&T and Telecom consortium. Clear Communications managed its own pathways to the United States.
Waikato University, relieved of the task of bandwidth provision, continued facilities management of the equipment installed at its premises, while the new players sorted out how they would manage their new gateway functions as wholesalers to the nations ISPs. A remaining challenge was maintaining harmonious interconnection arrangements with ISPs and network providers so New Zealand Internet traffic didnt end up bouncing around the United States, for example, before it reached the browser or email box of the intended user who might be across the street.
Waikato continued to provide advice on ways to encourage competition and improve both pricing and quality of service. Under the new deal the universities would buy capacity as if they were commercial ISPs. The prime concern during the handover was that no single company could take control of the Internet in New Zealand. There was understandable concern, particularly among the existing service providers, that the carriers would soon launch their own ISPs as they had in other countries.
ISPs push for fairer deal
Meanwhile the winds of change were blowing through the fledgling ISP market, which was rapidly evolving with Voyager far and away the strongest competitor and CompuServe, Iconz, and Ihug battling behind. ISPs had to lift their game to provide a better level of service to customers and sort out their arrangements with Telecom. Everyone was waiting for the carrier to make its move. The gigantic splash came when Telecoms Xtra jumped into the pool among the 50 or so other ISPs in May 1996, claiming 10,000 users within four months, and changing the whole dynamic.
In August 1996 Xtra cut its rates by 50 percent and its 0800 access prices to $1 an hour below Voyagers. Xtras entry price was below other ISPs and had a dramatic effect, with many smaller regional ISPs unable to compete and shutting down. IBM and CompuServe didnt even try to keep up; their international infrastructure and minimal local points of presence no longer gave them an edge. Voyager lost significant business during this period. It was unable to meet the new pricing regime due to its cost structure, which it complained was largely dictated by Telecom, from whom it purchased bandwidth and an 0800 service.
The water was further muddied when Clear signed a $2 million contract for dial-in modems and routers and several million more in establishing 15 points of presence (POPS) around the country. When it launched its ClearNet ISP in November it made no pretence at creating a separate branding; it was obviously a business unit of Clear with a strong focus on business customers. ClearNets pricing strategy was different to Xtras and based on peak service demand times ($5.95 an hour); its 0800-access service was unable to compete with Xtras toll-free pricing. However it quickly captured about 17 percent of the market.
Plain Communications managing director Robert Hunt ended up in conversation with Mark Frater from ISP Quicksilver and both agreed it was absurd that the individuals at the top of each ISP organisation didnt have a common body representing their interests. We thought if all the ISPs were to get together in a room we could push for a fairer playing field in the industry, Hunt said. A major catalyst was what he and others saw as targeted anti-competitive behaviour by Telecom and Xtra against the rest of the ISP community. The pressure on ISPs was extraordinary and the way Telecom was doing the wholesale side of things was unbearable. It was way beyond ridiculous. He claimed Telecom had allocated Xtra an 0800 dial-up service at a much lower rate than anyone else could get access to. Such concerns resulted in the formation of the original ISPANZ in late August 1996.
ISPANZ was not only seeking a better deal with Telecom; it was seeking a more constructive market overall as ISPs generally didnt feel their interests were being represented by the newly formed ISOCNZ. The fact that it had established a monopoly register with its company Domainz didnt help either. We saw the need for ISPANZ to be clear and distinct from all of that. While the Internet Society had some mention of ISPs in its charter it seemed their main focus was on the public or users with an absolute minimum of consultation or any kind of involvement by the organisations around New Zealand who were supposed to be delivering Internet to people, said Hunt.
Although the computer press reported about 90 percent of ISPs were members of ISPANZ, a large number, in Auckland in particular, did not join, either because they werent clear about the objectives, or felt it was a Telecom-bashing group. Then Voyager took Telecom to court over the 0800 access issue and got a last-minute backroom settlement. That signalled the end of the line for ISPANZ as Voyager pulled its membership. Things became very divided and it was clear we werent going to get the kind of unity we needed after that, said Hunt.
The Commerce Commission determined at the end of the year that Xtra was getting the same rate for 0800 calls as all other ISPs.
A cat or a fox?
In June 1996 the government reaffirmed its reliance on general competition law to achieve its objectives in telecommunications interconnection deals, which it said should be based on terms that would promote efficiency and deliver the benefits of competition to consumers. It also made it clear that it did not accept the Baumol-Willig ruling that Telecom had tried to push through the courts.
Although the Privy Council had claimed this rule, which would have compensated Telecom for losses made through competition, was not anti-competitive, the government opposed its use but failed to suggest an alternative. Subsequently, Clear and Telecom agreed on an interconnection agreement that served as a template for all other interconnection agreements. It was neither based on the Baumol-Willig rule nor cost-based. Arguably, access seekers were from that point reluctant to dispute interconnection pricing in case the Baumol-Willig rule was applied by the courts. In other words, in the absence of specific legislation requiring some other pricing principle to be applied, the courts might be obliged to apply the onerous Baumol-Willig rule.
Telstra New Zealand had been granted a licence to operate international services to local business in March 1996 and by mid-year had entered discussions for interconnection agreements with Telecom and Clear for national, international, 0800, and other services. The agreement with Telecom took until November 1997 to complete.
An old Arab saying: my enemys enemy is my friend was an apt way for Telstra New Zealand CEOs Peter Williamson to describe how he would compete here. He would start by winning ten percent of the $3 billion New Zealand telecommunications market over the next three years, be extremely price competitive and have a complete range of services. Telstra planned to be in the bandwidth re-sale market and use tails off its nationwide private switched network based around the massive $8 million Ericsson AXE switch installed in Auckland, along with multiplexing equipment, to provide bandwidth on demand. The switch and ancillary equipment were capable of not only delivering voice and leased line services but also ISDN. Williamson gave every indication Telstra would enter the market by stealth. There are more ways to skin a cat than are known to most people.
Telstra began focusing on the corporate market, with wholesale Internet offerings and plans to build fibre rings in Auckland, Wellington, and Christchurch. Williamson insisted the Clear-Telecom debacle would never have occurred in Australia, where interconnection rates were determined from day one. He likened the situation to leaving the fox in charge of the chicken coop. The company set up offices in Auckland, Wellington, and Christchurch, employed 60 staff, and had 1000 retail customers and four resellers of its bandwidth and services. Its main focus was large businesses. It was rapidly adding points of presence in Wellington to create a data backbone and had the ability to switch data between Auckland and Wellington with full access to Sprints frame-relay service. It carried large volumes of wholesale Internet traffic and a business product that gave corporate customers guaranteed access times.
Telstra New Zealand made a further $30 million investment in 1997 to achieve its aims. It just missed its targets for the first financial year but won a lot of trans-Tasman business. It expected to be profitable by year three and was benefiting from volume discounts by leveraging the best network capacity deals with Clear or Telecom. In effect it was now Telecoms second biggest customer behind Clear.
Telecommunications exports had surged ahead, major investment had been made in the economy by the big carriers, phone and cellular calls were cheaper, the customer premises equipment (CPE) market had opened up, international call-back operators were offering up to 50 percent discount to major destinations, and third parties were reselling everything from bandwidth to phone cards and satellite space. There were clearly identifiable benefits from deregulation but for those trying to deliver data services, and for frustrated business customers including ISPs, the market was nowhere near open enough. Research group Ovum declared in late 1996 that Telecom New Zealand had the highest interconnect fee in the world, at 3.37 cents per minute. In the highly regulated and price-capped United Kingdom, BT got only 0.78, Bell Atlantic (under a price cap) was getting 1.28, and in Australia Telstras fee was 1.85.
Clear had operated with a toll bypass agreement, including an 050 prefix, for four years. It had sought local service access through the courts in 1992 but Telecom took until September 1995 to agree and March 1996 to sign it off. A year later the agreement had fallen apart. Clear wanted to renegotiate after Telecom launched its $5 unlimited time weekends. It had been paying Telecom on a per-minute basis for access to its network and in protest stopped paying a portion of its bill. So it was off to court again. Clear and BellSouth had also rejected a number portability deal, which had meant Telecom would continue to be paid per call, even after customers had left its network.
The legal wrangling over access rights and conditions remained painfully prolonged, largely because the Commerce Act, administered by the under-resourced Commerce Commission, was seen to have no teeth and the courts no independent guidelines. The Act often relied on court action to determine an outcome which had in itself bogged down the process of competition. BellSouth and Clear were backing a call from TUANZ, which wanted the government to adopt the telecommunications principles and guidelines that Communications Minister Williamson had been threatening to introduce for some years. TUANZ chairwoman Judy Speight said a clear set of principles and an up-front agreement on what was required for interconnection were needed: (the government) should be driving those stepsthe Commerce Act needs more backbone.
Meanwhile Australia, which had entered full deregulation in July 1997, was taking a page from the New Zealand experiment but with more serious consequences for those who stepped outside the boundaries. The Australian Competition and Consumer Commission (ACCC), the general trade practices regulator, was given a role within telecommunications, signaling a shift from industry-specific regulation to general competition law. If a carrier was caught acting anti-competitively it could be fined up to A$10 million and A$1 million for each day the conduct continued. It had also pre-determined how interconnection would occur and what fees would be paid.
Despite five other carriers being registered as telecommunications companies, and 11 as international operators, the same delays evident from day one of new Zealand deregulation persisted. British Telecom had taken 100 percent control of Clear in 1996 and country manager Stephen Wilks quipped the Kiwis were the laughing stock of the industry, claiming Telecom had simply priced any competitor out of the market with high infrastructure fees. Between mid-1996 and 1997 there were 96 complaints and only one of them was investigated. Thats an absolute joke. You could not get a more damning indictment.
New Zealand had been assured it had all the international communications capacity it would ever need with PacRim East, PacRimWest, and Tasman 2 fibre-optic undersea cables linking us to the rest of the world. By mid-1997 this capacity was all booked up largely due to the growth of the Internet. Telecom was partnering with Australias Optus to build a new $1.44 billion big pipe across the Tasman before the Americas Cup, and the other carriers were seriously looking to extend their international capacity. Telstra Australia admitted it was already carrying more data than voice traffic on its network.
After Telecom was sold to former Baby Bells Ameritech and Bell Atlantic, we were told we now had a 98 percent digital switched network and that the days of exchange overloads between the major cities were over. However the once familiar Im sorry but this call cannot be connected right now because of overloading … was again commonplace and the Internet the darling of the information age had already experienced its first brown-out in New Zealand.
By mid-1997 the exchange at Waikato University was using 14 ports, with 10Mbit/sec and 100Mbit/sec capacity, on a 1.2Gbit/sec backplane and all but one were in use. In May New Zealand lost international access to the Internet for eight hours when Telecom-owned Netway Communications 512kbit/sec frame-relay connection into the United States went dead. A server in the United States crashed, and there was no alternative route to redirect Internet traffic on to its next destination. The Internet exchange server at Waikato University became overloaded and also crashed, leaving all outgoing traffic hanging in cyberspace. Netway promised it would add three more servers as alternative international routes for the local backbone.
Research group IDC declared one-third of New Zealanders homes had a computer, and there were around 175,000 ISP subscriber accounts, suggesting this would rise to 260,000 by 1998 and 425,000 by 2000. We had the seventh-highest number of Internet hosts or permanently connected Internet sites per 1000 people. Our Internet uptake had been growing much faster than most of the world. Whatever else that meant it signalled huge pressure on the New Zealand Internet Exchange (formerly NZGate) at Waikato, where the bulk of Internet traffic come into and out of New Zealand. Technology working groups were looking at how to build efficiencies into the nationwide backbone, including better management of traffic routed from the ISPs to ensure it took the most direct route to its destination.
In July 1997 the Internet Exchange (NZIX) at Waikato was upgraded from a single Cisco CAT 5000 to two fully redundant Cisco WS-2926 switches. Clear pulled its MCI circuits back to Auckland to avoid tromboning international traffic to Waikato and back. Clear, Telecom through subsidiary Netway Communications, and Telstra NZ were finally reaching some level of agreement on a peering exchange in Auckland. A triangle of 2Mbit/sec circuits was provisioned between the three telcos as an interim measure.
The proposal, initiated by Waikato and Auckland universities, was that exchanges be built to ensure more accurate routing of Auckland traffic from the carrier to the intended ISP. Wellington City Councils CityNet fibre-optic network, linking buildings in the central business district, would provide a central hub for routing Internet traffic in the capital. Such exchanges kept local traffic within a city where bandwidth was cheaper and also provided redundant paths so there was no single point of failure.
Roger Hicks, Clears representative at the peering meetings, said the talk of routing protocols using ATM circuits in Auckland was akin to the establishing of five Network Access points (NAPs) throughout the United States, which were now run by the major phone companies. Pacific Bell in California, Sprint in New York and New Jersey, Ameritech in Illinois, and Metropolitan Fibre Systems in Washington were routing all the US and European traffic but due to unprecedented demand were already choking up. Now numerous peer-to-peer arrangements between ISPs were springing up to ease that congestion. It was logical and practical, he said, that New Zealand carriers work towards a robust solution before our own network drowned in data.
There had been backdoor paths between ISPs for some time; several of the larger ISPs including Voyager, ClearNet, KC Internet, and Ihug had their own alternative international routes. A consortium of ISPs, including Internet ProLink, Iconz and KC Internet had already established virtual links to keep local traffic local. Internet ProLink director Craig Anderson welcomed the move by the telcos to establish an exchange but said virtual exchanges became too expensive and unmanageable, particularly when more than three players were involved. It would be ridiculous and very expensive for the 30 or so ISPs in Auckland to have links to each others sites. Everyone would prefer a real exchange, managed by an independent body in a basement somewhere in downtown Auckland.
Maintaining the high ground is everything when you are competing for coverage. A snapshot of several floors in Aucklands gleaming finger to the sky tells the story of how things rapidly evolved in 1998. The 328 metre Sky Tower hosts all manner of media with 360-degree coverage of Auckland city and inner suburbs. The top 93 metres is set aside for microwave dishes, transmitters, telecommunications, and broadcasting equipment and FM, VHF, and UHF broadcasting. Competing telecommunications, television, FM radio, communications, courier, and taxi companies had staked a claim on prime sites.
TVNZ, TV3, TV4, Telecom, Clear Communications, Telstra, Broadcast Communications, broadband wireless providers, and BellSouth were there, and Ihug saw it as the ideal site to supplementing its satellite coverage with broadband wireless across Auckland. It was also the ideal location for deploying wireless loop technology. Internet gateway manager John Houlker was a fan of the spread spectrum radio solution and believed the iconic location would be ideal for an independent Auckland peering exchange (APE). I had set up a wireless hub at Waikato University at the time and even had an 11Mbit/sec link running to my house. Locating an exchange in the Sky Tower opened up all sorts of possibilities for wireless.
New Zealands main connection into the international Internet backbone continued to come through NASAs FIX West gateway until Kawaihiko, the de facto science and research network, shut down at the end of 1998. Not coincidentally, Houlker, essentially having done himself out of a job, moved across to work for Telecom International Networks. His first project was to establish new landing points for its international circuits in Los Angeles and at the Palo Alto Internet exchange.
Anyone for seconds?
Telecom was doing a roaring trade in second telephone lines into homes and small businesses. In excess of 40,000 homes had taken up the option and about three times as many businesses, largely due to their desire to run Internet or fax machines while on the phone. At this time about 30 percent of New Zealand households had at least one PC. That was about on par with Australia, slightly behind the United States, which had 40 percent penetration, but way ahead of most of the world.
While IPNet might have taken some of the pressure off the voice network, Telecom was still creating sparks across the industry with its attitude, steep pricing for ISDN if you could get it and favouritism shown to its ISP arm. Xtra was without doubt the leader in the ISP market with 45,000 customers by the end of 1997, but the industry was pretty miffed about the special treatment it was getting from its parent company. Xtra boss Chris Tyler had openly stated it was getting a better price on bandwidth than other ISPs. The Commerce Commissions decision that it was within its rights to do so did little to alleviate market frustration.
According to a 1997 survey by the Australian-based telecommunications analysts NUS, New Zealanders were paying too much for their phone calls, and our phone system was lagging behind the rest of the world. While toll charges had dropped dramatically over the previous five years the cost of a phone call to New York, for example, had fallen 52 percent since 1992 we still had the most expensive international toll calls and the third-most expensive domestic calls in the entire OECD. The Ministry of Commerce concurred in its briefing paper:
In the continuing debacle over ISDN tariffs, Communications Minister Maurice Williamson issued a please explain notice to Telecom over its consistently high pricing. Its response wasnt satisfactory, and the minister suggested it had better come back with a better answer. This is a perfectly legitimate example of where were out of kilter with international comparisons. Theres no reason for it, and we kill our chance of international competitiveness if high-speed data isnt made available to New Zealand businesses, he said.
Williamson was great at making the right noises about new technology, but less willing to make a fuss when the market asked where it was or why it cost so much. Williamson can pour forth details of prices that have fallen, services that have improved. Usually these entail comparisons between the horse and buggy days when Telecom was a state monopoly, and the present day, said Listener writer Gordon Campbell. Moreover Williamson gives big ticks to the deregulated system that he oversees in New Zealand. Yet prices have tumbled worldwide in recent years, thanks to massive technology advances.
To many observers New Zealand has seen benefits in spite of, not because of, the system. Given the nature of telecommunications and the need for competing networks to interact with each other every other country in the developed world has some kind of regulator to ensure competition remains open and fair. Every other country sees danger in a situation where an incumbent can effectively set the terms and the pace at which its rivals can compete.
Slower than molasses
While the carriers were holding regular meetings on how to interchange Internet traffic, the old bone of contention, interconnection agreements, and an unwillingness to expose intimate details about each others networks meant things were moving like molasses in winter, to regurgitate a Maurice Williamson classic.
Call waiting and caller ID, which had been available in other markets for four to five years, were only just being introduced. Telecom had insisted on charging enormous fees for interconnection between networks and had released its controversial number portability proposals in 1997, which let people keep the same number if they changed networks. Telecom chairman Roderick Deane was naively optimistic a $30 entry charge and .5 cents a minute thereafter would be acceptable. Number portability was supposed to have been sorted at the time Telecom was sold. The offer, a decade on, simply added salt to the wound, but Telecom was in no mood for criticism. It reacted sharply to the howl of protest from the media and various industry groups. It felt it was being picked on and withdrew its membership of TUANZ because of a perceived bias.
ISPs were feeling like second-rate customers of the carriers, at a time when they needed to lift their game to meet increasing competition and the demands of their own customers, who were screaming for better service. In February 1997 Ihug announced it had been forced to stop signing up customers in the South Island due to delays in provisioning extra phone lines, a fault in circuits provided by Netway and a lack of national bandwidth. Delays over the year have caused a loss of around $200,000 in initial earnings, with unrealised income from lost customers running even higher, we have calculated a figure close to $500,000, claimed Ihugs Tim Wood.
Ihug had investigated options for satellite and wireless Internet delivery and became the first ISP to deliver broadband services. Through an arrangement with PanAmSat it began fast Internet services to 30,000 users in New Zealand and then ventured across the Tasman with its breakthrough approach. It was now transmitting 85 percent of its Internet traffic via satellite, with two 34Mbit/sec satellite feeds into Napa Valley in San Francisco to earth stations in Auckland, Wellington, Christchurch, and Dunedin and to Sydney, Melbourne, Brisbane, and Adelaide. Voyager installed its own satellite dish for better international links. Iconz , which already had videoconferencing and streaming video and was talking about voice over the Internet, brought three new Cisco boxes which each took 120 phone lines and announced grand plans to become a pseudo carrier. However by September 1998 Ihug and Xtra, unable to keep pace with the dial-up revolution, had booked up all new primary rate ISDN lines out of Aucklands Mayoral Drive exchange for the next six months. In fact the situation had become so dire Telecom had to stop accepting any new ISDN customers because it simply couldnt keep pace.
The carrier squabbles continued to slow the rate of competition but the pace of technology hadnt eased. Before ISDN had gained any serious grip in the market, alternative technologies were looming. While distracted for a time with plans to become a cable TV and HFC network provider, Telecom had been evaluating another technology that seemed promising.
It was thought from March 1998, depending on the success of Wellington trials, that Telecom and NEC would begin installing new DSL switches at exchanges around the country. Business would be targeted first but residences close enough to the business exchanges would also have access. The new technology, everyone hoped, would slash data charges for existing dedicated lines and ISDN customers. Among the services on offer would be pay TV.
I dont see much of a future for ISDN. Its too expensive for us. We try and pass it on to our customers at the best price we can but it costs us a fortune. These new solutions will offer much better price performance, said Telecom group general manager of broadband strategy, Vaughan Smith. Weve got a lot of increasingly frustrated customers out there because there just hasnt been a good solution at an affordable price. We think this will be fixed within the next 12 months. Were very optimistic. Were going to deliver a broadband network to as many people as we can.
Telecom spent about $810 million in 1997 on new plant and equipment, including extending its fibre-optic and cellular networks and a contract to future-proof its entire network. According to Telecoms broadband and interconnection manager, Alan Holden, the current network would eat itself. The old copper in the ground was coming to the end of its use-by date and modern business requirements needed a new fabric a more flexible way to distribute and manage voice and data traffic. To that end Telecom invested $50 million in high-speed switches and ATM, a cell-based switching network, seen as the way of the future.
The decision to go with ATM came after a three-year trial across the countrys academic backbone where Telecom sought to learn as much as it could about broadband networking. The Opera Network run by Telecom subsidiary Netway Communications in conjunction with the Tuia Society also provided vital information about the use of DSL and its performance over the copper wires of Telecoms network.
Telecom signed a deal with Newbridge-Siemens in May 1997 for an ATM network capable of 155Mbit/sec622Mbit/sec speeds. Several months later further investment was made to boost the capability to data speeds up to 20 gigabits per second. ATM, according to Telecoms Holden, was the beginning of the end for Telecoms old infrastructure, which would gradually be assimilated into the ATM umbrella. ATM is the beginning of something quite big. It is the common transport switching system over which a large number of services are supported. Previously you had a lot of dedicated networks to service. Voice would be the last to move across.
Businesses with big bandwidth requirements wouldnt be starved for choice as rival Clear already had an operational nationwide ATM network for 18 months. A number of customers including universities and ISPs were using its Stratacom switch to access speeds up to 155Mbit/sec. Clear planned to spend a further $14 million broadening the coverage of its fibre loops in the main cities. The Museum of New Zealand, Auckland University, Fisher & Paykel, WCC, and Mercury Energy were just a few of the locations that already had successful ATM networks running to meet future communications needs. Telstra was also planning to build an ATM network.
In November Telecom further reduced charges for services including ISDN and leased lines, resulting in data costs coming down between 20 and 50 percent overall for the year. Clear responded in kind, announcing it would launch a basic rate ISDN service in 1988, drop its business analogue line rates to below the Telecom offering, and publish its ATM rate card. Finally the pressure was on to bring down the cost of business bandwidth, which would ultimately be fought at the high end, using fibre-optic networks. Clear would offer bundled services based around 65Mbit/sec circuits. Telecom, however, claimed it was only interested in selling full-blown 155Mbit/sec ATM, and Graeme Rowe, Telecoms marketing manager for connectivity and computer communications, believed the market was too immature to set tariffs.
While Telecom was saying leased lines remained more affordable for the bulk of businesses, Jane Hindle, Clears manager of data services, insisted ATM was more affordable. A lot of businesses do not use their dedicated circuits all the time and want a service that gives them bandwidth when they need it and for only a bit more to have a guaranteed bandwidth availability By making the rate card public, ATM will no longer be viewed as an emerging technology but a real service that can be tailored to meet customer requirements.
The 20Gbit/sec boost to Telecoms ATM fibre backbone, the 26GHz frequencies it won at auction, and the 1Mbit/sec50Mbit/sec capabilities of the DSL overlay to its aging copper network would according to the hype, finally give New Zealand a fast lane on the information superhighway. As usual time soon proved that these sunrise technologies and the promised benefits were a lot further off than originally promised.
Wham bam, thank you, ma’am
It seemed the market was always crying out for more, but when questioned about why this wasnt forthcoming Telecom would counter by saying it would deliver the appropriate services when the market demands. Someone wasnt listening, or perhaps the industry wasnt making a loud enough noise to stir the government into the action it had promised time and again if the big boys didnt start behaving themselves and learn to share the sandpit.
An analysis of the market in November 1997 by Paul Budde Communications stated our telecommunications services cost too much, with Telecoms dominant position allowing it to dictate terms and conditions for competitors and raise prices, instead of bringing them down as in other deregulated economies. The fact was, without regulation or any government restraint, Telecom had acted like any business, fighting tooth and nail to keep its customer base, and in the end putting its shareholders ahead of its customers. It could easily have complied with all its competitors requests but Telecom chief executive Rod Deane was not being paid $1.2 million to give away market share.
Telecoms reluctance to offer more broad-ranging services at better prices by suggesting the demand wasnt there was purely and simply a snow job. It was obvious that Telecom would continue to stall, delay, debate, and argue with Clear, BellSouth, Telstra, and anyone else over commercial issues until it got the deal it wanted. In November 1997, just as real competition seemed to be emerging, both Baby Bell investors announced they wanted out.
New Zealand Herald financial columnist Brian Gaynor was among those miffed that the Americans had not only made huge profits from their short sojourn, but got away with under-investing in the core infrastructure. On close examination of the figures over their tenure he claimed the two American giants had not made a significant contribution to Telecoms profit growth. In the end the sale of Telecom to overseas interests probably had a negative rather than positive impact on the New Zealand economy.
In the modern world capital is easily transferred; it does not have emotional bonds. When overseas investors have maximised their profits, or found more attractive horizons, they get up and leave. The two North American telecommunication giants will have fond memories of New Zealand. They bought Telecom for a song in 1990 and have made a financial killing from the transaction. In September 1990 the New Zealand Government sold Telecom for $4.25 billion. The buyers were Ameritech (45 percent), Bell Atlantic (45 percent), Sir Michael Fay and David Richwhite (5 percent), and Alan Gibbs and Trevor Farmer (5 percent).
While it was obvious Telecom had no moral imperative to pursue public good objectives, the government, which is in the end answerable to the people, certainly did. It was responsible for ensuring communications technology is an engine of economic growth and job creation and a powerful tool for educating our children and expanding access to health care.
After nine years of non-interference, the chief architect of deregulation, Richard Prebble, finally chipped in, admitting the government was supposed to be involved in the process. I never imagined that the Crown would not be a party. He said litigation was a key element in his strategy, but Maurice Williamson had ignored his urging to have the attorney general join court action. As a result the courts didnt understand the Kiwi Share, put in place at Telecoms request to ensure equal access and pricing for residential customers regardless of where they lived.
What bothered many taxpayers and ultimately Telecom customers, was the years of rhetoric and unrealised consumer benefits, particularly in relation to data services. New Zealand was supposed to be the envy of the telecommunications world with its fully digital network, privatised phone company, and no rules. Instead of becoming a wired Utopia our national competitiveness was compromised in the vain hope that the free market would somehow create a beneficent pocket of self-regulation the world could admire. We thought we were special. We thought the world was watching and it was, but for other reasons. We were the guinea pig, a test case for telecommunications deregulation.
We had been subjected to the commercial equivalent of rape and pillage through the wham, bam, thank you, maam methods of multinational investors while the government stood on the sideline, determined to keep hands off. We thought we were opening our doors to competition and investment but got shafted.
 Keynote speaker at TUANZ conference, August 1993
 According to Statistics New Zealand average residential phone call pricing plummeted 50% between 1987 and 1993
 Paraphrased from a feature article for Communique Magazine, August 1993 by Keith Newman
 Keith Newman, Changes to WWW charges in pipeline, Network World, 17 February 1995
 Donald Neal, The Harvest Object Cache in New Zealand, Information & Technology Services, The University of Waikato
 Chris Barton, The Kiwi Connection, New Paths to The Net, PC Magazine, December 1995
 Christina Enright, Strategic Behaviour of Internet Service Providers in New Zealand and the Performance of this Market, March 2000: Updated document now at http://www.iscr.org.nz/f241,4849/4849_ausnz_020600.pdf
 Keith Newman, Competition at last?, PC Magazine, September 1996
 Adrienne Perry, Telstra to spend $30m in NZ, The Dominion, 11 August 1997
 Keith Newman, Regulation by litigation, LAN Magazine Australia, July 1997
 Keith Newman, Internet relief well overdue, PC Magazine, June 1997
 Keith Newman, Internet relief well overdue, PC Magazine, June 1997
 The proposal was shelved for a time but went ahead in 1999 with Wellington CityLink locating equipment on the 48th floor of the Sky Towers and an equivalent exchange established in Wellington (WIX)
 Keith Newman, Anyone for seconds, PC Magazine, June 1997
 Aardvark News, 26 February 1997
 Keith Newman, DSL trials, NZ Herald, November 1997
 Keith Newman, Telecom tests broadband, NZ Herald, 12 August 1997
 Keith Newman, Rewiring the nation, unpublished feature, 16 July 1997
 Keith Newman, Telecom prepares for ATM Network, NZ Herald, 20 May 1997
 Clinton-Gore, October 1997
 Revised from Keith Newman, The rhythm method of competition law, Metro magazine, July 1998