Telecom admits it is behind the times

By RICHARD BRADDELL

WELLINGTON - With remarkable candour, Telecom's submission to the Government's telecommunications inquiry admits that much of its network is not up to scratch.

It says it will cost at least $550 million to upgrade it to standards already regarded as unacceptably slow for internet traffic.

It concedes that its Neac switches, the exchange workhorses for the past 15 to 20 years, are outmoded, and it is looking at replacing them with new generation packet switches that will integrate voice with data.

But the problem goes much deeper - to the cables in its network which are often too old to handle internet traffic adequately.

How much of the network flunks the test is hard to determine from the pie charts Telecom provides. It is clear much of it is very old, with large parts of the cabling having been in the ground more than 30 years.

Age, it seems, is a problem in rural areas in particular.

"There is a high cost associated with maintaining, let alone upgrading, the rural network," Telecom said, adding that it would cost $230 million to make the network capable of serving all customers at 14.4k internet speed.

Modems that slow have not been sold for years, but even at 33k (not the current 56k standard), the cost more than doubles to more than $550 million.

In Telecom's view, that investment is not economic. The even higher cost of providing ubiquitous broadband access is another story. There are two ways of looking at Telecom's argument.

The first is that the quandary it finds itself in is the result of paying out much too much in dividends, often more than 100 per cent of reported earnings, when it could have been improving the quality of its network.

As Telecom admitted last week, "far-reaching renewal is needed. The investment task is substantial and it will take a long time to turn the ship around."

If that sounds like Telecom pleading its own shortcomings, then it has a counter-argument: much of the network may be old, but it is adequate for what it was designed for - the transmission of voice. No one had heard of the internet when the Kiwi share was drafted a decade ago, and, even five years ago, no one would have thought it would account for more than half the network traffic.

There is force to both points of view. Telecom's profitability in recent years has owed much to keeping costs under control, and that has not involved gold plating its network. There may be an element of overstatement in the upgrade cost, particularly given that Saturn built an entire broadband and voice network in Wellington for $250 million. But rushing in now could be money wasted. Alternative technologies such as wireless and satellite might do the job better.

Telecom is using the problems related with managing internet traffic to justify its separation from the Kiwi Share. Local residential voice traffic should stay free, but it deserves compensation for carrying local internet use since that accounts for more than half the traffic, and is growing.

Instead of making itself unpopular by imposing the kind of 2c a minute "internet tax" that has aroused discontent, it argues that charging should be invisible to consumers, and should be paid by internet service providers. They can work out how much to pass on to customers.

Clear Communications contends Telecom should subsidise free internet by paying interconnection charges to internet service providers on its network, just as in Britain interconnection revenue is split between the carrier and the internet provider.

That conveniently overlooks that in Britain local calls for voice and internet are charged on a timed basis, so it pays the carrier to encourage free internet service providers to get network time higher.

In New Zealand, the problem may well solve itself over time. Unless there is a dramatic improvement in dial-up speeds, most users will probably move to broadband services.

That will not happen until the price to consumers comes down dramatically, but no one has suggested broadband traffic should not be paid for.