Telecommunications: The Way Forward

Issue 3 | 10 June 2009

Paul Fletcher is Principal at Fletchgroup Advisers. Before this he was Director of Corporate and Regulatory Affairs at Optus, and worked for a time as Chief of Staff to the Minister for Communications, Information Technology and the Arts, Senator Richard Alston. Paul Fletcher has been involved in key changes and developments in telecommunications over a decade. He is also the author of Wired Brown Land: Telstra’s Battle for Broadband? (UNSW Press, 2009). Paul Fletcher addressed The Sydney Institute on Wednesday 10 June 2009

TELECOMMUNICATIONS – THE WAY FORWARD

PAUL FLETCHER

It is a great pleasure to be speaking at The Sydney Institute. I am a great admirer of its work and of its founders Gerard and Anne Henderson.

It is also a pleasure to be speaking with Henry Ergas, who taught me a great deal about telecommunications economics and regulation – although he is no doubt disappointed that I have rather gone off the reservation since those days. Undoubtedly he is “one of Australia’s best known microeconomists” as the flyer for tonight’s event states – and isn’t it a remarkable sign of how our country has changed that such a description can be used. Thirty years ago the notion of a “well known Australian microeconomist” would have been met with blank stares.

While I agree with much of what Henry has said tonight I want to focus my remarks on the areas of disagreement. Let me start by observing that the Rudd Government’s announcement in early April, of its plans for a national broadband network, was quite remarkable – for three reasons.

Firstly, the announcement committed the government to a $43 billion project to build the network. This compared to its previously stated position that it would spend $4.7bn on the network – consistent with the policy which Labor announced in March 2007 and took to the 2007 election.

Secondly, the government was announcing its intention to start up a new business – in competition with existing private sector players. This reversed the policy trend of the last decade – in which government has steadily reduced its ownership stake in telecommunications, and the amount of taxpayers’ capital at risk in this industry.

Thirdly, the announcement made it clear that the colossus of the telecommunications sector, Telstra, is facing very substantial regulatory intervention.

What has driven the Rudd Government to this remarkable decision – a decision which took most people in the telecommunications sector by surprise, and went considerably further than most commentators had previously thought was likely? And what does it imply for the future of telecommunications and broadband policy in Australia?

In my view the answers are to be found by looking at the saga of Australian telecommunications over the last two decades – and particularly in the last four years, since Sol Trujillo arrived at Telstra in mid 2005 and proposed that the company build a “national broadband network”. This is essentially what I have done in my book Wired Brown Land: Telstra’s Battle for Broadband.

In my brief time tonight I want to look at three issues. The first is the objective of Australian telecommunications policy – to drive vigorous competition and in turn deliver better services, lower prices and more innovation to consumers than they experience under the present market structure. As I will show, in fixed line telecoms and broadband the objective has been only modestly successful.

Secondly, I want to talk about the fact that in Australia as elsewhere we are facing a transition to a new broadband network, which uses much more optical fibre and delivers much higher speeds to end users than the current network. In every country this transition is raising tricky questions for public policy.

In Australia, Telstra has tried to use this issue to insulate itself against competition. Its attempts backfired – and its conduct over the last four years has in fact strengthened the political and policy case for more profound regulatory intervention than had previously been contemplated.

Thirdly I want to argue that the highest priority to drive competition – and in turn deliver benefits to consumers – is to split Telstra into two companies under different ownership. In the jargon of telecommunications policy, this is “structural separation”. In my view this should be done regardless of when and how quickly the new national broadband network is built.

Driving competition

There is one thing all parties in this debate agree on: a vigorous competitive market delivers the best results for consumers.

The results in the Australian mobile market have demonstrated this point very powerfully. Take up has risen relentlessly since competition came to the sector in 1992 with the arrival of Optus and Vodafone, and now mobile penetration exceeds 100 per cent of the population; prices have dropped steadily; and there has been continuing innovation in services, with first analogue AMPS, then GSM, then third generation mobile, as well as all kinds of value added services, content and so on.

This has happened because we have a genuinely competitive market structure, with until recently four separate operators each with their own network, and the largest player Telstra having only just over 40 per cent market share.

But in fixed line, by contrast, Telstra retains a market share of well over 70 per cent in most products. In my book, I argue that the root cause for this is that when competition was introduced in the early 1990s Kim Beazley, who was Minister for Transport and Communications, made a fundamental mistake. He put together two big companies, Telecom and OTC, and created a monster: Telstra. The new entrant Optus was a minnow – and it faced a very tough battle to win market share from Telstra.

Now, the government correctly recognised that given the network nature of telecoms, unless there was some regulatory support for the new entrant it would fail. So it established an “access regime”. The idea was that Optus could build some bits of its network and then rent capacity over Telstra’s network to fill in the gaps where it did not have network of its own.  For example, Optus built its intercapital fibre network, and it combined this with Telstra’s local network to allow it to offer long distance calls to end users. A very similar approach is used in most countries around the world.

When full competition was introduced in 1997, and Optus was joined by a whole bunch of other competitors in fixed line telecommunications, the access regime was continued. However, with the benefit of hindsight, it was deeply flawed, and too weak to allow substantial competition to flourish.

Even so, Telstra has always hated the access regime. That of course is because if it was not subject to an access regime it could make more money because competition would be even weaker than it is today. For example, Phil Burgess argued that the access rules were as absurd as Hungry Jacks being allowed to seize and resell a batch of fish sandwiches cooked by McDonalds:

If McDonalds … has to go to a regulator before it goes into business and declare what it’s going to do on its ovens … And Hungry Jacks comes across the street and says, I want to buy those fish sandwiches, by the way, at a discounted price and by the way, I get the first batch … and I’m going to sell a lot of them, and if you want more, you’ve got to build another oven and by the way, you build the oven at your own expense and by the way the cost of the air-conditioning, the gas and the electricity and the lights, those don’t go into the price.

If you believe in free markets as I do, why do you force a company like Telstra to rent access to its network to its competitors? The answer is that in fixed line telecommunications, the market is so dominated by one player that you do not get the benefits of competition. In the absence of regulation, that player could keep prices higher and quantities supplied lower than would be the case under a competitive market.

Indeed the arrival of consumer DSL in Australia has been an excellent case study of this. Telstra introduced consumer DSL in the year 2000.  But it managed the arrival very carefully. Telstra wanted to wring as much profit as possible out of the previous technology, dial-up Internet, before it allowed DSL to become a mass-market product. In 2001 Telstra had 1.1 million dial-up Internet customers, so there was a big revenue stream to protect. Hence it made sense for Telstra to charge a much higher price for DSL than for dial-up Internet when it first introduced DSL – to make sure that dial-up remained relatively attractive to many customers.

At launch, in 2000, the monthly charge for a 512 Kbps DSL service was $89 for a customer who took telephony from Telstra, and $105 for a non-Telstra customer. To give a sense of how high these prices were, in 2008, Telstra offered a service three times as fast for less than half the cost.

This was rational conduct for Telstra – but a bad public policy outcome because it meant that broadband take up was suppressed for several years. By the end of 2003, over three years after Telstra launched the service, there were only 241,000 consumer DSL services in Australia. In turn, for several years Australia’s broadband penetration was well behind most other OECD countries. In 2004 Optus entered the market however and prices dropped sharply – and in turn broadband growth rocketed through 2005 and 2006, with broadband services reaching 3.64 million by September 2006.

Given these facts, it is astounding that Telstra’s head of corporate affairs Phil Burgess went around complaining about Australia’s low broadband penetration. In my view this is akin to Robert Mugabe complaining about Zimbabwe’s high inflation.

Transition to fibre based broadband networks

So one factor in the Rudd Government’s recent announcements has been that fixed line competition has remained weak and Telstra has remained dominant. But another important factor has been the desire for Australia to upgrade from today’s telecommunications network, which still largely uses 100 year old copper wire technology to connect homes and businesses. Around the world these networks are being replaced with optical fibre networks – either going all the way to each home, in so-called fibre to the premises networks, or going halfway to the home, in so-called fibre to the node networks.

This offers many advantages. The new networks are more reliable and cost less to run. For consumers, they offer much higher broadband speeds. Today the practical limit for the copper network using ADSL2+ is around 20 megabits per second for those who happen to live close to a telephone exchange. A fibre to the node network can readily offer 50 Mbps and a fibre to the premises network can offer 100 Mbps initially and much higher over time. In Japan and Korea today, there are residential services at one gigabit per second. This is four thousand times faster than the 256 Kbps service which is sold as entry level broadband in Australia today.

In August 2005, shortly after Sol Trujillo arrived, Telstra proposed to the Howard Government that it would build a national broadband network offering a guaranteed speed of six megabits per second to 98 per cent of Australian premises. This was subject to a condition: that Telstra would be freed from the regulatory requirements to offer “access” to the network which I spoke about earlier. When the Howard Government rejected the proposal, Telstra came back with a new one in November 2005. It would build a network offering 12 Mbps to four million premises in the five big cities – around 46 per cent of all premises in Australia. Again, the offer was conditional on legislative change so that Telstra was freed from the requirement to offer access. Again this offer was rejected.

At the same time under Sol Trujillo continued to use its market power to the hilt. For example, during this period Telstra unilaterally increased the amount it charged to competitors for wholesale line rental by $5 per line per month – but made no change to the retail line rental price it charged its own retail customers. Another example was the way that Telstra continued to delay the arrival of high speed broadband services, over DSL, for as long as it could. So it kept a limit of 1.5 Mbps on all services until 2006, even when the first generation DSL technology it was then using could go up to 8 Mbps.

Telstra also refused to introduce the next generation ADSL2+ technology, capable of up to 20 Mbps, until 2008. The only exception was in exchanges where its competitors had already installed their own equipment so that they could deliver such a service themselves.

How did competitors do this, given that Telstra owned the copper wires connecting the exchanges to homes? They did so under a regulatory ruling called the “unconditioned local loop service” – otherwise known as “unbundling”. This let the competitors install their own electronic equipment in the exchange, and then rent Telstra’s copper line from the exchange to the customer’s home.

Unbundling was a bright spot in a landscape which otherwise struggled under the dull conformist yoke of the incumbent. As competitors increased their unbundled lines and expanded into more and more exchanges, a growing number of customers could now receive genuine alternatives to Telstra’s broadband products. There was real competition and choice on a number of dimensions: speed, pricing and download limits.

Customers could choose among attractive telephony and broadband offers from players such as Optus, Primus and iiNet. As one example, the Optus Fusion package offered customers broadband with a two Gigabyte download limit, plus unlimited local and long-distance calling and unlimited calls to Optus mobiles, for a flat $79 per month. Another example was iiNet’s offering of a “Naked DSL” service for $49.95 a month (that is, the customer did not have to also pay $30 line rental for a voice telephony service as Telstra required). Both of these services were much cheaper than equivalent offerings from Telstra. Today, prices have fallen even further.

Nevertheless, the most striking feature of Australia’s broadband market, even today, is the extent to which it is dominated by Telstra. In mid 2008, the total size of the consumer broadband market was 5.9 million services. Of these, 78 per cent of all customers were delivered by Telstra – by Telstra either directly at retail or indirectly via another retailer.

And this a product which did not even exist when Telstra was first exposed to competition in the early nineties. It has leveraged its dominance in other fixed line products into dominance in this new product. It is a powerful demonstration that the policy of seeking to achieve a competitive market in fixed line telecommunications has fallen a long way short of its objectives.

What Should be Done to Drive Competition

Which brings me to the third part of my comments today: what should we do?

The purist free market view says, if you want competition, get people to come and invest the money to build the networks and compete. If you give them a free ride on Telstra’s network that is what they will take; you will never see them actually investing and taking a risk. Henry Ergas has articulately argued this position tonight.

I certainly agree that, in a first best world, there are multiple network owners competing vigorously – and there is no need for regulation. But I believe that Australia is simply too large – and thinly populated – to support more than one fixed line telecommunications network. That is particularly so given the natural monopoly characteristics of the industry. In other words, the huge fixed costs and low variable costs of this capital intensive industry mean that the first entrant has lower per customer costs than subsequent entrants, and can therefore maintain its share and cost advantage.

This is not just a theoretical claim – we actually have the results of a real world trial in Australia. In the mid 1990s Optus spent $4 billion to build the hybrid fibre coax or HFC network to pass 2.2 million households in our largest cities. The network carries voice telephony, broadband and pay television. But it has never generated a commercial return – because it has been unable to win sufficient customers away from Telstra. Other potential investors will look at this example and conclude that it is not attractive to put their capital at risk to build a new network to compete with Telstra.

Therefore, if we are in a second best world where a new commercially funded fixed line network is very unlikely, what is the best way to stimulate competition?

Well, the current access regime was designed to do this. But after operating for twelve years, we can draw a clear conclusion – it is gravely flawed. And I say this in full acknowledgement of the fact that I was an adviser to Minister Alston when the regime was devised and introduced. I might add that Professor Ergas was also a frequent visitor to the Minister’s office – and we were always pleased to get the benefit of his advice.

There are two fundamental problems with the current system which need to be fixed up.  Firstly, the current system for setting access prices is far too slow and legalistic. Second, because Telstra is vertically integrated, it will always treat its wholesale customers as second class citizens compared to its own retail business. Telstra owns the only access network in most parts of Australia, and its incentive is always to supply itself in preference to supplying its competitors.

Telstra does sell to its competitors today – but only because it is required by law to do so by.  To give you one example, Sol Trujillo refused to even meet with Optus CEO Paul O’Sullivan – even though Optus was Telstra’s biggest customer, spending over one billion dollars a year with the company.

The solution, as I argued at the outset, is the structural separation of Telstra. If you have the network owned by a company that only sells wholesale services, then its management team will be highly motivated to do just that. In turn that will spark vigorous competition at the retail level. I believe that if you do this, and hence you get the market structure right, then a whole lot of other prescriptive regulation becomes unnecessary. In other words, government can back off and let competition work.

I would like to see a “grand settlement” in Australian telecommunications regulation. One half of the settlement would be major intervention – by breaking up Telstra. The other half of the settlement would be to remove a whole lot of unnecessary regulation: things like the retail price controls, customer service guarantee, network reliability framework, operational separation, rural presence plan, integrated public number database – the list goes on and on. Some apply only to Telstra; the majority impose cost and compliance burdens on all telecommunications companies.

Now the Rudd Government’s position is that the national broadband network will be built by a company which is majority owned by the government, and no retail telephone company can have a controlling stake in it. In fact Minister Conroy has said the share held by any one retail telco could be capped at levels as low as 15 to 20 per cent.

So once the NBN is built, it will automatically produce “structural separation” – that is to say, the network will be wholesale only, and it will not be owned or controlled by a company which also has a dominant share of the retail telecommunications market.

However, in the short term, before the NBN is built, the government is contemplating additional reforms to Telstra. They have said in a discussion paper that these reforms could include the functional separation of Telstra. That would involve the network business being under separate management to the retail business, with separate staff, premises, IT systems and so on. It would sell to the retail business on an arm’s length basis, just as it would sell to other retail telcos such as Optus and AAPT and Primus and iiNet. In my view this would be a useful step forward, but it does not go far enough. The two businesses would still be under the same ownership. By contrast, under full structural separation, they are under different ownership.

Now some people question how structural separation can be achieved as a matter of law.  Surely if Telstra has been sold to shareholders on the basis that it owns both a network business and a retail business, there must be restrictions on government’s power to force Telstra to exit one of those businesses. Indeed, some argue that any such forced exit would entitle Telstra to compensation under section 51(xxxi) of the Constitution. That section authorises the Parliament to pass laws for the acquisition of property, but only on ‘just terms.’

One reason the government failed to accept any of the proposals from non Telstra parties to build a national broadband network, under the Request for Proposal process conducted in 2008, was a concern that those proposals involved an “acquisition of property” from Telstra. In turn the government feared that it would be exposed to the risk of having to pay compensation. In my book I propose that the solution to this problem is to require Telstra to split itself into two parts, by means of issuing to shareholders two shares in place of each share they hold now.

One share would give an ownership interest in the network business; the other share would give an ownership interest in everything else. This scheme would avoid the issue of ‘acquisition of property’ because at the end of the process Telstra would still be owned by the same people who owned it at the start.

Conclusion

We are clearly at an exciting time for the Australian telecommunications sector. The Rudd Government has proposed change of an unprecedented magnitude. In my comments today I have sought to explain how we got to this point – and what I believe the key priority should be from here. Delivering a superior broadband infrastructure is only part of the exercise; in my view we first need to get the foundations right by establishing the right market structure in which competition can flourish. If that is done, then I believe there is enormous potential for a new broadband network to deliver a step change in competitiveness, productivity and connectedness for our nation.