Wed, 11 Mar 2015 - 22:00
Viewed 20 times

Speech to the ACCAN USO Forum


It is a pleasure to participate in this important forum to discuss the universal service obligation in telecommunications.

I commend the organisers on setting the ambitious but timely objective of ‘Rethinking the USO’.

In my former life as a senior executive of Optus, I spoke about the USO rather a lot.

So frequently in fact that we used to joke that people in Canberra thought USO stood for ‘usual submission from Optus.’

Now of course I bring a different perspective to the topic.

It is no longer my job to make the case for one particular company: instead my concern is that the USO does the job it is supposed to do, in the most efficient way possible.

And what is the job it is supposed to do? That is the first issue I want to cover today.

Next, I want to highlight some well known questions about the current USO arrangements.

Third, I want to argue that the previous government made USO policy considerably murkier. Finally, and most importantly, I want to raise some possible directions for reform.

What Job is the USO Supposed to Do?

Let me turn first to the job the USO is supposed to do.

In the Government’s 2004 review, the USO was described as a ‘regulatory safety net’ which underpinned the strategy of promoting competition in telecommunications:

An important element of the safety net is the universal service regime, which provides for access to standard telephone services (STS) across Australia, and particularly in unprofitable areas of rural and remote Australia.[1]

The original rationale for this policy was presumably that access to telecommunications is fundamental to participation in modern society.

But in regional and remote areas, where population densities are low and distances are great, it is very expensive to connect people to a telecommunications network.

So in the absence of policy direction, it is likely that people in regional and remote Australia would be charged a price which reflected the true cost of providing a service.

That would mean a high price – in some circumstances so high that the customer would not be able to afford to be connected, and in turn would be profoundly disadvantaged.

The most important practical component of the USO, and the one I intend to focus on today, is the obligation, presently borne by Telstra, to ensure that all people in Australia, wherever they reside, have reasonable access, on an equitable basis, to a standard telephone service – that is, a voice telephony service.[2]

In practical terms, Telstra gives effect to this obligation by connecting customers to its network, wherever they are in Australia, for a connection charge which is does not exceed a specified maximum amount.[3]

As my old friend Dr Phil Burgess used to say in his time as head of public policy at Telstra, this was the obligation the company faced even if the customer seeking the connection lived two hundred kilometres out of Alice Springs in an abandoned car.

To underline his point, Phil used to show a picture of the abandoned car in his standard presentation - and make the point that this particular connection had cost the company hundreds of thousands of dollars.

I should note that changes were made in 2011 to formalise what was a ‘reasonable’ request for a USO service to be provided to a premises. Since that time, for a USO service to be reasonably supplied, there has been a requirement for the premises to be habitable, secure, and to have a power supply.

A universal service obligation in some form has been a feature of telecommunications regulation going back at least as far as the Telecommunications Act 1975. This Act required the government owned monopoly telephone company, Telecom, to make telecommunications services available throughout Australia as far as it was reasonably practicable to do so.

Later, the Australian Telecommunications Corporation Act 1989 created a more specific requirement that the standard telephone service (STS) was to be made reasonably accessible to all people in Australia on an equitable basis, wherever their residence or business was located.

Of course mandating this requirement is one thing – working out how to pay for it is something else.

When all telecommunications services were provided by a government owned monopoly, this was relatively straightforward.

There was an internal cross-subsidy: loss making services in regional and remote Australia were subsidised out of profitable metropolitan services.

But the arrival of competition – authorised by the Telecommunications Act 1991 – made things more complex.

This Act provided that the newly renamed Telstra would bear the universal service obligation – but the new entrants Optus and Vodafone would be required to contribute to the cost of it.

This was achieved through the introduction of a USO levy, which was imposed on all three carriers, Telstra, Optus and Vodafone.

The purpose of the levy was to contribute towards Telstra’s net costs of providing the USO.

With the introduction of full competition in 1997, the USO levy was extended to all carriers.

Some well known questions about the USO

As a principle, the universal service obligation makes good sense.

But in practice, there are a number of well known questions about how the USO works.

What Are Its Real Costs?

The first question about the universal service obligation is how to calculate the true cost which Telstra incurs in delivering it.

The model initially used by ACMA – and previously the Australian Communications Authority – in an effort to determine the so-called ‘Net Universal Service Cost’ was widely criticised.

For many years setting this amount was highly politicised, leading to a situation where the use of a detailed cost model was largely discontinued, and the Minister of the day effectively imposed a negotiated settlement.

You need only look at the highly variable amount of this figure over the last fifteen or so years to see how arbitrary it is. In 1997, for example, it was approximately $252 million; by 2001 it was $299 million; and by 2008 it was $145 million.[4]

What About the Benefits?

One of the other problems with the formula initially used to determine Telstra’s USO costs – costs which were then recovered across the industry – was that it disregarded the benefits that Telstra may obtain from being the universal service provider.

These could include economies of scale and scope and, of course, the brand benefits which come from being seen as the ubiquitous provider of telecommunication services.[5]

The behaviour of incumbents in other countries suggests that these benefits are real and well recognised.

In Switzerland, for example, there was a competitive tendering process to award the status of universal service provider. The incumbent Swisscom bid for and won this right – and did not seek compensation or payment.[6]

Impact on competition?

A broader question about the universal service obligation is its impact on competition.

Telstra competes with Optus and Vodafone in mobile and with Optus, iiNet, TPG and many others in fixed services.

Over recent years, of the total USO subsidy of $145 million, typically about 40 per cent, or around $60 million, was paid by Telstra’s competitors.

So the USO arrangements require Telstra’s competitors to pay substantial amounts towards a company which is considerably larger and more profitable than any of them.

Telstra’s competitors have long argued that this arrangement is not consistent with a level playing field between competitors.

A related point is that Telstra’s competitors pay towards the cost of providing universal service – but they do not have the option of bidding to become the universal service provider themselves.

Apart from an unsuccessful trial in 2001, there has been no subsequent attempt to make the status of universal service provider contestable.[7]

With the passage of time, and the ever increasing share of mobile revenues as a proportion of total industry revenues, Optus and Vodafone have paid a rising share of the USO costs.

Which technology?

Another issue which has become more visible over time is that the USO arrangements have tended to entrench existing technologies.

The USO applies only to voice services, not broadband – and Telstra has generally chosen to supply it using a fixed line service rather than a mobile service.

While there is, strictly, no regulatory requirement to deliver the USO over the fixed line network, under the USO Telstra is required to deliver a standard telephone service and there is currently a requirement that the standard telephone service must support preselection.[8] To meet this requirement generally requires the supply of a fixed service.

The market has moved a long way in the last twenty years – there are now 31 million mobile services in operation, and only 9 million fixed lines[9], and the number of fixed lines has dropped by over 1 million in the last five years[10].

But even though this suggests a strong consumer preference for mobile as the means of voice communication, the majority of USO services are still delivered over the fixed network. While market offerings and consumer preferences have changed over time, the regulatory requirement has not been responsive to those changes.

The previous government made USO policy considerably murkier

I have argued so far that there are a range of issues with the USO arrangements that have been in place for many years.

But thanks to the previous government, the already murky area of USO policy became considerably murkier.

In June 2011, then Broadband Minister Stephen Conroy announced the creation of a new body, the Telecommunications Universal Service Management Agency, or TUSMA, to “administer the universal service obligation (USO) and other public interest services.”[11]

This occurred on the same day as the announcement that Telstra and NBN Co had signed the Definitive Agreements to allow the NBN rollout to proceed, with a net present value to Telstra of $9 billion.[12]

As Telstra disclosed to the market, the new USO arrangements had a net present value to the company of $700 million (in addition to the $9 billion).[13]

The rationale for the new arrangements was stated to be the introduction of competition in the provision of universal service. According to the explanatory memorandum of the Bill which established TUSMA:

... it is appropriate that the model for delivering universal service and other public policy telecommunications outcomes be reformed to facilitate the competitive supply of universal service and other public policy telecommunications outcomes.

But while this would appear to indicate that a provider other than Telstra could provide the USO, even saying that “the government will contract with service providers for the supply of these important services,” in reality it was a fait accompli – the Government had signed a contract with Telstra to provide the universal service for a further 20 years.

Under the contract Telstra would continue to do essentially the same things it was already doing; it would continue to receive broadly the same amount of money from the rest of the industry for doing them; but it would in addition receive $100 million a year that the government would now pay in each year.

Prior to the contract, the USO subsidy had been approximately $145 million per annum, of which approximately $50-$60 million per annum came from Telstra’s competitors.

After the contract, the USO cost was now determined to be $270 million – and there was now to be an additional $100 million a year being paid from taxpayers.

These elaborate arrangements presumably assisted Broadband Minister Conroy to land a deal with Telstra on the NBN.

Whether they constituted a useful reform of the USO arrangements is a very different question.

The TUSMA Agreement essentially locked in the existing USO requirement for the provision of a standard telephone service. This was to be delivered over the NBN fixed line network where it was available, and over Telstra’s copper network otherwise.

In other words, premises in the last seven per cent, which were to receive NBN wireless or satellite services, would continue to receive a standard telephone service over the copper.

The Coalition believes the establishment of TUSMA added unnecessary cost and regulatory duplication. We have announced our intention to abolish TUSMA and transfer its contract management functions to the Department of Communications, and legislation is currently before the Parliament to do this.

But the more serious legacy of the new contractual arrangements is that they make it more difficult than it might have previously been to move away from the current USO arrangements. The reason is that Telstra now has a set of contractual rights against the government which last for twenty years.

Some Possible Directions for Reform

Having spent some time talking about problems with the current USO framework, let me now turn to consider some possible directions for reform. In setting these out, I want to emphasise that the government is interested to start a discussion on these issues.

We will not find the answers overnight – and we are keen to hear the perspectives of stakeholders including of course those who will speak at today’s conference.

The government does not have a concluded view about what changes, if any, we would like to make to the USO. But we are certainly open to asking the question of whether the USO arrangements could be varied to do a better job of achieving their underlying policy intention: providing Australians reasonable access on an equitable basis to telecommunications services wherever they live.

Let me therefore run through some possible directions for reform.

How do we solve the legacy copper network problem?

Outside NBN Co’s fixed network, the contractual USO arrangements presently mandate the provision of a continuing voice telephony service over the copper network – apart from more remote places where there is no copper and Telstra uses alternative technologies such as High Capacity Radio Concentrator (HCRC) systems, wireless or satellite to deliver voice services.

So we have a strange situation where taxpayers’ money is being used to build out the national broadband network to serve people all over Australia, with a combination of fixed line, fixed wireless and satellite services – yet at the same time there is a separate subsidy being paid, so that a minority of premises, largely those intended to be served by NBN Co’s fixed wireless and satellite broadband networks, will continue to receive a legacy voice service on a separate network using decades-old technology.

One potential reform direction would be to consider delinking the USO funding from the legacy copper network.  Instead, the subsidy could be provided, to the extent necessary, for the provision of specified services over any suitable network, and this could in turn allow the legacy copper network in the last seven per cent to be withdrawn from service over time. 

I hasten to add that the existence of the agreement with Telstra means that such a reform could not occur unless changes were negotiated which were acceptable to that company.  And, having just successfully completed contract re-negotiations with Telstra as recently as December, we are not proposing to re-open negotiations.

However, over the longer term it will be interesting to observe how consumers behave.  Will consumers, at least in the areas served by fixed wireless, continue to demand a voice telephony service over the copper infrastructure?  Or will they prefer to use a voice service delivered over the NBN fixed wireless? 

In time, the answer to these questions may well signal opportunities for the current arrangements to be re-visited.

Certainly, the feedback to date from customers on the fixed wireless network has been very positive. So as the network gets built out, it may very well be that people increasingly choose to use this network for a voice service as well as a data service.

In that scenario, would it be sensible to have a legacy copper network remaining in operation because it was required under the USO, if customers were increasingly shifting off that network and onto the wireless network?

No doubt as the universal service provider and operator of the legacy copper network Telstra will have some particular insights on this issue, and I would encourage Telstra along with other stakeholders to come forward with its views about whether this scenario is likely and, if so, how we should respond to it.

What technology or service should be mandated under the USO?

When the USO was put in place in the early nineties, the question of which service to mandate was pretty straightforward. About the only thing people got from their telephone was a voice service, so that is what was mandated.

Today, though, there are many more options. Should it be a voice service or a data service?

To take one example, the service that NBN delivers over the fixed wireless network is a data service. However, most service providers are also offering a voice over IP telephony service. Therefore if the USO were mandated as a data service today, it would in practical terms also mean that the end user was able to receive a voice service.

Now a key question here is cost. If you had asked this question fifteen years ago, it would have been possible to extend the USO to mobile or to data – but that would have involved considerably greater cost. Today that may very well not be true. If the USO is framed in terms of a mobile service or a data service, that conceivably could cost less – on an individual service basis and a system-wide basis – than mandating the delivery of a voice service over the traditional copper network.

We do not know the answer with certainty, but as a direction for potential reform, it could be worth looking at. In particular, if advocates of change can demonstrate that change would not necessarily cost more, they are likely to get a more interested hearing from government!

Can the USO be delivered over mobile or wireless networks?

If the NBN fixed wireless network is one possible technology to deliver USO services, then another possibility would be the mobile networks.

As I mentioned, the majority of USO services are presently delivered over the copper, reflecting the requirement that a standard telephone service must support preselection. Because a mobile service cannot meet this requirement, Telstra typically uses mobile to deliver the USO only in circumstances such as new estates which are awaiting an NBN connection, and here it has a specific regulatory exemption from the preselection requirement.

As I have already mentioned, the market evidence shows that customers are increasingly taking up mobile in preference to fixed for voice services – which suggests that delivering the USO over mobile could potentially better meet customer preferences than the current arrangements.

There are some potential policy benefits from using mobile networks to deliver USO services.

To start with, there are three competing vertically integrated mobile network operators. Hence a USO subsidy to connect customers in particular categories could be delivered in a more competitively neutral way, with the subsidy available to any operator which provided the service to the customer.

A second benefit could be the opportunity to capture economies of scope – that is to say, savings from the fact that one network is used to deliver two separate services rather than supplying those services over two different networks.

Today we have a framework where the government is funding the construction of a fixed wireless network, and the three mobile operators are steadily building out their mobile networks.

Yet the USO is at the same time subsidising the provision of services over the legacy copper network. It seems likely from first principles that there would be economies of scope in using one wireless network – be it fixed or mobile – to supply both its existing services and also voice services subsidised under the USO.

A third benefit could come from the lower incremental costs incurred in connecting a customer to a wireless network as compared to a fixed network. Certainly, once the base station is built, connecting another customer within the coverage area incurs virtually no network capex – as compared to the cost of building a new fixed line connection to a premises which is not presently connected to the network.

Better alignment between the USO and the NBN

There is a broader issue to consider: how do we align the policy framework for the USO with the policy settings for the NBN, particularly as the Coalition moves to deal with the inherent cross-subsidy that the previous government built into its NBN policy.

Stephen Conroy’s big idea was that the NBN would make good profits in the cities and these would be used to cross-subsidise loss-making services in regional and remote Australia.

But this model has problems. It entrenches higher costs for the majority of Australians living in metropolitan areas; and it masks the true costs of delivering services in regional and remote areas.

It also piles complication on complication as you seek to defend the metropolitan revenues against competitive entry from operators who can provide services more cheaply, since they are not constrained by a requirement to serve unprofitable areas.[14]

The issue of subsidy models was raised in the Vertigan review of regulation, which said that any subsidies “should be transparent, sourced in ways that minimise inefficient burdens on taxpayers, delivered as efficiently as possible and sustainable in a competitive environment.”[15]

It went on to note that the cross subsidy model fails to meet this test:

"Entrenching an opaque cross‐subsidy, it offers no transparency on the extent of transfers and provides no benchmarks that would allow an assessment of whether future claimed transfers are reasonable compared with the actual costs of supply and revenues received."[16]

The review said that its preferred option would be an ongoing subsidy funded through consolidated revenue, and its back up option would be a USO-type industry levy. However, the Vertigan panel pointed out the same risk with industry levy arrangements as I have already discussed in reference to the current USO, saying:

"When the quantum of subsidy required is not capable of being market‐tested, there is a risk that the levy itself will become a means by which powerful suppliers extract benefits at the public’s expense."[17]

The Coalition Government recognised the problems with the NBN cross-subsidy in our response to the Vertigan recommendations, saying:

"This model is unsustainable in the long term and not in the interests of the consumers who ultimately fund the cost of the services under any model, and typically face higher costs where competition is reduced."[18]

In that same response document we said we would commission the Bureau of Communications Research, the policy research arm of the Department of Communications, to undertake an assessment of the costs of NBN Co’s fixed wireless and satellite services and provide options to Government for replacing the internal cross‐subsidy with more transparent funding arrangements.

The terms of reference for this assessment will be released shortly with a view to having more transparent funding arrangements in place under the new regulatory framework set to commence on 1 January 2017.

Now this work is looking at the cost of providing a network delivering designated broadband services to designated premises in regional and remote Australia.

This is not dissimilar to the question which the present USO arrangements seek to address: how much does it cost to provide a separate service – a voice telephony service – to designated premises in regional and remote Australia.

They are not the same services and not necessarily exactly the same premises: but clearly there is an overlap, at least potentially if not in reality. So a potential reform direction for the present USO arrangements is: can we better integrate them with the arrangements for funding the NBN?

How can the USO stimulate competition?

I have spoken about the implications of the USO for competition. A possible reform direction is to find ways in which we do not have one company, or one technology, entrenched as the recipient of subsidies, whether those subsidies are funded by government or by the recipient’s competitors.

How can the USO be lowest cost?

A related question is how the USO arrangements can be as low cost as possible. The present arrangements impose a significant cost on the telecommunications industry – which in turn is passed on to end users – and on government – which in turn is passed on to taxpayers.

In the present system, there are limited incentives to reduce the cost of providing the subsidised services, or to update the technology which is used to provide them.

A USO policy which built in incentives to reduce costs – including through competition between those receiving the subsidy – would be a marked improvement on present arrangements.

Conclusion and Way Forward

Today I have spoken about the policy rationale for the USO – and some issues with the current USO framework.

I have argued that there is a range of possible reform directions which could be considered when it comes to the USO.

The government’s immediate priority is developing our plans for transparent, pro-competitive funding of loss-making NBN services in regional and remote Australia to replace the current hidden cross-subsidy, based on the Bureau of Communications Research study. Options include, for example, a transparent contribution from operators of high-speed metropolitan broadband networks (including NBN Co itself), or a contribution from a wider industry base, and we will consult further with interested parties before taking any decisions.

But we are also interested in a broader dialogue about whether there is a case for reforms to the universal service obligation. I have sought to contribute to that dialogue today by identifying a number of potential issues which could be considered as part of a change process.

The USO forms one part of a complex set of arrangements to ensure consumer protection in the telecommunications sector. Much of the current consumer protection regime was designed to create incentives for Telstra to install, repair or maintain telephone services supplied over its network. As consumers switch from services supplied over copper and other legacy infrastructure to NBN Co’s fixed wireless and satellite services there is an opportunity to consider how these (or equivalent) consumer protections should be applied – if at all - in an NBN environment.

During the course of this year, we will consider the way forward on these issues. An immediate opportunity for stakeholders to offer their views will be the next Regional Telecommunications Review, which will be held this year as the Act requires. In coming weeks we will have more to say about the arrangements for this review.

Thank you for the opportunity to speak to you today on this important topic – and I hope that the deliberations at today’s forum are productive.

[1] Review Of The Operation Of The Universal Service Obligation And Customer Service Guarantee, Department of Communications, Information Technology and the Arts, 7 April 2004, p 14

[2]Telecommunications (Consumer Protection and Service Standards) Act s 9(1). The obligation also extends to payphones but I will not focus on those services in this speech.

[3] Strictly, there are two charges: one for the connection from the boundary of the Telstra network to the customer’s property boundary (network extension) and the other for the connection from the customer’s property boundary to the customer’s premises. But the key point is that there is a maximum charge for each component, which in rural and remote locations can be considerably less than the cost of the connection.

[4] Source: ACA, ACMA Annual Reports


[6] Optus Submission to Department of Communications, Information Technology and the Arts on Telecommunications Universal Service Obligation Review November 2007, p35

[7] There was a (not very well designed) attempt to introduce USO contestability to specific regions in 2001, but there has been no serious attempt to make the status of universal service provider contestable. See

[8] The Telecommunications Legislation Amendment (Deregulation) Bill 2014 (presently before the Parliament) will remove pre-selection requirements on the wireless delivery of a standard telephone service, including when a mobile network is used to provide a fixed service.

[9] ACMA Communications report 2013-14 p14-17

[10] ACMA Communications report 2009-10 p29



[13] Telstra, “Financial Summary of the Proposed Arrangements”, attachment to Telstra Media Release, 23 June 2011, “Telstra Signs NBN Definitive Agreements,”, downloaded 11/3/15


[15] Independent cost‐benefit analysis of broadband and review of regulation p21

[16] Independent cost‐benefit analysis of broadband and review of regulation p21

[17] Independent cost‐benefit analysis of broadband and review of regulation p21

[18] Telecommunications Regulatory and Structural Reform paper, December 2014 p4