Spectrum reform
The Secretary of the department, Peter Harris, delivered the following speech on spectrum reform at the RadComms 2012 conference in Melbourne on 6 June.
This speech is known in the trade as the death slot, as you settle back after lunch. And the mind wanders. I hope the seats are uncomfortable.
Much of what the department, as distinct from other arms of government such as the ACMA, does in the spectrum space attracts little attention.
And generally, I like it that way. There’s little enough reward for public servants in public speaking, even less for answering questions on the record.
But there is a gathering pace to regulatory reform in the spectrum area that makes it worthwhile to expose a few points about the direction of change, and perhaps help shape the discussion which will be necessary if we are to improve the efficiency with which spectrum is allocated.
And with that, support one of the faster growing areas of the Australian economy.
We all know data use is currently on an explosive growth path. The ABS statistics show growth rates of 20 per cent a year and more. CISCO released an estimate of 29 per cent per annum for last year, with internet traffic now 11 times the level recorded in 2005. Their outlook is for compound growth rates of 40–60 per cent, with consumer growth leading business growth.
While fixed lines still clearly dominate this growth in data use in aggregate, the pressure for more, faster mobile and wireless links means every historic area of allocation across the spectrum range will in future come under scrutiny, either by the government or by commercially-driven businesses looking over the fence at less efficient users. You’ll generally have one fixed connection but ever-increasing numbers of devices, all of which will want to connect wirelessly to each other.
Of course, we have a big digital dividend to offer the most voracious consumers of spectrum from clearance of analogue TV, and this should release pressure in the short term in the mobile telephony area.
But the longer term forecasts out there from various vendors and all the commentators, and the seemingly endless increase in consumer wireless products, together suggest that this will not be sufficient.
Thus the Convergence Review spent a reasonable amount of time and effort on our capacity to meet this demand, as broadcasting and telecommunications converge. And while the response of some of the affected parties was less than enthusiastic, the general conclusion of the review is unarguable: we will have to revise our approach to allocation, and increasingly apply market-based concepts to spectrum that has been previously treated as in some way special.
The Convergence Review provides a few synopses towards the way to address the challenge:
- It poses a compelling case for a basic overhaul of the current legislative framework for media and communications. The Content Service Enterprise concept should be looked at not just for what it will replace, which quite a few lawyers have already done; but for the opportunity that will lie in applying an effective model, that which regulates telecommunications, to broadcasting and online entertainment.
- Thus the review sets in motion a number of debates. It has opened up issues that could well have been left to sit for years. That is a proper role of a review. To start the debate.
- In the not too distant past, our entertainment options were limited, but the availability of spectrum was vast. Today it’s the opposite: spectrum is relatively scarce and entertainment options are almost overwhelming. Thus pressure for change will prove irresistible.
- The review argues TV broadcast spectrum is better placed by the government into a form that is market-responsive, encouraging efficiency of use by allowing a quicker response to demand and technology shifts. There are many policy complexities about the implications of that argument, but there were no sacred cows in this review.
- Further down the track, but within ten years, fibre will have replaced the copper telephone network. The NBN automatically changes the game because it provides new delivery options for converged service providers. Some broadcast services could choose to review their options for delivery by that time, or before it. The allocation system must be not be left in a form which impedes moves to more efficient delivery.
The spectrum allocation reforms of twenty years ago were about creating a legislative and regulatory structure that resisted pressures to continue to give spectrum away for free. As I note above, this was in a time when spectrum was vast and apparently unpopulated, and choice for consumers was limited. Whereas today the situation is the opposite: choice for consumers is virtually unlimited, and spectrum is becoming scarce.
Governments managed to keep up with spectrum demand throughout the late 20th century by increasing spectrum supply to accommodate technology and demand shifts. Recent years have seen demand accelerate. This spectrum crunch places greater pressure on regulators to correctly ration and balance spectrum demand between various competing uses. It raises the problem of how to balance certainty of tenure for users with the flexibility to put spectrum to its most valuable use. And it encourages regulators to revisit existing allocations, to see if they can be rationalised. The ACMA’s current review of the 800 MHz spectrum is a case in point.
The 1990s reform target—to ensure spectrum is valued and not given away—has now been achieved. It is well recognised that even spectrum allocated for the public interest—as the government is currently considering, for rail safety or emergency services use—needs to reflect its opportunity cost to some extent, lest the concession be exploited.
But viewed more broadly, the time for a new target—a regulatory structure better able to choose between competing users, each with a good case for ‘special’ treatment—will soon be upon us.
This morning the minister spoke in his recorded remarks about the NBN. It’s not an issue I will be addressing, other than this: the NBN itself will, in this country, create the mother of all cable networks, at consumer level. And fibre networks, to extent that under the NBN they dominate high value urban zones, will alter the options available to current and future spectrum users. Depending of course on how the spectrum is structured. The more location-based the licence, the more substitutable it may become.
The NBN’s current contracts mean we will now certainly have fully-networked fibre services across a couple of hundred locations around the nation within the next two years. And with legislation and competition undertakings from Telstra that mean in effect that this network could not be owned by it or any other telco, and must continue to be wholesale in nature, this network owner would have no reason to refuse business from any current spectrum users that chooses to go terrestrial.
A wide variety of factors—fiscal and competition—mean that the NBN business should remain in ownership terms structurally separate from any other telco. These competition reforms are once again bipartisan in nature, and so can be relied on for long term planning.
The supply side of data-handling capacity is not the only basis on which reform seems inevitable. The demand side—driven by the digitisation of things—is equally potent in creating the pressure for ever higher levels of data transfer.
The digitisation of things has reached the point where no less a journal than The Economist has very recently called it the third wave of industrialisation—equivalent in its impact on manufacturing to the industrial revolutions of 1820 and 1920.
And has pointed out that although manufacturing is now a small part of most developed economies, it is still the most powerful force for R&D. And thus the driver of much of formal innovation, patents and the like.
And this digital revolution in manufacturing is delivered by fibre. Business-to-business fibre to be sure, but the number of firms that can join in will be severely limited if investment remains at the proprietary network level of big corporates.
Without continuous growth in supply of fibre ubiquitously available in most locations, some part of innovation is likely to stall in economies not able to make the digital transition. In Australia, we are making the transition.
Returning to spectrum, the department completed early this year a tortuous process for renewal of the spectrum licences for mobile telephony granted nearly 15 years ago.
These licences were initially allocated by auction but the renewal was a negotiation.
In reality, the term negotiation is giving it a very polite description. The offer and counter-offer process was not really very elegant. All of us engaged in the process would have preferred a neater, more predictable renewal arrangement.
But, a bit like democracy itself, renewal by administrative action is deeply unattractive until compared to all the other options. While renewal by auction looks attractive, the reality is that it is also a costly disturbance to firm and consumer alike if the spectrum churns. And churn isn’t of itself a useful public policy outcome.
Yet in the future, if either government or licensee expects to benefit greatly from churn, then following this recent experience we may all decide to go to auction. Unless we find a better way.
The Convergence Review agreed that we should find a better way, but stopped short of solving the problem.
So, for discussion, I want to suggest that the only renewal concept I can see that offers greater certainty for both the public interest and the licensee, and still ensures that commercial factors drive the choice to renew or not, is to accompany a licence with a right to renew after 15 years, based on some factor related to the value of turnover in the period immediately preceding renewal.
There are strengths and weaknesses to such an option. Renewing a licence where turnover has been zero is probably not promoting efficient use. But where there is significant turnover, it would create some predictability of likely renewal cost in the late term of a licence, and this would probably be of value to all parties.
I should emphasise that there is no government policy commitment to this idea. It simply emerges from the department’s experience of the last year, in running the renewal process.
But this speech today is generally about launching a process to consider ideas. So there’s an idea.
Spectrum Square
In his remarks, the minister mentioned the strong collaboration between stakeholders, industry and government.
The path of further spectrum reform should involve active industry input. While the debate has been started by the Convergence Review, we need a forum to try out ideas without triggering, as far as possible, an overreaction.
The department has been working with ACMA to develop a process that will be used to consult with industry on a range of spectrum policy matters. Designed to make it less formal and effort-intensive than standard consultation processes.
More importantly, what I’m looking for with this initiative is to encourage a conversation about what the medium and long term shape of spectrum management in Australia should be.
I’m expecting there will be opportunities for a wide variety of interested individuals and groups to provide informed commentary, hold debates, or submit short or extensive statements and submissions on important spectrum management issues.
In the past, these sorts of processes have tended to be bilateral affairs where individuals and agencies directly tell us, the government, what you the affected parties, think. There’s little room for lively discussion and interaction.
That’s what I would like it to be. But what exactly is it?
We’re calling it the Spectrum Square. It will be a web-based discussion forum developed and curated by my department in consultation with ACMA.
It can’t and won’t replace our existing and extensive stakeholder engagement activities but Spectrum Square will offer an informal place to try out ideas.
Let me take you through how the Spectrum Square will work.
At the core of the initiative will be a series of discussion papers released periodically throughout the next 12 months or so. These we intend as conversation starters, small papers highlighting particular issues and asking central questions.
The first discussion papers, released today, are titled:
- The Spectrum Management Framework
- Enabling Regulatory Amendments, and
- Priority Review of Spectrum Band Use
These papers have a twofold purpose—to set the scene for the Spectrum Square discussions that will follow; and to propose some immediate spectrum management changes.
These changes are a starter pack—several relatively minor regulatory reforms, and priority spectrum bands for replanning and possible future sale.
The web address is up on the screen but for those not here today, it’s to be found at www.dbcde.gov.au/s2
Or s squared ...
The Spectrum Square has been established to give interested stakeholders and the general public the opportunity to be part of the policy development process. It will be an important part of the government’s ongoing dialogue with industry and the public on spectrum policy issues.
During the next year there will be a steady release of papers.
Following the Priority Reforms paper, we’ll look to open up discussion on the use of spectrum for non-commercial purposes.
As demand for wireless services increases, pressure also increases on the mechanisms that enable the efficient allocation and use of spectrum for non-commercial uses.
And while the default option in future years for allocation is likely to heavily favour market mechanisms, there must be scope to move to non-market choices when the public interest demands that. I mentioned spectrum for rail safety a little earlier, as an example. The quid pro quo will almost certainly be restrictions on the future trading of that spectrum; and the need to keep the asset in the hands of the public interest for which it is allocated.
And transparency.
If spectrum is to go to public interest purposes at less than market rates, this will need to be the subject of clear public justification.
Other papers could cover topics such as:
- Spectrum trading and secondary markets
- Access to spectrum, including allocation, pricing and licensing
- Convergence, and, finally
- The legislative and institutional framework supporting spectrum management in Australia.
Convergence Review—final thoughts
In the short time I have left I want to circle back to the Convergence Review and give you some final thoughts.
As I said earlier, the impact that communication sector convergence will have on the Australian and world economies represents a major structural shift. This is inevitable.
And alteration to spectrum allocation arrangements is inevitable within such a shift.
Either the regulatory structure moves to support this kind of demand, or demand bypasses regulation. So much of online related demand is already sweeping by black-letter law regulation—in copyright, in gambling, in broadcasting, in retailing planning.
Yet spectrum users are by definition in a regulated industry and will remain so. The concept of deregulation we will apply here is more about flexibility and the use of models drawn from other environments than about abandonment of regulation. Ultimately, every licensee wants to be protected in the use of their licence.
In my view, we will need new models if we are to bridge the inherent differences between the affected firms, each in different states of transition, and bring them into a coherent structure.
The model of regulation we use in future will not necessarily be loved by lawyers. It is likely to involve more discretion, more judgement as well as more market-orientation.
This will reflect the simple reality that change will surprise all of us. So we will need greater flexibility, to allow for the unforeseen.
Maybe that too will be a further topic for the Square.
