Speech to the Spectrum Review Workshop

I am pleased to join you at this industry workshop as part of the government’s review of how we regulate radiofrequency spectrum.

If you had predicted, fifty years ago, that large companies would pay hundreds of millions of dollars for the right to use an asset which you cannot see or touch, you would have been met with considerable scepticism. 

Look at the way television licences were granted in Australia in the nineteen fifties.  These rights to use the scarce broadcasting spectrum were the essential element to what would become extremely valuable businesses – yet in the main, they were simply handed out to existing media companies.

But in the nineteen nineties we adopted a very different approach, including using auctions to allocate spectrum. 

Today I want to argue that the system Australia adopted in the early nineties has served us well - but it is timely to look at whether we can now improve it.

In particular, I want to speculate that there may be some opportunities we are missing with today’s system – by making greater use of a market based approach, and by being more flexible.

A system which has served us well – but it is timely to look at improvements

In the early 1990s, Australia was a world leader in introducing market-based mechanisms for the allocation of spectrum.

This was driven by a growing recognition that spectrum is a scarce resource.

Demand for spectrum was growing; there were new uses and technologies emerging.

We began to see that the old approach of administrative allocation was not efficient, and did not promote competition or innovation in the market.

The new approach was designed to allocate this scarce resource in a way which maximised the return to society. 

Hence a key aspect of the regime we have today is to incentivise efficient use by assigning ‘property rights’ in the spectrum. 

This gives licensees the opportunity - and incentive - to optimise their use of the spectrum and respond to changing technologies and consumer demand.

The legislative regime also seeks to manage potential external issues, like interference to neighbouring users.

Australian consumers and businesses have benefited greatly from these market-based mechanisms, in terms of enhanced flexibility, access to new services and technologies, and ultimately improved productivity and standards of living.

But of course it is now twenty three years since the new Act was passed.

In that time there have been profound technological advances – which make it possible to use spectrum more efficiently, flexibly and intensely.

To mention just a few:

• Analog technologies have been supplemented or replaced by digital – in mobile phones, in broadcast radio and television
• Improved modulation and compression techniques bring greater capacity and reliability
• Interference mitigation techniques let spectrum be used and re-used more efficiently
• There are emerging technologies which let different users share spectrum without interfering with each other. 

The spectrum sharing possibilities are increasing, with new technologies on the horizon such as cognitive radio and dynamic spectrum access.

Some of these emerging technologies are hard to accommodate under the current regulatory framework.  To allow for the new ultra-wide band technologies, for example, we have needed legislative changes so spectrum and class licences could co-exist.

There have also been dramatic changes in the market.  The most obvious is a rate of take up of wireless based services that absolutely nobody predicted in 1992.

I know from speaking to people involved when Optus won the second licence in 1991 that they expected to make money from fixed services rather than mobile.  In fact if the government had not required that the winning consortium also acquire a mobile licence they would not have bothered with it!

More recently there has been another wave of change as Australians have enthusiastically taken to internet access over wireless devices.

In June 2009, just 9% of people used the internet via a mobile handset – by June 2012 that figure was 32%.[1]

ACMA estimates that mobile data usage will grow by 265% over a four year period to 2017, increasing from an estimated 22.2 petabytes in 2013 to 81.1 petabytes in 2017.[2]

But there are plenty of other market trends underpinning the ever more intense demand for spectrum, including:

• A growth in fixed wireless services including those from NBN Co and also from private sector players like Vividwireless, now of course owned by Optus.
• Machine to machine communications, the automated communication between connected devices that is driving increased efficiency, value, cost savings and innovation.
• Radiofrequency identification that is quietly revolutionising logistics and warehousing, retailing and even the health care sector, providing an estimated economic value in 2013 of US$130 billion in the USA[3].

Are we missing opportunities with today’s system?

Let me turn then to the question of whether we are missing opportunities with today’s system of spectrum regulation.

Certainly, the current system has often been successful in supporting the reallocation of spectrum to uses which deliver greater social benefit.

For example, the mobile operators have been able to refarm spectrum in the 800/900 MHz and 1800 MHz bands – in turn allowing them to transition their 2G services to 3G and LTE technologies.

But there are also some rigidities in the current system – which often mean that those to whom spectrum has been allocated have little incentive to pursue the most efficient use of that spectrum, either in their own hands or in the hands of others.

Today if you hold more spectrum than you need under a spectrum licence, you can sell the excess – but this has not happened as often as was hoped when the current Act was drafted.

Not all spectrum is substitutable, and how it is valued can differ based on its technical characteristics – such as propagation, or how far the signal will travel.

As well as technical barriers there are other factors which have tended to suppress the trading of spectrum.  One is that you are exposed to stamp duties – levied by the states – and capital gains tax.

Another is that if you want to sell an apparatus licence, its value to other users can be limited because such a licence comes with specific limits on what it can be used for.

There are even greater hurdles to trading broadcasting spectrum.

This in turn means that as a broadcaster there is little incentive for you to use your existing 7 MHz allocation of spectrum more efficiently.

Imagine if you decided that you only needed a fraction of the current digital delivery capacity of your 7 MHz multiplex – perhaps because new compression technology meant you could deliver your existing services with a fraction of the current data levels.

Today if you hold spectrum in the Broadcasting Services Band, there is no way that you can sell or rent out spare capacity.

But as Minister Turnbull has indicated, the government is interested to explore whether we can provide broadcasters with more flexible options in the way that they deal with their spectrum.

The US is pioneering reverse auctions of broadcasting spectrum, and while these are highly complex and have suffered delays, it is an example of the innovative market-based approach we are looking at.

There will be lessons we can learn that can be used to design effective processes for our market.

Let me ask a similar question in relation to non-broadcasting spectrum.

Should there be a mechanism to allow spectrum owners to hand the spectrum back ‘on consignment’ to government, so government can auction it off – with the proceeds shared between the owner and government in an agreed ratio?

Consider for example the 2.3 GHz and 3.4 GHz spectrum in regional Australia, which was held by Austar and unused for many years.

It was first sold to OPEL for $65m, and after that fell through, it was then sold to NBN Co for $120m.

Would it have been useful for Austar to have a mechanism to hand it back to government on consignment, so that potential new entrants knew it was available?

It’s not all about money

I want to make one more point about how we allocate spectrum. As many organisations have commented in their submissions, spectrum allocation is not all about money.  We need to make sure we weigh up the social and community benefits of using radiofrequency spectrum.

Of course that is right.  Spectrum is used for many vital community purposes such as police, fire and other emergency services; communications for many sectors such as marine and aviation; science; and of course national security and defence.

So understanding the market value of spectrum does not mean that we axiomatically then allocate spectrum in a way which captures that market value in the form of a cash payment to government.

But the fact that spectrum has a high market value is an important indicator of the social benefit which it can deliver. 

After all, the mobile phone operators are only willing to spend billions of dollars on spectrum because they can in turn use it to serve many millions of customers.

So it is always important to understand the market value of spectrum – even if not all of it is going to be sold in the market.

The Radiocommunications Act 1992 requires the ACMA, when managing spectrum, to maximise the public benefit through the efficient allocation and use of spectrum.

If the ACMA determines that non-commercial uses of specific spectrum will provide the best return to society, the ACMA may allocate spectrum administratively to them at below market prices.

A good example is the recent decision on re-issuing 1800 MHz licences held by the State Rail organisations, some of which they in turn will use to support rail safety services. A 50 per cent public interest price reduction was provided for 2x10 MHz of this spectrum.

In any changes we make to the current arrangements, it will be important that we continue to provide exceptions to the market-based approach where there is a sound public interest rationale.

However, it would be useful to have a way of better considering these alternate uses on a consistent basis - a ‘level playing field’ as it were.  It is interesting that the UK Government has proposed a similar approach as part of its spectrum strategy.

Another very important issue is that non-commercial users face incentives to use spectrum efficiently – just as we want commercial users to face such incentives.

The best way to make sure there is enough spectrum for all the important purposes for which it is needed – commercial and non-commercial – is to make sure that we do not have spectrum left underused.

Reflecting both of these issues, one question being considered by this review is the approach for pricing for administrative allocations.

Conclusion

Let me conclude by observing that this spectrum review is an opportunity to examine a system that has served us well – but has now been going for many years.

There are undoubtedly opportunities to better capture the potential of the scarce national asset which spectrum represents – in the interests of the Australian community.

Australia led the world with our approach to spectrum in the nineties – we now have the chance to do so again.

The government is keen for industry to think about the opportunities you could take up if the arrangements were more flexible.

The submissions we received in responding to our discussion paper took a pretty cautious approach. 

I hope today will be a chance for some freewheeling discussion about other possibilities. 

Good luck with today’s deliberations – and thank you for your input to this vital review.

[1] Deloitte Access Economics, Mobile nation-The economic and social impacts of mobile technology, 2013

[2] April 2014 – Research ACMA – ‘The economic impacts of mobile broadband on the Australian economy from 2006 to 2013 – Research report prepared for the ACMA by the Centre for International Economics’ - http://www.acma.gov.au/theACMA/Library/researchacma/Research-reports/economic-impacts-of-mobile-broadband-1

[3] Wifi Forward, Assessment Of The Economic Value Of Unlicensed Spectrum In The United States, 2014 http://www.wififorward.org/wp-content/uploads/2014/01/Value-of-Unlicensed-Spectrum-to-the-US-Economy-Full-Report.pdf