opinion/analysis by Renai LeMay
24 January 2014
“Technology-neutral policy” is a useful but treacherous concept that has led many Australian politicians into making poor decisions about important technology issues over the past decade. It’s still a very useful framework, but it’s time this theory was re-examined by our policymakers in light of those failures and the importance of learning from the past to better inform the future.
If you’re a politician or a policy advisor working for a politician, the concept of attempting to create “technology-neutral policy” must be an extremely seductive one.
The seductiveness of the idea lies in its simplicity. The concept of technology-neutral policy holds that when writing laws or otherwise setting up government regulations or the frameworks for government-sponsored technology projects, that policymakers should maintain a focus on the end goal of those initiatives, rather than on the specific technologies which will be used to deliver on those goals.
To understand how this may work in practice, let’s go through a few examples.
In the area of telecommunications policy, Liberal MP and current Communications Minister Malcolm Turnbull has stated many times, both in Opposition and in Goverment, that national telecommunications policy should focus on what benefits the Government wishes to achieve in terms of services to the public, rather than the specific technologies which will be used to deliver those services.
At a press conference in September 2013, shortly after taking office, the Minister stated about the Strategic Review (PDF) which the National Broadband Network Company was to conduct into its future mode of network rollout:
“… this strategic review is going to be a very, very rigorous and objective exercise in getting to the facts of the matter so you know the goal of the strategic review is principally the following: to ascertain what it will really cost in dollars and what it will really take in years and months to complete the project on the current specifications and then to assess what options there are to reduce that cost and time by using different techniques, different technologies.
As you know, as everyone knows we’ve canvassed an example of that in our policy document, but let me say again, as I’ve said to NBN staff today, I am – and the government is – thoroughly open minded, we are not dogmatic about technology, technology is not an ideological issue, we are completely agnostic about it. What we want to do is get the best result for taxpayers and consumers as soon as possible.”
The ‘Multi-Technology Mix’ rollout model which the Coalition’s Broadband Network is likely to be based on is a very good example of technology-neutral policy in action. It will likely rely on at least six separate technologies to deliver high-speed broadband services — Fibre to the Premises, Fibre to the Node, Fibre to the Basement, HFC cable, Fixed Wireless and Satellite.
Another example is the way that Australia’s Copyright Act is written. In mid-2013, the Federal Government asked the Australian Law Reform Commission to examine how copyright laws might be best updated to deal with new technologies as part of what it terms the “Digital Economy”.
The Framing Principles for Reform published by the ALRC explicitly state that the Copyright Act should be technology neutral as much as possible. The principles state: “As far as possible, the Copyright Act should be technology neutral and predictable in application in such a way as to minimise and avoid unnecessary obstacles to an efficient market, and avoid transaction costs.”
If you read the text of the existing Copyright Act, what you’ll find is that this principle of technological neutrality is already enshrined in its pages. Throughout the Act you’ll find hundreds of references to different types of copyrightable works, from literary pieces to music to computer games and broadcasts. However, you will rarely find direct references to certain technology platforms. Instead of directly mentioning YouTube or digital television, for example, the Act usually stipulated “broadcast”. It tries to avoid the issue of how books are distributed — electronically via Amazon’s Kindle platform, or via old-style paper books, and so on.
A third example comes from the Federal Government’s procurement principles. For contracts above a certain value, most Government agencies are required to put those contracts on the open market in the form of competitive tender processes.
Procurement regulations set by the Department of Finance stipulate that the aim of all government procurement is to obtain “value for money”, with one of the most important considerations to be taken into account during that process being “the business need”. Although government Request for Tender documents do sometimes stipulate that specific technologies be supplied as part of suppliers’ responses, it’s far more common for the concept of technological neutrality to be applied.
For, example, if a department or agency is seeking to purchase mobile devices such as tablets, it is far more common for tender documents to stipulate the category of devices being purchases — as in, tablet fleets — rather than the specific brand which the department may wish to purchase. In the case that a business IT system such as an Enterprise Resource Planning or Customer Relationship Management platform is sought, it is rare that a certain style of deployment (on-premises, managed services or cloud computing) is stipulated. It’s far more common that the Government will allow suppliers flexibility in how they respond to the tender document.
According to an influential essay on the subject of technological neutrality published in 2007 by Chris Reed through UK legal journal SCRIPTed, the concept of technological neutrality dates back at least to mid-1997, when the US Government published a document setting out what it called a Framework for Global Electronic Commerce. The document stated:
“… rules should be technology neutral (i.e., the rules should neither require nor assume a particular technology) and forward looking (i.e., the rules should not hinder the use or development of technologies in the future)”
According to Reed, the concept was then rapidly taken up internationally. However, as Reed writes, the concept is seldom questioned:
“The desirability of technology neutral regulation has become part of the general wisdom, and is rarely questioned. Alarmingly, this consensus among legislators seems to have developed in an almost complete absence of any clear understanding what the term “technology neutrality” might actually mean.”
In this article, I use three recent Australian examples will attempt to advance the theory that while technology neutral policy is still a very useful concept in the Australian policymaking environment, it should not universally be taken for granted. I will especially attempt to demonstrate that as a general rule, the capabilities of a specific technology cannot be disaggregated from that technology itself; and as a rule, this calls all so-called “technology neutral” policies into question.
The first example I wish to examine is that of the National Broadband Network project begun by Labor and now continued as the Coalition’s Broadband Network. As previously mentioned, Minister Turnbull is fond of stating that government policy with respect to broadband should be technology-neutral; setting aims in terms of speed and coverage but not in terms of the technology being used.
In the short to medium term, the Minister’s approach would appear to be correct. NBN Co’s Strategic Review document ably demonstrates that it is possible to use a variety of different technologies — FTTN, FTTB, FTTP, HFC cable and even alternatives such as 4G/5G wireless broadband — to achieve similar aims. All of these technologies can — and will, under the Coalition’s current plan — deliver download speeds to Australian consumers significantly greater than are currently available. These speeds will be likely to solve many of Australia’s broadband problems over the next five years, and perhaps much of the next decade.
However, it is when you examine the NBN project over the long-term that the concept of technology neutrality breaks down.
As the telecommunications industry has understood for around 50 years (and there were many experiments before that, including by Alexander Graham Bell as early as 1880), the inherent characteristics of light as a data transmission medium represent a step change compared with any other known form of transmission medium. The fundamental reason why fibre-optic cable has emerged as a preferred broadband delivery technology globally over the past decade is that data transmission needs are continually outpacing any other form of broadband delivery apart from that offered by light.
The Minister and NBN Co executive chairman Ziggy Switkowski are correct in several statements made recently that the Australian public does not currently need the gigabit speeds offered by FTTP in the short or perhaps even the medium term future.
However, there is absolutely no doubt that in the long-term, and certainly as we approach 2030 or debatably, even before that point, the Australian public will increasingly need speeds which sit above the capacity the copper network, or even hybrid fibre/copper networks such as HFC cable and FTTN, are able to deliver. The extremely rapid growth in bandwidth requirements, and the constant extension of fibre networks to replace copper networks globally, demonstrates evidence of a clear trend in this area.
Most Australians may not need 100Mbps broadband speeds right now. But every year, more will. In ten years, that statement is likely to focus around the need for even higher speeds: It will be something like: “Most Australians may not need 300Mbps right now. But every year, more will.” And as every year progresses, lower-level technologies such as FTTN, FTTB and HFC cable, which use less fibre (and thus less light) in their networks than FTTP, will become less useful.
An even more uncomplimentary comparison can be made between the upload speeds available on the various networks. This issue is particularly serious for HFC cable networks, due to their nature as a shared medium.
Unfortunately, there is no technology which delivers capability in between copper and fibre optic cables. These are the two best data transmission materials we have: Physicists have not yet discovered something suitable in between. It is true that in many other areas, comparable technologies exist that overlap in terms of their capabilities. However, when it comes to broadband, there are inherent differences between the technologies available which limit their uses.
As a concept, mobile networks are alluring as they use radio waves — a form of electromagnetic radiation, like light also is — to distribute data. However, evidence has consistently shown they are simply unable to handle all of the requirements of fixed premises due to the inherent nature of such platforms as shared mediums. We definitely need mobile networks — but those mobile networks are clearly a complementary broadband option to, and underpinned by, strong fixed networks.
With all of this in mind, it appears clear that there really is no real choice in terms of how Australia proceeds with its telecommunications policy. Eventually, our nation will be blanketed with fibre broadband, because we will need it and there is no alternative. Our nation will also be blanketed with high-speed mobile broadband, because we will also need that. In the long-term, there is no alternative. In this sense, the phrase “technology-neutral” means rather little, because there is no real choice between technologies.
The obvious rejoinder to this position is to argue, as Minister Turnbull has done, that FTTN potentially represents a step on the path to a full FTTP rollout. The Minister has argued that deploying FTTN instead of FTTP, the Government will preserve the “optionality” of deploying a full FTTP rollout in the future, taking fibre to neighbourhood ‘nodes’ first and then eventually all the way to the premises when it can be demonstrated that such a rollout is needed and cost-effective.
However, even discounting the unpalatable nature of this idea to an Australian electorate which has consistently signalled its desire for a full FTTP rollout, NBN Co’s own Strategic Review signals that such a stepped upgrade path would be inefficient. The preferred ‘Multi-Technology Mix’ option in the Strategic Review is slated to require $39 billion of peak funding and finish at the end of calendar year 2020; a redesigned full FTTP rollout would cost only $15 billion more and finish three years later.
In the long term, which is, logically, the only way to think about the rollout of national infrastructure that is expected to continue to serve the nation’s need for 50 years or more, the investment difference is almost negligible, especially when you consider that the Government would make a minor return on its investment in all scenarios; effectively neutralising the invested cost.
This example starkly demonstrates that you cannot disaggregate the capabilities of a specific technology from that technology itself. There’s nothing quite like fibre when it comes to the fundamental provision of telecommunications services. You cannot extrapolate the unique qualities of light-based transmission to other, non-light-based mediums.
A second area where the concept of technology neutral policy is increasingly breaking down is in the area of major Government ICT project and service delivery.
In Australia’s public sector, the governance of major ICT projects is currently in crisis. For example, in June last year, Queensland’s first whole of government comprehensive ICT Audit found that ninety percent of the Queensland Government’s ICT systems were outdated and will require replacement within five years at a total cost of $7.4 billion, as Queensland continues to grapple with the catastrophic outcome of years of “chronic underfunding” into its dilapidated ICT infrastructure.
Similarly, in November 2011, Victoria’s Ombudsman handed down one of the most damning assessments of public sector IT project governance in Australia’s history, noting total cost over-runs of $1.44 billion, extensive delays and a general failure to actually deliver on stated aims in 10 major IT projects carried out by the state over the past half-decade.
Other states are suffering similar problems, if not on the same scale as Victoria and Queensland. New South Wales, South Australia, the Northern Territory; there is ample evidence to demonstrate that the concept of major ICT project and service delivery has broken down in Australian Governments, especially at the State Government level.
When you get to the heart of the problems being experienced by such State Governments, much of it relates directly to one issue: The decision by ICT strategists within those Governments to leaving their procurement processes in a technology-neutral stance.
In many of the high-profile cases, this has led to major departments and agencies procuring major traditional IT application platforms in areas such as Enterprise Resource Planning, from major vendors such as SAP and Oracle, and attempting to customise those platforms to fit their business models. The reason these platforms were chosen in the first place was that they were the most comprehensive; seen as the most capable to meet the business needs of the departments concerned.
However, almost universally, what has been seen around Australia is that these traditional platforms have been too inflexible to be modified. In Queensland, the state’s Health department was unable to customise a SAP platform to meet its needs, and suffered a billion-dollar overrun when attempting to upgrade its payroll systems; in NSW, problems with complex ticketing requirements bedevilled the state’s Tcard public transport system; and in Victoria attempts to upgrade the state’s troubled LEAP police database have also failed.
The reason these projects failed not because the technology chosen was not up to the task; the reason they failed was that the departments and agencies implementing the projects did not modify their business practices to fit the demands of the platforms being implemented.
As a consequence of this extremely long and comprehensive string of disasters, Australia’s State Governments are broadly abandoning the concept of ‘technology-neutral policy’ during their IT procurement practices. Instead, Queensland, NSW and Victoria — the most populous states — have actively chosen a specific class of technology — cloud computing — to focus on in the mid-term, mandating a “cloud-first” approach to IT procurement.
The reason they have done this is fairly obvious: It forces them to standardise upon global best practice.
Cloud computing providers (whether software as a service, infrastructure as a service or platform as a service and so on) are able to leverage economies of scale in a much greater fashion than providers of traditional ‘in-house’ or even managed services platforms. Companies like Salesforce.com, Google, NetSuite, Amazon Web Services and others don’t look locally when building best practice into their products — they look globally. Consequently, they are forcing local companies to change their business practices to match the needs of their software; rather than customising their software to meet every customers’ needs.
The focus by Australian State Governments on just one class of technology solutions — cloud computing — represents the fact that their technology-neutral ‘fit for purpose’ approach was not working, because it depended on the individual judgement of their own employees, and not the global, expert judgment of the cloud companies about what represents international best practice.
Again, what we see here is the idea in action that the capabilities of one specific set of technologies cannot be extrapolated to others. Cloud computing technologies, on paper, provide the same functions as traditional on-premise facilities, or can. However, in practice, they are radically different and force transformative changes in business practices that are helping State Governments avoid the IT disasters of the past.
The last area I want to explore, briefly, is copyright law. As mentioned above, this is a key area where the concept of technological neutrality is constantly questioned.
To a large degree, the concept is useful when it comes to copyright and intellectual property. Technology-neutral policy, written into the Copyright Act, stopped the book industry from collapsing when books started being read on Kindles rather than on printed paper; ensured musicians were still getting fair recompense when their music started to be distributed online; allowed universities to continue to fairly use copyrighted resources for academic purposes even when most journals went online, and so on.
Without the concept of technology-neutral policy, the Copyright Act would have collapsed under the weight of its own constant modifications a decade ago.
However, the system is far from perfect, and a constant string of exceptions are being tested in the courts. iiNet’s High Court defence against the Australian Federation Against Copyright Theft is one example. The legal ruckus around Optus’ now-banned ‘TV Now’ online streaming service is another. And there are hundreds, perhaps thousands of other examples. Pretty much nobody thinks the Copyright Act is correctly set up at the moment. As the ALRC’s Framing Principles for Reform of the Act state:
“Some stakeholders submitted that the existing legislation is increasingly imposing costs through being out of date and unsuited to the digital environment. For example, rapid change in technology and consumer behaviour is creating a ‘growing rift between platform-specific provisions of the Copyright Act and the ways in which Australians are increasingly using copyright materials’. The Australian Interactive Media Industry Association submitted that, despite all the opportunity offered by the digital economy, ‘the Copyright Act is too technology specific and inflexible and as a result is unable to support today’s and tomorrow’s innovations’.”
What this indicates is that the concept of technological neutrality in the Copyright Act is a constantly evolving one. The underlying framework of the Act will always be necessarily technology-neutral, but Parliament needs to act constantly to keep it fresh and able to deal with technological challenges. In short, technological neutrality is not a black and white issue; it’s a useful concept that should be applied when necessary, but not over-applied constantly.
In his 2007 paper, Reed wrote that in the real world, it was apparent that policymakers need to make a choice between technology-neutral and technology-specific approaches, using each one where appropriate. He writes:
” … the lawmaker ought to consider whether a technology specific approach would produce better regulation. Specificity is often rejected at the outset because technology neutrality has been hailed so loudly as the only correct way to proceed in the ICT arena, but it has a number of advantages. It creates substantial certainty as to the scope of the regulation and what needs to be done in or to comply with it.
As a consequence, the level of regulatory compliance is likely to be high, and the compliance costs lower than would be the case if advice were needed as to the meaning and application of the regulation. A further benefit of technology specific regulation is that its effects will not spill over into other fields of activity, forcing unwanted behavioural change in those fields. Finally, specificity forces the lawmaker to reconsider the regulation at regular intervals, thus ensuring that the regulation keeps pace with technological and other changes.”
According to Reed, there is, obviously a cost in choosing technology-specific policy over technology-neutral policy — technology-specific policy will necessarily be short-lived and need to be updated as technological change occurs. However … isn’t this just reality? As Reed’s paper states: “Perhaps, though, this is an inevitable consequence of a world in which technological change occurs rapidly and in a seemingly unpredictable direction.”
Regardless, one thing is certain: It’s time for Australian policymakers and policy advisers to stop taking one side, and one side only, in this false dichotomy. It is very clear from recent examples in Australia that policy and law must be a combination of technology-neutral and technology-specific policy. To consistently take one side over the other is to risk falling victim to bad policy and, ultimately, failed IT projects and poorly structured industries. As with most policy settings, a more pragmatic method using all the tools available for their fit purposes would be a more ideal approach.