Devising the right NBN policy
It's been an interesting 2012 for the NBN and NBN Co with plenty of progress on the ground. Sure it's not rolling out as quickly and smoothly as many of us would like it to, but every connection made between now and the federal election will be priceless.
Most importantly, the prospect of a fibre network for all Australians is now pretty much set in stone whatever the outcome of the election. However, when it comes to policy we only have one side of the picture and therefore reliant on our imagination to ponder the implications of a Coalition victory.
There's a simple methodology to assess any public policy and while we don't have time machines we can still have a pretty good stab at answering some of the key questions about the competing NBN policies. The value of such an exercise may not seem apparent but it is handy way of remembering just what's at stake.
When pondering policy one needs to consider the four main areas:
- Competition
- Regulatory Environment
- Technical Solutions
- Commercial Arrangements
Competition
The "last mile" of any utility network, the customer access network (CAN), is a natural monopoly. The right number is of networks is one, if your objective is to minimise cost and maximise public utility. Allowing competition leads to poor, expensive services, where no player gets good returns discouraging investment and upgrade. We saw the Optus/Telstra HFC Cable TV debacle that destroyed billions in shareholder funds and failed to create a viable cable TV industry. The problem is "how do you keep them honest and motivated? Not "fat, dumb and happy" or complacent, inefficient and possibly corrupt.
How does a society grant anybody, public or private, a monopoly on anything, let alone essential infrastructure, and expect them to operate efficiently, selflessly and altruistically for decades on end?
It's an impossible task when real people with their many failings and drivers, like greed and ambition, are involved. Certainly some degree of regulation and oversight/governance will stop, or eventually detect, most excesses.
Perhaps a natural monopoly could benefit from the following two things:
• Private ownership, under the close control of a regulator, to allow incentives for management and prevent business being "captured" by trade unions. Making a profit, while preserving a fixed asset in good order, is a discipline missing within public bodies.
• Micro-competition could be encouraged with new entrants allowed to enter niche markets where new technologies or business/funding/ownership approaches can offer users a substantial benefit. The regulator needs to prevent "cherry-picking", where an operator runs the same technology past the same premises, but only in the cheapest deployment areas. Providing a universal service at a consistent, affordable price is a social benefit that implies some degree of cross-subsidy.
The on-going challenge is encouraging the operator to both maintain their assets and invest in upgrades, but not to "game" the regulatory system with over-investment or other strategies, such as we've seen with New South Wales retail electricity pricing in the last 5-10 years.
We know from experience in Australia, courtesy of PMG/Telecom/Telstra, that a self-regulated monopoly does not put the national interest first, that some form of external discipline is needed to replace that normally provided by free market competition. We also know, from the break-up of AT&T into a number of local "operating companies", a number of which recombined, that geographical monopolies are not competitive.
Regulatory environment
While a natural monopoly exists for the CAN and no, or tightly controlled competition, is in the best national interest there, strong competition should be enabled and encouraged for the provision of full retail services. To maximise both the public utility from the CAN and the economic viability of the operator, necessary for efficient operations and minimising customer pricing, there cannot be wide-scale alternatives, they will undermine the network, destroy the economies of scale and inevitably lead to higher prices, limited availability and poorer service.
The problem the regulator has is preventing near-monopoly players, like Telstra, from manipulating the market and driving out competitors, or using predatory pricing in high-demand/low-unit-cost markets to use other services to undermine the profitability of the CAN operator. Telstra can over-dimension a 4G mobile network in dense urban networks and, via "special bundles", price it to undercut the CAN in just those areas.
Technical solutions
NBN Co has currently defined a single network topology, layer 2 Ethernet VLAN-in-VLAN, and is currently rolling-out three CAN technologies:
• Direct-fibre, GPON, to 93 per cent of premises with up to 25 per cent strung on poles, i.e. "aerial".
• Fixed wireless, or 3G/4G mobile to fixed base stations in premises for marginal areas.
• Satellite delivery to remote areas outside the economic areas of GPON and Fixed wireless.
In 30 years, the end of the current NBN plan, still having a copper CAN analogue phone service, ADSL, VDSL and limited Fibre-from-the-Node (FttN) in addition to GPON/Fixed-Wireless will produce an catastrophic environment, very similar to the National Railways until the 1960's: fractured, expensive, inefficient, unattractive to new operators and new investment only limited to the already well-served, advantaged "low-cost" areas.
Deliberately fracturing the market and undermining the economics of a full-fibre GPON roll-out now, will inevitably lead to being trapped in a digital ghetto: prices will be high, speeds low and coverage poor, with no commercial operator willing to invest because competing providers will always destroy their business model.
From the first century of Australian Railways, we know that "good" short-term decisions can, in the long-term, be not just a "drag" on the economy, but a positive millstone, actively preventing wide-spread economic development.
From the Optus/Telstra cable TV rollout, we know that once it is clear a business model is destroyed, no further investment will be made.
Like it or not, the only technically correct and viable customer access network in 15 years is one technology per service area. After the sunk costs are recovered, the most profitable, least expensive to run and most upgradeable/extensible is a full-fibre GPON for the majority of premises.
Commercial arrangements
Transferring ownership of NBN Co from pubic to private hands only after it has established a reliable positive cash-flow, so that it can be reasonably priced by the market, is essential. Preventing Telstra, or any other retail service provider from controlling a privatised NBN Co is also essential. This suggests wide-ownership and a preponderance of small shareholders is in the national interest.
The core objection the Coalition has to the current NBN plan is centred around costs and funding: it is unclear from their many statements just what they object to...
Is it solely government funding, is it that $30 billion is too large for their tastes, do they want to limit ownership to a select few large investors? Do they want Telstra to once again control the access network or is it a solely partisan political stance?
What may be more useful to explore is non-centralised funding and ownership models:
• NBN Co, if it offered good, guaranteed returns, could access some of the large pool of "DIY" superannuation funds, either as fixed-rate bonds, equity (shares) or pre-paid instalment costs - a lease/buy-back scheme offered to small- and mid-size investors, not the large end of town.
• Promotion of local Co-Op building and ownership of small regional access networks to NBN Co specifications. Operation of the Co-Op Networks would be by NBN Co. Maintenance could be contracted out or undertaken by locally employed staff.
So what sort of a digital network infrastructure do we want to see by 2050? Well here are a few things I would like to see and a few we could do without.
Desirable
• A single, national solution offering ubiquitous, universal digital network access with simple pricing regimes, internationally competitive pricing for basic and high-end bandwidth, latency and volume.
• A privately owned and viable major Access Network Operators with possibly other small niche Access Network Operators.
• A highly competitive, multi-player Retail Network Services sector.
• Per service area, an economically sustainable choice of connection types.
• A competition, regulatory and commercial regime that encourages efficiency, investment, good asset stewardship and non-predatory competitive behaviour.
Undesirable
• Telstra still as the dominant, controlling player across all sectors.
• A fractured, inefficient, divided and undercapitalised/under-maintained access network and retail services.
• Highly variable and complex local charges, unavailability of basic or premium services enforced by geographical boundaries.
• Highly variable services and "preferred areas": an entrenched social and digital divide.
• Disincentives or perverse incentives for investment, upgrade and asset maintenance.
Unfortunately, neither of the political parties currently has a policy that adequately addresses all of the issues and for the time being that situation looks unlikely to change.
This is an edited version of a blog post originally published on November 25. Steve Jenkin has spent 40 years in ICT in wide variety roles including large and small software projects, 7 years writing real-time Exchange software in a Telco and Admin, Software and Database work on PC's Unix/Open Source software and mainframes.