Section 30 R review of Chorus UBA Service

30 November 2016

View the PDF of this submission here


InternetNZ welcomes the opportunity to submit on this draft determination under Section 30R of the Telecommunications Act.

We welcome further opportunities to discuss our views. Please contact Reg Hammond on 021569980 or

The focus of this submission is on one key aspect of the S30R review - the service description of the UBA service and how it can be made "fit for purpose" for all New Zealand end-users.

InternetNZ's vision is "A better world through a better Internet"

Our mission is to promote the Internet's benefits and uses and protect its potential. We act as part of the New Zealand Internet community, seeking wide input on what is important for the Internet in New Zealand and how best to respond.


In InternetNZ's submissions into the Final Pricing Principle (FPP) Determination, which set the regulated prices of copper broadband services, we made the strong point that any price determination made without a revised specification of the service(s) being priced would lead to distortions. The most obvious distortion is that the price calculated was based upon a hypothetical service of a minimum 450kbps increasing by 50% per annum. A service standard that Chorus could not provide for many New Zealand endusers and in the case of ~20,000 rural end-users are unlikely ever to provide.

Our expectation was that in the absence of a service specification as part of the FPP determination that this belated S30R review would correct the anomaly created whereby Chorus is subsidised to provide a service that it does not provide. In this draft determination, rather than correcting the anomaly the Commission has likely chosen to perpetuate it.

Commission's draft determination

The Commission's draft determination has concluded that the UBA service should be fit for purpose and should deliver appropriate quality of service suitable for a range of general Internet use and is dynamic and should improve incrementally to meet end users current and future needs. That largely conforms to InternetNZ's interpretation as well - but the Commission avoid saying in any measurable terms what fit for purpose and appropriate quality of service actually mean.

The Commission goes to some length to emphasise that its determination is consistent with the purpose statement of the Act and its decisions are intended to best promote competition - this is particularly used to discuss the issue of inefficient investment by Chorus and is the basis upon which it has carved out the legacy (rural) ATM network from the draft determination. The Commission considers that the Government's RBI initiative may better address these rural issues and it is effectively deferring any decision in this area until there is greater clarity about the RBI. We interpret this to mean the Commission do not want to force Chorus via regulation to make inefficient investment that might be overtaken by future RBI decisions.

In regard to the service description rather than attempt to define what a fit for purpose or appropriate quality of service actually means the Commission has opted to focus on line congestion between the DSLAM and Chorus' first data switch (referred to as the local aggregation path or LAP) and believes that the removal of congestion will mean that end-users will get a fullspeed/full-speed service with no requirement for up or down stream speed to be specified. Monitoring of the congestion will in essence mean that Chorus will have to continue to upgrade this section of the network to prevent congestion as anticipated consumer demand increases. This decision seems to rely on two major assumptions which we consider to be flawed - first that congestion can be accurately defined and measured and secondly that congestion is the only major problem. We will look at those issues separately later.

Commission response to InternetNZ submission

In reaching its decision on "fit for purpose" the Commission considered two options: the InternetNZ recommendation to replace the current 32kbps minimum throughput requirement with an updated number, such as 450kbps increasing by 50%pa, or; a new service metric that requires Chorus to maintain uncongested links - they have chosen the latter, their view being that the STD should be updated to more directly link end-users' needs with Chorus service requirements. In rejecting the suggested minimum throughput requirement the Commission said that while InternetNZs approach would ensure that end users' current bandwidth demands are met on average, the actual throughput provided by Chorus might not keep pace with end-users needs on specific links; it might result in inefficient outcomes, either by requiring Chorus to invest in additional capacity where it is not required or alternatively lead to congestion where end-user demand has grown at a faster rate than the minimum.

We interpret that as the Commission saying Chorus will be incentivised to build the network to the minimum specification which might lead to inefficient outcomes. It almost certainly would lead to inefficient outcomes, if it happened, but InternetNZs submission called for a combined option – that is a full-speed /full speed service combined with a minimum specification – not one or the other.

While hypothetically a combined option might result in Chorus having to make some inefficient investment where it may not be required (although it is difficult to see an end-user not wanting a minimum of 450kbps) it would certainly not lead to increased congestion. In any case, that inefficient investment has effectively already been accepted by Chorus (which proposed the 450kbps minimum in the first place) and the Commission with the "averaged" FPP price being set at a level to compensate Chorus to make that inefficient investment.

If the Commission can conveniently carve out the rural ATM network from the service specification during this S30R review, it should have also carved it out from the pricing determination - this was one of the key reasons InternetNZ and others wanted the STD review to take place in parallel with the FPP determination.

The net-result of carving out the rural network from the determination is that subsidies paid to Chorus (through regulated price averaging and the FPP price determination) in order for them to provide a “fit for purpose” service 4 are not being used to that end. ~20,000 end-users are once again left in limbo – worse still, by tying the decision to the Government’s choices on its RBI contract allocations it has the potential to skew those contract decisions. Chorus is already being subsidised through the FPP price while its competitors are not.

As a minimum perhaps Chorus should be required to disclose its investment intentions and show where the FPP subsidy is being applied so that it can be discounted from any RBI application.

For consistency the Commission should also explain how carving out ATM from the service specification while continuing to subsidise Chorus via "averaged" pricing improves competitiveness.


Returning now to discuss the two assumptions the Commission has made, namely that congestion can be accurately defined and measured and secondly that congestion is the only major problem and it will be addressed by the new service specification.

Our reading of the amended STD indicates that individual LAPs will be monitored on a monthly basis and any LAP in any 15 minute period during that month must not reach 95% of utilisation and where the monthly report shows peak utilisation is greater than 85% Chorus must provide plans. As discussed above the Chorus ATM network is excluded from these requirements.

While it is regrettable, but partly understandable, to remove the ATM network from any implementation requirements there would seem to be no reason to remove it from the measurement and reporting requirements. It would enable parties to see for themselves how poor the service actually is.

It is also not clear to us what meaningful sanction there would be for breaching the 95% threshold and what incentive there is for Chorus to do anything other than report and provide a plan if utilisation is greater than 85%. There needs to be some tangible and binding requirement on reaching both of these measures.

We also have some concerns about possible interpretation or definition of terms. The amended STD seems to be a reasonable standard for a congestion free LAP however the actual wording will be important. For example is it a peak throughput measured in the 15 minute period or the average throughput? How is an average derived? - if the measurement device has boundaries of 0% and 100% then any instance of congestion over 100% will only be recorded at 100% and distort the average. Also, the impact on an end-user of a single instance of congestion, however short, is likely to be far more significant than a sustained average. As an analogy a single instance of a driver passing a speed camera at 110 KPH is a much more significant event than driving a long distance at an average speed below 100KPH.

We also have concerns about how the ATM network is defined and what it actually refers to in terms of the physical network. At the Commission’s workshop Chorus and Spark, as the previous Telecom partners, appeared to have the same understanding of what they meant by the ATM network while few others did. It also seemed clear at the workshop that the ATM network was changing day by day. The Commission may need to specify this in much more detail and specify a process for managing LAPs moving onto or off the ATM network.

In regard to the assumption that congestion is the only, or even major, issue that needs to be addressed our understanding is that there are large numbers of rural users who do not currently get the minimum 32kbps and that there are even more rural users who know that any service they get will be so impaired because of loop length and interference problems that they do not even apply for a service. Removing congestion on LAPs will not solve loop length and interference problems.

There are ~20,000 rural end-users for whom an un-congested full-speed/fullspeed service will be of little or no benefit whatsoever. If the Commission specified the service standard as an un-congested, full-speed/full-speed service with a minimum throughput of 450kbps increasing by 50% per annum those end-users would be able to get a service that was genuinely “fit for purpose”.

About InternetNZ

InternetNZ's vision is for a better world through a better Internet. We promote the Internet's benefits. We protect its potential. And we focus on advancing an open and uncaptureable Internet for our country.

We provide a voice for the Internet in New Zealand and work on behalf of all Internet users across the country.

We are the designated manager for the .nz Internet domain. And through this role we represent New Zealand at a global level.

We provide community funding to promote research and the discovery of ways to improve the Internet. We inform people about the Internet and we ensure it is well understood by those making decisions that help shape it. Every year we bring the Internet community together at events like NetHui to share wisdom and best practice on the state of the Internet.

We are a non-profit and open membership organisation.

Be a member of InternetNZ and be part of the Internet community. You can keep a close watch on the latest tech and telecommunications developments and network with other like-minded people at cool events. Being a member of InternetNZ only costs $21 per year. Find out more at

Thursday, December 1, 2016