Spark and all have lots to say to Commerce Commission on why it needs to revisit wholesale copper prices for broadband

Spark says Commerce Commission model for cost flawed, warns of higher prices for retail consumers.

Spark wants the Commerce Commission to have a second think about plans to go ahead with a new wholesale pricing structure for telecommunication companies using Chorus’ copper network.

The industry’s input costs – which influence the retail prices for broadband services and landline phones – could rise by at least $12 a month per customer, should the Commerce Commission go ahead with proposed new charges for access to the Chorus copper wholesale network, Spark New Zealand says.

In a submission to the Commerce Commission, Spark New Zealand has called for the Commission to rework its numbers and consult further with industry players before it finalises the new charges later this year.

General Manager of Regulatory Affairs for Sparks, John Wesley-Smith, says Spark New Zealand believed the large increases in wholesale charges proposed by the Commission in a draft decision published last December were unnecessary and should be reversed when the Commission sets the final charges later this year.

He said Chorus – which is laying down the ultrafast broadband infrastructure – has no competitors and it is the Commerce Commission’s primary duty to protect consumers against high prices in monopoly markets.

Chorus, on its side, has also voiced its side of the story, challenging a November 2013 wholesale pricing model used by the Commission. In September 2014, the Court of Appeal dismissed Chorus’ appeal for a review of the High Court’s decision on the pricing principle. 

A very apt picture from Chorus' annual report on the tricky business of regulation

InternetNZ, TUANZ, Vodafone

Submissions made by InternetNZ which represents internet interest groups, and the Telecommunication Users Association of NZ (TUANZ), among others note that: “For every dollar that the price goes up or down, the impact of the Commission’s decision is around $140 million. Even movement of just one variable, such as the valuation approach to reusable assets, could see movement exceeding $1 billion.” The submission also highlights why pricing should not be backdated.

Vodafone's submission states that work by consultants WIK-Consult and Network Strategies found the model used (TERA) had inaccurate assumptions, materially overstating the cost of delivering UCLL (unbuddled copper loop) and UBA (wholesale broadband) services in the country. Vodafone also does not agree with the backdating of prices.

The Commerce Commission’s final pricing is expected to be determined by September 2015, according to the Commission’s website. Other submissions can be found on the Commission’s website.

Model challenged

Spark is also taking objection to how the Commission had modelled the cost structure used to decide on the new charges proposed.

It reckons the charges for landline access (required for broadband and/or voice services, and known as the UCLL Price) is 80% higher than the median price in 14 other comparable countries – and 60% higher than the next most expensive country.

“This is not because of any New Zealand specific factors – it does not cost 80% more to provide landline access in New Zealand than everywhere else on a like for like basis,” Wesley-Smith says.

He adds: “This is the result of choices made in the draft modelling, and in a number of cases we think there are better choices it could make that avoid this significant divergence from international prices.

“Our experts’ analysis shows that the Commission’s draft model loads an enormous amount of unnecessary costs onto New Zealand consumers. We think we’ve shown enough to put the onus on the Commission to change its model to exclude these costs - or if not, to explain why it would be in New Zealand’s best interests to hold broadband prices up by setting wholesale charges 80% higher than in other countries.”

Different views

Spark engaged two international expert firms to review the Commission’s cost model. “By making some conservative adjustments to the Commission’s model, one expert firm came up with a charge for landline access of $16.64 a month – compared with $28.22 proposed by the Commission,” Wesley-Smiths says.

The other expert firm reckons this charge could be reduced even further if the model accounted for use of newer technologies such Fixed Wireless Access (FWA).

“With these adjustments the wholesale charges in New Zealand would be much more in line with those applied overseas – and this would translate into lower retail broadband prices than we have today,” Wesley-Smith says.

Spark increased its prices for home phone and broadband packages by between $2.50 and $4 a month from 1 February 2015, in a move that only partially offset the expected increase in Chorus wholesale charges.

Spark's case so far

In Dec 2014: The Commission released two draft decisions in December 2014, setting out proposed charges for Chorus unbundled copper local loop network service (UCLL) and wholesale broadband service (UBA).  Changes proposed: $28.22 for UCLL (up $4.70 from $23.52) and $10.17 for UBA.

Spark’s challenge: The Commerce Commission’s draft charge for UCLL is 80% higher than the median of 14 comparator countries, including those countries that New Zealanders like to compare to in social and economic terms. It is almost 60% higher than the second-most expensive country.

Spark on why the Commission’s model is flawed:  

a) Does not use modern FWA technology and incorrectly models potential FWA coverage and so overstates the cost of serving non-urban New Zealand by 37%;

b) Compensates Chorus for lead-in costs (which make up 26% of overall network costs) that are actually separately funded by end-users;

c) Makes unrealistic assumptions that an efficient operator building a modern network today would not re-use any existing ducts and trenches, or seek to share the cost of trenching with any other utilities in order to save costs – even though both practices are used by operators in our market today; and

d) Assumes no population growth in New Zealand over the next five years, and no further high or medium density housing projects – despite policies being put in place by central and local Government to drive exactly this sort of urbanisation in Auckland and other areas.

Correcting for those two assumptions alone reduces the wholesale charges in the Commission’s model by almost 10%, Spark says.