Summary

  • New Zealand’s Commerce Commission says mobile tower ownership has moved to specialist TowerCos, but the statutory access-provider definition still points to the company operating the mobile network.
  • The Commission has started an investigation into whether to remove the regulated service. The hinge is whether commercial tower access and network sharing now constrain the market well enough without a rule written for an older ownership map.

A rule looking for its old landlord

The tower did not move. The legal subject did.

New Zealand’s mobile co-location service has sat in Schedule 1 of the Telecommunications Act since 2001. It lets one mobile network operator mount transmission and reception equipment on another operator’s tower. In the old arrangement, the operator running the radios and the operator owning the steel often occupied the same corporate box. A rule could name the network operator and still reach the person who controlled the site.

That box has split. The Commerce Commission’s February 2026 decision records that all three mobile network operators sold their passive tower assets to specialist infrastructure owners. Those sales included long-term rights to occupy tower space. Radio networks stayed with the operators; towers and the business of filling them with tenants moved elsewhere.

The statutory definition did not make the same trip. Schedule 1 calls the access provider “every person who operates a cellular mobile telephone network.” The Commission says that wording does not capture the TowerCos that now own the towers covered by the obligation. Under the present description, those tower arrangements are not regulated. The Commission says that if regulation is still needed, the service description—potentially the definition of access provider—would likely need amendment.

That is an unusually clean market-design problem: the obligation kept the old address after the asset changed hands. Parliament’s sentence still rings the operator’s doorbell; the tower keys are elsewhere.

Why the old rule mattered

Co-location is not simply a way to avoid buying a patch of land. It can change whether an entrant can put equipment on a useful site at all. The Commission’s account of earlier reviews says regulators kept the service because mandatory access could help a third network enter or expand, speed competing infrastructure deployment and reach remote areas.

The physical alternatives carry a cost. An access seeker can place equipment nearby rather than on the same tower, or build a new site. The Commission says those options take more time and money; the best locations may already be occupied, leaving a new build with weaker coverage. A tower-sharing right can therefore matter most where the map is sparse and a second site is an expensive substitute for a first-class one.

The old case has also changed. Rural coverage has expanded through the government-appointed Rural Connectivity Group’s shared radio access network sites. The Commission says those sites give all three networks access to rural coverage and expects little future demand for co-location as an economical way to extend coverage. It sees commercial tower owners as having a reason to add tenants: an empty position earns nothing from another operator.

That logic has force. A specialist owner selling space to several tenants has a different incentive from a mobile operator deciding whether to host a rival on a site it also uses. The new owner can make filling the mast its core business. A network operator can treat the same request as a strategic nuisance. That difference is why tower separation can weaken the case for a rule aimed at operators.

But the commercial-incentive argument does not answer the statutory mismatch by itself. It says why a TowerCo may want another tenant. It does not show what happens when a location is scarce, a request is technically awkward, or the terms make access unattractive. The Commission’s public decision lays out the incentive and the alternatives; it does not publish a comparable record of tower-access requests, refusals, prices or disputes under the new ownership structure.

The regulator has opened an investigation, not closed the file

On 24 February 2026, the Commission found reasonable grounds to investigate whether Mobile Co-location should be omitted from Schedule 1. That is the start of a separate investigation, not the deregulation decision itself. The Commission’s current project list continues to describe the Mobile Co-location Investigation as work on whether continued regulation is necessary to promote competition.

The distinction matters because the Commission has two live questions, not one. First: do RAN sharing and commercial TowerCo access constrain the market enough that a mandated service has outlived its purpose? Second: if the answer is no, can the rule be redrawn to name the parties that actually own or control the sites?

The first question concerns competition. The second concerns plumbing. A well-designed access obligation can still be useless if it attaches to the wrong legal person. Conversely, a corrected definition can still be unnecessary if commercial supply is contestable and the alternatives are adequate.

The Commission’s own record gives both sides material. It says network coverage has become extensive and overlapping, rural RAN sharing provides another route to coverage, and tower owners have a commercial incentive to maximise tenants. It also says co-location can help a new network reach sites where another build is slower, costlier or inferior. Neither case can be settled by looking at the tower company’s ownership chart alone.

The missing table

The decisive evidence would be mundane, which is why it matters: requests for tower access; how often they are accepted, refused or delayed; the reasons for exceptions; the terms offered; and the outcomes of disputes. Compare those records across locations and over time, and the investigation can test whether the new commercial market behaves like a competitive supplier market or a collection of indispensable gates.

The published decision does not provide that table. Without it, a reader can see the changed incentives and the legal mismatch, but cannot measure the strength of commercial access in practice. A clear record of easy entry on predictable terms would strengthen the case for removing the rule. Repeated refusals, long delays or site-specific terms that defeat entry would strengthen the case for a revised obligation that reaches TowerCos.

New Zealand has not yet chosen between those outcomes. It has identified the problem correctly: the towers have a new owner, and the legal guest list still starts with the old one. The next report must show whether the new landlord opens the gate because business says so—or whether the market still needs a rule with the right name on it.

Sources