Rationing consents is not an energy policy

The genie and the grid

Green Party co-leader Chloe Swarbrick has called for a one-year pause on resource consents for new AI data centres until stronger national rules are in place. Her trigger is the Datagrid facility at Makarewa in Southland, a $3.5 billion, 78,000 square metre AI data centre that already holds consent.

Swarbrick has a point about the scale. The Datagrid centre would draw 280 megawatts of power, making it the second-largest single electricity user in the country behind the Tiwai Point smelter. That is roughly 6% of New Zealand’s total electricity supply in one building. In March 2026, University of Auckland computer science lecturer Ulrich Speidel noted that 280MW represents about 70% of Christchurch’s entire electricity consumption, warning “you can’t just add major power users without eventually running into generation limits.”

He is right that the constraint is real. The trouble is that a consent freeze does nothing about it.

A freeze doesn’t fix the market

A consent moratorium is a blunt instrument aimed at the wrong target. It would not build a single megawatt of firm generation, nor reform a single market structure. It substitutes political rationing for price signals.

That matters because the underlying diagnosis is structural, not a matter of too many consents. A government-commissioned Frontier Economics review found that New Zealand’s electricity market is not equipped to support the scale of new data centre demand without major structural reform. Frontier warned of “irrevocable harm” unless bold changes improved security of supply, addressed dry-year risk and drove investment in firm, dispatchable generation. It recommended consolidating the country’s 29 electricity distribution businesses into super-distributors and merging the Electricity Authority with the Gas Industry Company.

None of that is delivered by pausing consents. The regulatory vacuum is the real story. A McGuinness Institute technical paper published in January 2026 examined exactly this gap. New Zealand has no national framework requiring large power users to co-invest in firming generation, no transparent connection pricing reflecting the true cost of grid stress, and no demand-side signals that would force hyperscale operators to internalise their grid impact.

Price the grid, don’t ration it

The smarter policy is to make big users pay for the stress they create and reward them for easing it. Ollie Hill, New Zealand country manager for Schneider Electric, has argued that data centres could become active participants in grid stabilisation rather than passive consumers, but only with “smarter use of energy, better visibility over demand, more flexibility across the system, and practical investment in storage, efficiency and resilience,” not blanket restrictions.

There is capital ready to move. Mercury Energy has taken a 12.7% minority equity stake in Datagrid NZ for $53 million, following a 140MW power purchase agreement signed in March 2026. That is a generator putting its own balance sheet behind the demand, the kind of co-investment a proper framework would encourage rather than block.

The $30 billion at stake

Here is where the moratorium gets expensive. Invest New Zealand is targeting up to $30 billion in offshore data centre investment over five years. New Zealand currently runs 56 facilities consuming around 0.6% of total power, so Datagrid alone would lift the sector’s footprint dramatically.

Swarbrick cites Australia’s new Office of AI and New York’s hyperscale pause as precedents. But both paused to set standards and then resumed consenting. Neither treated the freeze as the destination. The risk in New Zealand is that a one-year moratorium, with no clear framework process behind it, becomes the default through political inertia, chilling investment while the actual market problems sit unresolved.

What this means for business

Three things are on the line. First, whether New Zealand can credibly host digital infrastructure at scale or gets filed as a second-tier destination for hyperscaler capital. Second, whether grid stress from large new users flows through to higher commercial power bills if it isn’t managed through proper pricing mechanisms. Third, whether Southland’s regional development case, already fragile with Tiwai Point’s contract running to 2044, is undermined by a policy pause that signals ambivalence about big investment.

Swarbrick is right that New Zealand needs rules. She is wrong that a consent freeze is one. The fix is transparent connection pricing, firming obligations on large users, and a distribution market that actually functions, not a year of doing nothing while $30 billion waits at the border.

Sources

• Urgent need for rules and protections on data centres, Greens say (2026-07-26)
• A massive new power user could put extra pressure on the South Island’s electricity system (2026-03-13)
• Exploring Data Centre Infrastructure in New Zealand – McGuinness Institute Technical Paper (2026-01-12)
• NZ’s electricity market needs reform to meet data centre demand
• Data centres could help stabilise New Zealand’s power grid, Schneider Electric boss says
• Tech NZ aims to dispel fears as Crown looks to build more in NZ