Solution space

Front-loaded three-waters and transport infrastructure unlocks land for development at density. The bottleneck is fiscal: a typical greenfield growth area needs $30k–$80k of trunk infrastructure per dwelling before a single building consent can be issued, and council balance sheets cannot fund this at the scale the housing supply-side constraint requires without a new revenue instrument.

Concrete tools available: Infrastructure Funding and Financing Act 2020 (IFF) levies — used for Tauranga's Tauriko West and Wellington's Eastern Porirua, with a debt-service levy on the benefiting properties for up to 50 years; targeted-rate value capture (Auckland's Watercare Infrastructure Growth Charge); development contributions (Section 199 LGA, currently capped well below true marginal cost). Crown Infrastructure Partners and the National Infrastructure Commission can underwrite trunk infrastructure ahead of demand. Three-waters reform (now Local Water Done Well) is the structural pathway for consolidating the small-council fiscal gap: Council-Controlled Organisations or shared-service entities can borrow at scale where individual councils cannot.

Trade-off: IFF levies and targeted rates work, but transfer cost visibly to specific properties; political durability has been mixed. Pure rate-funded build is more equitable but slower. Crown grants are fast but compete with health and education in the same fiscal envelope.

Auckland's 2016 Unitary Plan upzoning is the largest natural experiment available. Greenaway-McGrevy and Phillips (2023) found that the upzoned tier produced roughly 5× the consent rate of the previous decade, with rents in upzoned suburbs growing 14–35% slower than control suburbs over 2016–2021. The Medium Density Residential Standards (MDRS, 2021) extended the principle nationally; the 2024 Government partly retracted them, leaving councils to opt in.

Concrete tools: reinstate or strengthen MDRS at council level (Christchurch's "City Plan 2024" went further than the central mandate); abolish parking minimums (already national since 2022); remove minimum lot sizes; enable up-to-six-storeys around transit stations (the original MDRS three-storey rule was a partial measure); fast-track consenting for projects above a density threshold. The cost to government is near zero; the political cost lands on incumbent owners whose viewshafts and neighbourhood character protections are removed.

Trade-off: zoning reform addresses the housing supply pattern directly but cannot work where infrastructure is already at capacity (paper upzoning). It is the cheapest lever and the politically hardest: there is no fiscal cost to spread, only a capital-gain redistribution from owners of currently-protected single-house land to owners of newly-developable parcels.

Cyclone Gabrielle's $9–14B national damage and the Auckland 2023 Anniversary floods exposed the compound hazard exposure pattern: property is being built and re-built in locations where physical risk is rising faster than insurance and rates can absorb privately. Toka Tū Ake EQC caps natural-hazard cover at $300k (residential), with the private market layered above; private insurers are already withdrawing or risk-pricing flood cover in coastal Hawke's Bay and Tairāwhiti.

Concrete tools: extend Toka Tū Ake EQC's risk-based premium pricing (currently flat by region); mandate hazard disclosure on property sale via the LIM (currently inconsistent across councils); fund a managed-retreat statute (the Climate Adaptation Bill, 2024 select committee, was the vehicle and is currently stalled); allow councils to refuse new consents in modelled high- risk zones (NZ Coastal Policy Statement 2010 already permits this; few councils use it). The low adaptive capacity pattern is largely about these tools sitting idle.

Trade-off: every step transfers cost from future general taxpayers to current homeowners in identified hazard zones. The precedent in NZ — Christchurch red zone, Hawke's Bay and Auckland voluntary buyouts — is for the Crown to cover roughly 100% of pre-event valuation; risk pricing would shift that closer to market value and lose owners equity. There is no version of this lever that does not generate that distributional fight.

Effective density (jobs and workers reachable within 30–45 minutes) drives the productivity gain from city scale that the agglomeration deficit pattern identifies as missing. Auckland's Northern Busway demonstrates the mechanism domestically: opened 2008, by 2019 carrying 5.4M annual trips and supporting roughly 30% population growth in catchment suburbs above the regional average. The City Rail Link (open 2026) doubles inner-rail capacity and is forecast to lift central-city employment access by 30% within 30 minutes by rail.

Concrete near-term moves available: complete the Northwest Busway (Auckland) — currently consented, awaiting funding allocation in the 2027–30 GPS; reinstate Auckland light rail in some form (cancelled 2024) or commit to a heavy-rail extension to the airport; deliver the Wellington Airport-to-Hospital mass transit proposal (Let's Get Wellington Moving residual); progress the Christchurch Mass Rapid Transit business case (Greater Christchurch Spatial Plan 2024). Each is shovel-ready or near-ready; each requires a 10-year capital commitment that crosses electoral cycles.

Trade-off: rapid transit is high-capex, slow-payoff. Capital cost runs $200M–$400M per kilometre for surface rail/light rail, with break-even on agglomeration benefit measured in decades. Politically, rapid transit projects compete for the same NLTF and Crown capital as state highways with shorter payback periods and clearer regional vote distributions.

The primary sector concentration pattern and the tourism volatility pattern describe regions where one or two price-taking sectors dominate employment: Southland's Tiwai Point smelter (~1,000 direct jobs, 13% of NZ electricity consumption); Marlborough's wine (~28,000 ha, ~$2B exports); the East Coast's plantation forestry; Otago's Queenstown tourism (~2M pre-COVID visitors). Each is a regional concentration risk, not a national one — the policy lever has to be regional too.

Concrete tools available: R&D Tax Incentive (15% credit, ~$650M annual uptake by 2024) for companies expanding non-commodity capability; Callaghan Innovation grants (Project Grants up to $400k, R&D Loans, Student/Career grants) — Callaghan was announced for restructure in 2024 but the funding envelope remains; NZ Trade and Enterprise Better by Lift programme for export-ready firms; Regional Strategic Partnership Fund ($150M reset 2024) for anchor-project investment in transition regions like Southland and the West Coast. Specific national success cases — Rocket Lab, Xero, A2 Milk, Fisher & Paykel Healthcare, Pushpay — show that high-value export specialisation does scale from NZ; the question is whether the policy stack reproduces it deliberately or only by accident.

Trade-off: regional diversification works at decade-plus horizons; the affected workforce needs interim income support. Just Transitions Partnership (Taranaki, Southland) is the closest existing template — a regional taskforce co-funded by central government with iwi and industry seats — but it has delivered limited measurable workforce transition to date. The digital connectivity gap is a precondition: any cloud-software, design, or remote-services sector needs symmetric fibre and 5G coverage that the Rural Capacity Upgrade and RBI extensions are still rolling out.

The four patterns added in 2026 — AI/automation labour exposure, AI productivity capture asymmetry, ageing workforce shortage, and rural population decline — are not independent. They interact in ways that change which levers above are still available.

Ageing reduces working-age population (the old-age dependency ratio rises from 26 to ~42 per 100 working-age by 2048); AI offers a productivity offset, but mostly to frontier firms in Auckland/Wellington; rural regions face the steepest demographic contraction with the least capacity to capture the AI offset; net migration lands disproportionately in Auckland. The four patterns reinforce each other.

Concrete tools, ordered roughly by readiness:

• SME AI diffusion. Extend the R&D Tax Incentive to cover off-the-shelf AI tooling adoption (currently excluded as "routine"); fund Callaghan management-capability vouchers for firms with <100 staff; replicate Singapore's SkillsFuture credit at NZ scale (~$500/working-age adult/year) as a portable training entitlement. Cost: ~$200–400M/year for meaningful coverage.
• Regional immigration loading. The Skilled Migrant Category already includes a "regional area" bonus (30 points). Deepen it: time-limit Auckland-region work rights for some visa categories (already done partially under Accredited Employer Work Visa); create a streamlined regional residency pathway for health, education, and trades workers outside Auckland. Cost: low; this is mostly Immigration NZ rule-change.
• Active labour-market policy. The Income Insurance Scheme (deferred 2023, design ratified) would provide 80%-of-wage cover for up to 6 months following redundancy with active reskilling tied in. Specific design choice: whether levy is on employers, employees, or general taxation determines who bears displacement cost. Cost: actuarially ~1.4% combined levy.
• Health workforce. Te Whatu Ora's Health Workforce Plan needs a near-term capacity injection: bonded scholarships for nursing/medicine in exchange for 3-year rural posting; recognition of overseas qualifications for registered nurses (currently a 6–18 month bottleneck); immigration fast-track for senior medical officers (already on the green list). Cost: low compared to the cost of unfilled positions.

None of these tools is a New Zealand original; all are represented in the OECD policy literature. What is locally distinctive is the constraint set: a small economy with thin institutional capacity outside the main centres, a fiscal trajectory that is tightening not loosening, and a political economy in which 25-year structural levers compete for attention with three-year electoral cycles.