Cases for the Practical State explores what good public service looks like in practice: the places, relationships, and operating choices where outcomes are actually made real. This week turns to Kaikōura. When the 2016 earthquake severed Aotearoa’s main north–south transport corridor, the recovery could have defaulted to siloed agencies and slow procurement. Instead, government, iwi, and contractors built an Alliance that showed what happens when capability is matched to complexity: decisions were integrated, risks were shared, and legitimacy was sustained through community-facing governance. The result wasn’t just a rebuilt highway and railway. It was a demonstration of how the state can choose to organise itself differently when the challenge demands it.
On 14 November 2016, a 7.8 magnitude earthquake severed Aotearoa's primary north-south transport corridor. State Highway One and the Main North Line railway disappeared beneath massive landslides along the Kaikōura coast.
The immediate damage was evident: 200 million tonnes of rock and debris blocking the route, 21 major slips, and bridges twisted beyond repair. The economic implications were equally stark: $1.2 billion in annual freight movements were halted, tourism access to Marlborough and the West Coast was cut, and the small town of Kaikōura was isolated, except by helicopter.
To understand what happened next, and why it succeeded where similar projects often fracture, requires examining the response through what the New Zealand Institute of Economic Research calls the "Three Lenses" of durable policy: political feasibility, policy effectiveness, and administrative workability (Nixon, 2016). †
According to the framework, policy decisions endure when they meet all three conditions simultaneously. Fail on any one, and the odds of collapse rise sharply. The Kaikōura recovery, and the land transport infrastructure elements in particular, offer a clear demonstration of how these lenses operate in practice, and what happens when the government organises itself to address all three simultaneously.
What followed over the next twenty months after the quake was one of the most complex infrastructure recovery projects ever attempted in Aotearoa. But it was also something rarer: a demonstration of how the state can organise itself differently when complexity demands it.
The conventional response would have been predictable: parallel workstreams, separate contracts, careful delineation of agency responsibilities. Waka Kotahi would manage the highway, while KiwiRail would manage the railway, with coordination through inter-agency meetings and shared reporting to ministers. Iwi and community consultation would proceed through established channels. Risk would be allocated according to standard templates.
Instead, the lead agencies made a different choice. In early 2017, they established the North Canterbury Transport Infrastructure Recovery Alliance, with a single, integrated Board that brought together the state agencies, Te Rūnanga o Kaikōura as a Te Tiriti partner, and private contractors under one decision-making structure.
The Board held unified budget authority across both road and rail recovery, with delegated powers that bypassed normal procurement and approval processes. Risk was pooled rather than allocated, with cost overruns shared proportionally across all partners.
This was not crisis expediency but deliberate design. The Alliance operated under an explicit kaupapa of "people first", requiring all major decisions to be assessed not only for technical merit but for community impact and cultural appropriateness (NCTIR, 2021).
Regular public meetings were held in Kaikōura township. Local suppliers were prioritised where possible. Local tikanga was integrated into site management, including karakia before major earthworks and restrictions during culturally sensitive periods (NCTIR, 2021).
The Board met weekly throughout the recovery, with full cost and progress data shared in real-time across all partners. When unexpected discoveries required design changes, as they did repeatedly, decisions could be made within days rather than months. When community concerns arose about noise or dust, responses were coordinated immediately, rather than being filtered through separate agency hierarchies (NCTIR, 2021).
The Alliance's design choices become particularly revealing when examined through each lens of the NZIER durable policy framework in turn. Rather than treating political feasibility, policy effectiveness, and administrative workability as competing demands to be balanced, the Alliance's governance and operating model was deliberately crafted to reinforce all three simultaneously.
Political feasibility was secured through constitutional recognition and community engagement. By treating Te Rūnanga o Kaikōura as a governance partner rather than a consultee, the Alliance acknowledged the Iwi's constitutional status while avoiding the parallel consultation processes that often slow major projects. Regular community meetings maintained public support, and therefore political support, even when the work created significant disruptions (Rennie et al., 2018). The unified governance structure also meant ministers received consistent advice rather than competing recommendations from different agencies.
Policy effectiveness was built into the design through integrated planning and shared objectives. Rather than optimising separately for road and rail recovery, the Alliance sequenced work to minimise total disruption and maximise the use of shared infrastructure. Community outcomes were treated as equally legitimate to engineering outcomes, preventing the relegation of Treaty obligations or local needs to afterthought status. The "people first" kaupapa ensured that technical solutions were tested against community expectations throughout the process.
Administrative workability emerged from the Board's unified authority over budgets, timelines, and approvals. Cost variations could be approved immediately, rather than being subject to separate agency processes. Procurement could prioritise speed and integration over rigid compliance with standard frameworks. Risk pooling meant partners focused on solving problems rather than protecting individual positions.
This does not mean there are no downsides to the governance model: the unified Board concentrated significant power in a relatively small group, and the focus on speed and integration, with hindsight, means that less time was spent on detailed option analysis that might have identified superior technical solutions.
However, despite these downsides, the success of the governance and operating model was primarily due to the quality and alignment of individual leaders. Had key personalities clashed or pursued different agendas, the integrated structure could have become a source of paralysis rather than efficiency, but it didn't.
The Alliance won the 2020 Public Service Award for Leadership in Governance due to its integrated, transparent, and community-facing approach. The judges called out the Alliance for setting a new standard in how large, complex infrastructure projects can be governed by uniting agencies, iwi, and contractors under one kaupapa, delivering results on time, on budget, and with the community at the centre.
The immediate outcomes were clear. The highway reopened in December 2017, thirteen months after the earthquake. The railway followed in September 2018. Total recovery costs were $1.2 billion: well within the original estimate despite the project's unprecedented complexity. Independent assessment found that traditional procurement approaches would likely have taken 30-50% longer and cost significantly more (NCTIR, 2021).
But the deeper test of the Alliance model lies in its subsequent adoption. Since 2018, the Alliance approach has been used for billions worth of infrastructure projects across Aotearoa. Sure, each adaptation requires significant modification because the governance and operating models that work for earthquake recovery do not always translate directly to planned infrastructure development. However, the core principles of integrated decision-making - people-first, outward-facing, and shared risk - have proven durable.
What happened at Kaikōura was not simply good project management. It was a demonstration of how the state can organise itself when capability is matched to complexity rather than constrained by institutional boundaries. The Alliance succeeded not because the technical challenge was easier; it was among the most difficult civil works ever attempted in Aotearoa, but because the Alliance was designed to sustain coordinated action over time, with active attention to the community and the Alliance’s social license.
This matters beyond infrastructure. Most of the challenges facing our nation, from climate adaptation to housing affordability to social inequality and inequity, require exactly this sustained, integrated response across traditional boundaries.
Our shared problems are too complex for single agencies, too interconnected for sequential treatment, too urgent for lengthy coordination processes, and too complex for ivory-tower treatments.
The Kaikōura recovery offers a template, not a blueprint. Its lessons are about holding multiple lenses in view simultaneously, about accepting that government work is always a constitutional act that shapes the nature of the state itself, and about choosing deliberate coordination over accidental fragmentation. These are choices available to the government whenever complexity demands them.
The question is whether we choose to make them.
Notes
† With thanks to Derek Gill, who not only peer-reviewed the original NZIER paper (Nixon, 2016), but also called my attention to its core framework. The "three lenses" model of political feasibility, policy effectiveness, and administrative workability has become one of the most valuable tools I've encountered for diagnosing why policies hold or fail. It gave shape to something I was discovering in my doctoral work: that durability isn't just about technical design, but about maintaining complexity in motion across systems, relationships, and time. Additionally, the NZIER framework subtly yet significantly improves upon Mark Moore's public value triangle. Where Moore places public managers at the centre, balancing authorising environments, value propositions, and operational capacity, the NZIER model puts the policy system, elected officials and their constituents at the centre, and names feasibility as a constraint, not ambition. It resists the idealism baked into public value, and instead offers a subtractive logic: one that asks what won't work and removes it, leaving behind the space where real solutions can live. Most importantly, it shifts attention from strategic intention to practical execution. It puts implementation, which I have been arguing is long treated as the poor country cousin, back at the heart of public policy. In doing so, it centres what I call the practical state: the place where discretion, capability, and relational delivery determine whether anything the state promises can be made real. And perhaps most usefully of all, it offers a quiet reminder to those of us doing this sort of research: much of what we discover has, in some shape or form, been found before. The work is not to claim novelty, but to see clearly, connect honestly, and contribute meaningfully to the lineage we're all part of.
*Noting too that the wider Kaikōura recovery and rebuild also did not appear to share benefits evenly: an imbalance later compounded by the impacts of the COVID-19 lockdowns (Rennie et al., 2018; Fountain & Cradock-Henry, 2023).
References
Fountain, J., & Cradock-Henry, N. A. (2023). We’re all in this together? Community resilience and recovery in Kaikōura following the 2016 Kaikōura-Hurunui earthquake. New Zealand Journal of Geology and Geophysics, 66(2), 162–176. https://doi.org/10.1080/00288306.2023.2167842
New Zealand Transport Agency. (n.d.). Kaikōura earthquake response. Waka Kotahi. Retrieved 8 August 2025.
Nixon, C. (2016). Durable policy approaches: Framework development and brief literature review (NZIER Public Discussion Paper 2016/2). New Zealand Institute of Economic Research.
NCTIR (North Canterbury Transport Infrastructure Recovery). 2021. Moving mountains to reconnect communities: the story of the restoration and improvement of Kaikōura’s transportation network. Wellington: New Zealand Government.
Rennie H, Simmons D, Fountain J, Langer ER, Grant A, Cradock-Henry N, Wilson T. 2018. Post-quake planning: tourism and surfing in Kaikoura. In: Hendtlass C, Borrero J, Neale D, Shand T, editors. Shaky shores coastal impacts & responses to the 2016 Kaikoura earthquakes (Special Publication 3rd ed.). Wellington: New Zealand Coastal Society; p. 37–38.
Stevenson JR, Becker J, Cradock-Henry N, Johal S, Johnston D, Seville E. 2017. Economic and social reconnaissance: Kaikōura earthquake 2016. Bulletin of the New Zealand Society for Earthquake Engineering. 50(2):343–351
Thank you for this thoughtful and interesting article.
Is this 'template' being used for multi- site developments? I couple I was thinking about were:
- Building new, and upgrading existing Aged Care Residential Facilities (Almost all are in private ownership, but are struggling to keep up with demand( where there is more than 12,000 bed shortage and growing
- Social, Affordable, or Accessible Housing? Not the 100,000 homes type policy, but concerted efforts to address the human right of an appropriate dwelling for every citizen
Both of these would need multiple sites right across the country. But the opportunity to work constructively within Government, Private, and Community seems a great setting and could be replicated (With modifications) to different regions, and for different complex problems that we currently aren't able to solve.
Thanks Deb. Another really interesting piece.
I'd suggest that the attitude to funding deserves a mention as an important background variable. And I mean that in two ways.
First, as the road was the key freight and lifeline route for the region, something was going to have to be done, even it meant shifting the whole road or strengthening the inland alternative. Neither was likely to be much less expensive.
Second, the PM basically said, what is a public surplus for if not this kind of situation?
But also, within a few days, it was also clear that the National Land Transport Fund had the headway to debt fund the work - as it was still doing with the Christchurch recovery works - if necessary, too.
Easier to align when you take away a possible source of competition.