Making decisions without perfect answers

Resilience does not begin at the design stage. It starts earlier, with investment decisions about which risks to accept and which to address.

When infrastructure fails, attention often turns to what broke. More useful, however, is understanding earlier decisions and how priorities were set. Many outcomes are determined well before design or construction begins. Decisions about risk, cost and acceptable disruption ultimately define how infrastructure performs over time.

Resilience requires decision makers to confront uncertainty head‑on. We cannot predict exactly when a major event will occur or how it will unfold. What we can assess is likelihood, consequence and tolerance for failure, and that is enough to guide informed investment decisions.

Too often, resilience is treated as a technical issue to be addressed later in delivery. By then, the most critical decisions have already been made.

Resilience planning sits between science and judgement. There is a natural tendency to seek more information. More hazard modelling, more ground investigation, more scenarios and more refinement can feel like progress. In practice, however, this can delay decision‑making rather than support it.

At the same time, acting on limited information brings its own risks. Under‑investigation can lead to misplaced confidence, poorly targeted interventions, cost escalation, delivery delays or solutions that address the wrong problem. The challenge is not choosing between more or less analysis but knowing when there is enough understanding to move from studying risk to managing it.

This is particularly important in a constrained funding environment. Not every part of the State Highway network, for example, can or should be protected to the same standard. The key questions are where risk matters most, what level of disruption is acceptable and when intervention is justified. These are not purely technical questions; they are value judgements that require clarity, transparency and, at times, courage.

GHD’s work on the State Highway 3 Manawatū Gorge transport corridor illustrates how uncertainty plays out in practice.

On 24 April 2017, two moderate to large landslides occurred almost simultaneously within the Manawatū Gorge, around 2 km apart, during a relatively dry period. Both occurred on ageing cut slopes and each site had a history of similar slope activity several decades earlier. However, this history was only widely recognised at one of the sites – Kerry’s Wall.1

Events like this do not represent a failure of analysis or process. They highlight the unpredictable nature of complex natural systems. Even with rigorous investigation, we cannot forecast exactly where, when or how a major failure will occur. Slopes behave differently under real‑world conditions than in models and small changes can trigger disproportionate outcomes.

The lesson is not that risk assessment is futile, but that it has limits and decisions need to reflect that reality.

If we accept that not all events can be predicted, then resilience requires a broader definition. It is not only about preventing failure, but also about how systems respond when failure occurs.

This is where adaptive planning becomes critical. Resilient infrastructure networks are designed with options: alternative routes, staged interventions, scalable responses and decisions that can adapt as conditions change. Flexibility is a deliberate choice, not a compromise.

In the case of the Manawatū Gorge, the triggering event was unpredictable, but the response reflected earlier thinking about network vulnerability, redundancy and long‑term corridor function. The closure was disruptive and costly, but it also prompted a shift from reactive maintenance towards more strategic investment decisions, with a greater focus on network performance and resilience over time.

Perhaps the most challenging aspect of resilience planning is accepting that decisions made under uncertainty may sometimes look wrong in hindsight. A lower-ranked risk may eventuate, and an investment may appear misplaced. That does not mean the decision was flawed, but that the future unfolded in ways that could not be fully foreseen.

Good resilience decision‑making, aligned with the principles of the NZ Transport Agency Waka Kotahi Decision Led Approach, supports this process. It focuses on making informed and defensible choices based on the best available information, clear risk tolerances and an honest understanding of what is not known.

The experience of the Manawatū Gorge, along with GHD’s recent work within the NZ Transport Agency’s South Island Resilience programme, reinforces a simple but powerful message: We cannot predict every failure, but we can prepare for disruption by prioritising the risks that matter most.

For those in New Zealand responsible for keeping people and goods moving, the most resilient infrastructure may not be the one that never fails, but the one that adapts, recovers and continues to serve communities when the unexpected occurs.