Trust as infrastructure: What disaster risk reduction still needs to measure

Disaster risk reduction (DRR) practice routinely tracks physical readiness through warning coverage, shelter capacity, evacuation routes, drainage and response times. Those indicators matter, but they can show that a system exists without telling us whether people will rely on it when the pressure is real.

In DRR practice, trust often sits under communication, engagement or inclusion. Yet it can shape whether a protective system is used at all. 

I use "infrastructure" here deliberately. The claim is not that trust can replace sound engineering, accurate forecasts or competent institutions. It is that DRR systems depend on relationships between institutions and the people expected to act on them, and those relationships can strengthen or weaken performance.

Early warning makes the problem visible

Early warning systems are a useful place to see this gap because the final step is behavioural. A forecast can be accurate, an alert can reach the right place, and a response plan can exist. None of that guarantees that people will regard the warning as credible, understand what action is expected, or believe the recommended action is feasible.

UNDRR has begun to make this issue more explicit. A 2025 case study on the Community Trust Index for Early Warning Systems describes an IFRC tool designed to measure trust in multi-hazard early warning systems, the institutions behind them and the information they provide. Pilot work in Mozambique and Nepal identified strong trust drivers, including risk understanding and knowledge, while also finding concerns around transparency, inclusion and participation.

The practical value is straightforward: planners do not have to wait for a failed warning to discover that people distrusted the system. Trust can be examined before an emergency exposes the gap.

The same point appears in evacuation research. After the 2012 earthquake in Banda Aceh, Indonesia, McCaughey and colleagues surveyed 202 residents living near tsunami vertical evacuation buildings. Only 26 per cent had evacuated to one of those buildings; 74 per cent evacuated horizontally. The study found that people who preferred horizontal evacuation did not trust the safety of the vertical evacuation buildings and believed they could reach a safe inland destination in time.

The result should not be read as proof that distrust alone determined evacuation behaviour. Route choice, perceived safety, mobility, social influence and the physical location of shelters also matter. But the case shows why technical provision cannot be treated as equivalent to effective protection. 

A structure may meet its engineering purpose and still be used differently from what planners expect.

Trust is a performance condition

For DRR, trust is not a general measure of whether communities "like" an institution. It is more practical than that. Do people regard the warning source as credible? Do they understand why a recommendation is being made? Have previous failures or false alarms been explained? Can people question information they believe is wrong? Do local experiences influence how preparedness measures are designed?

These questions matter because the same system can be experienced differently across groups. UNDRR’s inclusive early warning and early action guidance addresses governance alongside risk knowledge, monitoring and forecasting, communication, and preparedness, with gender and disability inclusion built across the system. 

Trust cuts across those elements as well. 

Information can be technically available while still being difficult to access, understand or act on for some of the people most exposed to risk.

Trust also needs to be separated from compliance. People may follow an instruction because they have no practical alternative while still viewing the institution behind it as unreliable. They may also rely on community leaders or informal social networks [SS1]   because those channels have proved more dependable in past emergencies. A 2026 Journal of Risk Research study of rural disaster communication found that credibility hinged on source trust and relational embeddedness, especially where formal systems or institutions were mistrusted. For practitioners, those patterns are part of how the system operates, not background noise.

This does not mean adding a single "trust score" to every DRR dashboard and considering the problem solved. A more useful approach is to treat trust as a performance dimension that can be investigated alongside technical indicators.

What should be measured?

A trust-sensitive DRR assessment could ask whether people recognize and believe the source of a warning; whether they understand the action being requested; whether the recommended action is realistically accessible to different groups; and whether institutions explain errors, changing forecasts or failures after they occur.

The answers will not replace measures such as lead time, coverage, shelter capacity or response speed. They add information those measures cannot provide.

The same approach applies beyond early warning. An evacuation shelter, heat action plan, flood relocation scheme or community preparedness programme can be technically well designed while depending on forms of cooperation that are rarely captured in completion reports. In each case, the relevant question is not simply whether an intervention was delivered, but whether the people expected to rely on it see it as credible, usable and responsive to their circumstances.

That also changes the role of participation. Consultation should not be treated as proof that trust exists. IFRC's Community Engagement and Accountability guidance makes the distinction explicit: building trust depends on enabling people to engage and acting on their feedback and needs. A community meeting may reveal useful local knowledge, but the test is what happens afterwards: whether concerns alter decisions, whether responsibilities are clear, and whether institutions remain answerable when something goes wrong.

People-centred has to be tested in practice

The language of people-centred DRR is already well established. The harder task is to test whether systems are actually experienced that way.

Trust cannot make a bad forecast accurate. It cannot make an unsafe shelter safe, and it cannot compensate for weak institutional capacity. A technically sound system may still underperform, however, if the people expected to use it do not regard it as credible, understandable, accessible or accountable.

The emerging work on trust in early warning systems offers a practical direction. It suggests that gaps in transparency, inclusion or participation can be monitored before a crisis reveals their consequences, and treated as operational weaknesses rather than vague social concerns.

If disaster risk reduction is serious about how systems perform under pressure, trust belongs in that assessment alongside technical performance. Protective capacity depends on both: what institutions can deliver, and whether people can rely on the systems they are asked to use.

Rama Nassif, ORCID, is an Environmental Systems Researcher whose work examines how governance, institutional capacity and implementation shape environmental and urban outcomes across diverse contexts. Her research focuses on environmental recovery, urban governance and the institutional conditions that determine whether technical solutions work in practice.