For example, a risk assessment of a coastal floodplain might identify a particular area as “low-risk”: few buildings, low asset value, sparse population. A neighbouring community might consider that same area among the most important places on their landscape, a site of customary food gathering, seasonal ceremony, and intergenerational relationship with land and water. This contrast means that in the technical assessment's terms, there is little to lose, but in the community's terms, that place is really important. 

Climate risk assessments are now a standard part of how countries, cities, and communities prepare for climate change. They identify which communities are exposed to hazards like flooding, drought, and extreme heat, and they inform decisions about where to invest, what to protect, and how to adapt. The assumption built into most of them is that risk can be measured in terms that apply equally to everyone in the assessment area.

In our recent publication (led by Marcelle Scadden, a PhD student working in my research team), we examine this assumption by reviewing 45 city-level climate risk assessments from the United States, Canada, Australia, and New Zealand. These reports were submitted to the Carbon Disclosure Project, the most comprehensive public collection of city climate documentation available. The question we sought to answer was how are Indigenous communities and their perspectives represented?

We found that over a third of reports made no mention of Indigenous populations at all. Only six of the 45 included Indigenous worldviews or values in any substantive way. Where Indigenous communities appeared, it was usually as a category of “vulnerable people,” rather than as holders of knowledge about what is at risk and why.

Not only does this present an equity and colonisation issue, but it's also fundamental for risk science: risk is not a feature of the physical world alone. Whether something counts as a hazard at all depends on your relationship to it. For communities with a long-standing seasonal relationship to flooding, where periodic inundation is expected, part of how land and community have coexisted for generations, the categorisation of that flood as a hazard to be managed is itself a value judgment. Technocratic risk frameworks treat it as a problem to be solved. But that framing is not neutral, instead risk is a relationship between an event/change, what is exposed, and what matters to the people affected. In turn, what matters (e.g., what counts as a loss, what must be protected, what a community cannot afford to lose) depends on values. And these values differ across people, cultures and communities. Indigenous worldviews often understand risk through long-term relationships between people and land, through intergenerational responsibility, and through connections between environmental, social, and cultural systems that technocratic frameworks treat as separate.

When those values are absent from the assessment, the output is not a risk picture for that community. Instead, it is a risk picture for someone else, applied to them.

The issue with this is that adaptation options then may not reflect what a community values, may fail to gain support, or may address one aspect of the picture while damaging something the community depends on or really cares for. Several of the assessments we reviewed acknowledged Indigenous communities as disproportionately exposed but did not examine why, essentially ignoring the structural and historical conditions that produce that exposure. Naming vulnerability without addressing its source does not help communities adapt.

What needs to change is not just who is consulted, but when and how risk itself is framed. “Procedural inclusion” (asking Indigenous communities to engage in an already set process or review a finished document) cannot fix a framework that was designed without their values in it. Meaningful engagement requires rethinking what counts as risk from the start, and being willing to hold multiple frameworks alongside each other rather than fitting all perspectives into one.

In some cases that may mean parallel Indigenous-led assessments. In others it may mean genuinely collaborative processes that surface different understandings of consequence and use them to inform decisions. Either way, it requires recognising that a risk assessment conducted through a single cultural lens does not produce universally relevant or applicable information to enable action.

Read more: Scadden, M., Warren, G.W., Glenn, D., George, L., & Logan, T.M. (2026). Risk Without Values: Including Indigenous Perspectives in Climate Risk Assessments. Risk Analysis. https://doi.org/10.1111/risa.70309