How do climate disasters affect pro-environmental behaviour?

Seeing others affected by extreme weather events like flooding, storms and heatwaves is not enough to change people’s behaviour. For the average person, a global problem leads to action only when it gets personal: we respond much more when our own postcode is affected by climate disasters.

Floods, storms and heatwaves are widely covered in the media each time they occur. While many recognise these extreme weather events as major threats that are exacerbated by climate change, this awareness does not always translate into action.

This is one of the central puzzles in climate policy. If people already know that climate change is real and harmful, why is public engagement in climate solutions still often too slow?

New research suggests a simple human explanation: seeing others affected by climate disasters is not enough to change behaviour; instead, people respond much more when their own postcode is affected.

The ‘only in my backyard’ phenomenon

This localised response can be shown by looking at people’s charitable donations in response to flooding in the UK. Using a decade of donation records shared by the Charities Aid Foundation (CAF), it is possible to link donors to detailed flood records and track how these same individuals change their giving before and after floods. This analysis shows that:

Direct flooding increases donations to climate-related causes. After a flood affects someone’s postcode, they become about two percentage points more likely to donate to environmental charities. This is a substantial increase given that only around 6% of donors supported these charities before any flooding. Crucially, the effect compounds: the more often that people experience flooding directly, the more likely they are to take green action.
The ripple effect stops at the front door. If a flood devastates a nearby neighbourhood, even within 200 metres, but a person’s own postcode remains dry, their environmental giving does not change at all. This remains true no matter how often those neighbours are hit. Simply witnessing others endure repeated climate shocks is not enough to change behaviour.

Figure 1: Estimated effects of flood proximity on environmental donation

Source: Charities Aid Foundation
Note: Distance categories are defined as intervals in metres. Zero indicates that floods directly affect an individual’s postcode. Panel (a) presents coefficient estimates from a standard two-way fixed effects model, and Panel (b) plots estimates obtained using the imputation method of Borusyak, Jaravel, and Spiess (2024).

This is the ‘only in my backyard’ pattern: people act when climate shocks cross the line from ‘out there’ to ‘here’, but not when the effects are down the road, even if very close. This trend suggests that global or even regional climate threats often feel too abstract: it is the personal, tangible shock that prompts individuals to re-evaluate their actions. In fact, the data show that people directly affected by floods are more likely to see their own environmental efforts as insufficient, effectively raising the bar for what they consider ‘green enough’.

How do we show this with data?

Measuring actual changes in behaviour is notoriously difficult. Surveys can suffer from social desirability bias, where people might claim that they act in a more sustainable way just to look good to the researcher.

To avoid this, the research uses actual financial commitments – donations from nearly 90,000 donors to more than 55,000 charities across England. Because donors are identified at the six-digit postcode level, it is possible to link them precisely to local flood records and observe how their giving changes after a flood.

The study does not just compare flooded towns to dry towns. Instead, it compares individuals before and after a flood against people in the same area who face the exact same underlying risk of flooding but simply get lucky that year. This confirms that the change in behaviour is driven by the lived experience of the disaster, not just living in a flood-prone place.

This is also why the ‘neighbouring floods’ result is so shocking: people living just 200 metres apart share much of the same local context and media environment, but they do not change their behaviour. It is only those whose homes are flooded that act.

Self-interest as a catalyst for climate action

This hyper-localised response is not an isolated quirk. It aligns closely with a growing body of evidence showing that self-interest is often a powerful, and sometimes necessary, driver of climate action.

For example, one study that uses survey data from 137 countries highlights that people whose livelihoods depend on the weather are more likely to respond to climate shocks with greater concern and show stronger support for climate policy (Gazmararian and Milner, 2025).

Similarly, research shows that public support for green policies is rarely driven by abstract environmental concerns alone (Dechezleprêtre et al, 2025). Instead, people are far more likely to back climate action when they see direct, tangible benefits to their own lives – whether that means lower energy bills, better local air quality or new jobs.

This aligns with findings on the implications of flooding: general messages about climate disasters may not change behaviour unless they connect to concrete household consequences.

So are people entirely selfish? Not exactly. The data from the flooding study show that individuals with strong universalist values – people who naturally prioritise the wellbeing of strangers – do increase their green donations when their neighbours are flooded. This echoes earlier work that links universal moral values to stronger climate commitments (Andre et al, 2024). But for the average person, a global problem leads to action only when it becomes a personal one.

Communication implications: ‘floods exist’ is not enough

The UK is not a country where climate change is widely dismissed – surveys show that around 70% of the population see it as a serious threat. Yet far fewer people believe that it will cause serious harm to them personally.

That gap – between general concern and personal relevance – helps to explain why communication that simply reiterates that ‘floods are happening’ often fails to change behaviour. This research suggests that what really matters is whether people can picture the consequences landing on their own doorstep.

This does not suggest resorting to scare tactics. Rather, it means explaining – clearly and concretely – how climate risk shows up in everyday life, even for those who have never been affected themselves. For example, flooding can bring repair costs, disruption to transport, losses for local businesses and higher insurance premiums. And climate shocks elsewhere can still hit households through food prices, energy bills, supply-chain disruption and pressure on public finances.

The implications for policy-makers and campaigners are clear. Broad warnings about melting ice caps, distant wildfires or abstract temperature targets are unlikely to move the needle for most citizens. If we want to bridge the gap between climate awareness and climate action, we need to change how the crisis is framed – shifting the narrative from a global crisis to a personal one.