In 2024, behavioural scientists published findings from from one of the largest ever experiments on climate messaging.

Tested across 59,440 people in 63 countries, they found that doom and gloom was the single most effective strategy for driving social media shares. It was also the worst for motivating real-world action.

People exposed to catastrophic framing were measurably less likely to change their behaviour than those who read nothing at all. The findings were widely circulated in climate circles, a lot of people said thoughtful things, then promptly returned to business as usual (AMOC anyone?).

Three things explain why the research hasn’t shifted practice:

1. First, reach. Catastrophism performs online. Organisations and influencers measure success in impressions and shares, and doom delivers both.

2. Second, identity. For a movement that has spent a decade warning of catastrophe, any retreat feels like concession.

3. Third, inertia. Messaging infrastructure, fundraising copy, campaign strategy — the climate sector has spent years building muscle in these areas, and is understandably reluctant to give those skills up. To paraphrase Upton Sinclair, it’s hard to get someone to change their mind when their job depends on them not changing it.

Most climate messaging is still being aimed at people who already agree, delivered in a register that is driving the unconvinced further away. The audiences that are the most important for climate communicators to reach, regular voters, are checking out, and the cost is visible in the polling. Climate concern has softened across most of the world’s liberal democracies.

What the study found actually worked was moral framing and scientific consensus. Messages that leave people feeling like they’re part of something bigger, and capable rather than overwhelmed.

But that stuff doesn’t travel well online, which is probably why it hasn’t been widely adopted. That's the problem: what works doesn't spread. What spreads doesn't work. It’s time for a change.