Climate messaging is everywhere, urging people to recycle, vote, protest or rethink their daily habits. But which types of messages actually motivate people to take collective action on climate change? A study by Stanford’s Climate Cognition Lab set out to answer that question, testing 17 different climate message approaches to see which ones most effectively drive people to take action.
The study, published in January in PNAS Nexus, recruited 31,324 U.S. residents to undergo short “interventions” through online surveys. The prompts asked participants to complete randomly assigned exercises, from writing about the importance of protecting the environment for patriotic reasons to watching a video on three new climate policies. Researchers then offered opportunities to take different actions, which they grouped into three buckets: public awareness actions like signing up for newsletters, political actions like letter-writing to representatives and financial actions like donating.
“We came at this question from a broad need and desire to understand the state of the field,” said environmental behavioral sciences assistant professor Madalina Vlasceanu, a senior author of the paper. “There’s so much research… and we were still unsatisfied with the answer to our question: what works best to incentivize and motivate people to take climate action?”
The 17 interventions were selected through an open call that welcomed submissions from academics and practitioners. After receiving about 60 proposals, an advisory group helped narrow them down based on theoretical relevance and predicted effectiveness.
Across outcomes, the most consistently effective approach was what the researchers labeled collective efficacy and emotional benefits, which “tells people that collective action has worked in the past, it can work again, and participating can make you feel good and connected,” said lead author Danielle Goldwert, a Ph.D. student at NYU and Stanford research affiliate.
This intervention and the binding moral foundations intervention, which framed climate action as protecting the “purity and sanctity” of our land, were successful in motivating Republicans to take financial actions, such as donations and committing to divestment. This was a particularly noteworthy result, Goldwert said, for a group that is “notoriously difficult to engage on climate-related issues.”
Across the board, messaging focused on positive emotions outperformed those which focused on negative emotions. The guilt and historical responsibility intervention, for example, presented participants with information from a New York Times article about the collective responsibility of the global North for mitigating emissions. It did not significantly impact any advocacy outcomes.
“A lot of people, including myself, assumed for a really long time that you would need to scare people into action,” Vlasceanu said. “But it doesn’t work for these types of [collective action] outcomes.”
For Students for a Sustainable Stanford co-president Shreya Ramachandran B.S. ’25 M.S. ’26, the results felt immediately relevant.
“I would love to see everyone who works at a major nonprofit to look at what their current advocacy strategies are [and] look at who their target audience is,” Ramachandran said. “What is the type of reasoning and appeal that I’m making, and based on the literature, does that show that that works?
The research team hopes that their results will help policymakers and practitioners in the climate advocacy space make more informed decisions about what types of messaging they use. They created a web tool to display their data interactively, showing effectiveness filtered by audience characteristics.
“Different interventions will be effective depending on which outcome you look in, and also which population you look in,” Goldwert said. “You have to be specific about what exactly you’re trying to change.”
On campus, although SSS initiatives like zero waste and sustainable airport shuttles already have large uptake, Ramachandran plans to consider whether they could implement these different interventions to reach new populations of students.
She also pointed to potential challenges to bringing these results into the real world. “Most interventions that I see or put together are a combination of a lot of these [interventions],” she said, describing how groups could mix moral framing, bipartisan cues and co-benefits in a single pitch.
Vlasceanu and Goldwert say that translating controlled experiments to real life practice is the main obstacle ahead. The lab has worked to understand what drives people to take collective action in the moment. However, a bigger question remains: what happens when collective action actually takes place, and under what conditions does it change outcomes?
“What does it take to make progress on climate? What are the main barriers?” Vlasceanu said. “The next step is: what would it mean for a lot of people to take collective action? What else needs to happen to coincide with this movement?”