Ask most people whether they would back a tax that doubles the price of beef, or a ban on short-haul flights, and the response is rarely warm. These are exactly the kinds of climate policies that decarbonisation demands and that voters tend to resist: They ask for real sacrifice, impose visible costs, and spread their benefits thinly across everyone and no one in particular. Governments know they need public backing to pass them, and the harder question has always been how to build it.
Education is the answer governments often reach for. It is written into Article 12 of the Paris Agreement, and countries from Italy to Argentina have built climate learning into schools and public institutions. Yet the evidence that it actually shifts attitudes has long been thin, and at times discouraging. Large research reviews have tended to find that, on its own, education does relatively little to change environmental behaviour, and a study spanning 63 countries found that informational campaigns barely moved the needle on policy support.
A new paper published this month in PNAS suggests the problem may not lie with education itself, but with the kind of education researchers have tended to test. The study, Educational policies can strengthen climate coalitions, follows what happened when 1,845 students across ten French universities took part in a three-hour interactive climate workshop called 2tonnes – an interactive simulation initially devised by French climate educator François Laugier on the model of the Climate Fresk workshop, and now run by a dedicated non-profit, which has delivered it to more than 100,000 people in schools, companies and public institutions.
The result was striking: students who had just completed the workshop were, on average, seven percentage points more likely to support tough climate policies than those who had not. The paper is the work of four former PhD researchers – Max Bradley, Nina Lopez-Uroz, Rens Chazottes, and Susanna Garside – who developed it together during their time at the EUI Department of Political and Social Sciences.
To understand the result, it helps to step inside one of the workshops, which unfold nothing like a traditional class. As Nina Lopez-Uroz explains, the experience begins before participants even arrive. “Before you join, you're sent a tool to estimate your carbon footprint, and you get to see how you compare to the French average,” she says. “So, you come into the workshop already knowing how much more you emit than the two tonnes of CO2 per person we should reach by 2050.”
From there, participants spend three hours playing a simulation that runs forward in time to 2050. Across eight rounds of decisions, some individual and some collective, they spend “action cards” to bring down their emissions. “You can play the card of becoming vegan, or vegetarian, or flexitarian,” Nina says. “And then, depending on your decisions, you see after each round how effective your choice was at decreasing your footprint.” The collective rounds turn into something closer to a miniature cabinet meeting. “It's a bit of a simulation of a ministerial coordination meeting, where you're supposed to discuss the best decisions to take and convince the others. So, there's a lot of peer interaction.”
That combination – immersive, interactive, sociable, and grounded in each participant's own data – is precisely what led the researchers to suspect this intervention might outperform the discouraging average of other educational interventions. “What hinted at that for us, before we even had the results, was this immersive, interactive part,” Nina recalls. “Really active learning, where you're confronted as a participant with the effect of your decisions on your own carbon footprint. It's not just passive, where you're exposed to some findings from the IPCC. It's a bundled treatment that combines different elements of active learning.” The team also had early reason to expect the workshop would register measurable effects. Max Bradley, one of the paper's four authors, notes that they ran a pilot study six months before the full experiment. "We couldn't say it with statistical validity, but the workshop seemed to be shifting certain things," he says. "It gave us the confidence that the design could detect an effect, if there was one to find."
A seven-point shift can sound modest to anyone outside the social sciences, and Max is quick to push back on that reading. “In social science terms, these are quite large effects – definitely on the higher end of what we observe across different types of interventions,” he says. The point is easy to miss, he explains, because it is an absolute increase rather than a relative one. Take the beef tax: Only about 34% of students who hadn't done the workshop supported it. “It's not 7% of 35. It's seven points on top of 35, so you're going from 35 to 42. In relative terms, that's closer to a 25% increase.”
But he is equally clear about what the finding is not. “Our argument is not that this workshop is a one-shot tool to build a lasting, durable coalition for costly climate policies. It's one tool – low-cost and effective at meaningfully shifting attitudes – that policymakers can use alongside others.” Support, he points out, is only ever half the equation. “It's one thing to support a flight ban. It's another to get people to actually engage with it, because if there aren't viable alternatives in terms of trains or buses, you can increase support as much as you want and people still won't change.”
Perhaps the most intriguing finding sits just beneath the headline number. The researchers had expected the workshop to work through two channels: by making the climate crisis feel psychologically closer and more personal, and by strengthening people's belief that these policies actually work. In the end, almost all the effect ran through the second channel – perceived effectiveness – and very little through psychological distance. Nina traces this back to the design of the workshop itself. “It really has this interactive element where you recalculate your footprint depending on your decisions, and you see the effect,” she says. “Feeling closer to the climate crisis is a by-product – you talk about it for three hours. But the core of the workshop is the footprint effect of your decisions.”
Max sees a broader lesson in this. “The workshop literally asks participants to implement policies and then shows them: Put this in place, and emissions come down by this much. That's the definition of seeing effectiveness,” he says. “And policy effectiveness is something that's often very far from the average citizen. Governments should do a better job of communicating the effectiveness of the policies they implement, and they often don't – so they don't get credit for the successes, and people only notice the failures. The nice lesson here is that when effectiveness is clear and legible, people are more likely to support a policy. It seems intuitive, but there isn't much evidence showing it.”
Not everything moved in the expected direction. Strikingly, although students became more supportive of costly policies, they were no more willing to donate to a climate NGO – a reminder that shifting what people think is not the same as changing what they do. Bradley points first to a well-worn pattern in the literature. "There's a very well-documented attitudes-behaviour gap, where interventions shift attitudes but not necessarily day-to-day behaviour,” he says. But the design mattered too. Students in both groups donated generously to begin with – around 34 of a possible 100 euros – and the money on offer was a lottery prize rather than real cash. “We ran this experiment with basically zero budget,” he says. “In an ideal world, we'd have given each student 100 euros of real money, but with 1,800 students, that's 180,000 euros. With real money, you might have seen different dynamics.”
Equally notable was how uniform the effect turned out to be. It held across political affiliation, income, and personal cost – even students who fly often became more supportive of flight restrictions. Nina again credits the intensity of the format, and the constant repetition of information through discussion and a final wrap-up. “My high-school teacher always said teaching is repeating,” she says. “You're confronted with the facts, repeatedly, and it hammers the information in.” Max explains it in more empirical terms: “If you have a really strong treatment, unless there's a strong theoretical reason for it to moderate, it makes sense that we don't see much variation across subgroups.” The sample matters too: These are young people, in what political scientists call the “impressionable years,” before attitudes harden.
The authors are candid about the study's limits. The sample is French university students, who lean more toward pro-climate than the general population, and the medium-term follow-up – encouraging, with effects persisting six weeks later for two of the three policies – relied on a smaller group of respondents. The obvious next step, Max says, is to test the workshop in settings where the transition genuinely costs people something. “If you did it with beef farmers in rural France, I'm not sure you'd see a seven-point increase for a beef tax, for obvious reasons,” he says. A workplace would be a natural setting, or, more provocatively, a government ministry.
“It would be really interesting to go into the French ministry for energy and run it with the policymakers actually designing this stuff. If you shifted their views, they'd have real influence.” As it happens, Nina notes, a programme already exists to run the workshop with French civil servants.
Even within their existing sample, the authors found reasons for cautious optimism. Max points to students in courses tied to high-emitting sectors – aviation, agricultural engineering – who will graduate into jobs that these very policies threaten. “Among those students, we don't see lower effects,” he says. “It's a net positive that you can shift the attitudes of people going precisely into the sectors that need to decarbonise.”
There is an uncomfortable possibility lurking in findings like these: that policymakers might seize on cheap workshops as a way to sidestep the harder, costlier work of carbon pricing or building green infrastructure. Both authors are emphatic that this would be a misreading. “I really hope that's not the takeaway,” says Nina. “This looks at the demand side – how to build support and avoid backlash. But if people become more supportive of stringent policies and still have no alternative to fossil-fuel mobility or heating, that won't lead to anything substantial. Building workshops instead of railways or heat-pump subsidies – that is not our message.” Max makes the case in plainer terms “Building public support doesn't actually reduce your emissions. It's a tool to pass the harder policies you need to meet your targets,” he says, noting that EU countries face real penalties for missing their 2030 commitments. “Even if a government did this instead of the harder policy, they'd still be hit with fines. The rational incentive to do that just doesn't make sense.”
So does the research leave them optimistic, at a moment when the political space for climate action seems to be narrowing? Cautiously, yes. “It makes me optimistic that people respond to this,” says Nina. “If they hadn't, that would be quite depressing.” She also points to a ripple effect the study could not capture: “People who take this kind of workshop talk to their peers, their families, their colleagues. There's a sort of network effect.” Bradley frames it as one front in a longer fight. “The political space for strong climate policy is getting smaller, and anything that opens it up a tiny bit is a cause for some optimism,” he says. “Am I optimistic it'll solve the problem? Of course not. But political actors who want to defy a strong lobby against decarbonising need popular support behind them to be brave. This helps open up that space.”
The paper is also, in its own quiet way, a story about how research gets made. It did not begin as a funded, top-down project with a principal investigator handing out tasks. It started on a Slack channel for researchers interested in environmental and climate politics, when co-author Renz Chazottes, who had taken the 2tonnes workshop floated the idea of studying it. The four authors were at different stages of their doctorates, and they came together as equals. “What I really liked about our team is that the hierarchy was quite flat,” says Nina. “There was no formal PI. We all felt we could express our ideas freely, and we'd take each other seriously.” For her, as a primarily qualitative researcher, it was a first foray into field experiments. “I felt my feedback would still be taken seriously, and that was super cool. We weren't afraid to voice things, because we knew the others wouldn't judge.”
Max makes a similar point. “When you're working on a senior scholar's project, you can be a bit trepidatious about suggesting something, because you don't want to seem out of your depth,” he says. “Here, we were all at the same level. We could talk as equals, try things, watch them fail, go back and forth in hour-long debates. That's how you learn – by making mistakes, and the more room you have to make them, the more you learn.” The four had guidance from supervisors along the way, and ample chances to present their work and gather feedback across the EUI's research working groups, but the project was theirs. “When you're working on someone else's project, you care, but you're not as invested. Here, the four of us were all PIs. That makes you want to put the time in.”
What began as a side project on a Slack channel became, over three years, a paper in one of the world's leading scientific journals. The authors have since defended their theses and moved on to postdoctoral positions at leading universities. Which, in the end, is its own small piece of evidence: that given a flat structure and the freedom to fail, a group of PhD researchers can produce work that genuinely advances what we know – about climate, and about how good research comes together.
Max Bradley and Nina Lopez-Uroz are former researchers at the EUI Department of Political and Social Sciences. Their paper, 'Educational policies can strengthen climate coalitions', co-authored with Rens Chazottes and Susanna Garside, was published in PNAS in May 2026.