The phrase belongs to Christiana Figueres and Tom Rivett-Carnac, who used it as the subtitle of their 2020 book on the climate crisis. What they meant by it is narrower than the word optimism usually carries, and the narrowing turns out to matter more than the sentiment does.

Writing for the Cambridge Conservation Initiative’s Earth Optimism project in October 2021, Rivett-Carnac defined stubborn optimism as “a mindset based on grit and determination in the face of daunting but unavoidable challenge.”

Not cheerfulness. Not a belief that things will work out. Determination held in place while the facts stay ugly.

That is a stricter claim than the phrase usually gets used for, and the research on hope and climate engagement treats the strict version reasonably well and the loose one roughly.

Two kinds of hope, pointing opposite ways

The most direct evidence comes from Maria Ojala at Orebro University, who published a study in Environmental Education Research in 2012 covering two Swedish samples: 723 teenagers and 381 young adults.

She separated hope into two kinds. Constructive hope rests on a belief that people, institutions and technologies can act, held by someone who accepts the problem is real. Hope based on denial rests on something else: a sense that the threat is overstated, or that it will resolve itself without anyone doing much.

From the outside the two can look similar. Both produce a person who says they feel hopeful.

They did not behave the same way. Constructive hope had what Ojala describes as a unique positive influence on pro-environmental behaviour across both samples. Hope based on denial was negatively correlated with that behaviour in both, and among the teenagers it was a significant negative predictor.

Her own summary carries the qualification: hope “could also work as a motivational force, if one controls for denial.”

The full paper is paywalled and we worked from the abstract in the ERIC record. Both samples are Swedish, the participants are young, and the design is correlational. Nothing in it shows that hope causes anything.

Worry did more of the work than hope

Ojala returned to the question with Jala Rizeq in a 2024 paper in Ecopsychology, comparing young people aged 16 to 19 in Sweden and Qatar. The Swedish sample was 277, the Qatari 134. Both were convenience samples.

They measured worry and hope alongside climate-friendly behaviour and wellbeing, and the two emotions came apart.

Worry was associated with climate-friendly behaviour in both countries, strongly so in Sweden, where the correlation reached .57. In Qatar it was .36. Worry held its positive relationship with engagement even after hope was controlled for.

Hope did not perform nearly as well. In Qatar it correlated with behaviour at .29. In Sweden there was no significant relationship between hope and behaviour at all.

The interaction is the part worth sitting with. In the Qatari sample, hope predicted behaviour only among those who were already highly worried. Hope by itself was not doing the work.

What hope did predict, in the Swedish sample, was wellbeing. Less negative affect and higher life satisfaction, with no measurable difference in what anyone actually did.

That is a genuinely awkward result for anyone recommending optimism as a strategy rather than as a way to feel. It suggests the two can come apart, and that the pleasant one is not automatically the useful one.

The authors are direct about the weaknesses. Both samples were convenience recruited, the design is cross-sectional, and worry and hope were each measured with a single question with no established validity. The direction of causation is entirely open. People who already act on climate may simply feel more hopeful as a result.

What the best messaging actually shifts

Underneath the argument about mindset sits a plainer question: how much any framing changes what people do.

A registered-report megastudy published in Nature Climate Change in 2026, led by Jan Voelkel with a large international team, took the ten most-cited climate messaging strategies from 157 research papers and tested them on 13,544 US adults in spring 2024, following five replication studies with a further 3,216 people.

Six of the ten moved preregistered attitudes. The effects ran between one and four percentage points. Averaged across all ten, belief in climate change rose 1.16 percentage points. The strongest performer was a scientific-consensus message rather than any framing built on hope or fear.

No message increased donations to environmental organisations.

Effects also barely differed across party lines, which cut against what the field had predicted.

Voelkel told the Cornell Chronicle in January 2026 that short-term messaging works “like a session of physical therapy” rather than a permanent fix. For a shift of one to four points, that seems a fair description. The study covered US adults only, and a single exposure to a short message is not the same thing as a sustained public argument.

What survives of the claim

Read strictly, the stubborn optimist’s position holds up better than most slogans do. Rivett-Carnac’s definition is about persistence rather than expectation, and Ojala’s constructive hope has the same shape: a belief that action is possible, held by someone looking straight at the problem.

The version that does not survive is the one where optimism stands in for concern. That is close to what Ojala measured as hope based on denial, and among Swedish teenagers it predicted less action rather than more.

Whether stubborn optimism is our best hope is a larger question than this evidence can settle. What the studies support is narrower. Among the young people surveyed, worry predicted action better than hope did, hope contributed mainly where worry was already present, and no message yet tested at scale has moved anyone’s money.

The gap worth watching is between what people say and what a change costs them. The messages shifted the first. None of them shifted the second.

Edited by Lachlan Brown