Renewable Energy · Strategy · Peer-reviewed

Saying "this is wrong" didn't make the message stronger. Or longer lasting.

Moral convictions are supposed to be stickier than ordinary opinions, so this academic study tested whether adding explicit moral claims to messages about fossil fuels would create more durable persuasion. In a three-wave experiment with 4,488 Americans, it didn't. But every version of the message worked, and held.

Academic research by XandY's co-founders and colleagues · Science Communication, 2025 · 4 minute read

Attitudes with a moral basis tend to be more stable and harder to change. That suggests a tactic. If you want persuasion to last, tell people the issue is a matter of right and wrong. This study put that idea to the test on renewable energy, using short animated videos about the harms of fossil fuels. It was conducted by Abel Gustafson through the University of Cincinnati with Matt Goldberg and colleagues at the Yale Program on Climate Change Communication, the University of Cambridge, and Ball State University.

The researchers recruited 4,488 U.S. adults, quota-matched to the population on party, age, and gender, and showed each one of five videos. Two described how pollution from fossil fuels harms innocent people (the "harm" appeal). Two described how fossil fuels contaminate the purity of nature (the "purity" appeal). Within each pair, one version added explicit moral claims ("This is simply wrong. It is unethical to hurt innocent people...") and one did not. A control group watched a video about robotic surgery. They measured support for a transition away from fossil fuels immediately, after about 12 days, and after about 11 more. A pilot study confirmed that people noticed the moral framing: the explicit versions were rated as far more morally emphatic than the plain ones.

All four messages increased support for renewable energy (d = 0.19 to 0.21 right away), and all four decayed the same way: a drop of roughly half over the first two weeks, then a plateau. Adding explicit moral claims changed nothing. Not the initial effect, and not the durability. The lines sit on top of each other. The study also found no evidence that the harm and purity appeals worked differently for liberals and conservatives, and no evidence that durability depended on how much people like to think things through or how engaged they are with the issue.

Four messages, one curve Effect of each video on support for renewable energy, immediately and at two follow-ups 0 0.1 0.2 0.3 Right away ~12 days ~3 weeks Harm 0.21 → 0.09 → 0.08 Harm + moral claim 0.19 → 0.07 → 0.06 Purity 0.19 → 0.09 → 0.09 Purity + moral claim 0.20 → 0.11 → 0.10 Effect on support (SD units) GUSTAFSON ET AL. (2025), SCIENCE COMMUNICATION · N = 4,488 AT TIME 1 · EFFECTS VS. CONTROL, SD UNITS
Figure 1. Standardized effects on support for a transition to renewable energy, compared with the control group, at three time points. Solid lines are the plain harm and purity appeals; dashed lines add explicit moral claims. All four effects are statistically significant at every time point, and none of them differ from each other.

Why no difference? One clue is in the mechanism. In the pilot, the explicit moral claims made people see the issue as somewhat more of a moral issue. In the main study that effect nearly vanished. The moral framing was noticed, but it did not change how people held the attitude, which is the step that is supposed to produce durability.

The moral emphasis was noticed. It didn't take hold. How much the explicit moral claims changed each perception, compared with the same message without them Harm appeal Purity appeal 0 0.3 0.6 0.9 1.2 Effect of adding explicit moral claims (Cohen's d) Message seen as more moral pilot study, manipulation check d = 1.11 d = 1.12 Issue seen as a moral issue pilot study d = 0.21 d = 0.20 Issue seen as a moral issue main study, N = 4,488 d = 0.05 d = 0.04 GUSTAFSON ET AL. (2025), SCIENCE COMMUNICATION · MANIPULATION CHECKS · COHEN'S D
Figure 2. The explicit moral claims were clearly perceived as more moral (top). But their effect on whether people saw fossil fuels as a moral issue was small in the pilot and close to zero in the main study (bottom). Without that shift, there was no reason to expect more durable persuasion.
Adding "this is wrong" to the message changed neither its immediate effect nor its durability. Every version worked, and about half the effect held for three weeks.

For communicators, the takeaway cuts two ways. Moral language did not help, but it also did not hurt or polarize, so it remains available for other goals, like signaling an organization's values. And simplicity turned out to be a feature. A plain, factual message about harms performed as well as a carefully moralized one. That means time and money are usually better spent on reaching more people than on perfecting the appeal.

Source. Gustafson, A., Goldberg, M. H., Lee, S., Remshard, M., Luttrell, A., Rosenthal, S. A., & Leiserowitz, A. (2025). Testing the durability of persuasion from moral appeals about renewable energy. Science Communication, 47(6), 800–827. doi.org/10.1177/10755470241311011

About this research. This study was carried out through the authors' academic positions at the University of Cincinnati, the Yale Program on Climate Change Communication, the University of Cambridge, and Ball State University. This page is XandY's plain-language summary of the published findings.

Method. Preregistered three-wave longitudinal experiment with a 2 (harm vs. purity) × 2 (explicit moral claim vs. none) design plus a control. U.S. adults recruited on Prolific with quotas matching the population on party, age, and gender; N = 4,488 at Time 1, 3,640 at Time 2, and 3,121 at Time 3. Attrition analyses found no threat to internal validity. Effects are OLS estimates versus control, adjusting for pre-test support. Data, materials, and preregistration: osf.io/xjk4z.

More renewable energy research Follow us on X