The Five Canadas of Climate Change: Using audience segmentation to inform communication on climate policy

communicating around policy design, especially among less engaged audiences (i.e., the Con- cerned and Doubtful segments). That these framing effects were mostly found among moder- ately engaged groups echoes Zaller’s [78] two-moderator model of persuasion, and is in line with research by Chong and Druckman [79] highlighting the role of strongly held prior atti- tudes in moderating framing effects. Both of these studies suggest that framing effects are most likely to be found among audiences with relatively weaker opinions. A negative effect of the “Invest in solutions” and “Relative price” messages among the Alarmed is also broadly consistent with previous studies documenting “boomerang effects”, occurring when exposure to a given message moves beliefs in a direction opposed to the origi- nal intent [39, 80]. However, in contrast to these studies–which typically found boomerang effects among groups that were more skeptical of climate change—our results demonstrate that boomerang effects may also occur among the most Alarmed segments of the population. This result among the Alarmed could be explained by the relative sophistication of this audience. Recall that of all segments, the Alarmed are most engaged with climate change and are much more likely than national averages to report being both knowledgeable about the cli- mate change issue, and very interested in politics. As a result, the Alarmed may be more moti- vated by accuracy goals and to scrutinize the credibility of messages, thus making them less

inclined to believe that such a low (2 cents per litre) carbon price can have any real impact on green infrastructure or the relative price of different energy technologies [c.f., 3]. To test this explanation, we created a dummy variable equal to 1 for respondents who received the "Invest in solutions" or the "Relative price" treatment (i.e., those with a negatively signed coefficient at a low price stringency) and 0 for those who received the "Equal dividend" treatment. We then examined whether these treatments had a different effect on perceptions of policy effective- ness, and whether policy stringency moderated this effect. As shown in the S4 Table in S1 File, we found a negative interaction effect between the "Invest in solutions" and the "Relative price" messages and the level of policy stringency. Consistent with the idea that the Alarmed are more sophisticated and therefore more likely to scrutinize the coherence of carbon pricing messages, we found that messages alluding to substantial earmarking or leveling the playing field for alternative energy sources decreased perceptions of policy effectiveness among the Alarmed, but only under the lower price treatment. Overall, our results suggest that attempts at devising uniform, one-size-fits-all communica- tion strategies around climate policy are likely to face important challenges. For instance, our findings suggest that governments and communicators should be careful when communicat- ing with more sophisticated and engaged climate change audiences (like the Alarmed), who

may be more motivated by accuracy goals when processing messages around climate change. As a result, careful consideration ought to be given to the credibility of messages, so as not to alienate more engaged audiences as they seek to win the support of less engaged groups. More broadly, the different framing effects found at different levels of policy stringency further sug- gest that the overall cost structure is important when crafting messages about climate policy. As the costs of climate policy and concerns with cost of living evolve over time, communica- tions may need to be adapted so they better fit the context in which messages are received, while meeting the informational needs and motivations of different audiences. Conclusions Analyzing data from a random probability survey of adult Canadians collected in October 2017, this study examined how unique audience segments within the Canadian population think and act toward climate change and explored whether and how the level of audience engagement moderates the effect of various messages on support for climate policy. We found PLOS ONE The Five Canadas of Climate Change PLOS ONE | https://doi.org/10.1371/journal.pone.0273977 November 23, 2022 16 / 21