Protest Under Uncertainty: Evidence from a Survey Experiment

ABSTRACT

Environmental disasters generate uncertainty, which is a crucial element of post-disaster political dynamics. Does communication of uncertainty affect public willingness to participate in political activism? This article first provides a content analysis of news coverage to show that uncertainty framing is prevalent in the aftermath of environmental disasters. The article then examines the effect of such uncertainty on public willingness to protest, presenting a survey experiment of over 3,600 Americans recruited through Amazon Mechanical Turk. In theory, uncertainty framing may trigger emotions like anxiety, which reduce people’s willingness to engage in protest. Since in environmental communication uncertainty frames are often used to reify the status quo, which is more easily compatible with conservative beliefs, the dampening effect should be stronger among those who are ideologically conservative. Our experiment confirms the latter expectation but does not find support for the former. These results help explain why some protest coalitions may have more breadth than others after environmental disasters.

KEYWORDS:

• Environmental disasters
• protest
• uncertainty
• emotions
• political ideology
• framing

Data availability statement

The data supporting the findings of this study are available within the article and its supplementary materials. Some data are available from the corresponding author upon request.

Human research participants

This research was approved by the University of British Columbia’s (UBC) Behavioral Research Ethics Board (BREB), BREB Number H18-00999.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Articles that did not contain a substantial section on the disaster of interest were not coded.

2 They, however, focus on climate change and include conflict frames, potentially overstating the presence of uncertainty frames compared to other, less contested environmental issues.

3 We omit a manipulation check for two reasons. First, we are concerned that any such question would affect respondents’ perceptions of the experimental treatment by drawing their attention to the manipulation of interest prior to the measurement of the outcome measures (Fayant et al., 2017; Hauser & Schwarz, 2015; Hauser et al., 2018). Second, we expect measurement error to be a concern for any such question in this context, thereby undermining its value (Hauser et al., 2018), and indeed it is likely not appropriate to use manipulation checks when randomizing intrinsic message features (O’Keefe, 2003). We do include a pair of attention checks in our survey and find that our null findings extend to those who are highly attentive to the survey (see Table 11 of Appendix 4).

4 We conducted a power analysis to determine the needed sample size. We have 80% power to detect 0.11 effect sizes at the 0.05 significance threshold between the control conditions and either the neutral or high cost uncertainty conditions when pooling human safety conditions. We can detect 0.16 effect sizes at the same power and significant level looking exclusively at the conditions that omit human safety framing. More details can be found in Appendix 5.

5 The means for each outcome across all six experimental conditions can be found in Table 10 in Appendix 4 and Table 11 presents the full 3X2 factorial estimation for all three outcomes. The null results for H1 and H2 do not depend on the presence of human safety frame nor the type of uncertainty.

Additional information

Funding

This work was supported by the University of British Columbia’s Faculty of Arts and Masaryk University's Faculty of Social Studies.

Notes on contributors

Miriam Matejova

Miriam Matejova is an Assistant Professor of Political Science at the Masaryk University and a fellow at the Norman Paterson School of International Affairs, Carleton University. She holds a PhD in Political Science from the University of British Columbia. She has published articles on energy and environmental security, global environmental activism, foreign intelligence, and international conflict management.

Eric Merkley

Eric Merkley is an Assistant Professor of Political Science at the University of Toronto. He holds a PhD in Political Science from the University of British Columbia. He has published articles on political communication and public opinion with a focus on how citizens engage with expert consensus and the role of elites in shaping public attitudes.