![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
Compassion-evoking and widely circulated news images seem capable of affecting public opinion and change history. A case in point is the picture of the drowned toddler Alan Kurdi, who died while trying to escape to Europe with his family in September 2015. We theorise how these types of photos affect public opinion. By relying on panel data with sequentially embedded survey experiments conducted in the aftermath of the picture’s publication and while large numbers of refugees were coming to Europe, we demonstrate that the image of Alan Kurdi had context-varying effects on policy preferences. Initially, the upsetting photo of the toddler increased support for liberal refugee policies across ideological divides. However, when individuals had the time to think about policy implications, they started processing even this highly upsetting picture through their ideology. Consequently, our results indicate that compassion-evoking images shift meaning over time.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 While Hungary received a large share initially, many of these individuals eventually moved on to other countries such as Germany and Sweden.
2 Towards the end of the survey when we asked respondents if they had seen the picture of Alan Kurdi before taking part in the study, 95% reported that they had seen the picture previously.
3 The image polarisation hypothesis can only be tested with experimental data since we are expecting effects based on renewed exposure.
4 Both pictures are listed as “100 Images That Changed the World” (Time Citation2017).
5 There are potential workarounds to this issue, but neither solves the basic problem. First, the compassion measure could have been embedded in a battery of other questions that serve as decoys. However, this would be questionable since there is a risk that such a battery would attenuate the effect of the experimental treatment on the policy measure. Second, the compassion measure could have been asked after the policy question. Yet, this is not a good option because the policy question could have influenced the compassion question, either by priming considerations that go into answering a question on the policy or by offsetting the effect of the treatment because of the time it took to answer the policy question.