Far-Right ‘Reactions’: a comparison of Australian and Canadian far-right extremist groups on Facebook

ABSTRACT

Little is known about which features of Facebook’s interface appeal to users of far-right extremist groups, how such features may influence a user’s interpretation of far-right extremist themes and narratives, and how this is being experienced across various nations. This paper looks at why certain ‘Reactions’ appealed to users in Australian and Canadian far-right groups on Facebook, and how these ‘Reactions’ may have characterized user decisions during their interaction with far-right extremist themes and narratives. A mixed methods approach has been used to conduct a cross-national comparative analysis of three years of ‘Reaction’ use across 59 Australian and Canadian far-right extremist groups on Facebook (2016–2019). The level of user engagement with administrator posts was assessed using ‘Reactions’ and identified themes and narratives that generated the most user engagement specific to six ‘Reactions’ ( ‘Love’, ‘Haha’, ‘Wow’, ‘Sad’, ‘Angry’ and ‘Thankful’). This was paired with an in-depth qualitative analysis of the themes and narratives that attracted the most user engagement specific to two popular ‘Reactions’ used over time ( ‘Angry’ and ‘Love’). Results highlight ‘Angry’ and ‘Love’ as the two most popular ‘Reactions’ assigned to in-group-out-group themes and narratives, with ‘ algorithms having propelled their partnership in these groups.

KEYWORDS:

• Far-right extremism
• Facebook
• reaction
• algorithm
• Australia
• Canada

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 We acknowledge Dr Brian Ballsun-Stanton, Macquarie University, for his assistance with the method and analysis used in this paper.

2 Auxiliary accounts enable political groups on Facebook to represent their group identity across multiple accounts, with individual accounts branded as ‘chapters’, ‘charters’, ‘clubs’ or ‘fronts’ in a specific city, town or region. In this study, every group page account was submitted as an individual data point during data analysis.

3 It must be highlighted that (Peucker’s et al., 2018) ideological heuristic typology is a means of arranging Australian and Canadian based far-right groups into ideological-level categories for data analysis. The authors recognise that, unlike the typology, far-right groups rarely represent or remain in one or another distinct category overtime. It has been brought to the author’s attention and it is reasonable to suggest that (Peucker’s et al., 2018) typology ought to include other categories relevant to far-right extremism in these countries, such as anti-Semitic or ‘ethnopluralist’ clusters, or amend the language used to label its categories, such as changing anti-Islamic to anti-Muslim. However, collaborating with Peucker and colleagues to justify and amend the typology was beyond of the scope of this study and outside the intention of its use.

4 It is important to clarify that the ‘Reactions’ measured in this study are representative of those ‘Reactions’ user-assigned to administrator posts in Facebook’s platform. We do not measure ‘Reaction’ or any other emojis in Facebook, such as those published in comments, in other posts generated by users, or in private correspondence within Facebook’s Messenger application.

5 Facebook temporarily or periodically features other Reactions among the more consistent set. For instance, illustrated as a flower, the ‘Thankful’ Reaction re-appears periodically on the platform for Mother’s Day while the ‘Care’ Reaction, illustrated as a person embracing a love heart, appeared in response to the COVID-19 pandemic to afford users with an economical means to convey their solidarity (Lyles, 2020). The ‘Thankful’ Reaction was included in this study because it periodically featured during the timeframe of analysis, while other Reactions like ‘Care’ are not represented here because they were present on the platform during data collection.

6 During data selection, 30 far-right Facebook accounts were chosen from both the Australian and Canadian context. However, it is not understood why one of the Canadian accounts, ‘Unit 14 (Blood & Honor Alberta)’, was not captured by Facebook’s online software program ‘Crowdtangle’ during data collection. During the time between selecting Facebook accounts for analysis and capturing online content, this account may have been suspended for violating Facebook’s terms of service. Facebook had de-activated or removed certain far-right group page accounts from the platform following the Christchurch attack in New Zealand (Lamoureux and Mack, 2019). If the group page account was removed from the platform following data collection, the linguistic content and links were available, but the photos and videos contained with specific posts were, at times, unable to be viewed during the thematic analysis.

7 Sahani (2018) findings were based on a multi-variate analysis of the Profiles of Individual Radicalization in the United States (PIRUS) dataset collected by the National Consortium for the Study of Terrorism and Responses to Terrorism (START). Sahani (2018) states that these findings were significantly correlated with younger persons who had an active foreign fighter status at the time of exposure, and who executed violent extremist behaviour following the popularisation of social media and relevant updates to their personalisation algorithm (2005+).

Additional information

Notes on contributors

Jade Hutchinson

Jade Hutchinson, PhD Candidate, Department of Security Studies and Criminology, Macquarie University.

Julian Droogan

Julian Droogan, Associate Professor, Department of Security Studies and Criminology, Macquarie University.