‘If you don’t know, vote no’: Symptoms of Destructive Polarisation in the 2023 Voice to Parliament Referendum in Australia

Paper presented by Axel Bruns at the IAMCR 2024 conference, Christchurch, 1 July 2024.

1. ‘If you don’t know, vote no’: Symptoms of Destructive Polarisation in the 2023 Voice to Parliament Referendum in Australia Axel Bruns, Tariq Choucair, Sebastian Svegaard, Samantha Vilkins, Katharina Esau, Laura Vodden Digital Media Research Centre Queensland University of Technology Brisbane, Australia

2. Indigenous rights and recognition: Complex topic since European arrival in 1788 Indigenous Australians remain severely disadvantaged Persistent lack of formal consultation Voice to Parliament: Endorsed in 2017 Ulu r u Statement from the Heart Commitment to referendum on a Voice in Anthony Albanese’s 21 May 2022 election victory speech Referendum design revealed in March 2023 Constitutional referendum held on 14 Oct. 2023 Proposed Constitutional Amendment: Chapter IX Recognition of Aboriginal and Torres Strait Islander Peoples 129 Aboriginal and Torres Strait Islander Voice In recognition of Aboriginal and Torres Strait Islander peoples as the First Peoples of Australia: There shall be a body, to be called the Aboriginal and Torres Strait Islander Voice; The Aboriginal and Torres Strait Islander Voice may make representations to the Parliament and the Executive Government of the Commonwealth on matters relating to Aboriginal and Torres Strait Islander peoples; The Parliament shall, subject to this Constitution, have power to make laws with respect to matters relating to the Aboriginal and Torres Strait Islander Voice, including its composition, functions, powers and procedures. Case Study: Voice to Parliament Referendum

3. Referendum Vote Voting modus: Compulsory for all registered voters Actual turnout: 89.95% Requirements for success: Majority of voters overall Majority of voters in majority of states (4 of 6) Results: Overall: 40% Yes, 60% No 0 of 6 states Yes win only in Australian Capital Territory By Teratix - Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=131601888

4. Case Study ( https://www.abc.net.au/news/2023-10-08/voice-polls-show-support-lower-than-republic-vote/102942468 )

5. Forms of Polarisation Polarisation on what attributes? Issue-based: disagreements over specific policy settings Ideological: fundamental differences based on political belief systems Affective: political beliefs turned into deeply felt in-group / out-group identity Perceived: view of society, as based on personal views and media reporting Interpretive: reading of issues, events, and media coverage based on personal views Interactional: manifested in choices to interact with or ignore other individuals/groups (and more…)

6. Agonism? Polarisation? Dysfunction? How bad is it, exactly? All politics is polarised (just not to the point of dysfunction) Much ( most ?) politics is multipolar, not just left/right When does mild antagonism turn into destructive polarisation? We suggest five symptoms ( Esau et al., 2023 ): breakdown of communication; discrediting and dismissing of information; erasure of complexities; exacerbated attention and space for extreme voices; exclusion through emotions. Image: Midjourney

7. (And also, are these the right / the only symptoms of destructive polarisation?) But: How do we operationalise these symptoms in empirical research?

8. Dataset and Methods Posts on Facebook addressing the Voice referendum: Retrieved from CrowdTangle (i.e. public pages and groups only): "Voice to Parliament", "Uluru Statement", (Voice) AND (Referendum), VoteNo , VoteYes , voicereferendum , ulurustatement , yes23 1 Jan. 2023 to 13 Oct. 2023 (last day before the referendum) 89,101 posts from 9,433 unique public pages and groups (i.e. public spaces) Analysis: Practice mapping to identify clusters of participating spaces Topic modelling to verify practice mapping cluster interpretations Zero-Shot Classification Model ( Laurer et al., 2023) applied to post texts: universal classifier based on Natural Language Inference (NLI) trained on datasets predominantly related to political communication ‘true’/’untrue’ probabilities for 14 label sets (e.g. “This text is divisive” / “This text is not divisive”)

9. Referendum question announced Ulu r u Statement anniversary Referendum bill passed Referendum date announced

10. Facebook offers only very limited network data, so we take a practice mapping approach (Bruns et al., forthcoming): Quantify specific aspects of space activities; Identify spaces with similar practices; Interpret their practice choices; Identify symptoms of destructive polarisation. Similarity analysis: Vector embedding of each space’s activity practices Cosine similarity comparison of practices across spaces Network analysis and cluster detection * With particular thanks to Kateryna Kasianenko. Inputs: Post content (word choices, etc.) Use of sources (domains, YouTube videos) Space-to-space content on-sharing … Facebook Analysis: Inventing Practice Mapping*

11. Nodes: public Facebook pages and groups addressing the referendum Node size: volume of posts (spline applied), minimum 5 posts Node colour: Louvain modularity algorithm cluster detection Edge weights: domain sharing similarity + YouTube sharing similarity + on-sharing similarity + Vertex AI text embedding similarity, minimum 1.0 Sky News Australia No Campaigners Anti-LNP ABC Pages and On-Sharers Uluru Statement from the Heart Indigenous Organisations Yes23 Yes Campaigners, Unions, ALP SBS Pages and On-Sharers YES NO (agonistic discursive alliance) (antagonists)

12. Zero-Shot Classification Model ( Laurer et al., 2023)

13. Destructive Polarisation? Erasure of complexities? Reduction to ‘LNP vs. ALP’ Reduction to ‘Voice of Division’ (+ necessary reduction to Yes vs. No) Extreme voices? Dominant role for Sky News Hyperpartisan party activists Breakdown of communication? Few spaces for actual open debate Engagement predominantly with No spaces (could still have genuine debate in comments) Yes discourse alliance disjointed

14. Many of the most widely shared videos from influential conservative news source Sky News Australia made explicitly conspiracist claims. Discrediting and Dismissal of Information ( https://www.skynews.com.au/opinion/peta-credlin/transfer-of-power-voice-has-very-little-to-do-with-supporting-indigenous-australians/video/597252c79e59d25cf3bfe0c423768dc1 , https://www.skynews.com.au/australia-news/sky-news-host-peta-credlin-exposes-labors-lie-on-the-uluru-statement-from-the-heart-under-freedom-of-information-act/news-story/f1539032a44c6658c2feb352b2ddea45 , https://www.skynews.com.au/opinion/andrew-bolt/youve-been-misled-real-agenda-of-the-voice-exposed-in-a-brawl/video/d2255bf53cb4c0223e990cabd1461f14 )

15. “The campaign to sink the Voice has instructed volunteers to use fear and doubt rather than facts to trump arguments used by the Yes camp.” ( Sydney Morning Herald ) Exclusion through Emotions ( https://www.smh.com.au/politics/federal/no-campaign-s-fear-doubt-strategy-revealed-20230910-p5e3fu.html )

16. Assessing Destructive Polarisation Key questions: Do these patterns represent the symptoms of destructive polarisation? (Or: do they represent a new pattern that might be seen as destructive – a new symptom?) How severe are the differences (i.e. how deeply and destructively polarised is the situation)? How does this compare with other cases of polarisation ? Image: Midjourney

17. Thank you! Image: Midjourney

18. This research is supported by the Australian Research Council through the Australian Laureate Fellowship project Determining the Dynamics of Partisanship and Polarisation in Online Public Debate . Acknowledgments