Signs of Democratic Resilience Following Aotearoa New Zealand’s March 15th Mosque Attacks

Abstract

Political responses to extremist attacks and other forms of social trauma might be deemed ‘resilient’ insofar as they restore order and security but can undermine democratic ideals and practices. This article offers an evaluative framework for democratic resilience to make sense of public discourse in the aftermath of extremist attacks. We argue that a democratically resilient response entails five necessary dimensions, each of which is required for dialogue across difference capable of ‘working through’ social trauma within the democratic polity. We then apply this framework to analyse the March 15 2019 mosque attacks in Aotearoa New Zealand, using mixed methods to analyse critical moments of national public discourse to illustrate how and to what extent the public sphere demonstrated democratic resilience.

On Friday 15th March 2019, mass shootings by a far-right extremist at Masjid Al Noor Mosque and Linwood Islamic Centre in Ōtautahi Christchurch during Friday prayers resulted in the death of 51 people and injury of 40 more. The attack was by far the deadliest in Aotearoa New Zealand’s recent history and came as a major shock particularly to Pākehā (European-descent New Zealanders), signalling that the country is not isolated from a world where white supremacist violence has been on the increase for the past decade (Salman, 2021).

Resilience in response to trauma is conventionally linked to being able to recover what well-being and stability was disrupted (Holloway and Manwaring, 2023). Although the term ‘resilience’ is frequently used in Countering Violent Extremism (CVE) discourse and policy (Grossman, 2021), its definition and the responsibility for fostering it remain debated (Holloway and Manwaring, 2023; Humbert and Joseph, 2019; Walklate et al., 2012). CVE policies generally focus on individuals or communities as key players or locations of resilience in combating extremist radicalisation. The goal in this context is to enhance resilience, making individuals or communities more resistant to extremist ideologies and narratives (Stephens and Sieckelinck, 2020). This is no different in most policy and scholarship of resilience following extremist violence: CVE discourse and policy predominantly focuses on the capacity of individuals and communities to resist radicalization, and on strengthening social cohesion (Grossman, 2021, 2023; McSwiney et al., 2024; Stephens and Sieckelinck, 2020). On balance, most responses to traumatic events by nations around the world have been efforts to reinstate security and recover prior conditions as swiftly and fully as possible. Yet such responses could signal political resilience while potentially undermining democratic ends, from freedom of movement and expression to the flourishing of civic spaces and cultures. What, then, might a distinctly democratic resilience to extremist events look like?

To ask this, some might look to the resilience of formal institutions such as free and fair elections and representative democracy, which are sometimes taken as synonyms for democracy. Yet these formal features are recently innovated mechanisms for realising much older and more fundamental democratic aims. Defining democratic resilience this way also misses much of democracy as ideal or practice in civil society more broadly. Others might look to the fortitude of constitutional features such as protection of rights or rule of law (Holloway and Manwaring, 2023; Volacu and Aligica, 2023). Yet these are a distinctly liberal and thus limited approach to democracy and are insufficient to illuminate more active dimensions of a flourishing democracy such as civic participation and collective decision-making.

This article offers a new heuristic for understanding and evaluating democratic resilience in the public sphere as national communities respond to extremist events. Drawing on democratic, deliberative and solidarity scholarship, we outline five necessary dimensions of democratic resilience (1. listening; 2. reflection; 3. responsiveness; 4. integration; and 5. solidarity) and then apply this evaluative framework using a critical discourse analysis (Fairclough, 2023; Wodak and Meyer, 2001) to make sense of the quality of public dialogue in the aftermath of the Mosque attacks in Aotearoa New Zealand on 15 March, 2019. Through this mixed-methods case study analysis we can observe the presence or prevalence of particular markers for each dimension of democratic resilience at critical moments of pronounced public discourse regarding the attacks. While this case demonstrates strong signs of a democratically resilient public sphere, this is neither evenly distributed nor ongoing across the polity; we therefore end the article with reflections on potential contributing factors as well as the complexity of realising these democratic features in diverse polities.

Five Dimensions of Democratic Resilience

To define democratic resilience, we turn to what at core sets democracy apart from other political systems and holds across all its iterations: democracies empower the irreducible plurality of the demos to govern. As Sheldon Wolin (1996: 98) writes,

democracy should not depend on elites making a one-time gift to the demos of a predesigned framework of rights …. Democracy is about the continuing self-fashioning of the demos.

This distinguishes democracies by a fundamental valuing of internal difference within the polity and its role in collective decision-making. ‘Rule by the people’ leaves the locus of power empty by design (Lefort, 1988: 19; Przeworski, 1991: 13), rather than tied to particular ideas or identities. Public power is instead exercised through ‘the free discussion among equal citizens’ (Cohen and Rodgers, 1994: 137) and remains open to contestation, revision and change – through free and fair elections, but also through empowered forms of civic dialogue and decision-making in the public sphere. As Claude LeFort (1988), John Dewey (2016) and others observe democracy always has the potential to lose this definitive commitment to the ongoing and open-ended activity of dialogue and decision-making across difference and become totalitarian. Groups and their values can dominate and muffle dissenting views; systems and institutions can develop that simplify decisions and ease implementation, but also prevent the active nature of democratic life. Thus, to be democratic, sites of public dialogue and decision-making must provide conditions for authentic and ongoing dialogue across difference Dryzek (2002), Dahlberg (2005), Chambers (2003), Mansbridge (2006) and other deliberative scholars characterise an authentically deliberative public sphere as one that meaningfully includes and considers different views in discussion, to develop well-informed and reasonable opinions and decisions. Others outline various norms that make this possible, such as equality, reciprocity, respect and reflexivity (Young, 2000; Gutmann and Thompson, 2004; Bächtiger et al., 2018; Della Porta et al., 2020). Informed by deliberative as well as broader democratic scholarship, this article develops an evaluative framework for democratic resilience that, we argue, has five necessary dimensions: (1) listening; (2) reflection; (3) responsiveness; (4) integration and (5) solidarity.

The first of these dimensions, listening, concerns the extent to which different perspectives and experiences, particularly those of historically marginalised communities, are being valued and attended to in the public discourse. Common efforts to reaffirm political resilience in response to extremist attacks can reinforce self-aggrandising national narratives and foreclose counter-narratives and positions that nuance or trouble simpler or more familiar frames (Schick, 2011). By contrast, a distinctly democratic resilience within the public sphere would express a fundamental care for difference, engage with nuance and complexity and even encourage greater opening to minority and marginalised vantages. This is why Susan Bickford (1996) and Tanja Dreher (2017) among others argue that listening is a demanding and altogether unguaranteed process, as vital as it is rare in political life.

While listening ensures diverse and even marginal views are engaged and understood, these views must also be meaningfully considered, with the potential for influence on thought and practice. Authentic dialogue across difference must also have the potential to prompt reflection (Chambers, 2003; Dryzek, 2009; Ercan et al. 2019; Mansbridge, 2006). This is why Susan Bickford (1996) argues that listening is an active, complex and altogether unguaranteed process. We are therefore also interested in signs of capacity and interest in questioning received terms for identity and politics, or current rules and relations organising the polity, particularly by dominant groups whose culture, values and perspectives are the taken-for-granted backdrop to social life.

Democratic resilience will also show signs that such public reflection has had meaningful impacts, in relation to immediate events and beyond them, what Dryzek, Niemeyer and others calls the necessary ‘consequentiality’ of deliberation (Dryzek and Niemeyer, 2010; Levine et al., 2005). We are calling responsiveness any action reflective of listening and learning across difference. Reflective public dialogue may not always call for change within such actions, but where it reveals previously un(der)acknowledged issues in the polity this will include revision and shift in prevailing discourses or frameworks regarding the event. Here we also count as shift and change the increasing use and normalisation of new vocabulary; we take inspiration here from Deva Woodly (2015: 5) who shows how ‘social movements have their most lasting and permanent effect not through particular policy victories but instead by . . . redefining what is at stake and what can and ought to be done about a politicized problem’. But democratic resilience would also entail what we call integration: signs that learning and revision in light of the event have diffused into prevailing discourse in such a way as to affect perception and response more broadly and are applied to interpret and respond to subsequent events.

Each of these four facets is integral to engaging across difference within a polity as democracy demands. Yet they also resonate with key stages required for healthy responses to social trauma. Kate Schick offers an evaluative frame for national responses to terror attacks using trauma scholar Dominick LaCapra’s (2014) distinction between ‘acting out’ and ‘working through’. Whereas unhealthy responses to trauma truncate collective processes of mourning and reflection in the name of clarity, strength or moving on (Schick, 2011), ‘working through’ responses enable healthy processing of trauma by attending to ambiguity, vulnerability and complexity, through four stages: (1) expression of grief, (2) public narration and its witness, (3) critical reflection, or what Moses-Hrushovski and Moses (2000) calls ‘soul-searching’ and (4) ‘political risk’, which is action that addresses ‘the underlying structures of power that facilitated those losses . . . and reflecting upon what these might mean more generally for institutions and law’ (Schick, 2011: 1853). Significant for our purposes, Schick (2011: 1840, 1853) argues that ‘working through’ social trauma by opening to uncertainty, vulnerability and risk could also be described as, and require for its exercise, an agonistic or ‘radical democracy’, in sharp contrast to national responses that seek to preserve ‘politics as usual’. Schick (2011), drawing on LaCapra (2014), emphasises that there is no set duration or order for these stages, and how it may look for a society to fully undergo each stage will vary. Yet each of these stages is required for a community to collectively process a traumatic event.

We add to this picture of democratic resilience a fifth dimension of solidarity. Solidarity is, in Juliet Hooker’s terms, ‘the reciprocal relations of trust and obligation established between members of a political community that are necessary in order for long-term egalitarian political projects to flourish’ (Hooker, 2009: 4). Similarly, Della Porta et al. (2020) maintain that solidarity and collective concern for each other’s well-being, inclusion, safety and equality are key indicators of the quality of democratic deliberation. To overcome challenges of citizenship in the wake of crises and violent extremism and to keep democracy alive, dangerous calls for cultural sovereignty of the white ethnic majority (Appadurai, 1996; Bauman, 2017) are countered by developing participatory and deliberative spaces where solidarity can be rebuilt and social cohesion revitalised (Mason, 2020). As such, solidarity is essential for democratic life, perhaps best conceived as the essential relational and affective conditions for the ongoing critique and revision democracy requires. Without it, these other dimensions of democratic resilience could be an intellectual or abstracted exercise that leaves actual relationships unformed or unhealed, and therefore could lead to polarisation or fragmentation. Understandings and intensities of solidarity might themselves be challenged and changed through democratic inquiry, but solidarity remains a fundamental expression of care for difference – the irreducible plurality of actual lives – that, in turn, enables listening, reflection and revision to be ever-oriented to the living community. However, relational care without social critique proves equally problematic in democratic terms, as it can obfuscate or simplify complexities and ambiguities or collapse significant differences in positionality, experience and perspective in the name of restoring unity and harmony. Solidarity, then, adds an essential fifth dimension to an account of democratic resilience.

Methodology

Determining democratic resilience in the wake of a violent extremist attack is fraught with methodological complications, as a nation’s collective response to a traumatic event cannot be easily measured. One starting point is Oliver Escobar’s (2022: 392) work on methodologies conducive to studying public deliberation: he calls for creative mixed methodological approaches that can grapple with complexity, with ‘attention to both patterns and cases, statistics and narratives, measures and meanings, numbers and words’. We have taken a mixed-methods approach where findings and inferences are derived through the methodological integration of news media data analysed through political theory frameworks.

News publications allow citizens direct access to the public sphere, as a key site where people communicate about issues of common concern, construct and contest collective identities and negotiate solidarities (Gastil and Levine, 2005). Deva Woodly (2015: 19) maintains that the public’s engagement with media is a key driver for democratic change, because

the way that we talk about issues in public both reflects and determines what solutions are considered desirable or plausible in the commonplace logics that shape the politics of a particular moment.

Accordingly, we focused our study on national and regional newspapers, with particular attention to editorials and everyday commentors, as representative of mainstream political discourse. We built a data set of all (N = 4031) articles that referenced the attacks from first public mention to 5 years on, in March 2023. We used Proquest to search terms related to the mosque attacks (including Christchurch or Mosque and attack* or terror* or shooting*) to collect a data set that was as comprehensive as possible.

We used descriptive statistics to examine the full data set of (N = 4031) to highlight critical incidents within a chronological narrative. We employed a conversion Mixed Method Research ( ‘MMR’) design (Escobar, 2022), where the news article text was analysed both qualitatively and quantitatively. Moments of heightened engagement that deviated from the trend or ‘spiked’ were selected through adaption of a quantitative event analysis method that links spikes to specific events, aligning with increased reporting. This revealed 10 key moments of heightened national discussion about the 15th March attacks (see Appendix Figure 1), as well as the events at the time prompting this: (1) Arraignment and proposed reforms of firearms legislation on 18 March 2019; (2) Open air prayer service at Hadley Park on 23 March 2019; (3) National memorial service on 30 March 2019; (4) Mosque open day 26 April 2019; (5) Report on hate crimes and Stuff investigation ‘see no evil–hear no evil’ on 14 September 2019; (6) Anniversary Commemorative events on 14 March 2020; (7) Criminal sentencing and victims testimonies on 28 August 2020; (8) Royal Commission Report released to public on 9 December 2020; (9) Christchurch Hui on 16 June 2021 and (10) Anniversary Commemorative events on 16 March 2022. We focused our qualitative analysis on media from these ten key moments of heightened public discourse on the attacks, which reduced our initial data set to N = 576 for further synthesising through descriptive coding.

Critical discourse analysis (Fairclough, 2023; Wodak and Meyer, 2001) is a particularly useful lens for examining tensions and transformations across difference within news media because discourses structure whose perspectives are centred or marginalised. Following Deva Woodly’s (2015) analysis of transformative democratic discourse, we sought to reveal patterns that are present in news discourse. We elected to assess the articles at the paragraph rather than mere headline level, to generate a careful and systematic catalogue of content-in-context. We explored patterns and relationships in the data set in relation to our five dimensions of democratic resilience. We came to our findings by qualitising the coded data and highlighting patterns of engagement across difference that indicated reinforcement or divergence from democratic resilience. This evidence-based mixed-methods approach helped us derive a set of ‘negotiated and warranted’ (Bazeley, 2018: 277) findings through reflection and revision of our qualitative frameworks based on the underlying data set. Importantly, the salience of these frames is indicative of the political work required to repair tears in the social fabric in the wake of terror attacks and offers a methodology for discerning presence and pathways for democratic resilience.

This study has its limits: often we are seeking to observe qualitative and diffuse features of public dialogue and must use their most observable proxies within such a vast data set, for instance, using changes in representation of marginalised authorship or quoted speakers as an indicator for changes in public attention or the presence of open-ended questions as an indicator for reflection. Our focus on moments of pronounced public dialogue about the attacks also means we are unable to speak to broader trends or long-term patterns, and we exempt events and discussions outside these heightened moments, potentially missing important context or developments that could influence the analysis. Acknowledging these limitations, this novel mixed-methods approach for the study of the quality of public dialogue at national scale still has significant value it contributes to democratic resilience studies. Ultimately, we chose this approach because it allows for an in-depth examination of resilient responses during times of collective significance, with particular attention to the intricacies of engagement and power dynamics that were especially important to us.

Signs of Democratic Resilience in Aotearoa’s Public Sphere

Listening

We have a duty to listen, hear, and respond with words and deeds that support you . . . words will be one of the most powerful ways we can show that March 15 has, or hasn’t, changed us. (PM Jacinda Ardern at Ko Tātou, Tātou (We Are One) Memorial Service, One News, 29 March 2019)

Listening across difference in this context can be seen in any efforts to acknowledge or extend representation to diverse voices, particularly those most affected by events and otherwise absent or underrepresented in the public sphere. Leaders’ and prominent public figures’ speeches in the immediate aftermath helped define events and shape consequent discourse (Vatnoey, 2015). Often these early responses conflated internal difference within the polity: for instance, framing events as an assault on ‘all of Christchurch’ (by Rt Hon Winston Peters) or ‘all New Zealanders’’ (by Rt Hon Simon Bridges), denying pre-existing social issues in describing events as an ‘end to innocence’ (Stuff, 2019), or PM Ardern’s ‘They Are Us’ narrative that became so prevalent in public discourse.

News cycles were particularly fixated on PM Ardern’s empathetic response (McConnell, 2019). There were mixed reactions within Muslim communities to Ardern, as some critiqued her ‘They Are Us’ narrative and wearing a hijab when first comforting victims’ families as surface-level gestures (Ghabra, 2022; Rahman and Emadi, 2018; Tharoor, 2023). Yet Ardern’s language was distinguished from other leaders at this time by its acknowledgement of complexity and difference, even in gestures of solidarity (Fitzpatrick, 2022, 2023). For example, her initial address to victims’ families hours after the attack stated that the nation ‘cannot know your grief but we can walk with you at every stage’ (The Guardian, 2019); her speech at the national vigil 2 weeks later also emphasised the particular loss and impacts on victims and survivors and the need for others to take responsibility for hate speech in their own words and actions. On the first Parliament sitting since the attack (19 March), parliament as a whole opened space to social difference in a similar way: members of parliament used their Question Time to offer condolences to those directly affected by the attacks, Speaker Trevor Mallard broke with parliamentary protocol to lead a procession of various faith leaders into the House, and for the first time in a Christian-dominant country, a parliamentary session opened with a reading from the Holy Qur’an.

Within public discourse more broadly, our analysis found that Pākehā (European-descent New Zealander) authorship and framing continued to dominate the news on the Mosque attacks. Yet at key moments of mourning and judgement we found evidence of increased representation of Muslim voices, whether through ‘passing the mic’ to Muslim authors on media platforms or through interviews and quotations. For example, across our key moments of public discourse about the attacks, Muslim authorship increased from 1.5% in the first week after the attack to 18.9% by the first commemorative event at Hadley Park. This suggests that commemorative events play an important role in extending authorship. We also observe an overall trend of more Muslim and Māori perspectives being represented over time, such as through interviews and detailed coverage of speeches by Muslim leaders at commemorative events. News media also focused on giving greater voice to affected Muslim communities at times of mourning and commemoration through online memorials on honouring the lives of those lost and story-telling by survivors (Closs Stephens et al., 2021).

Another key moment of listening was the criminal sentencing on 28 August 2020, where the percentage of Muslim perspectives represented peaked at 81.82% and victim impact statements expressed a range of perspectives and emotions. News media collectively refused to give a platform to the perpetrator and his white supremacist beliefs, choosing instead to centre victims and survivors, illustrating the power of deliberate silence as well as voice (Roy and Jong, 2019). During the perpetrator’s trial in particular, the public was more concerned with listening to the voices of the affected Muslim communities (Graham-McLay, 2020).

Listening is more than making room for voices; it ‘depends upon the creation of safe spaces in which communication can take place’, (Schick, 2011: 1850) and demands openness and vulnerability from those who often do not want to hear. Susan Wardell (2022) reports in her qualitative study of online discourse following the attacks an unusual shift in the ‘feel’ of social media, as everyday users adjusted their normal patterns of online communication: ‘quieter’, with fewer responses to posts, more reliance on emojis than words and those known for more argumentative or provocative posts ‘gone quiet’; ‘slower’, including time between posts and responses; less argumentative or judgmental and more ‘tentative’; and (temporarily) ‘kinder’. Multiple articles on 18 March 2019 also overtly discuss the need to amplify Muslim perspectives, and that ‘the non-Muslim communities of Christchurch should pause and listen to these stories and others, while also respecting those who do not wish to share their stories’ (The Press, 2020).

Simon Koschut (2019: 154) argues that inside a nation’s ‘grief culture’, members ‘are woven together in asymmetrical power relationships’. He suggests that national grief becomes politically contested when individual or collective manifestations of attachment and identity conflict with dominant codes for grief within state-led responses. We see in this case ongoing challenges of Muslim voices not being listened to, from critiques of the first national conference on Countering Terrorism and Violent Extremism for still marginalising Muslim and Māori voices, to the Muslim community in Christchurch feeling unheard around the non-Muslim nature of initial commemorative events and the city’s planned memorial park.

Reflection

I have done nothing so why do I feel guilty?

Maybe I feel guilty because I have done nothing. (Letters to the Editor, The New Zealand Herald, 30 March 2019)

A democratically resilient response to social trauma entails critical reflection on the conditions that allowed the suffering to take place. In settler-colonial Aotearoa New Zealand where Pākehā dominance and ongoing racism and discrimination persist despite the country’s ‘superdiversity’ (Spoonley, 2015) there was much within the broader political context on which to reflect in the context of the attacks. This was made evident by the differing reactions to the Mosque attacks, where dominant Pākehā groups expressed a sense of shock and disbelief (with ‘#This is Not Us’ trending on Twitter), while Muslims expressed frustrations over ignored forewarnings and shared views ‘distinct from that of the mainstream in that the attacks [were] seen as a culmination of, rather than an exception to, the everyday lives of Muslims’ (The Royal Comission, 2020). One opinion writer observed 2 weeks after the attacks that

if we’re not willing both to honestly examine our own attitudes, and to confront the extent to which white supremacist thinking exists in New Zealand, we’ll never work out how it’s able to hide in plain sight among us, which is essential to rooting it out (Royal Commission of Inquiry, n.d.).

Critical reflection on the broader context in which terrorist attack occurred and differences in lived experiences, particularly by dominant groups, became an integral part of the process of working through social trauma in this case. As the New Zealand Herald, reporting on the National Memorial Service, stated, ‘For many ordinary New Zealanders, with no previous connection to Islam . . . The biggest wake-up call, after the security failures, has been the revelation of how marginalised Muslims in New Zealand have felt’ (The Royal Comission, 2020). Examples of questioning include Hawkes Bay Today’s ‘Facing Hard Questions’ and The Press’s ‘Questions Left After the Terror Attack’, which addressed the warning signs ahead of the mosque attacks, such as growing racism, Islamophobia and rising white supremacy online. Of 574 articles, 322 demonstrated questioning (58% of the dataset). Overall, this is an encouraging indication of critical reflection in the public sphere.

Reflection was, in the main, not common as a first response: we observed less questioning in the immediate aftermath of the event (with only 25.98% of articles reflecting on the underlying causes of the attack). This aligns with the first necessary but largely ineffable phase of expressing grief when ‘working through’ social trauma. However, this shifted over time. Key consequent moments of mourning are also linked to increased reflection: questioning significantly increased to 78.83% on 23rd March 2019, corresponding with the first open air memorial service in Hadley Park, as well as on 30th March 2019 at 70.12%, following the national memorial service. This shows that mourning and commemorative events in particular prompt more critical reflection as well as listening.

Key moments of judgement showed the greatest correlation with reflection in national discourse. Questioning spiked on 14 September 2019 (84.38%) with the simultaneous release of the Royal Commission Inquiry Hate Crime Report, Stuff’s seven-part investigation on the attacks ‘See no evil–hear no evil’ (Bingham, 2022) and legislation requiring New Zealand history in the school curriculum. Reflection hit its highest on 28 August 2020, the day of the perpetrator’s criminal sentencing, with 84.85% of articles including questioning. Yet these patterns of questioning do not hold when we control for positionality. Overrepresentation of Pākehā authorship across the country means that these patterns observed are much more linked to Pākehā perspectives at these key moments. By contrast, questions raised regarding possible reasons for the attacks and connections to broader societal factors were present almost immediately and continuously in Māori and Muslim-authored media. This shows that Muslim, Māori and other minority authors still carry the burden of questioning for society more broadly, in addressing hate crimes, discrimination and racial inequality.

Responsiveness

The white-supremacist terrorist attack in Christchurch has given New Zealand cause to reflect on the roots of racism in this country. (Letters to the Editor, The New Zealand Herald, 30 March 2019)

Listening and reflection are crucial to a democratically resilient response to traumatic events, but they are insufficient without signs of learning and action in response. To observe shifts in thinking within public discourse in response to immediate events, we identified key terms used across articles to interpret events – racism, Islamophobia, hate speech, colonialism and white supremacy – and observed any changes in their usage, as well as considering the positionality of authors. We found that Māori and Muslim authors were more likely to use the term ‘racism’, and Māori authors were the only group to speak about colonisation to discuss the attacks within our 10 key moments of heightened discourse. Muslim and Pākehā groups were more likely to refer to ‘hate speech’. Interestingly, Pākehā authors were the group that most commonly referred to the issue of ‘white supremacy’, indicating self-reflection as a group within national discourse.

We observed three further significant changes in the national discourse about the attacks over this time. First, more Muslim voices in national media also meant pre-existing experiences of exclusion and discrimination specific to Muslim communities and past ignored efforts to alert leaders to increasing threats to Muslim communities were given greater voice and public attention. Tied to this, we see an increase over time of Pākehā-authored media referencing white supremacy (from 6% to 29%), Islamophobia (from 5% to 38%) or hate speech (from 2% to 67%), in the articles from our sampled periods. These shifts in public discourse troubled narratives of national unity and highlighted the erasure and complicity of mainstream New Zealand, opening possibilities for listening, learning and change across unequal positions in the polity. Second, many Muslims also shared in early weeks that ‘things have changed since the terror attack’, with more experiences of curiosity and warmth from non-Muslims (Miller and Beausoleil, 2023). Finally, changes in national memorialisation practices demonstrate increased listening and responsiveness to difference: at the request of the bereaved families and survivors, the third anniversary passed without a public memorial in accordance with Islamic mourning practices and was replaced with Islamic Awareness Week to provide a platform for members of the Muslim community to share their experiences and perspectives.

There were other forms of responsiveness to events noted in media from these key moments. On the same day on 30 March 2019, gun owners voluntarily relinquished their firearms, and news reported many civic initiatives already underway to build cross-community understanding and ‘help minorities feel welcome’ in Christchurch. Letters to the Editor and reports on attendee perspectives from vigils often included calls to actions: to better support Muslim women; make changes to school curriculum; open minds and learn; become more tolerant and ‘truly become the people we think we are’. In short, this time was observed to be ‘a wake-up call for New Zealanders’ to racism and xenophobia (Campbell/Christchurch, 2019). In an interview 1 year on, Riaz Rehman and Rana Naser report changes in public attitudes towards Muslims: ‘People are very welcoming now. They’re telling us we are accepted and telling us we are valued as part of society’ (Jacobs, 2020). Yet there is still unresponsiveness noted: on 9 December 2020 media report that the Royal Commission Report observes that ‘while the courts have dealt appropriately with the terrorist, the affected families, survivors and witnesses have not had their questions answered, particularly around the question of accountability’. Human Rights Commissioner Paul Hunt (2019) reflected that while there is still much work ahead, since the Christchurch terror attack.

I now feel there is a movement, for one reason or another, to call out racism, to give nothing to racism, nor to Islamophobia, nor to anti-Semitism. I think there is a shift and it is a very welcome long overdue shift.

Integration

Something’s changed. Something’s fundamentally changed.

And it’s a lot bigger than just firearms policy. (Letters to the Editor, Waikato Times, 30 March 2019)

Democratic resilience in the face of social trauma requires ‘political risk’: as critical reflection calls into question the sociopolitical context surrounding an attack and asks ‘who will we become in the response’ (Schick, 2011: 1852), a democratic public sphere will ideally show signs that reflection and learning have applications beyond the immediate context. There were a number of examples of interpretations and responses to later and broader events informed by reflection on the attacks. Prevailing in national media were reflections from citizens that, as Ardern also observed, ‘we have been changed’. The lessons of March 15th are discussed in the context of many later events and debates: the importance of New Zealand history in the school curriculum; several investigations into white supremacy, racism and hate in the country; debate over changing the name of Christchurch’s Crusaders rugby team; changing the usual focus of ANZAC day to the loss of life and need to take action to prevent ‘hatred and violence’; InternetNZ’s open letter about changes they have made to create an ‘Internet for good’ and ‘Internet for all’ in light of events; challenges to the trope of ‘average Kiwi’; attention brought to the failure to replace the previous Race Relations Commissioner; exposing poor housing conditions of Muslim refugees and quick disapproval of racist comments by MP Shane Jones regarding the Indian community, all with reference to the attacks.

News also reported on institutional responses that demonstrated application of learning to the broader context. Striking but by no means exhaustive examples include the Ministry of Justice’s National Action Plan Against Racism, Ardern’s transnational Christchurch Call to eliminate violent extremist content online, the addition of ‘Social Cohesion’ as core theme in the Royal Commission of Inquiry and consequent establishment of the Social Cohesion programme in the Ministry of Social Development, the Ministry of Justice initiating hate speech law reforms, and changes to national security surveillance to include white supremacist and far-right groups. One community-level example to note is the #TheyAreUsShutDown hashtag movement, where a 75,000-strong petition initiated by the National Islamic Youth Association eventually led to the suspension of a proposed film They Are Us centred on PM Ardern (NIYA, 2021).

Yet in this case, the shift from reflection to responsiveness and integration – what Schick (Schick, 2011) calls ‘political risk’ – is still far from complete, with ongoing work still required of government, leaders and civil society. As of August 2023 only six recommendations of the Royal Commission of Inquiry report are marked ‘complete’, and hate speech law reforms were tabled in 2023 by Leader of the Labour party Chris Hipkins as too ‘contentious’ and not a ‘bread and butter’ issue for the country (McClure, 2022). When Hipkins made this announcement, Islamic Women’s Council of NZ Coordinator Alyia Danzeisen (Manhire, 2022) said that in the 2 years since the Royal Commission issued its Report, Muslim women and children had ‘faced an onslaught of hatred’:

How many more people need to suffer from hate and for how long before our legislators will have the courage to do what is right? There is no more time or consideration needed.

In the same period, the Human Rights Commission reported a 20% increase of racist acts since 15 March. Some most affected by the attacks or advocating for the most vulnerable have lost faith in institutional action in light of how slow and even stalled programmes like the Ministry of Justice’s National Action Plan Against Racism or the Ministry for Social Developments’ Social Cohesion programme have been. Aotearoa poet Mohamed Hassan described the very notion of ‘they are us’ as ‘a promise made but not kept’ (Hassan, 2024). Whether this case truly represents an example of democratic resilience will depend on whether ‘political risk’ results in meaningful changes, so that this promise is fulfilled.

Solidarity

We are determined to not let anyone divide us . . . We are determined to love one another and to support each other …. We are here in our hundreds and thousands, unified for one purpose. That hate will be undone and love will redeem us. (Imam Gamal Fouda, Open air prayer service at Hadley Park, Christchurch, The Guardian, 23 March 2019)

A democratic public sphere is inclusive, reflective and responsive, as deliberative democrats often define it (Dryzek, 2002). Collective concern for each other’s well-being, inclusion, safety and equality are key indicators of the quality of democratic deliberation (Della Porta et al., 2020). But it is also sustained by solidarity, the relational ties and affective sensibilities of reciprocity, trust and obligation across the polity that make the difficult, uncertain work of ongoing democratic practise possible. Expressions of solidarity from both PM Ardern and the general public were often reported and praised in national and international media, from the thousands spontaneously standing vigil at Mosques and record-breaking vigils around the country, to non-Muslim women wearing headscarves 1 week after the attacks, to sports teams wearing ‘They Are Us’ on their uniforms, to hundreds of school children in Queenstown using their bodies to form the word ‘Kotahitanga’ (togetherness), to major community-sourced funding rapidly raised for survivors.

Yet democratic expressions of empathy or community must also be attentive to social difference. Some gestures of solidarity, as discussed earlier, collapsed or erased difference in the name of a simpler notion of political unity. Yet overall this case demonstrates an increasing and, compared to other cases overall, greater acknowledgement of internal difference: what Ghassan Hage (2019) described in this case as ‘the more difficult love, the love that is able to cross cultural boundaries and encompass multiplicity and difference rather than remain entrenched within the boundaries of oneself’. The presence of the other facets of democratic resilience make this possible, as listening, reflection and revision encourage features observed above in public discourse: focus on those most affected by the attacks, acknowledgement of differences in positionality and responsibility between Muslims and others and critical attention and response to broader social conditions of Islamophobia, xenophobia and racism. This moves the public beyond a surface level empathetic identification towards a deeper, reflexive understanding and complex forms of solidarity. This may strike some as a particularly big ask given contemporary challenges of increasing polarisation and disengagement with formal politics, and yet this is nonetheless necessary to realise democratic ideals of truly meeting across difference. As many Muslim and political leaders expressed over this time, this means that solidarity is explicitly aspirational, an unfinished project that calls for a commitment to continue to listen, question and change. In Ardern’s (2020) words,

A year on I believe New Zealand and its people have fundamentally changed. . . But the challenge for us will be ensuring that in our everyday actions, in every opportunity where we see bullying, harassment, racism, discrimination, calling it out as a nation. . . That could be the legacy of those who were killed.

Conclusion: Conditions and Implications for Democratic Resilience

This case shows strong indications of what we have outlined as key facets of a democratically resilient response: greater inclusion of Muslim voices; increased reflection about who ‘we’ are as a community and different responsibilities therein; growing normalisation of terms like white supremacy and hate speech; and application of learning to future events and broader phenomena. There are exceptions throughout, of course. A polity is by definition a plurality of positionality, experience and perspective, and therefore, any given response will not be uniform. These moments of democratic resilience coexist with their opposites from others in the social body – continued neglect of marginalised people and claims; refusal to reflect, acknowledge or act to address societal factors connected to the violence; recourse to simpler and exculpating narratives – and can also provoke backlash to further entrench undemocratic responses, as in the pushback from Christian groups that the nation had ‘gone too far’ after the first Muslim prayer in Parliament. Nonetheless, public discourse demonstrated distinctly democratic dispositions, commitments and practises, at times in sharp contrast to conventional national responses.

Some factors likely played a critical role here. First, as the first nationally experienced major terror attack, these events sent shockwaves through society that likely link to the strength and quality of the response. Second, national leadership was also pivotal. Broad uptake of PM Ardern’s various phrases and framings and impact of her decisive actions attests to her particular influence: for example, refusing to name the perpetrator and instead centre the victims; calling the shootings ‘terrorism’ within hours; her phrase ‘They are us’, voted New Zealand’s quote of 2019; and swift launch of the Royal Commission, hate speech law reforms, gun law reforms and the Christchurch Call. Notable unity among politicians across the spectrum at this time also meant these frames of unity, solidarity and disavowal of violence as a nation were widely shared, and events were not used for political gain.

Third, institutional or organisational actions to respond to the attacks can be seen here as ‘carriers’ (Eyerman, 2011) of public discourse, as the news reported on and thus encouraged public discourse about them. Facebook’s open letter naming white nationalism and racism, for instance, increased the use of these terms over this time, just as advocacy efforts by the Human Rights Commission and the Islamic Women’s Council affected the content and nature of public discourse. The Royal Commission of Inquiry report stimulated public discussion regarding core causes and implications, while memorials played a key role in the extent to which communities publicly grieved and reflected on events together.

Fourth, despite the shooter’s intense efforts to seek public attention via livestreaming the attacks, releasing his manifesto and representing himself at his trial, media reporting here was exceptional in how minimally it reported on the shooter himself. Early on, media agencies made a formal pact to report on events in ways that prevented the media becoming a platform for the terrorist or ‘any message, imagery, symbols or signals’ (Greenfield, 2019) that promoted extremism. This was also supported by the swift decision to categorise video footage of the attacks and the manifesto as ‘objectionable’ material subject to strict regulation of viewing or distribution of the shooting footage, and how seriously the courts took this, with over 35 prosecutions. This restriction prevented the amplification of the terrorist’s message and both extension and sensationalism of harm through traumatic footage. Throughout national coverage the shooter remained both nameless and faceless. This reflected a profound care for victims and national solidarity and commitment to making changes that prevent future harm. Media focused instead on the specific perspectives and experiences of those affected by the attacks, as well as broader factors such as the far-right movement, racism and hate speech, the role of social media and gun control laws over this period far more than the terrorist. Ultimately, ‘the media’s restraint in not sensationalizing the perpetrator and his views appears to deviate from the normal trajectory of media reporting in such events internationally and within New Zealand’ (Every-palmer et al., 2021: 278).

Some of these external factors are important to consider in fostering and protecting democratically resilient responses within the public sphere. But we also propose that the five features of democratic resilience outlined here support as well as express democratic resilience. For instance, greater media representation of most affected and marginalised voices proves integral to how much a given polity is able to attend to them; critical reflection and revision within national media or coverage of gestures of solidarity likewise reinforce how a given community interprets and responds to events, normalising and fostering democratic qualities in the public sphere. Likewise, this heuristic for democratic resilience provides terms with which to explain cases where democratic resilience is lacking: did a case rush past collective listening to and reflection on the complexity and diversity of experiences and perspectives in relation to events? Did a push to return to ‘politics as usual’ foreclose meaningful interrogation, learning and change of broader societal conditions connected to the violence? Did shift from reflection to responsiveness and integration – what Schick (2011) calls ‘political risk’ – lead to tangible changes that bring a society closer to democratic ideals?

Ultimately, this analysis highlights that although public dialogue may reflect democratic resilience, it is far from enough to demonstrate fully democratic responses to such events. Institutional actions are shown to often be major precipitators and leaders of public discourse. Without following through on ‘political risk’ and change of social conditions implicated in events, we have not truly ‘worked through’ social trauma as a community or realised democratic ideals. This study also highlights that while these five dimensions support democratic resilience in the wake of extremist attacks, they may not emerge immediately or consistently across time and society. The presence of any one of these features may occur in fits and starts, and unevenly across a polity, even, as in this case, in the presence of opposing discourses. Indeed, earlier phases that fail to reflect these five features can also be what provokes them, as in this case where initial framing of events as an ‘end of innocence’ and disruption of national unity prompted contestations that led to greater collective learning regarding the complexities of the polity and the implication of mainstream New Zealand therein. The degree to which each dimension is required will also vary in relation to a given attack: in this case, white supremacist violence in a settler-colonial, Pākehā-dominated context called for significant ‘soul-searching’ and ‘political risk’.

Whatever form they take, these dimensions will nonetheless be integral to a truly democratic response, as without them public dialogue and decision-making in a diverse polity could not occur. Moreover, by attending to the diversity of its polity in asking core and complex political questions in light of extremist violence, a democratically resilient response will generate more equitable and effective measures to address root causes, meet diverse needs and prevent future extremist events. Yet this response must entail ongoing work, in dialogue and action, that is by definition open-ended and uncertain, and produces material change. As Serena Chen (2020: 90) reminds us,

Democracy is not a thing we have, it is a thing we do. It requires ongoing vigilance to be maintained and kept healthy, to balance due process with real action …. To protect ourselves, our rights and freedoms, we must protect democracy. But to protect democracy, we must protect each other.