Trajectories towards Political Engagement on Facebook around Brexit: Beyond Affordances for Understanding Racist and Right-Wing Populist Mobilisations Online

Social media are frequently implicated in the racist and right-wing populist mobilisations that found voice in support for Brexit. However, research tends to focus on platform affordances and fails to provide a sociological account of individuals’ actual experiences with these media, and how they interact with broader social and political experiences to impact attitudes. Based on interviews with newly passionately engaged pro-Brexit Facebook users, this article examines the trajectories by which individuals came to be so engaged. The findings demonstrate that the technological opportunities provided by social media were only significant in the context of offline experiences and socio-political factors. These include racist discourses that predate social media, a loss of trust in traditional media and government, and the opportunity provided by Brexit to articulate and activate pre-held attitudes.


Introduction
In the UK, the 2019 European and General Elections illustrated the reverberating effects of the 2016 EU referendum and enduring support for 'Brexit' (Cutts et al., 2020). Characterised as a defining moment in the international 'rise of populism' (Norris and Inglehart, 2019), social scientists have attempted to explain the appeal of Brexit. The 'left behind' has been used to describe a group of disaffected and previously politically disengaged citizens who seized the opportunity to 'push back against the expectations of privileged elites' by voting to Leave (McKenzie, 2017: 204-205; also Gilbert, 2015;Watson, 2018). However, this has been challenged by scholars who point out the key role played by racialised opposition to immigration (Bhambra, 2017), with the campaign said to have 'carefully activated long-standing racialized structures of feeling about immigration and national belonging' or 'a reservoir of latent racism' (Virdee and McGeever, 2018: 1804, 1807. Social media platforms like Facebook have also been implicated, for instance through the use of 'psychographic microtargeting' by the firm Cambridge Analytica to influence voters (Risso, 2018) or as part of a net-borne 'post-truth' phenomenon (Marshall and Drieschova, 2018). It is online also where support for Brexit is linked to a transnational resurgence of conservative, authoritarian and occasionally white supremacist nationalism (Davis, 2019), with some of the most popular pro-Brexit social media personalities concurrently propagating Islamophobia, far-right conspiracy theories, and expressing support for Trump and other international right-wing populist actors.
Unfortunately, while a significant body of research on online racist and right-wing populist mobilisations has emerged, some of this addressing Brexit specifically, this field has tended to neglect close-up methodologies. Consequently, there is a dearth of research that seeks to understand users' own experiences with and interpretations of pro-Brexit and related racist and right-wing populist content online. Furthermore, few sociological analyses of support for Brexit address social media, meaning sociological understanding of the role of social media in mobilising this support is extremely limited. Such investigations are increasingly crucial, amid the international mainstreaming of hateful ideologies, much of the groundswell for which takes place online (Winter, 2019).
This article contributes to this research gap, using interviews to examine the routes or 'trajectories' by which individuals became passionately engaged in pro-Brexit politics on Facebook. Here 'engagement' refers to not only activities in, but also attention to, political issues and processes (Berger, 2009), thus capturing both the activities individuals undertake online (and offline) and the meanings they make of them (or their ideological consequences). A focus on individuals' accounts of becoming engaged is particularly crucial to understanding the role of social media in mobilising huge support for Brexit before and after the referendum, and thus in the mainstreaming of racist and right-wing populist ideology. The findings of this study challenge researchers of online racism and right-wing populism to move beyond a reductive focus on platform affordances. While technological opportunities were indeed significant, these could not have mobilised individuals without the ideological and discursive opportunities found in the offline realm.

Existing Understandings
The majority of research into the role of online media in mobilising racism and rightwing populism focuses on the affordances of the online environment and their 'polarising' or 'radicalising' potential, with little offered on how this is embedded within offline contexts. There is heavy focus, for instance, on the presence or effects of 'echo chambers' or 'filter bubbles', through which social media are said to filter out opposing views and polarise attitudes (Boulianne et al., 2020;Pariser, 2011). This, combined with an increasing reliance on social media for news and information (Kaspar and Müller-Jensen, 2021), and the large volume of far-right content available on alternative news websites that seek to promote a 'culture wars discourse' (Davis, 2019: 133), is said to be contributing to the spread of racist and right-wing populist sentiments.
Since 2016, racist and right-wing populist manifestations like the Brexit referendum result have come to be described as part of a 'post-truth' phenomenon to which social media use has contributed (Hannan, 2018;Marshall and Drieschova, 2018). Although constantly reinterpreted and commonly conflated with the problem of 'disinformation' or 'deep mediatisation' (Cosentino, 2020;Cvar and Bobnič, 2019;Kalpokas, 2019), strictly speaking 'post-truth' is defined as a condition in which 'objective facts are less influential in shaping public opinion than appeals to emotion and personal belief' (Flood, 2016: n.p.). Notwithstanding the obvious problem with the proposition that engagement with politics is newly affective (Moss et al., 2020), this concept appears to attribute an unrealistic, and empirically unverified, level of significance to the media environment. Such explanations obscure the importance of racist ideology to events like Brexit, severing online behaviour and content from the offline context in which they are experienced. A number of socio-political factors clearly contributed to support for Brexit, including decades of 'backlash' to ethnic, cultural and religious diversity and improvements in minority rights and representation (Hewitt, 2005); the resurgence of English nationalism (Fenton, 2012); tabloid and other media representations of the EU, Eastern European and Muslim immigration (Fox et al., 2012;Kundnani, 2014); terror attacks, the 'migrant crisis', and growing Islamophobia (Abbas, 2019); and contributions of right-wing populist parties to mainstream political discourse (Gupta and Virdee, 2018).
Meanwhile, some scholars have focused on social media's purported affinity with right-wing populism. It is theorised that their mass dissemination capabilities, subversion of mainstream media filtering, direct linking of charismatic figureheads with their followings, and ability to recontextualise and reinterpret news, are conducive to populism's promotion (Engesser et al., 2017;Gerbaudo, 2018). Within this literature, only Krämer's (2017) work addresses the actual, individual-level processes through which users may become politically engaged in right-wing populism. He outlines the ways social media may provide an environment conducive to users' socialisation into rightwing populism, such as selective exposure to content that consolidates a populist worldview and the opportunity to express and intersubjectively confirm a right-wing populist 'identity'. Unfortunately, this work is not based on empirical evidence, offering only theoretical suppositions.
Indeed, empirical attention to user experience is thin, with the field generally relying on scraping of online data divorced from the context in which it arises. Among the few empirical studies into Brexit and social media, the focus has been on either pro-Leave material being circulated online, or users' observable online 'interactions' with this content (e.g. Bonacchi et al., 2018;Bossetta et al., 2018;Bouko and Garcia, 2020;Del Vicario et al., 2017;Rosa and Ruiz, 2020). While these studies can tell us what behaviours or discussions are taking place on social media, they provide little insight into the lived realities within which these arise or their actual ideological consequences. Owing to an over-reliance on computational or 'big data' approaches to the study of political social media use (Fuchs, 2017), and a reluctance to engage in close-up research with 'distasteful' groups (Esseveld and Eyerman, 1992), research that connects these online manifestations with the users behind them is lacking (Hall, 2021). This leaves a stark gap in our understanding of the actual role of online realms in the mobilisation of support for Brexit and related issues.
Furthermore, there has been considerably less focus on the political social media use of those who did not grow up with the internet vis-a-vis 'digital natives' (e.g. Ekström and Shehata, 2018;Storsul, 2014), despite generally being assumed to be distinct (Palfrey and Gasser, 2008). The political social media use of 'non-digital-natives' in Britain clearly warrants examination given older populations were statistically more likely to have voted Leave (Ashcroft, 2016) and social media's implication in the referendum result.
The current study seeks to begin to address this gap, focusing on non-digital-natives' own narratives of how they became passionately engaged in pro-Brexit politics on Facebook. This user-focused, interpretive approach provides the first sociological account of this transformational process of online politicisation.

Methodology
While the majority of studies into online racism and right-wing populism centre on Twitter, this study investigated Facebook users (for the methodology of the wider study, see Hall, 2021). Facebook is by far the most popular social networking site in the UK (O'Dea, 2018), and its user base is less concentrated in the young and highly educated segments of the population than Twitter (Mellon and Prosser, 2017;Sloan, 2017), making Brexit voters statistically more likely to use Facebook (Ashcroft, 2016). Support for Brexit on Facebook is significant; long-time Brexit campaigner Nigel Farage's page has garnered over 1,000,000 likes, and numerous other pro-Brexit accounts (e.g. United Kingdom Independence Party (UKIP), For Britain and Make Britain Great Again) have enjoyed large followings there (Lee, 2019).
Responding to the above-described research gap, this study took a qualitative, individual-level approach. The target group for the research was non-digital-natives in England or Wales 1 who had publicly shared pro-Brexit material to Facebook. I manually identified users who had recently publicly shared a post from pro-Brexit pages like 'Leave Means Leave' or 'Get Britain Out', and used Facebook's Messenger function to invite them to participate. Gaining individuals' attention and consent in this way was challenging; a total of 287 were contacted before the target of 15 consenting participants was reached. A consequence of this approach was that the majority of participants were heavy Facebook users (1-5 hours per day) who were intensely politically engaged on the platform, valuing it highly and using it to the exclusion of other social networking platforms. This meant the study was not able to consider more casually engaged users, but it also revealed the existence of a significant milieu of passionate pro-Brexit Facebook users and led to the question of how they became so engaged.
The participants occupied a range of social classes and geographical locations, and were aged between 40 and 73. Eight were male, seven female. However, those who agreed to participate were exclusively White British. While by no means all Leave voters occupied this category (Martin et al., 2019), the majority did (Ashcroft, 2016), and given the significance of racist and nativist tropes to the Leave campaign (Virdee and McGeever, 2018) and the importance of whiteness to these (Valluvan and Kalra, 2019), deliberate attempts to recruit from minority ethnic groups were decided against in favour of a more focused investigation.
This article is based on data from 30 semi-structured interviews (two with each participant). The fieldwork was conducted between July 2018 and May 2019, at the height of critical media coverage of then Prime Minister Theresa May's negotiations with the EU. Initial interviews were mostly conducted online using Facebook video chat, centred on general questions about how participants used social media, and lasted about one hour. Follow-up interviews, mostly in person, were conducted after my having observed their Facebook posts for one month. These asked more specific questions, including about participants' political views and the content I saw them share, and were typically longer (up to two-and-a-half hours). Online observations were not for the purpose of systematic analysis of content, but were ethnographic in nature, following links to other sites and examining the content there to deepen my understanding of the broader context of participants' online political engagement. As the focus of this article is participants' accounts of becoming engaged, the role of online observations here was limited to informing follow-up interviews. All data were stored on password-protected, encrypted media, and participants were anonymised from the outset.
Building trust with participants was required to encourage them to openly share their interpretations, an essential step towards understanding the phenomenon in question. However, given the socially harmful nature of some of the attitudes expressed -which also differed starkly from my own political positions -developing close or empathetic relationships would clearly have been ethically problematic and risked legitimising views (Esseveld and Eyerman, 1992). Pilkington (2019) argues one may research and attempt to understand a group without 'siding' with them or 'faking' friendship. I chose a strategy of 'deflection' (Waldner and Dobratz, 2019) and 'vigilance' (Foste, 2020), whereby I demonstrated active listening but neither feigned agreement with nor actively challenged participants' problematic views. While some researchers advocate a more transparent approach (e.g. Blee, 2002), this would have risked driving participants away and/or caused interviews to deteriorate into conflict. Indeed, these participants were particularly sensitive about being told their attitudes were 'wrong', this being one of the very reasons for their online political engagement. Although my high education level did generate some scepticism among participants, with several alluding to their assumption that I was on 'the other side' (Kirk, 67, self-employed manual labourer), I was likely helped by my white appearance and disclosure of my Australian upbringing; participants spoke with surprising candour about race, migration and Islam, and some appealed to my heritage when inferring a shared concern between our nations on immigration. My position as a young woman also potentially affected willingness to participate and the content discussed (one male participant even explicitly noted this after our second interview). However, participants generally seemed eager to justify themselves and their views to me, which yielded rich meaning-making data.
The study provided insights on a number of aspects of participants' engagement, many of which are explored elsewhere (Hall, 2021, forthcoming). This article, based on thematic analysis of interview data, discusses specifically participants' narratives of their trajectories towards becoming politically engaged on Facebook in order to reveal the human stories behind this transformation. While privileging the interpretations of participants does not preclude the potential for them to reinterpret past experiences (Pilkington, 2016), understanding what these experiences mean to individuals is key to investigating (and ultimately ameliorating) the impact of social media on racist and right-wing populist mobilisations (Geertz, 1977). Those interpretations are reflected here not with the intention of condoning them, but for the purpose of critical analysis.

Findings
Although when I spoke to participants they were passionately spending several hours daily discovering and sharing Facebook content around Brexit and political issues they saw as related, this engagement was new to them. As little as six months (and as long as four years) prior, they had been more sporadic, primarily social, users of Facebook, most of whom identified as not being particularly interested in politics. More than half of participants identified that this politicisation was prompted by the issue of Brexit, whether before or after the referendum. However, no two stories of this transformation were the same. Some participants recounted a traumatic life event that had propelled them into investigating particular issues online, while others described a more gradual change. A variety of personal and local experiences were narrated as the motivations behind this process.
Below I use four participants' narratives to illustrate this diversity of contributing experiences, and to demonstrate that becoming politically engaged online around Brexit was not merely a product of online platform affordances but rather was influenced by these individuals' offline experiences, and the socio-political context within which they were situated. Thus, I argue that online racist and right-wing populist mobilisations cannot be understood by focusing on technological opportunities alone but must take into account the ideological and discursive opportunities found in the offline realm. These cases are not intended to be 'typical' or 'representative' of the participant group, as every individual experience was unique (Holdsworth and Morgan, 2007). Rather, they allow situated examination of the experience of becoming politically engaged online and contextualised discussion of aspects that were salient to the broader group.

Beatrice
A retired special needs teacher in her 60s, Beatrice got 'dragged into' using Facebook around 2013, merely as a way of staying in touch with friends in South Africa, where she and her husband had a retirement home. Beatrice's Facebook use became political when she joined a local UKIP Facebook 'group' in the lead-up to the referendum. Now she was spending several hours each day posting dozens of items of content around Brexit and Islam. She felt so strongly about May 'signing away the country' with the negotiated exit deal that she had booked a mini-bus to travel to a 'Brexit Betrayal March' in Londonher first time protesting -inviting others from her local area in North Wales whom she had only met online. She chuckled before saying, with no little sense of irony, 'I am not politically minded (. . .) I have always, I'm not for one party or another, right? I vote for the party of the time who is best answering what I believe in.' That party was the Eurosceptic and populist UKIP, of which she became a paid-up member.
It was also around the time of the referendum that Beatrice began to disengage from television news owing to a perceived blatant anti-Brexit bias. However, when I asked what had sparked her interest in UKIP, she told me it was likely long-time leader Nigel Farage's television interviews and his stance on 'immigration and getting the country back' that had appealed to her. There was no 'Eureka moment', rather 'a gradual process and listening to [Farage]'. Thus, not online content alone, but a mutually constitutive media ecology contributed to mobilising Beatrice. Traditional media outlets have laid considerable groundwork for anti-immigration and anti-establishment sentiments, and inflammatory narratives circulated online 'often feed back into the legacy press' (Bennett and Pfetsch, 2018: 244) constituting a cycle of 'disinformation-amplification-reverberation' (Bennett and Livingston, 2018: 126).
At the same time, Beatrice's turn away from traditional news outlets reflects the way deteriorating trust in and satisfaction with these media have contributed to the increasingly prominent role of social media in news and information seeking (Fawzi, 2019). According to participants, 'the media has become totally dishonest' (Fred, 60s, retired administrator), and social media were described by participants like Beatrice as 'the only fair way that you could actually (. . .) find out what was really going on in the world'. Having been a teacher, Beatrice told me she felt it was her duty to share this information on Facebook to educate young people in particular. They were being brainwashed by the 'mainstream media' and today's education system, which was 'feeding them left-wing anti-democratic stuff all the time', a perception she articulated as being shaped by firsthand experience. Indeed, Beatrice was one of three participants who had worked in education, perhaps not unrelated to this strong motivation to share the 'truth' publicly on Facebook.
Although Beatrice had not been 'political' before encountering Brexit, her interest in the issue of immigration began years earlier, stemming from her anxieties about the changing demography of her local community and interactions with Muslim students and parents at the school where she had taught. She told me it was their perceived unwillingness to 'integrate' since 9/11 that bothered her. This was compounded by media coverage of the 'grooming gangs' 2 scandal, cases of which were uncovered not far from her neighbourhood. Beatrice came to believe that '[Muslims] are the only people who have come over to our country and want to dominate, (. . .) they just want to install their way of life.' In this sense, the politicising content Beatrice encountered on social media served less to provide new ideas, but rather confirmed her existing perceptions. These were informed by her lived experiences, which she interpreted through Islamophobic frames garnered through offline media representations over many years.
Despite this, Beatrice stressed that, 'I really wasn't that interested in politics until this issue of Brexit came about, and it's only because of that that (. . .) I've started on this mission.' This comment reinforces the significance of the issue of Brexit to Beatrice's politicisation. For many participants, Brexit provided a comparatively legitimate discourse with which to express culturally racist concerns. Mark (50s, secondary teacher), for example, said, 'it's only [since] this Brexit issue and everything where people have actually had the balls to talk about immigration, because you can't mention immigration without being called a fascist or whatever by certain people'. That is, while Facebook provided the forum in which to express these sentiments, for many it was Brexit that provided both the discursive frame and the impetus.

Eileen
Eileen was a 61-year-old property entrepreneur who spent most of her time at home in rural North-West England, and was, like Beatrice, conscious of her political transformation around the issue of Brexit on Facebook. She said she had once looked down upon those who used Facebook, and even after her stepdaughter created an account for her, had only used it to connect with people she knew in real life: 'I completely misunderstood what Facebook was about, (. . .) up until probably prior to Brexit, I just used it like anybody else.' Here, echoing Beatrice, Eileen alludes to what many participants now saw Facebook to be 'about': finding and sharing 'information' to 'open people's eyes to what's going on' (Lawrence, 40s, steel worker). The ability to do this afforded by Facebook was significant only within the context of participants' profound distrust of media and government, which they felt did not represent their conservative and Eurosceptic concerns.
In the past, Eileen said, she was oblivious to issues like the 'migrant crisis': 'I lived in cloud cuckoo land, and I didn't know anything about politics.' However, unlike Beatrice, Eileen vividly recounted a specific life event that she claimed transformed her attitude around politics and propelled her into full-time 'blogging' on Facebook. Eileen told me she and her husband were driving back to Britain from their holiday home in Austria one day when they reached the entrance to the Channel Tunnel in Calais. There they witnessed 'five or six migrants armed with iron bars and sticks hanging off this truck'. Traffic was almost at a stand-still when, according to Eileen, one of the men came up to her car window and threatened her, which left Eileen and her husband extremely shaken. For Eileen, 'that was my epiphany'.
Eileen began to research issues like cross-channel stowaways online. She set up a group to share 'evidence' she gathered from truckers to combat what she felt was a deliberate government cover-up. As a truck driver's daughter, Eileen told me the issue of safety on the motorway was personal to her. She soon discovered claims online that these conditions had been deliberately created by German Chancellor Angela Merkel in line with globalist visions and the 'Kalergi Plan'. 3 This was followed by ever deeper investigation, which led her to identify the European Union as a major source of enforced societal change in Britain and across the western world. Thus, similar to Beatrice's case, 'information' discovered on social media provided an explanation for prior negative experiences and existing concerns, and Brexit became the central issue through which to articulate these.
In the lead-up to the referendum, Eileen dedicated herself to 'explain [ing] what Brexit was really about', to garner support for Leave. As she posted more and more 'information' and analyses on Facebook in her sarcastic style, Eileen began unwittingly to build up a following. Now she was spending around five hours a day managing a public page with tens of thousands of followers, as well as posting publicly about Brexit and the post-referendum 'breakdown of democracy' on her personal page. Eileen's case exemplifies how lived experiences, involving family, trauma and a desire for answers, prompted active processes of net-based research, in which algorithms surely played a part but did not simply render participants passive victims of 'echo chambers' or 'filter bubbles'.

Carl
Carl was a self-employed van driver in his 50s in North-West England. He too said a particular event, namely the publicised murder of humanitarian volunteer Alan Henning by ISIL in 2014, had acted as the catalyst for his research into Islam and subsequent engagement with anti-Islam and pro-Brexit material on Facebook. Carl told me the incident had particularly affected him because he had personally known Henning: So that's why I got into, basically, (. . .) I sort of like read the Qur'an (. . .) and um I started getting into politics a bit cos I was getting a bit twisted about the way they murdered him like that.
This led him to claim that he now knew 'more [about Islam] than a lot of Muslims'. By the time we met, Carl's research had driven him not only to become a frequent public sharer of material on Facebook and occasional conspiracy theorist, but also a highly active member of UKIP in his local area. He used Facebook 'just for politics', and the only 'friend' he had on there whom he knew offline was his partner, who was also 'very political'. In contrast to participants like Beatrice and Eileen, Carl was less focused on educating others than educating himself, something he said he did 'all day' online: 'when I'm not working, I don't drink, I don't smoke, (. . .) I just relax and I just read politics and (. . .) educate meself on certain subjects'.
While Carl's main 'problem' was 'with Islam, or radical Islam anyway', he also narrated an array of concerns around Britain 'let[ting] too many people in', the ineffectual Labour council where he lived and 'all this gender crap'. Like others, the issue of Brexit had provided a significant outlet for Carl's political frustrations, which he now expressed primarily through UKIP-based activism. Like Beatrice, Carl had witnessed demographic change in his local community, which he interpreted within a variety of racist and rightwing populist frames he had been consuming online or offline: The culture, the diversity, may be OK for some people. It isn't for me. Because, I'm losing my identity (. . .) like, I stood there the other day, (. . .) waiting for me missus in that shop, (. . .) and I couldn't count one white person. (. . .) I've got nothing against these people, I just think, as you know, (. . .) it's all been done on purpose to bring them all in so we can all knit together as one unit of people, but, I'm not racist -I just don't want that.
The complex relationship between support for Brexit, cultural racism, Islamophobia and far-right conspiracy behind Carl's comment is beyond the scope of this article to unpack (see Hall, forthcoming). However, Carl's narratives revealed how his personal experiences had contributed to the 'sense' of white victimhood in the form of this culturally racist trope expressed here. He told me his childhood had been very difficult, including being 'abused as a kid' and 'chucked out' when he was 15. Since then, Carl had 'looked after meself', going from not having 'a pot to piss in' to now owning his own home and business. He contrasted this with his perceptions of what was being offered to new arrivals, telling anecdotes about moving 'brand new' furniture and PlayStations for 'African' migrants whom he claimed had received these and other expensive items for free. Although he described his current situation as 'comfortable', Carl felt he worked long hours 'to pay for migrants to come over here and get free benefits. Why can't I get benefits? (. . .) I'm working my fuckin' arse off here, and I'm literally born here.' While 'facts' uncovered online were often relayed in interviews, Carl's case is illustrative of the way participants also invariably recounted individual offline experiences to justify their political positions. Importantly, the reproduction of decades-old culturally racist tropes about migrants and privilege (Solomos, 2003) and their relationship in Carl's trajectory to a securitised Islamophobia (Kundnani, 2014) points to the central role of the historical circulation of such discourses within British society, external and prior to the popularisation of social media. It also hints at the importance of offline political events, such as terrorist incidents, to the increasing salience of these as frames within which to interpret local experiences (Abbas, 2019). Coupled with the explanatory power of Brexit, these provided the ideological and discursive opportunities for racism and right-wing populism to mobilise individuals online.

Helen
A 57-year-old mother of two young adults in the Midlands, Helen had used Facebook for 'looking at cat pictures and playing scrabble' until she started to find information about politics and Brexit. Now Helen's activity on the platform centred heavily around political posts, and her primary concerns were motherhood, 'grooming gangs' and gender politics. Though she 'never thought I'd ever get into anything like that online', she did not see her concerns as new. 'I think before (. . .) the word "Brexit" (. . .) I was never comfortable with being in the EU', she said. 'I've learned so much over the last five or six years, but I was pretty much of this mind-set anyway with the limited knowledge I've got, cos it was a feeling inside me' (Helen's emphasis).
Like other participants, Helen's politicisation, though prompted by encounters with Brexit-related content online, was informed by many years' worth of experiences or doubts, which were in turn interpreted in light of this content. She described occasions when her daughter had experienced harassment in a 'Muslim-heavy area' and bullying from left-wing 'Corbynistas' at school; her son's difficulty finding work as a 'white, heterosexual male'; and her own experience of being harassed almost 40 years ago by a 'Pakistani taxi driver'. These now culminated in a strong sense of white victimhood, a concern that 'anything to do with the white population doesn't matter' in the political and media agenda and that 'whites' were constantly being told they were 'bad'. This interpretation was compounded by the visible white poverty she regularly encountered on the streets of her hometown alongside what she perceived to be the increasingly confident presence of Asian and Muslim cultures.
Despite Helen's long-held 'feeling', she had in fact previously voted Labour. This was because, as a full-time carer for a disabled family member and coming from a workingclass background, she was opposed to austerity policies. In fact, her first experiences of politics on Facebook had been on the left: I did find myself on a few pages that were sort of quite left-wing, and I sort of came off those because I just found that they were very information-light (. . .) name-calling and mudslinging (. . .) I do find the right-wing pages far more informative (. . .) and fact-based, rather than feelings and hysteria (. . .) and the more I listened to it, I thought 'God, why did I ever vote Labour?' This exemplifies the limitations of concepts like 'echo chambers' and 'filter bubbles' to explain online racist and right-wing populist mobilisations -contact with, not insulation from, 'left-wing' discourses contributed to Helen's mobilisation. Many participants recounted negative experiences of interactions with 'the left', online and offline, which had impacted their perception of liberal politics, describing these as 'vitriolic' (Deborah, 50s, retired, unspecified occupation), 'aggressive' (Olivia, 57, property entrepreneur) or 'abusive' (George, 73, retired civil servant). Furthermore, the value Helen (and most other participants) placed on what she interpreted as 'fact-based', rather than emotional, political discourse challenges in-vogue 'post-truth' claims that these mobilisations are part of a privileging on the right of emotional rather than 'rational' claims.
Finally, Helen's '360-degree about-face' in her party and political allegiances was also evidently a product of changing political conditions in Britain. As the left and Labour moved to focus more on issues of social justice, rather than evolve her conservative attitudes towards family and gender, Helen simply found a place for them on the right, where she was promised that achieving Brexit would address all her social concerns. Thus, while social media inarguably provided Helen with access to the racist and right-wing populist frames within which to interpret her experiences, a focus on these affordances of access ultimately obscures the important role played not only by individual offline experiences but also surrounding socio-political conditions.

Discussion and Conclusion
Based on semi-structured interviews with 15 passionately engaged pro-Brexit Facebook users, this article has sought to provide a sociological account of how individuals become engaged in racist and right-wing populist politics online. Each individual's trajectory was unique and situated within their lived experiences. The narratives and interpretations of the individuals themselves demonstrate that not only platform affordances but also offline experiences and socio-political context must be considered if we are to understand online racist and right-wing populist mobilisations. The affordances of social media clearly provided opportunities for individuals to become politically engaged for the first time, but would not have been salient without an array of ideological and discursive opportunities found in the offline realm.
Cultural racism and Islamophobia, circulating discursively and structurally, played a significant role, and individual offline experiences were central to the operationalisation of these discursive frames in participants' new racist and right-wing populist political interpretations of the world around them. Participants reinterpreted experiences -major and minor, personal and political, at work, in the community and further afield -within these racist frames, which although also encountered online, undoubtedly predate the advent of social media. Their groundwork has been laid offline over many years, not least in traditional media and party-political discourse. Thus, to focus solely on platform affordances is not only to provide a superficial and fragmented account of the human, embedded experience of social media use. It also obscures the importance of racist ideology and its long history of interaction with politics.
Within a mutually constitutive media ecology, loss of trust in mainstream media, developed over many years and contributed to by an array of socio-political factors, played a significant role in participants' engagement with political content online. Other socio-political factors at play included the state of party politics in Britain, and the effects of austerity politics on seen and experienced economic hardship. This is not to validate 'left behind' discourses, nor to justify or legitimate the racist and right-wing populist ideologies engaged with by participants. Rather it is to acknowledge that neither a sense of being abandoned by news and politics, nor an 'echo chamber' or 'filter bubble' effect online, was sufficient to politicise participants alone. Experiences online should be understood as interacting with what Pilkington (2016: 74) has described as 'a complex web of local environment and personal psychodynamics and family dynamics' that form the context surrounding these trajectories. Likewise, structural explanations like 'the disinformation order' (Bennett and Livingston, 2018), 'post-truth' or 'deep mediatisation' are insufficient to explain the phenomenon in question as they do not address extramedia experiences.
The results highlight the power of Brexit to explain and articulate, and thus extend or mobilise, such concerns, through and around (both new and traditional) media. Brexit had become an extremely important issue to participants because it acted as a vehicle for explaining and expressing discontents, as well as empowered them. Arguably, a crucial conjuncture can be identified between the increasing availability of racist and right-wing populist content on Facebook, the platform's growing popularity among older segments of the population and the rise to prominence of the issue of Brexit, triggered by the announcement of the 2016 referendum. It was only at this crucial conjuncture that participants who had otherwise been largely politically disengaged were drawn to active politics, their prior experiences reinterpreted and their existing racist attitudes validated.
This is not to discount the significant role of social media in providing a means for encountering these Brexit-based explanations and empowerments. Participants admitted they generally relied on whatever 'comes up' (Beatrice) in their 'newsfeed' or 'notifications'. Many had limited awareness of the effects of algorithms on this, reflecting social media's 'technologically inflected promise of mechanical neutrality' (Gillespie, 2014: 181). While participants in this study may have previously held socially conservative, Eurosceptic or even racist views, they experienced significant transformations to active political engagement through their encounter with political social media, and political use of Facebook had become extremely important to them. As Kirk described: Social media probably changed my son's views as well as mine. (. . .) If you was thinking it ain't that bad in the EU, and then you watch some of the stuff on social media and you think 'bloody hell, I didn't know it was that bad'. (. . .) That gives you the impetus to (. . .) push on and get out of it.
While Kirk interpreted social media as the catalyst for his politicisation, what is in fact revealed here is the interaction between the technological opportunity provided by Facebook and the discursive and ideological opportunities provided by Brexit. The power of this conjuncture to propel individuals into online political engagement around Brexit and related racist and right-wing populist politics was remarkable. Although participants' engagement did not necessarily follow a linear 'career', constantly evolving and at times ending owing to strain on personal relationships and mental health, it had important political consequences. These were not only in the propagation of racist and right-wing populist content online (to which the participants contributed considerably) but also in the offline realm -at the ballot box, in street and party-political activism (as seen in Beatrice and Carl's cases above).
The problem of why racist and right-wing populist social media use appeals to people and the consequences of this is a far more formidable research question than could be addressed by this article, and many more insights on this were garnered in the current study than could be explored here (see Hall, forthcoming). However, investigating individuals' accounts of how they came to be engaged in this sphere is a crucial step towards understanding the real role played by social media in mobilising support for Brexit and racist and right-wing populist ideology. Although the importance of situating trajectories within individual lives has meant that only four cases could be fully described here, the importance of offline experiences and socio-political context was observed throughout the participant group. Each unique account demonstrates that no one factor, online or offline, can explain how and why individuals become engaged with racist and right-wing populist ideologies online. The findings thus challenge researchers to move beyond a focus on describing or measuring the presence or effects of platform affordances, and point to the limitations of studying phenomena like 'echo chambers', 'mediatisation' or 'disinformation' in isolation from the offline lives in which they may be experienced. It is only by speaking to users that we can gain a holistic understanding of the complex ways in which individuals come to be engaged with contentious politics online. This observation should not be novel to sociologists studying media or engagement in contentious politics. However, in some realms this nuanced perspective appears to have been lost owing to the popularity of (both 'sexy' and accessible) approaches that privilege the collection of behavioural trace data on social media, and the lure of techno-determinist explanations for our current political predicament.
Finally, although Britain may have now officially left the EU, backlash to ethnic, cultural and religious diversity and progressive politics shows no signs of abating. The 'scavenger ideology' of racism is incredibly adaptable (Solomos and Back, 1996: 213), as recently exemplified by the transformation of Nigel Farage's 'Brexit Party' into the anti-lockdown 'Reform UK'. Even as major social media platforms have introduced increasingly strict measures, large racist and right-wing populist groups and pages still thrive thanks to the ability of these ideologies to toe the line between acceptable 'patriotism' and dangerous hate speech. Future research must continue to investigate, both inductively and interpretatively, how and why social media are used by individuals to engage with these kinds of politics, in a range of historical and national contexts.