Abstract
Public engagement is a gendered experience, whether offline or online, something which is reflected in women’s experiences of social media. In this article, we seek to systematically explore the experiences from politically engaged women Twitter users in New Zealand in order to draw some lessons, through a thematic and interpretative analytical approach, at four different strategic levels on how to deflect intimidating and aggressive behaviour. We conclude that understanding strategically how structural social locations like gender effect the ability to contribute to political participation and engagement, if addressed, can produce more inclusive and productive online political and policy spaces. Further, this strategic approach involves connecting together different levels of response to online negativity such as platform tools, space-curation, and monitoring, having these made coherent with each other, as well as with this strategic understanding of how structural social location plays into access and use of online political and policy spaces.
Keywords
Introduction
Previous research has demonstrated that engagement in politics and government is a gendered experience, both online and offline (Åström & Karlsson, 2016; McGregor et al., 2016; Evans & Hayes Clark, 2016). This has consequences for women’s participation in politics and government, as they are not only engaging in spaces where they often represent a minority, they also encounter negative online practices with the aim of intimidating, or even evicting, them from social media realms (Graham, 2017; Vochocoa et al., 2015; Citron, 2009; Eckert, 2018; Vitak et al, 2017). Among these practices we find clap-backs, bandwagoning/pile-ons, gaslighting, intimidation, doxing, stalking, ‘man-splaining’, and trolling. Despite the prevalence of online political spaces, the solutions offered to deal with online negativity have remained at the individual level, rather than the strategic and structural (Citron, 2009; Jane, 2017; Vitak et al., 2017; Eckert, 2018; Barker & Jurasz; 2019).
This article thus aims to analytically present some of the lessons of a group of women actively engaging in online politics, at four levels of strategic approach. It presents a qualitative case study based on interview data with women Twitter users of how politically active women engage with a gendered online world, and how they develop certain defensive strategies in their use of social media platforms. Guided by previous studies on gender and online political participation, as well as research on approaches to online negativity, this case study shows that being a politically active woman in a gendered online world requires a balance between deflecting bad behaviour through technical and strategic choices while maintaining engagement.
The article will primarily look to how the women in this case study managed participating in politics in digital spaces with the negativity they experienced there, and how they perceived it structurally as gendered (and raced, etc.). The research question for this paper is thus in the lessons learned in systematically exploring the experiences of political engagement from women Twitter users in New Zealand on how to deflect structural intimidating and aggressive behaviour at four different strategic levels.
Furthermore, these findings demonstrate an awareness of the dichotomy of the openness and unregulated nature of social media being a ‘double-edged sword’, as it were. These women saw the potential in hearing voices spoken that they would not otherwise hear, but also that a similar potential existed for multiple axes of structural oppression to be reproduced unchallenged. While the findings here refer to more traditional political online space, the results are valid for all kind of online interactions requiring political engagement and voice, and provide lessons for other forms of digital politics.
The conclusions here are aimed at an audience of anyone intending to set up or manage political and policy spaces online, where they are concerned about how to promote engagement. This includes a wide set of potential stakeholders, from those politicians creating social media spaces for community discussion and contribution, to those policy professionals wanting to gather perspectives online, to those that are working to regulate, moderate, or censor online political spaces. Particularly, this paper intends to provide the above as knowledge in a useful manner that is applicable at different strategic levels.
Empirically, the study is based on 25 unstructured interviews using prompt-style questions with women active Twitter users in New Zealand, providing a range of participants from different ethnic backgrounds, locations around New Zealand, and levels of political involvement. A theoretical framework of narrative analysis (see section three) was used during this analysis to look for the understandings and social meanings that the participants were invoking in their constructions.
The remainder of the article is structured as follows. In section two, we review the literature around gender, politics, and online experiences, then go on to look at approaches to online negativity. In section three we present the analytical strategy and methods for our study. Our findings are presented and discussed in section four, followed by our discussion relating the literature to the findings in section five, our final remarks.
Gender, politics, social media, and negativity: A literature review
Various studies have looked at women’s use of the internet for political purposes. Both Morahan-Martin (2000) and Sutton and Pollock (2000), looking at pre-social media internet use, have looked at gender-specific inequalities in internet-based political activism. Morahan-Martin (2000) argues that women, as opposed to men, often are victims of harassment online simply because they are women, due to gendered power structures in wider society. The experiences of negativity online are thus not because of a political position that any individual woman may hold, but rather in simply holding any political position. Sutton and Pollock (2000) point out that gendered analysis was not brought to bear on how online political spaces were resourced or structured, and hence political participation online was gendered as a result of replication of wider gender inequalities. In research on blogs and their use by young women, it has been found that through blogs, teenage girls have been actively reframing what it means to participate in feminist politics (Keller, 2012) echoing Schuster’s work (2013) on young feminists in New Zealand. Further on this creation of alternate spaces for political participation, it has been argued that contemporary social media operate as spaces “where young women can express both personal and political views and connect with diverse others” (Harris, 2010, p. 480).
Whilst not being explicit in this body of literature, it is probably not wrong to categorise the political engagement described in this literature, as well as our approach, as part of a post-parliamentary tradition. The sharp distinction between ‘politics’ (as in government politics) and civil society (like Putnam, 1995) is here replaced by a broader definition of political engagement as ‘the handling of differences, diversity and dispute’ reflecting every-day political engagement (Bang & Sørensen, 1999, p. 326).
Women politicians and social media
In research exploring the gendered differences in political communication among blogging politicians in Sweden, Åström and Karlsson (2016) found that there were substantial differences in how men and women politicians communicate in the blogosphere. Women politicians use blogging to foster a stronger connection with their readers, as well as receiving enquiries on policy and perspectives. Other research has noted the personalised nature of how women gubernatorial candidates in the US presented themselves on social media (McGregor et al., 2016) as opposed to men. Personalisation, or individualisation, in politics has been part of wider shift in society blurring the line between what is public and private (Enil & Skogerbo, 2013), and is characterised by an increased focus on personal traits, identity, and so on, rather than traditional political institutions (Holtz-Bacha et al., 2014; Van Aelst et al., 2013). While this shift in politics is far from new, the speed of change has accelerated, particularly since social media is specifically designed for the sharing of personal information (McGregor et al., 2016).
Other researchers have found that women political candidates on Twitter adopt a different type of communication during campaigns to ‘both combat stereotypes and to distinguish themselves from male candidates’ (Evans & Hayes Clark, 2016, p. 329), in part due to women’s out-group status. Women, in other words, need to identify avenues to both differentiate themselves as individuals from the group ‘women’, but also against the individualised norm of men as politicians (Evans & Hayes Clark, 2016, p. 327).
Women citizens and political social media use
In terms of bottom-up political engagement by citizens, observing how men and women politically engage in social media, some research suggests that there is not a large amount of difference (Bode, 2017). However, in terms of visible political behaviours, the same research shows that women seem to strategically engage in less visible political behaviours. For instance, Bode (2017) found that women are less likely to compose their own posts about politics, but are more likely to comment on, or ‘like’, the political postings of others. This has been attributed to the fact that women are more likely to use social media for relationship maintenance (Muscanell & Guadagno, 2012), and so must balance this with their traditional political engagement online (Bode, 2017).
Research from Brazil into the discursive strategies of men and women on Twitter has shown that men were more likely to use imperative verbal forms in hashtags, while women were more likely to use declarative forms, ‘Western men are expected to use linguistic forms that assert their power to a general audience and that Western women are expected to adopt more neutral forms’ (Cunha et al., 2014, p. 5). Again, both of these findings show that, while subtle, gendered approaches to politics online do exist. The online nature of politics does not preclude gendered experiences from occurring, and that this (like for politicians above) functions along traditional structural gender lines.
Negativity, misogyny and harassment
Given that the experiences of women doing politics online, whether they be politicians or individual political participants, are different from that of men, it is not surprising to find that women’s experiences of online negativity are also different from that of men. Popular news reporting on the experiences of New Zealand women political journalists (Graham, 2017) has revealed how prolific the harassment is (see also Martin, 2018). A study from the Czech Republic has been demonstrated not only difference in how women participate online, but that there was a gendered construction framed around ‘good girls do not comment on politics’ (Vochocoa et al., 2015, p. 1321), suggesting a moral aspect (i.e., being a ‘good’ girl).
Much of the extant research has been emphasising these problems as individual rather than systemic. The problem is framed as a matter of ‘resilience’ or ‘personal troubles’ despite actually being rooted in social structure akin to sexual harassment and sexual assault offline (Graham, 2017; Lumsden & Morgan, 2017; Sobieraj, 2020). Numerous researchers have also touched on this individuation of what really is a structural problem (Citron 2009; Jane, 2017; Vitak et al., 2017; Eckert, 2018; Barker & Jurasz; 2019). The structural nature is particularly evidenced in how its impact is more keenly experienced when women occupy multiple layers of marginalisation, such as racial, or sexuality, etc. and/or do not conform to traditional gender norms (Sobieraj, 2020).
What makes this individuation problematic is that research consistently finds that negativity in the online political sphere alienates women from active political participation (Citron, 2009; Eckert, 2018; Vitak et al., 2017) as women permanently leave these spaces (or reduce their participation). Participating in political spheres takes a considerable toll (Jane, 2017), requires modification of behaviour and freedom of expression (Megary, 2014), and undermines autonomy (Citron, 2009). And while there is some correlation between the effort expended to deal with negativity online and a reduction in women’s experiences of that negativity (Vitak et al., 2017) that is nonetheless an approach that deals with individual impact, not systemic.
One side of the problem is that online negativity is conceived as a form of abuse rather than as a form of communication (Lewis et al., 2017). It is conceived as a communication act, i.e., its effects are experienced as a part of how discussion (and hence participation) operates, not merely as abuse of, or towards, an individual, particularly so when the impact is minimised (Citron, 2009). The public sphere of political online space is then impacted, in a similar way to how hate speech operates; it delineates who is an appropriate member of the space, and who is not.
However, much research suggests systemic and multi-faceted solutions (Martin, 2018; Barker & Jurasz, 2019; Lewis et al., 2017; Jane, 2017; Megary, 2014; Eckert, 2018; Vitak et al., 2017), rather than framing this as an individual responsibility. Platform accountability has been suggested as a solution (Sobieraj, 2020; Barker & Jurasz, 2019; Eckert, 2018) as well as state-level policy interventions. The latter is often framed as including more effective legal ramifications for perpetrators (Eckert, 2018). Lipton (2011) provides one such legal framework, looking at different online acts, including cyber-bullying, cyber-harassment, and cyber-stalking, and the legal constructions of these. However, leaving the state responses to reactive ex-post ramifications (albeit certainly necessary) leaves a gap. This paper is thus located in this gap, in the lessons learned in systematically exploring the experiences of political engagement from women Twitter users in New Zealand on how to deflect structural intimidating and aggressive behaviour at four different strategic levels.
Methodology
Analytical approach
We have in this study adopted an interpretivist approach in which ‘stories’ and ‘narratives’ play a central role (Somers, 1992, 1994; Plummer, 1995). By posing open-ended questions about events, stories, changes, institutions, conventions etc., the researcher is provided with a ‘snapshot’ of how the actors, for example receive and interpret conventions and rules in practice, how actual power structures are understood, the constitution of boundaries between levels and organisational entities.
Participant codes and demographic data
Participant codes and demographic data
The participants, and target population, for this research had the following characteristics/sampling-constraints: they identified themselves as women, they permanently resided in New Zealand, and were active Twitter users (please note that it was not our intention to obtain a representative sample). This sample tried to fill the empirical gap in systemic research (as well as covering for a general lack of bottom-up studies of online politics in New Zealand).
‘Active’ was defined in terms of how recent participants’ last posts of a political nature were, how frequently they posted, and also if they were interactive in their approach (i.e., their posts included replies as well as one-off posts). While no quantifiable line was taken by the researchers on what was sufficient, a holistic appraisal of their engagement with Twitter was taken.
Twitter as an online platform was chosen as a case study for political participation and engagement construction primarily because it is one of the primary spaces of political discourse online in New Zealand. And given that context matters online (Papachrissi, 2014), removing platform particularity as a variable was important, and participants were restricted to this platform. It also afforded a useful and efficient way to garner participants, in that a selective snowball sample particularly lends itself to the networked nature of social media. Our study can also be conceived as an extreme case, i.e., a case having strategic importance to the general problem, and providing the opportunity to generalise: ‘If it is valid for this case, it is valid for all (or many) cases’ (Flyvbjerg, 2006). In other words, representativeness was not the focus for us in this research, rather instead it was to find a population where these kinds of strategic approaches to online political negativity are particularly salient, so that we can produce knowledge that gets to specifically this issue.
For the purpose of this article, 25 unstructured interviews were conducted by the researcher, using prompt-style questions, either in person or via video-call software (with the majority being conducted via video-call software, in allowing the research to capture a broader selection of participants). Selective snowball sampling was used by the researcher as a recruitment strategy, providing a range of participants from different ethnic backgrounds, locations around New Zealand, and levels of political involvement. This was achieved via the interviewer gathering a range of possible interviewee suggestions from the prior interviewees, and then looking to their Twitter profiles to ensure that diversity was maintained, and a high degree of similarity avoided. The sample size was chosen as sufficient when no substantively new responses had been received from several participants, and it was determined by the researchers that no new material would be gathered by interviewing more. Nonetheless, readers should bear in mind the small sample size in regard to representativeness and in light of the ‘extreme case’ approach above.
The participants were from a range of locations around New Zealand, though they did predominate in the main centres: 10 from the Wellington region, four from the Auckland region, four from Southland, four from Canterbury, and the rest were spread across more rural regional areas. Their ages ranged from their 20s to their 50s, and their socio-economic backgrounds varied from those who identified as working class and those on a welfare benefit, through to employed with much higher income bands. There were two who identified as Pasifika (descending from the Pacific islands), six who identified as Māori (New Zealand indigenous population), three who identified as Asians, and the remainder identified as Pākehā (European) or a variation thereon. Given that confidentiality was stipulated in their consent forms, a table aligning exact demographic details to each participant would allow for them to be too easily individually identifiable. However, a general indicator table of ethnicity, age range, employment status, and geographic region has been provided.
During the original research the unstructured interviews were analysed by the researchers using a narrative analysis theoretical framework (Plummer, 1995; Somers, 1992, 1994), looking for the underlying meanings that the participants were bringing to their explanations for what they were saying were their behaviours and reasonings. Themes were identified by the researchers in these narratives by qualitative reiterative coding in NVIVO software, and the common themes aggregated into the narrative arguments that the participants made. The themes were emergent through the commonalities between participants and informed by the literature.
They were themed by us via the varying levels of tactical or strategic approaches that the participants employed to manage them. Specifically, the ‘levels’ of the participants’ behaviours (Online Tools, Efforts, ‘Management’, and ‘Realisations’ – more below in Section 4) developed as the recoding progressed, and the common particularities were identified. These themes were then analysed in comparison to the literature above to provide the lessons in the discussion section.
The primary limitation, however, is sample size, and that this had a case study approach. A wider generalisable survey should be undertaken to determine population representativeness. Further, analysis should be done comparatively between the experiences of both men and women to determine if the recounting from these women as gendered is corroborated by comparative analysis in additional to constructive analysis.
Findings
The four themes we identified were 1) ‘Online Tools’ where the participants speak about the specific tools of the platforms they use to negotiate online negativity, both specifically what they do as well as what they think about them; 2) ‘Efforts’ where the participants speak about behaviours they took to deal with the negativity, and how it impacted them; 3) ‘Management’ where they talk about what they do proactively to manage safety, their energy levels, their emotionality, etc in the face of the downsides of being political online; and 4) at the most strategic level, ‘Realisations’ where the participants described how they developed their understandings of how online negativity worked, how they might contribute to such themselves, as well as how it fitted with being women. As is mentioned above, these themes are organised from the tactical to the strategic level, but it is particularly this last point that will be woven throughout these groupings: namely, how our participants saw this relating to gender, and/or how this fitted with other research on women’s experiences with politics.
Online functions
In this first theme, the participants talked about the platform tools which they employed to manage negativity. This does not just relate to the actual functionalities, and how they used them, but more importantly, how they related to these functionalities consciously as women. One of the most common responses was to ‘block’ the offending party, not only making the abusive post invisible, but also disabling the abusive user from seeing their own posts.
However, the participants expressed an ambiguous relationship to online blocking. Despite the tool being available, the mere provision of the tools was not sufficient.
I used to feel really bad about blocking people. I’ll feel bad about it now if they are someone that we have engaged on stuff and we might have followed each other for a while, but then they do something that I really disagree with. (Participant #11)
The respondents spoke about negotiating the use of that functionality with the norms of their gender and being socialised to interact with other people. They expressed that they felt reluctant to block early on in an interaction, despite knowing that they should. If they did decide to block another user, they felt guilty, or that they had to justify it somehow.
Yeah, so you know I have no problems in muting men in particular, blocking them, and telling them to just piss off. And it’s like … I’m not like that in real life. I think people think I must be like that, but I’m not. (Participant #10)
However, they often simply abstained from blocking. Participants attributed this reluctance to being socialised as women to put the feelings, wishes or perspectives of others ahead of their own, and hence merely stating that they find it easy to block transgressed gender norms for them. Or they would whisper this to the interviewer, as though they would be overheard admitting to breaking existing gender norms. The tension between this gendered socialisation, and the tools available to them, meant they had to expend much more energy than they would otherwise in order to fully participate.
I think it is a gendered thing and I think that men would call it being straight-forward and women would interpret that perhaps as being rude. There is lots of research to show that when women behave the same as men, they are perceived as being aggressive. (Participant #1)
However, others would do the reverse, in boasting about pre-emptive blocking (i.e., blocking someone before any interaction, based on recommendations from trusted others, or looking up the user’s profile). Another strategy was the ‘soft-block’. It is not an official functionality tool on Twitter, rather it is a side-effect of the blocking tool on the platform. If they identified unwanted ‘followers’, they would block them, but then immediately unblock. The result of this was that it removed the unwanted ‘follow’ with leaving that party unblocked. This released them from their conflicted feelings around blocking, while still removing themselves from the abusive user’s timeline.
Other possible strategies would be to lock down their own accounts to private, which would restrict who could access our participants’ tweets, and subsequent interaction. They would rather degrade their own interactional experience than open themselves up to abuse and have to use the block feature reactively.
So, I actually have replies set that I only see replies from people who either follow me or I follow them … I don’t get the drive-bys anymore, and that’s actually been really good. I mean, I’m sure I’m missing things, but you know I’m not missing all that much. (Participant #17)
Another available tool was muting, whereby they would simply not see the abusive person’s tweets, while that person could still see all their content without being aware that they had been muted (unlike a block). However, given that muting still allowed an abuser to observe their content, it was considered less than ideal.
Some participants took a different approach, namely creating a ‘Little Twitter’. Like the ‘soft-block’, this is not an official piece of platform functionality, but a user-generated innovation, or a side-effect of the official tools. By creating a network of locked accounts, accounts where only those that have received permission to follow can view tweets, as opposed to the default open nature of accounts (known as ‘Big Twitter’) a private network was created, where conversations of a personal nature could occur within the network. These small networks were considered to provide protective, and guarded, access to the discourse filtered through informal and trusted relationships.
The existence of Little Twitter means that you can quote tweet someone and say, ‘Oh this fucking dumb person’ and send that way and not have that direct confrontation with a person because you know them. (Participant #15)
While achieving this aim, they also disconnected conversations and participation from the intention of the platform. Some participants even deleted their ‘Big Twitter’ accounts and went entirely ‘Little Twitter’.
Efforts
This second theme focuses on the methods that our participants employed in a wider sense than merely platform tools against online negativity. The impact of the behaviours outlined here was largely associated with retracting themselves somehow from online engagement. This involved strategies like intentionally moderating the content of their own posts (or responses to other posts), based on assessing their risk appetite at the time. Further, they expressed how they assessed how much of their personal self was revealed in a post, or made vulnerable in a post, for similar reasons.
Maintaining the separation of different spheres of life as an approach to minimise impact from political involvement was something that several participants mentioned. This entailed not sharing their real names with other users, being careful sharing personal details, or in the case of users employed as civil servants, how much they could disclose and retain neutrality.
I think I’m very careful, you know? I live in a world where I am competing for jobs in a very sort of tight sector, and you know there’s politics within the industry that I’m involved in as well, but I think I don’t usually go all out unless it’s really important, because I just, I can’t afford to be seen as that political person anymore. (Participant #4)
In addition to the offline consequences of online negativity, some participants also noted that simply engaging politically in New Zealand was viewed with judgement. Given participants also spoke about how small the networks were in New Zealand, and how intertwined they were, they thought this had an inevitability about it, despite their efforts.
Here in New Zealand we don’t talk about politics every day. It’s not a, you know, it’s not a safe subject. […] if we start having those argument about it, we would very quickly butt up against very uncomfortable things that we’d have to talk about. (Participant #16) I think in a very small community, and definitely [NZ city], and just New Zealand more widely, you have multiple interconnections. You know? … And that’s why I just think you have to keep your tone because … yeah, you’re really got to [be] careful. (Participant #22)
Contrary to the social media behavioural ideal of open discussion and engagement, participants spoke of taking a more closed, cautious, and self-protective approach as a default. Seeking political common ground was not their immediate impulse, but rather safety. Expecting the worst from those that engaged with them was seen as the safest route. This was not to say that they would not try to assess whether interaction was safe, though. They would do so in the form of looking through a person’s profile to see the nature of their posts, as well as ascertain how they treated other users. However, giving people the benefit of the doubt was not a default position:
Unless I’m interacting with them on a personal level, I just don’t really interact with their politics, cause, I don’t know, it just opens you up to so much stuff, and because even, I don’t know, I just don’t think it’s worth it, because it’s often. (Participant #5)
Part of this assessment approach was deciding if it was worth getting involved in a political issue. This was driven by the emotional drain the users felt in being involved in an issue that was important to them, as well as what might be the potential for online abuse should they get involved. Declaring an opinion online, contributing to an online discussion, or even being involved offline, varied between different political topics.
Like why it is my responsibility to put in the emotional labour to educate other people on why what they’re saying is discriminatory, or is racist, or is prejudiced, or you know anyway, like it’s just, you know, kind of, yeah, I just don’t wanna feel that burden anymore. (Participant #5)
One way of mitigating such effects was to get involved collectively in an online action, allowing a mass of people to have an impact, and so contributing in a smaller way collectively meant they were at less risk. Speaking up on an issue individually could be thought of as too risky endeavour, compared to partaking in a chorus of voices.
Management
Moving up from specific actions within online networks to more strategic behaviours, this section refers to how the participants managed their online lives in respect to online negativity, or just the potential of online politics having an impact on their lives. This theme differs from the previous one, ‘Efforts’, via moving up strategically from reactive approaches (i.e., the specific actions), to more proactive approaches. These proactive approaches step up again in strategic orientation in the following theme (4.4) ‘Realisations’.
One type of behaviour that some of the participants particularly mentioned was the ability to leave a conversation and not reply. They expressed this through the lens of gendered expectations, where the breaking of a conversation was considered ‘impolite’; following the same kind of privileging the emotions of others, rather than their own, as was mentioned above. To act against this gendered socialisation was considered to be a radical act deviating from the norm.
I don’t need to be responsive to everyone, I don’t need to answer them and as a woman that is quite a challenging thing to do because we are socialised to be polite. (Participant #1)
Some took this further and expressed that they would leave the online networks entirely. They conveyed this as a last resort in order to take care of themselves in light of negativity online. Others spoke about becoming overwhelmed and taking temporary breaks from social media. This was not merely about becoming overwhelmed due to explicitly negative behaviour from others online but was also simply from the number of negative news posts being viewed about topics they cared about. This included living in fear of being ‘doxed’ (i.e., the online harmful practice of researching and publicly broadcasting private information about other users), which like the inevitable permeability of spheres of life concern mentioned above, was considered to be a risk.
The backlash, and the hatred, and the threats, and the … it’s just too much to take from those people. So, you know, you can have one aspect of these platforms almost immediately whipping around to undermine another potential aspect of it. (Participant #6)
To an extent this was something that they would consciously think about; on how much information they could share about themselves online, both in terms of direct personal contact details and other personal background information.
Like I remember when [political journalist] just got absolutely smashed and then she was; wasn’t she doxed as well, I don’t know? And you know, and feeling personally unsafe, I don’t have the spoons for that at all, I don’t. You know, I’ve gotta be really careful with my own mental health, and so yeah, I can’t put myself in those situations. (Participant #19)
As discussed in the literature above, women politicians feel pressure to locate themselves in more personal issues in order to negotiate gendered expectations, and it is this kind personal location within issues that the participants felt so impactful on their emotional energy by being politically involved and engaged:
I think that honestly, I think if you’re going to be in the activist space, particularly if you’re a woman, particularly if you are active in areas that have to do with your own lived experience, like some of the stuff around #MeToo stuff, is not just a theoretical kind of exercise for most women, it’s actually practical. Like an actual thing that has affected our lives. And we are, and so activism in that respect is incredibly tiring to do. It can be emotionally very draining. (Participant #16)
They stated that they needed to take temporary breaks from social media, or from issues, when they felt overwhelmed. This kind of managing energy levels, and how, and when, to expend in political engagement, was a consistent narrative in the interviews.
You have to kind of pick your battles, because if you don’t you will end up with this massive fatigue, where you’ll just kind of like ‘I genuinely, I can’t, I can’t do this, I can’t speak out every time’ you know? (Participant #16)
One related strategic approach was to first determine who was worth expending energy on, and who to ignore.
I don’t have to look for trolls; there they are. I don’t have to look for things to be offended by; they’re right bloody there. What I have to do is filter out all the stuff, all the trolls, all the big egregious bullshit that’s going on. (Participant #23)
The respondents also expressed that despite intentionally selecting issues that mattered to them personally, or they felt strongly about, outrage around those issues was itself fatigue-inducing, and how ‘soft-punches’ from confrontations, and news around those issues would add up for them.
I find myself just scrolling and reading stuff that is soft punches, you know what I mean? And they just add up over the course of the day and at the end of the day I am just like ‘ahhhh!’. And so, I probably have removed myself from more discourse as a result of trying to filter. (Participant #7)
One participant talked about a particularly gendered ‘soft punch’ she had regularly experienced and seen:
Very often this is gendered, a guy says something stupid, women say this is why it’s stupid, and the guys go no it’s fine, until a guy comes along and goes dude what you said was stupid, and then it’s like oh you’re right, how did I not think about it. (Participant #2)
They also compared this to the offline, where they felt they experienced similar emotional drains and fatigue from political involvement, but had less of an ability to manage this, or create safety for themselves, than they did offline.
But I get fatigued by the outrage and the alarm and the fear, and I get worn out by it, with all the tools that I have to process and filter it, with all of the power I have to feel like I can still act in the world, despite it. I think it’s probably having to a much more detrimental effect on disengagement by people who don’t have those tools or resources. (Participant #6)
One way they did this was through the creation of filter bubbles. Not for the echo-chamber effect, but rather strategically so that they could protect themselves. This differed from the ‘Little Twitter’ approach mentioned above, as it was more about curating a space that would allow the users to minimise the negativity.
There never really were bubbles, there were just little places you’d get away from the horrors occasionally. I mean, yeah, no one can really bubble away society. (Participant #17)
Rather than shutting out different perspectives, this space curation allowed them to include voices that many of our participants said were unheard in offline politics, particularly marginalised voices, which they characterised as being dominated by privileged voices. This was something they highly valued.
Realisations
The most strategic grouping here includes, as the title suggests, the realisations that the participants came to about how their own online behaviour, other users’ online behaviour, and how to manage negativity in online politics proactively rather than reactively. This involved things like exploring, or rather conceptualising, motivations, incentives and stimuli – both their own and others’. A statement that particularly characterised this theme, and which for several of the women almost had become a mantra, was ‘you don’t owe anyone attention’. This was something that our participants tended to connect to their gender, with the anticipation that women should always be polite, not interrupt, prioritise the feelings of others over their own in a traditional compassionate role. This, thus, making the above statement radical and disruptive for the participants:
We will say to people ‘you don’t owe anybody your attention, you don’t have to educate that person, you don’t have to …’ but if they’re coming up to you and they’re hostile, saying ‘well show me, prove it to me, tell me’ you know. You don’t own them attention, you don’t owe them space on your timeline. And that’s really powerful and it’s not something that we’re generally taught as women. (Participant #10)
This freedom to refrain from responding was something they felt strongly, and again some even whispered about it, as was mentioned above regarding leaving conversations. But realising that they could break existing gender norms and just exit a conversation, or choose not to respond to a post, was something they found transgressive and empowering. They also felt that this was something that they could less easily do offline, where a similar ‘in real life’ interaction would still have required politeness of them.
I am a nearly 40-year-old woman with two children of my own and yet my mother still tells me to say thank you, so that is very deeply embedded, that sense of being polite. So actually, I have had to transgress some of those social norms which were sort of very strongly instilled in me in order to stay safe essentially on Twitter. And I suspect that is true of quite a lot of women, we are socialised to be polite and men are not socialised to be polite and so, you get this kind of response I think. You will see a set up where men are rude on Twitter and women then respond politely. (Participant #1)
Understanding the behaviours of other users assisted them in distancing themselves from the emotional impact of the issues they were involved in, they claimed. By contextualising the motivation behind negative online interventions, understanding blunted their impact, the participants claimed.
I’m trying to imagine being a man on Twitter. What must that be like? ‘I have an opinion and I’m going to shout about it and if you don’t like it, I don’t care.’ Whereas I think women are much more measured and interested in keeping ourselves and other women safe and interested in social issues. (Participant #11)
The participants also asserted that those voices that would deny having social privilege (whether that be race, gender, sexuality, etc.), were individual users who felt that they were losing their privileges. This allowed them to see the patterns in the political interactions they participated in or witnessed.
At a different level, and according to our participants, these realisations would manifest themselves through seeing that science and evidence often do not suffice for convincing people in entrenched positions. Rather, appeals to personal experiences or emotional connections had a far better effect. However, this represented a more precarious approach, as it required a higher level of emotionality from our participants. Participants would sometimes ask themselves if their contribution or participation was even needed. Had someone else made the point already? What would reiterating the point contribute? Is this a space where their own social privilege could be used to speak up for others, or would it be better for them to be quiet?
I guess in the last few years I’ve stepped back from [always having an opinion] you know, and that’s become, you know, as I’ve developed my own knowledge and understanding of my place in the world and you know, and I choose to amplify other’s words now rather than … you know, unless something actually really directly affects me, I will, whereupon I’ll comment on it. (Participant #20)
Another pattern that the participants noted was something they felt was most likely to impact them emotionally, and that was not having lived experience validated. Rational and distanced claims to objectivity were an ideal of political engagement that they felt alienated them. This was because they felt most strongly about those issues that were embedded in everyday lives. They also similarly characterised thought experiments or playing the Devil’s Advocate.
A growing awareness of not doing something I consider quite gendered, which is treating important issues thought experiments, which I find is something that men tend to do much more […] that is not the conversation that we’re going to have, we can actually discuss important issues without belittling them or treating them in the abstract, and I think that’s a big change. (Participant #2)
Which is interesting given that traditionally these techniques of a more rational approach to political discourse are seen as acceptable.
for male activists there are not, the issues are different, a lot of the things that they are active about may be don’t touch them in that same fundamental way. (Participant #16)
They also expressed realisations about their own standing in the political discourse. Those with a socially privileged position (a white racial identity, heterosexuality, being cisgendered [non-transgender], fully-abled, etc.) realised that they could speak up under such a mantle, providing a degree of relative safety that those with less privilege could not afford. They also expressed that acting collectively with supporting teams around them, made their political space safer, rather than acting individually:
I think there is something intensely political about women supporting women actually in terms of their … just kind of … what’s the word I am looking for? Just validating their lived experience. There is something uniquely political in the way that women do. I kind of think that the everyday is political and I see a lot of validation of other women’s kinds of experiences of just being a woman frankly, which I think is quite political actually. (Participant #1)
Nonetheless, they realised the limitations with respect to expressing themselves, including expressing ‘half-formed’ thoughts. Online political spaces were not considered safe spaces to explore things ‘verbally’ in cases where the participants had not completely developed a reasoning. Voicing drafts of individual reasoning would inevitably make them more vulnerable to the online community.
now it seems everyone has a ‘Little Twitter’: so, you have a Big Twitter persona where you kind of can push those updates out and you have Little Twitter where you talk to a select group of people about your thoughts and feelings and gripes. (Participant #15)
Something which is quite contrary to the ideal of a political engagement space, online or offline.
So … you have this sort of incredible potential democratisation of platform, and then these platforms are not safe for many people of colour, people with a disability, women, gender-queer people, etc. (Participant #6)
Finally, while this research was performed solely with women participants, excluding a comparative analysis with men, the gendered nature of these experiences is revealed not just in how tightly it aligns with previous literature of women’s experiences of online politics and negativity, but also in the participants’ nascent structural understandings of gender. For them, it was not merely their experiences as women, but also about interpreting those experiences in systems of gender (and race, etc.).
We would argue this is a gendered phenomenon not in the sense that we are doing comparative analysis on a sample of men vis-à-vis a sample of women. It is a gendered phenomenon because our participants themselves articulate their tactics and strategies via a gendered lens. They spoke specifically about gender in relation to online negativity consistently and were constructing political spaces and realities with gender being fore fronted. Their experiences are gendered because they construct/perceive them as gendered, just as they did along axes of race, colonialism, sexuality, et cetera.
Final remarks
One observation that is immediately apparent from the statements from the participants is how individualised the ‘solutions’ (or at least the mitigations, for they are not addressing the causes) are, something that particularly aligns with the literature (Jane, 2017; Vitak et al., 2017; Citron 2009; Barker & Jurasz; 2019; Eckert, 2018). The contributive political milieu (for they did indeed see it as contributive and impactful) they are engaging in is one that is not structurally set up to mitigate the negativity, beyond the provision of tools that users have to manage themselves. Thus, the burden of managing the ‘space’ is predominantly placed on the users of that space, rather than with those that provide the space, with the resultant impacts of various levels of disengagement, and along perceived gendered lines (Vochocoa et al., 2015; Sobieraj, 2020). Or as with Jane (2017) even if these women remain, it takes a considerable toll.
Regarding the first theme, ‘Online Functions’ the lesson framing is that the mere provision of tools is not sufficient to ensure that a) they will be employed, and b) that their use will not result in demotivation. If more inclusive digital politics is to be realised, then reactively operating as facilitators and support in development and delivery is not sufficient. Using online platforms for digital politics means acknowledging the different social locations, and how those spaces get managed with the presumption that all groups will operate in tandem. Curation of an online political space so that a) the tool use becomes normalised, but b) that the primary responsibility, but with facilitators with a clear mandate.
The second level of analysis in the findings section is around monitoring. It is about addressing how leaving individual citizens to manage negativity on their own individualises the problem, resulting in them prioritising self-protection over engagement, something noted by Megary (2014). Why this is challenging for political engagement is that at best you receive limited contribution to your projects from the involved citizens (Citron, 2009; Eckert, 2018; Vitak et al., 2017), or even worse, more vulnerable social groups are excluded (as is evidenced in both the findings and the literature review section, or given that politicians often use social media platforms for outreach to their constituents, just holding those platforms to account (Eckert, 2018; Barker & Jurasz, 2019; Sobieraj, 2020). Addressing this involves the positive actions taken in the previous section, but also monitoring who is involved in online political activities, who is excluded. Merely focusing on the outcome of digital politics externalises and individualises potential impacts on the citizens involved.
The third level involves returning to the curation of online co-production spaces mentioned above. This involves, given the way in which the participants stated the impact being involved in political spaces took on them and resulting fatigue (Jane, 2017), doing the kind of work stated that the participants felt they had to themselves. This included managing what is required to respond, providing a degree of buffering between online political/policy contributions and offline life. Taking on these things as a default curation of a space, so that they are not individualised down to the politically engaged citizens, ensures that the contemporary mode of facilitator for policy makers in interactive digital governance realms.
Finally from the fourth grouping, it is apparent from the strategic approach taken by the participants, that the response from any organisers of digital political spaces also needs to be strategic. For the participants, online politics and policy are located in amongst intersectional social locations, where power, voice, and access were mediated through such. Understanding strategically how social locations like gender (Vochocoa et al., 2015; Sobieraj, 2020) effect the ability to contribute to a political discourse is part of ensuring that the shift in the role of citizens from passive spectators to a more active and producing role is fully realised. It also provides a more systemic structural approach than requiring platforms to deindividualize solutions, which is often the approach taken in the literature.
This is particularly the case due to the acknowledgment that experiences from the political sphere have expanded into domains once considered administrative and technocratic. This kind of strategic approach involves connecting together the different levels of response above together, with tools, space-curation, and monitoring (before, during, and after engagement), having these made coherent with each other, as well as with this strategic understanding of how structural social location plays into access and use of online political and policy spaces. In addition to the general lessons learnt from the study, which are relevant to a broad group of actors (including women active online), there is a particular message to any organiser of online politics and policy spaces to take into account the structural gendered aspects of the online during the design process, as is there for any agencies regulating speech in online contexts.