Violence against Women in Politics from an Intersectional Perspective

Violence against women in politics is a specific form of gender-based violence that is directed at women politicians. We rely on a broad understanding of violence, defined as intentional behavior that is unwanted, nonessential, and harmful (Hamby Reference Hamby2017). In the context of violence against women, this violence is gendered, meaning it targets women because they are women, and not because, for example, the offender opposes their political views. In other words: “Its central motivation is … not to gain the upper hand in a game of partisan competition, but rather to exclude women as a group from public life” (Krook Reference Krook2020, 4). Gender-based violence is not the only form of violence that politically active women face, but it is a form that is directed only at women (and at men who do not adhere to hegemonic norms of heteronormative masculinity). Women face both gender-based and issue-based political violence, explaining why surveys suggest that women politicians experience more violence than men, both in the UK (Collignon, Campbell and Rüdig Reference Collignon, Rosie and Wolfgang2022) and in other countries (Håkansson Reference Håkansson2021; Herrick et al. Reference Herrick, Thomas, Franklin, Godwin, Gnabasik and Schroedel2019).Footnote ³

Violence can take different forms; it can be physical, psychological, sexual, economic, and semiotic (Krook Reference Krook2020; Krook and Sanín Reference Krook and Juliana Restrepo2019). Physical violence produces bodily harm and injury and includes unwelcome physical contact or involuntary physical confinement. Psychological violence is the most common type in politics (IPU 2016, 2018). It inflicts trauma on a person’s mental state or emotional well-being (e.g., death and rape threats, or otherwise insulting, taunting, or scaring the target). A large part of this psychological violence is online abuse. In the UK, for instance, an overwhelming majority of MPs report that they have experienced online harassment and intimidation (Akhtar and Morrison Reference Akhtar and Morrison2019; Collignon, Campbell, and Rüdig Reference Collignon, Rosie and Wolfgang2022). Sexual violence refers to sexual acts and attempts at sexual acts by coercion, as well as unwelcome sexual comments and advances. Economic violence includes behaviors aimed at denying, restricting, or controlling women’s access to financial resources. Finally, semiotic violence mobilizes sexist and degrading words and images to injure, discipline, and subjugate women. This last type of violence is distinct from the other four types because it focuses on influencing how the public views politically active women as a group, rather than targeting women individually. Acts of semiotic violence “seek – through their circulation among citizens at large – to send a message that women as a group are unworthy” (Krook Reference Krook2020, 67) by rendering women invisible or incompetent (Krook Reference Krook2022). The opportunities to create and distribute negative and harmful portrayals of women have dramatically increased due to technological advances. Qualitative in-depth analyses of Twitter and YouTube data, for instance, have demonstrated how social media is used to discursively construct the political arena as a fundamentally male-oriented space, and to reinforce stereotypical and sexist representations of women in politics (Esposito and Zollo Reference Esposito and Sole2021; Harmer and Southern Reference Harmer and Rosalynd2021). These forms of violence are gendered if they are shaped by attitudes and beliefs about gender, including about roles and norms attached to notions of femininity and masculinity (Bardall, Bjarnegård, and Piscopo Reference Bardall, Bjarnegård and Piscopo2019). For instance, women are more likely to be attacked sexually, men physically. Online, women MPs do not necessarily receive more abusive messages than male MPs, but they are subject to significantly more sexist abuse (BBC 2022; Southern and Harmer Reference Southern and Emily2021; Gorrell et al. Reference Gorrell, Bakir, Roberts, Greenwood and Bontcheva2020; Ward and McLoughlin Reference Ward and Liam2020). In practice, these types of violence can overlap; however, we use them as analytically distinct categories to identify which types of violence the Muslim women in our case study were primarily confronted with.

The perception of political violence by the public can be gendered, too. Violence may, for example, deter more women than men from running for office (Bardall, Bjarnegård, and Piscopo Reference Bardall, Bjarnegård and Piscopo2019). However, we do not know whether that is true. In both the UK and the US, for instance, record numbers of women ran for office in the last elections, despite high levels of sexist harassment and abuse (Collignon and Rüdig Reference Collignon and Wolfgang2021, 429; Krook Reference Krook2020, 30). Therefore, violence against women in politics may have both a deterring and a mobilizing effect on women. Gorrell et al. (Reference Gorrell, Bakir, Roberts, Greenwood and Bontcheva2020), for example, found that politicians who had received more online abuse than their peers were more likely to decide against re-election. Surveys in the UK have shown that more women than men experience fear for their safety on the campaign trail (Collignon, Campbell, and Rüdig Reference Collignon, Rosie and Wolfgang2022; Akhtar and Morrison Reference Akhtar and Morrison2019), which makes them more likely to adapt their campaign strategies, such as avoiding canvassing, which in turn diminishes their chances to be elected (Collignon and Rüdig Reference Collignon and Wolfgang2021). Violence also negatively affects the political work of standing MPs. According to a Danish survey among municipal officeholders, the experience of psychological and sexual violence negatively affects the feeling of having freedom of speech (Kosiara-Pedersen Reference Kosiara-Pedersen2023). Online abuse, indeed, spikes after public contributions by MPs to contested debates (Ward and McLoughlin Reference Ward and Liam2020). Overall, the more visible politicians are, the more heavily they are targeted (Kosiara-Pedersen Reference Kosiara-Pedersen2023; Collignon and Rüdig Reference Collignon and Wolfgang2021; Håkansson Reference Håkansson2021; Rheault, Rayment, and Musulan Reference Rheault, Rayment and Musulan2019). Our research contributes another important element that, so far, has been largely neglected: We suggest that the appreciation and support politicians receive contributes to the decision of why they choose to go into or stay in politics, despite abuse. Nurturing this support and appreciation, we argue, represents an important counterweight to violence and should, therefore, be considered in the development of political strategies that seek to promote women’s political participation.

A further gap in the research on violence against women in politics is intersectionality (Kuperberg Reference Kuperberg2018). Intersectionality as a concept was coined by Kimberlé Crenshaw (Reference Crenshaw1991), who, with respect to violence, has challenged feminist and anti-racist discourses for marginalizing women of color, because feminists tend to focus on white women, and anti-racists on men of color. Ultimately, the concept of intersectionality can be used to critique the idea of taking “one group at a time” to solve discrimination (Collins and Bilge Reference Collins, Hill and Bilge2020, 7). Even before Crenshaw (1989), black scholars and activists argued “that multiple forms of discrimination are not ‘purely additive’ but instead function as simultaneous oppressions with a multiplicative effect” (King 1988, 46, cited in Kuperberg Reference Kuperberg2018, 686). In politics, as well, women are not a homogeneous group; their experiences are not limited to their gender alone. However, “the experiences of minority females is often lost in single-axis analyses that focus solely on gender, race, or religion” (Al-Rawi, Chun and Amer Reference Al-Rawi, Hui, Chun and Salma2022, 1919). Intersectionality has not yet been incorporated widely in theorizing violence against women in politics, even though the intersections between gender and other biases are present in emerging data. Research on the UK, for example, suggests that women of racial, ethnic, religious, and sexual minorities receive more abuse in politics than other women (Collignon, Campbell, and Rüdig Reference Collignon, Rosie and Wolfgang2022; Collignon and Rüdig Reference Collignon and Wolfgang2021; Amnesty International 2018).Footnote ⁴ Despite these observations, we do not yet know much about the experiences of violence among women of non-dominant groups in politics and about the underlying dynamics of their experiences. An intersectional perspective allows us to address questions of how sexism and Islamophobia intersect in the experiences of Muslim women in politics and what that tells us about the perpetrators of that abuse – a group largely neglected in research on violence against women in politics so far. To our knowledge, there have been only two studies that focus on these questions, revealing that, in the case of online violence in the UK, women of ethnic minorities are confronted with more racist than sexist or intersectional abuse that portrays them as incompetent and disloyal to the British nation-state and questions their experiences of discrimination. This implies that the abuse is informed primarily by far right and nationalist attitudes (Harmer and Southern Reference Harmer and Rosalynd2021; Kuperberg Reference Kuperberg2021). Our analysis confirms this finding for online abuse, but also goes beyond it by revealing that Muslim women, due to being members of multiple disadvantaged groups, are confronted with violence from different communities and in different arenas. Both types and origins of violence are shaped by the specific local context of the – online and primarily offline – constituency.

To operationalize local dynamics with respect to violence against Muslim women in politics, we draw on Krook’s (Reference Krook2020, 97–102) use of three additional, interconnected dynamics of violence: Structural violence (originally coined by Johan Galtung) is rooted in the social structure and enacts harm through the stratification of access to basic needs. In the context of violence against Muslim women in politics, these are the systemic barriers that limit the political participation of Muslim women, including unequal access to political and other resources based on their ascriptive group memberships. These structural disadvantages and discrimination are justified and normalized culturally, for example, through cultural norms, beliefs, and practices, such as ideas denigrating and disparaging women and Muslims who enter the public sphere. Such structural and cultural dynamics, finally, produce symbolic violence (coined by Pierre Bourdieu), an invisible mode of domination upheld by symbols, language, and discourse, as well as different forms of physical, psychological, sexual, economic, and semiotic violence that put marginalized groups who deviate from prescribed norms “in their place” (Krook Reference Krook2020, 101).

Methodology

This article on violence against Muslim women politicians grew out of a larger project on the political participation of Muslims in the city of Bradford. In the context of that project, participant observation and qualitative interviews were combined to delve into the local dynamics around three general elections (2015, 2017, and 2019) in one specific constituency in the UK. The theme of violence against Muslim women politicians emerged from these observations and interviews.

Participant observations were carried out at four pre-election campaigning events that included at least one of the four Muslim women candidates (one candidate event in 2015, three hustings in 2015 and 2017); during election day in 2015 at the Bradford University polling station; and at three post-election events with the new MP in 2015 and 2016. These events were observed by one of the authors to gather information about the local context, including key players in local politics and a sense of the general atmosphere during the elections. The observation notes were also used to inform and shape subsequent interviews.

We conducted thirty qualitative interviews with locals who lived, studied, or worked in the city of Bradford and who were interested and/or engaged in politics. They self-selected at our project stand outside the Bradford University polling station in 2015, responded to research flyers, or were recommended by other interviewees through subsequent snowballing (Bergin Reference Bergin2018). To ensure anonymity, we have changed all names. In terms of ethnic identity, 46.6% of interviewees identified as British Pakistani, 6.6% as British Asian, 43% as White British, and 3.3% as Pakistani. Over half identified as Muslim and one as a cultural Muslim. In terms of gender, we interviewed 46.7% women and 53.3% men. Interviews were conducted in the weeks running up to the first of the elections studied in this research (2015) and immediately after the election. They focused on the question of how interviewees perceived the election campaigns, the candidates, and local dynamics around voting and representation in the city. While political and cultural violence were not the only themes discussed in the interviews, they featured prominently. We also interviewed the MP of Bradford West several times between 2015 and 2023 about her experiences during the three elections; and we had informal conversations with Muslim women activists and journalists. The formal interviews were transcribed and then coded using thematic analysis (Clarke and Braun Reference Clarke and Braun2013) and cross-checked by two of the authors to make sure that themes were interpreted consistently.

We complement the participant observation and interview data with the analysis of 3,378 Tweets sent to Muslim women candidates, and compare how they were attacked in both the offline and online spheres. We included the three Muslim women candidates who competed in one or several of the three elections (Naz Shah, Amina Ali, and Salma YaqoobFootnote ⁵) and analyzed all Tweets sent to these three women around the time of their nominations, the hustings, and the elections (for each, one week before and one week after the specific event). We scraped the data from Twitter (now known as X) using the professional service of ScrapeHero, including only Tweets sent in English and directly to one of the three politicians using their Twitter handle (@…), but excluding those just mentioning their names in the text or in Hashtags. We included original Tweets as well as replies to capture reactions from users to relevant Tweets, including those issued by one of the three politicians. We also included Tweets with links, and we included the content behind these links in the analysis if they were still active. The coding of the Tweets, done by three different coders with considerable overlap to ensure intercoder reliability, sought to identify, on the one hand, whether the Tweets were “violent” (e.g., obvious threats to physical integrity), “derogative” (e.g., name-calling, stereotyping, swearing, silencing, questioning of intelligence, accusation of lying, questioning of position in politics), “critical” (criticizing the politician’s views or actions using appropriate language), or “supportive” (e.g., congratulation); on the other hand, whether they referred to the identity of the politician (e.g., gender, religion, age, race, national/cultural heritage), their political stance (e.g., policies, opinions), or personal characteristics (e.g., corrupt, hard-working, decent). The lack of context in many Tweets, however, made this differentiation often difficult. To ensure as much consistency as possible, we discussed each unclear Tweet among the three authors and, if we could not agree on the coding, excluded the Tweet from further analysis. Where possible, we also tried to factor into the analysis the origin of the Tweet, primarily if whether from the far right or an ethnic minority community (e.g., Pakistani, Muslim). Yet, the fact that users do not have to use their real name makes this information unreliable. In connection with the arguments they use, this nevertheless allows for some tentative assumptions about the provenance of some of the Tweets. We present exemplary Tweets in the form of “soundbites”, small-scale “forms of expression” presented in tables (Intezar and Sullivan Reference Intezar and Sullivan2022; Sullivan Reference Sullivan2012). Due to the lack of explicit consent from individual Twitter users, we paraphrase tweets which constitute harassment or abuse, keeping to the spirit of their meaning but changing the words so they cannot be attributed back to any specific account.

We combine the observation, interview, and Twitter data with information found in the news media to trace experiences, perceptions, and local dynamics of violence against Muslim women in (Bradford) politics in as much detail as possible, which we will present in chronological order of the three elections. First, though, we provide some background information on the constituency of Bradford West.

Case Study: Bradford West

The constituency of Bradford West is an instructive case study because it has had consecutive elections with multiple female Muslim candidates. This deep dive into a single constituency allows us to attend to multiple intersectional differences across time and in a single space in a way not previously explored.

Bradford is a city that has placed identity, in particular Muslim identity, at the heart of contentious politics for nearly thirty years (Husband et al. Reference Husband, Alam, Huettermann and Fomina2014; Samad, Reference Samad1992). According to the ONS data from the 2021 census, 30.5% of residents in Bradford were Muslim, and 32.1% of Bradford residents identified their ethnic group within the “Asian, Asian British, or Asian Welsh” category, up from 26.8% in 2011. While the majority of the city’s population (64%) identify as White British, the fact that the city has the largest proportion in England of people of Pakistani ethnic origin (20.3%) can be seen in the make-up of specific electoral constituencies. Ethnic concentrations in particular neighborhoods mean that over 60% of the constituents of Bradford West describe themselves as Asian, Asian British, or Asian Welsh compared to just 9.6% in the UK population as a whole. The ethnic and cultural make-up of the constituency is a significant factor in the experiences of political violence, set within the context of largely segregated communities.

The constituents of Bradford West have voted in a Labour MP in every election since 1974 except in 2012, when George Galloway, standing as a Respect Party candidate, won a spectacular majority in a by-election (Akhtar, Reference Akhar2015b). Three years later, in the 2015 general election, the constituency of Bradford West achieved notoriety for high levels of personalized slandering and abuse levelled at political candidates (Akhtar Reference Akhtar2015a; Guardian, May 2015). The constituency had two young Muslim females from different ethnic communities as prospective parliamentary candidates for the Labour Party. The initial candidate Amina Ali (a Somali Muslim female candidate who had been brought in from London) stepped down three days after being selected (Akhtar Reference Akhar2015b; BBC News 2015), succeeded by Naz Shah, a local Muslim woman of Pakistani decent, who finally secured a majority to win the election.

In 2017, as the country went to the polls for another general election, the constituency once again saw a particularly bitter election make the national headlines (Financial Times, 2017). This time, a female Muslim campaigner with Pakistani heritage from Birmingham, Salma Yaqoob, stood as an independent, left-leaning candidate against the once again successful Labour candidate Naz Shah. Few constituencies in British elections have had two female Muslim candidates standing against each other to achieve political office. The two candidates, however, differed in other respects: One wore a headscarf (Yaqoob) while the other (Shah) did not. This symbol of religiosity allowed us to assess how important public markers of faith impacted the type of political violence experienced. Even though they shared the same ethnic heritage and were of similar age and political experience (Shah had been an MP for two years by then and Yaqoob had been a local councillor in Birmingham previously), they came from very different class backgrounds. Prior to coming into politics, Yaqoob had trained and practiced as a psychotherapist. Shah, on the other hand, took on factory jobs after leaving school so that she could support her siblings. In 2019, Naz Shah stood once more for the Bradford West parliamentary seat to, again, win a resounding victory.

Due to her prominence, we primarily focus on Labour candidate and MP, Naz Shah, who was subject to most of the abuse. Shah is a controversial figure in British politics. Her Labour membership, for instance, was temporarily suspended for sharing antisemitic content on Facebook prior to her election. She also liked (accidently, according to her) and shared a Tweet that belittled the sexual abuse of girls in Rotherham.Footnote ⁶ She withdrew the post after just eight minutes and apologized, and the suspension was lifted (Guardian 2016). However, both incidents led to a barrage of abuse both online and offline. The abuse came from different sides. She was called a “Jew” for now supporting Israel’s right to exist whilst also being viewed as a “grooming apologist” (Yorkshire Post 2020; BBC 2017), despite her decades-long activism against abuse and discrimination.

Experiences of Violence against Women in Three UK Elections

To capture the nuanced dynamics of each election, we analyzed political violence across the three contests chronologically. This let us consider whether and how experiences of political violence change over time. Primarily, our Twitter analysis confirms the findings from previous research that political visibility usually increases the chance of being targeted (Kosiara-Pedersen Reference Kosiara-Pedersen2023; Collignon and Rüdig Reference Collignon and Wolfgang2021; Håkansson Reference Håkansson2021; Rheault, Rayment, and Musulan Reference Rheault, Rayment and Musulan2019), and that violence against women in politics has intensified over time (Collignon, Campbell, and Rüdig Reference Collignon, Rosie and Wolfgang2022; Gorrell et al. Reference Gorrell, Bakir, Roberts, Greenwood and Bontcheva2020). In 2015, for example, both Amina Ali and Naz Shah received relatively low levels of abuse on Twitter; we found that around their nominations, 2.9% (Ali) and 3.2% (Shah) of Tweets directly sent to the two candidates used derogative language. Around the following elections, online abuse for Shah increased to 20.9% (2017) and 28.4% (2019). Salma Yaqoob, by contrast, was faced with 1.7% of derogative Tweets in 2017. In the following, we look at the three elections separately to identify prevalent forms of violence and the arenas in which they appeared.

The 2015 General Election

In the run up to the 2015 election, the constituency of Bradford West made national headlines due to the highly personalized and abusive nature of the election campaign. For the first time in the history of the constituency, a female Muslim candidate stood against an experienced white male politician, George Galloway. Galloway was elected as a Labour Party MP in 1987, but was expelled in 2003 for bringing the party into disrepute over actions related to his stance against the War in Iraq 2003. Galloway has a track record of using global issues, such as the Israel-Palestine conflict, to galvanize the Muslim vote, and with some success.

In 2015, Galloway was seeking to defend the seat of Bradford West as a Respect party candidate. Amina Ali, a Somali Muslim candidate from London, was selected as the official Labour candidate. She faced hostility from within the local Labour Party because she was seen as an outsider imposed on the local party by the central Labour Party. Informal discussions with Labour Party members at hustings pointed to Ali’s position as an “outsider”, as someone parachuted into the constituency and, as such, unfamiliar with the needs of the constituents. This finding confirms that violence and intimidation can occur within parties and are not necessarily restricted to inter-partisan differences (Reeder and Seeberg Reference Reeder and Seeberg2020). Ali stepped down after three days, citing personal reasons, and Naz Shah took her place.

On the national stage, Shah had been a complete unknown and common criticisms levelled at her were that she was the “second choice candidate” (Tweet, 21 Feb 2015), inexperienced and therefore incompetent (“What do you know” [Tweet, 24 Feb 2015]), and chanceless against Galloway (“I don’t think you have any chance of victory against @georgegalloway” [Tweet, 3 March 2015]). The turning point for her campaign came when she wrote about her backstory in a local newspaper (Urban Echo, March 2015). Her background is extraordinary, characterized by the intersection of gender, religion, and class, and was picked up by national newspapers and personalities. The Guardian ran a piece with the headline “Naz Shah’s story is one of survival. Politics needs women like her” (Guardian, March 2015). Growing up, Naz Shah was poor and, at times, destitute after her father left her pregnant mother and two children for the neighbors’ 16-year-old daughter. Shah was then sent to Pakistan by her mother, who feared she would be abused by the man they were living with. In Pakistan, Shah was forced into an arranged marriage at the age of 15. Her mother, meanwhile, suffering long term abuse at the hands of the man she was living with, ended up poisoning him to death. Along with her brother and sister, Shah had spent much of her youth trying to free their mother from prison, arguing that she had acted in self-defense (The New Statesman, July 2015). Shah’s journey into politics, thus, could not be further away from the Oxford-educated template of the traditional upper-middle-class career politician in the UK (Durose et al., Reference Durose, Richardson, Combs, Eason and Gains2013).

Shah faced a barrage of mostly psychological and semiotic violence, ranging from a dead crow put on her doorstep, its beak stuffed with grass, to personal attacks on her integrity during hustings, including an online smear campaign alleging that she was only selected because she had sexual relations with those in power and fake pictures of her circulating the internet. The Conservative Party candidate, George Grant, captured the “wild west” nature of the 2015 campaign by likening it to a 19th-century rotten borough rather than a 21st-century parliamentary democracy (Akhtar, Reference Akhtar2021).

At a University of Bradford hustings, Shah faced personal attacks questioning her integrity. Galloway, for example, purported that Shah had lied about the age at which she had been forced into a marriage in Pakistan (Guardian, April 2015). The accusations created a narrative of distrust and deceitfulness which was amplified on Twitter (Table 1).

Table 1. Soundbites questioning Shah’s integrity

Numerous Tweets questioned Shah’s trustworthiness and called her a liar. Rather than focus on policies, they sought to weaken her position in the political race. This, however, backfired, as the publication of Shah’s story led to an outpouring of support from the general public and proved to be an obstacle in Galloway’s supposed clear run of the election. As noted by Surraya, one of our interviewees:

Online, fake pictures circulated the internet showing Shah in a gym wearing only a towel, and in another picture wearing a mini skirt. These attacks were levelled against her identity as a Muslim woman, characterizing her as having loose morals and questionable character. Although such pictures would have no significance in wider British society, they were meant to send specific cultural signals to a socially conservative Muslim community that Shah was not a “good Muslim woman” to weaken support for her in her predominantly Muslim constituency. However, neither of these tactics succeeded, as Shah achieved a spectacular win in 2015, becoming the first Muslim to represent the constituency.

The 2017 General Election

In 2017, Shah ran again and entered an election campaign which she says was the most difficult of her life. Once again, Shah’s Muslim identity became center stage; however, the abuse came from very different audiences and was expressed through varying channels.

Shah was met with a lot of resistance, primarily online, for her statements about being Muslim in the UK. After a piece in the tabloid The Sun by Trevor Kavanagh, in which he asked for a solution to “the Muslim Problem”, Shah sent an open letter to the newspaper editors, backed by more than 100 other MPs, in which she criticized the Nazi-like language and Islamophobic tropes used in the comment piece (Independent 2017). For this, and other, interventions into the public discourse, she received much Islamophobic opposition online (Table 2). In 2015, doctored images attempted to paint Shah as a “bad Muslim woman”; ironically, in 2017, she became a visible symbol of “Muslimness” (Shah, Interview 2023) (Table 2).

Table 2. Soundbites of attacks against Shah’s Muslim identity

Most of the Tweets did not target Shah directly, but denigrated Muslims per se and denied the existence of Islamophobia. Nonetheless, they indirectly targeted Shah’s identity as a Muslim and, thus, suggested she was unqualified to represent British society in parliament because of her identity. Some members of Shah’s constituency, accordingly, felt they needed to remind her of the importance of remaining unbiased: “just make sure you do good for all communities and not just the Pakistani’s. Bfd has gone downhill over the yrs” (Tweet, June 9 2017). They also questioned whether Shah was attempting to play at being the underdog: “As one of your potential constituents i think you are playing the devalued victim card. You will not be getting my vote’ (Tweet, June 2 2017). In Shah’s own words:

Our analysis confirms Kuperberg’s (Reference Kuperberg2021) finding that, online, Muslim women politicians receive more Islamophobic than sexist or intersectional abuse, suggesting that it is primarily informed by far right and nationalist attitudes. For instance, most Tweets sent to Naz Shah around the elections in 2017 that we classified as derogative were of an explicitly Islamophobic nature (>70%). Interestingly, in 2017, Shah’s main competitor, Salma Yaqoob, also wrote an article concerning the rise of anti-Muslim sentiments (Guardian 2017). Although Yaqoob was questioned about broader cultural practices, such as: “are men and women sitting separately a part of diverse culture” (Tweet, May 18 2017), the majority of the 2017 Tweets Yaqoob received were not derogatory. Instead, she was celebrated as a “voice” for the “oppressed” and “helpless” at a global level (Tweet, May 14 2017). This confirms that more prominent politicians are more likely to receive more online abuse.

Our intersectional approach interrogated a unique facet of the 2017 campaign: it was essentially a contest between two Muslim women (Salma Yaqoob and Naz Shah) of similar ages and the same ethnic background (British Pakistani). Yet, they differed crucially in two respects: class and religiosity. Yaqoob was university educated and had had a career as a psychotherapist, while Shah did not have further education and had worked her way up in factory work. The second core difference between the two Muslim female candidates in Bradford West was the hijab, the Muslim head covering, and its representation of religiosity. Yaqoob covered her head with a hijab; Shah did not. This led to centralizing religion in the campaign. Within parts of the Muslim community, Shah’s religious authenticity was questioned. Discussions in campaign meetings questioned whether Shah was Muslim enough, and she was called a “dog” with the wrong “pedigree”.Footnote ⁷ Shah was identified as a “bad Muslim” for not wearing a headscarf while Salma Yaqoob was branded a “good Muslim” for wearing one. This “hijab-shaming” was specific to the ward, and again an appeal to the mainly Muslim electorate. For Shah, this tactic was psychologically damaging:

Leaflets were posted to private homes to prevent constituents from voting for Shah, for example, claiming her affiliation to Shia clerics (BuzzFeed.News 2017; Interview, Ifat) within the context of a majority local population of Sunni Muslims. Shah occupied a unique space in 2017; she became a symbol of ‘Muslimness’ for the wider public while simultaneously not being Muslim enough for some in her constituency. The atmosphere during a 2017 husting at the University of Bradford was particularly fraught and at times very uncomfortable, as this Tweet reporting from the events illustrates:

However, these tactics, again, had the opposite effect to the one intended. Many of the Muslim constituents at whom this kind of religiously framed political campaigning was targeted were critical of how the head covering was being manipulated to win votes:

This quote sums up the general feeling amongst many of our Muslim interviewees that the focus on religious identity and the subsequent shaming of Naz Shah for the way she dressed constituted “moral harassment”, and that the attacks on her personal life were significantly misjudged. From the interviews and observations, it was clear that the majority of constituents, Muslim and non-Muslim alike, wanted an MP who would represent their interests, and not necessarily someone who would mirror their religious-moral code.

Naz Shah herself is convinced that Yaqoob’s decision to stand and her campaign tactics were orchestrated by men:

This was about attempting to maintain hegemonic control of women in politics. Here, gender and religion intersected to shape the abuse levelled toward Shah with a local contest where local politics had been dominated by patronage politics as we will explain further below.

The 2019 General Election

The 2019 snap election was atypical because it took place in winter and because, in Bradford West, there was no serious competition from another candidate. They were largely “paper candidates” and there were no hustings, either. Naz Shah described this campaign as a much more positive experience:

Her assessment confirms what Collignon and Rüdig (Reference Collignon and Wolfgang2021) have found, namely that violence deters women from canvassing, likely reducing their chances to be elected, because electioneering, the campaign trail, and hustings are part and parcel of the road to become elected to political office. In this case, this worked to Shah’s advantage, despite the increase in abuse on Twitter from 2017 to 2019. It also suggests that, at least for Shah, the offline violence experienced on the campaign trail, and being visible, affected her more than the violence experienced online.

In 2019, online abuse against Shah had spiked to more than a quarter of all Tweets sent to her being derogative. However, the topics and characteristics picked up on by the abusers were far more mixed than in our sample for 2017. Most Tweets interpreted as derogative by us targeted her policies and public statements on racism, the situation in Bradford, and the above-mentioned child sexual abuse scandal in Rotherham, or denigrated her personality, calling her, for example, a liar or hypocrite. Only 13.5% of the derogative Tweets (also) addressed her identity, primarily her gender, religion and/or race, or Pakistani heritage (in equal numbers), mostly separately, but sometimes also in combination. For example, whereas Tweets from 2017 treated Shah as a general “symbolic form” of Islam, the Tweets from 2019 are more personalized in the sense that the Tweeter is directly addressing Shah with the use of “you”, “you’re”, and “your” (italicized in Table 3).

Table 3. Soundbites of identity-based Tweets sent to Shah in 2019

Similarly, the 2019 Tweets illustrate an additional layer of intersectionality in the sense that they combine gender, religion, and race. While the Tweets from Table 2, for instance, display direct derogatory language towards Muslims, thus only indirectly attacking here, the 2019 Tweets condemn Shah for being a Muslim women who is against “White people” or “young White girls.” Yet, we found only a few allusions to class, none to age, and hardly any sexual violence. We also found only two overt threats in our Twitter sample, one of them adding, after expressing “disgust” about some of her statements, that they know where she lives with a link to a picture of a house (Tweet, December 6 2019), and another one “daring” her to come to their door while canvassing (Tweet, December 10 2019). However, we only included Tweets directly sent to the candidates; therefore, we may have missed violent threats that circulated Twitter but were not sent directly, such as the one circulated via Twitter in 2017 from a now closed account that incited physical violence against her by accompanying the exact rally start time and route with a picture of Naz Shah’s face behind a stylized hairline cross. Other threats may have been sent through other social media platforms, such as Facebook. Twitter may also have removed some of the most explicit and particularly hateful posts before we were able to retrieve them, based on their policies of deleting tweets considered hate speech and suspending or deleting offending accounts. Threats are also often sent by email rather than through Twitter, exemplified by the two cases in which perpetrators were jailed after emailing offensive content and death threats to Naz Shah (Guardian 2022a). Shah summarises the situation using the following words:

Our case study of Bradford West suggests that politicians from intersectional groups, such as Muslim women like Naz Shah, due to their membership in different groups, receive different types of abuse coming from different directions. Different groups of abusers, in turn, tend to use different channels. While our results confirm Kuperberg’s (Reference Kuperberg2021) finding that, in the UK, online violence against women of Muslim minorities is primarily informed by far right and nationalist attitudes, we also find that in the offline world of electioneering, women of Muslim minorities are confronted with the local context of their constituencies. In Bradford West, as we will explain in the next section, this was characterised by a politics of patronage in a predominantly Muslim community.

Local Context Matters

In the following section, we explain how structural, cultural, and symbolic violence interact in the production of abuse against Muslim women politicians in Bradford West. While previous research has primarily focused on abuse from mainstream society levelled at Muslim women due to their “Muslim identity”, the data from our research also shows the types of violence Muslim women politicians face from within the Muslim community. This is, in large part, due to the demographics of the constituency of Bradford West. Like in many former industrial inner-city areas across the UK, the structural and cultural context of the constituency of Bradford West is shaped by concentrated ethno-religious communities which, in turn, shape the political experiences of women. All South Asian women we interviewed believed misogyny to be a factor in local communities. In interviews, women sometimes pointed to the “people of Bradford” who “don’t think women can be politicians” (Interview, Robina). Other times, it was even more specific to people in certain city postcodes, e.g., people in BD5 or BD7, who did not think women could or indeed should be involved in politics. Both postcodes have majority minority populations.

In Pakistani Muslim communities of Bradford West, the political sphere is governed through kinship ties which structure South Asian Muslim communities (Akhtar Reference Akhtar2013). These community-specific structures actively exclude women from both voting (for who they want) and running for office (Akhtar and Peace Reference Akhtar and Peace2018; Peace and Akhtar, Reference Peace and Akhtar2015). In February and March 2016, the BBC’s flagship news programme Newsnight broadcast a series of features on Muslim women in local politics and national politics who alleged that they had been deterred, in some cases through intimidation and the threat of physical violence, into abstaining from politics, including standing for office, by Muslim men (BBC Newsnight, 2016a and b). Within this structural context, images of the political sphere have been normalized that construct politics as a man’s world, as one of our interviewees explained:

Both modes of semiotic violence identified by Krook (Reference Krook2022) – rendering women invisible and incompetent – are visible here, and indeed both modes appear to work together in an iterative manner to normalize structural barriers to women’s political engagement.

Repeatedly, our research showed the gendered nature of politics as perceived by our Muslim women interviewees: how deeply rooted the idea of politics as a man’s domain was; every Muslim woman interviewed mentioned it, as did some of our White British female respondents. One of our interviewees told us: “I talked myself out of [politics] because I said, ‘They only vote for men anyway so no one’s going to vote for a little Asian girl’” (Interview, Surraya), stressing powerful intersections of exclusion based on gender, ethnicity, and age. She recounted her “surprise” at Naz Shah’s success in the 2015 general election, “because I thought they will vote for the man…because from the families I’ve met the men decide who votes and they all vote for the men, and women are seen as up themselves if they’re in a higher up position” (Interview, Surraya).

Another female councillor in the city told us that the primary explanatory factor in understanding politics in the city is that “a lot of people in Bradford don’t think women can be politicians, women can be local councillors, women can be MPs” (Interview, Robina). Going one step further and perhaps the logical conclusion of the idea of politics being a man’s world was that “women shouldn’t be involved really. Like, there’s no point educating them in terms of politics, because politics is a man’s domain” (Interview, Misba).

How socially accepted this view is within the Muslim community in Bradford West is confirmed by the fact that the majority of our male interviewees did not see any barriers for women to get into politics, facilitated by the view that one or two Muslim women had “made it” in politics by gaining office as local councillor or MPs. Only one male interviewee mentioned that some Muslim women may not want to go into politics because of their home situation. Asked to elaborate he explained:

There may not be visible or formal structural barriers preventing Muslim women from taking part in politics, but cultural factors can be equally constraining. The consequence of overriding these cultural barriers would be to put oneself firmly outside the bounds of what is culturally acceptable and so become “’fair game’ for sexual and other forms of gender-based harassment” (Krook Reference Krook2020, 99, drawing on Mantilla 2015).

Constituents Responses: Support amongst the Violence

Our case study revealed another aspect that is not often discussed in the emerging research on violence against women in politics: support and appreciation. Most research on the topic warns that violence might deter women from running for office. Without denying this threat to gender equality and democracy, we found that this sort of violence mobilizes many women (and some men) to engage in politics and to vote for women who receive abuse. Some of our interviewees stressed that they voted for Naz Shah precisely because of the abuse and the personal attacks she received on the campaign trial, impressed by her perseverance despite the abuse:

Interestingly, this was also the reaction of some men:

This support can be based on the perception of the individual person and their circumstances, as the examples above suggest, or focus on their identity, as the experiences of one of our interviewees illustrate:

Our Twitter analysis reflects the support for the three Muslim women politicians, as well. The proportion of supportive Tweets almost always exceeded that of derogative Tweets, ranging from 33.1% (Shah, 2017) to 35.2% (Yaqoob, 2017), to 65.7% (Ali, 2015) and 75.8% (Shah, 2015). The only exception constituted 2019, when only 2.8% of Tweets sent to Naz Shah were of a supportive character, mirroring the fact that the online environment she operated in had become increasingly toxic. Only a minority of these supportive Tweets referred to the politicians’ identity, primarily gender, sometimes in combination with ethnic identity. We interpret this as positive, as it shows that their supporters perceived them primarily as politically competent individuals rather than just representatives of a demographic group.

We do not claim that support outweighs the abuse that politicians encounter. But we argue that it is important to consider to, first, understand why women enter and stay in politics, despite abuse, and second, identify ways to further strengthen women and minorities in politics.