Abstract
This study examines the intersection of political social media use, online harassment resulting from political expression, and the perceived physical, social, and emotional costs of political engagement and how these dynamics differ across religion and gender. Using survey data in India, we find that frequent political use of social media is associated with online harassment, especially among religious minorities and most intensely among minority women. Although political social media use alone does not heighten perceived political costs, these costs increase indirectly through experiences of harassment. Further probing reveals that this indirect effect is more substantial for religious minorities, with no significant gender differences. These results highlight how online political engagement is not experienced equally, as identity-based vulnerabilities amplify the emotional and psychological burdens of political engagement. Ultimately, rather than serving as inclusive public spheres, social media platforms reproduce existing offline hierarchies, placing the heaviest burdens of visibility on marginalized groups.
The rapid increase in social media users worldwide has transformed traditional political engagement, providing a new venue for political discourse, mobilization, and deliberation (Boulianne, 2015). At the same time, this digital expansion has raised significant concerns about online political incivility and harassment (Gorwa et al., 2020; Ordoñez and Nekmat, 2019; Theocharis et al., 2020). In democratic contexts, social media platforms are often heralded as equalizing forces, granting marginalized voices access to political participation without the traditional gatekeeping mechanisms (Tufekci, 2017). However, digital spaces are increasingly marked by aggression, polarization, and targeted harassment, particularly for those at the intersection of multiple marginalized identities (Im et al., 2022; Sobieraj, 2020).
Engaging in political discourse on social media frequently exposes individuals to hostility and abuse, particularly when they express dissenting views or represent marginalized viewpoints (Vogels, 2021). This is particularly acute in polarized environments, where minority groups who participate in online political discourse face heightened levels of harassment (Banaji et al., 2019; Saha et al., 2021). Gender further compounds these dynamics as research shows that women in politics and public life are routinely subjected to character defamation, sexism, and physical threats (Amnesty International, 2020; Sobieraj, 2020). This pattern extends into digital arenas, where female political figures and social activists face orchestrated trolling aimed at delegitimization and silencing their voices (Ahmed et al., 2025; Gurumurthy and Dasarathy, 2022; Im et al., 2022; Sobieraj, 2020). Consequently, individuals already marginalized in society, like minorities and women, are more likely to experience online harassment.
Furthermore, while we see that political engagement on social media often increases users’ exposure to online hostility, the latter can, in turn, elevate the perceived cost of participation, defined as perceived personal burdens associated with engaging in politics for ordinary citizens (Smith et al., 2019). Instead of the narrow economic costs of engaging in politics for ordinary citizens that political scientists have long assumed, we examine broader costs from social, emotional, and psychological perspectives, such as relationship strain, mental health issues, and emotional disruption (Smith et al., 2019). Digital platforms are frequently celebrated for lowering structural barriers to political expression, their affordances such as amplified visibility and algorithmic virality, thus reducing costs of political participation from traditionally narrow perspective; however, they also make users more vulnerable to harassment, trolling, abuse, and inequity in participation (Pan et al., 2024; Rheault et al., 2019), escalating broad social, emotional, and psychological costs on active citizens. For politically active users, especially those expressing dissent or unpopular views, this can result in reputational harm, emotional distress, and safety concerns (Gardiner, 2018). These risks contribute to what has been described as the “chilling effect,” the suppression or withdrawal from political discourse in anticipation of hostility (Kim and Shin, 2022). Online harassment thus functions as a central mechanism linking political social media use to rising political costs. Regular political expression online increases visibility but also increases the likelihood of hostile encounters, which may erode individuals’ willingness to engage (Rheault et al., 2019). This dynamic reveals a paradox: while platforms promise inclusion and democratized participation, their failure to effectively moderate public activities may ultimately deter the very civic engagement they advocate for (e.g. Gorwa et al., 2020).
This “chilling effect” may particularly suppress political expression among marginalized groups who face heightened risks for simply being visible online (Gurumurthy and Dasarathy, 2022). For example, for women and religious minorities, political participation in digital spaces often entails high personal costs, undermining the democratic promise of inclusion. In other words, these perceived political costs are not evenly distributed; they are patterned by identity markers such as religion and gender. Those positioned at the intersection of these identities, such as religious minority women, face compounded marginalization when politically active online. The concept of intersectionality offers a critical lens for understanding how overlapping identities contribute to cumulative disadvantage (Cho et al., 2013). In digitally mediated political environments, gender and religion operate as structuring forces that shape users’ visibility, vulnerability, and capacity for participation (Barthwal and Shukla, 2022). Therefore, women from religious minority communities often face dual layers of discrimination and harassment, which amplify their perceived cost of politics and curtail their political agency over time (Amnesty International, 2020; Gurumurthy and Dasarathy, 2022).
Thus, this study empirically examines the relationship between political social media use, online harassment, and the perceived costs of politics, with particular attention to how these relationships are moderated by religious status and gender. Using large-scale survey data collected from respondents in India, the analysis proceeds in several stages. First, we focus on online harassment by investigating (a) whether political use of social media is associated with experiences of online harassment (H1), and (b) whether this relationship is moderated by religious status (H2) and gender (H3). Next, we turn to the personal consequences of political engagement by examining (a) whether political social media use is related to the perceived cost of politics (H4), and (b) whether this relationship is mediated by online harassment (H5). Finally, we examine the conditional indirect effects of this pathway, focusing on the intersection of religious and gender identities (RQ1).
Study background
India provides a compelling context for examining these theoretical propositions. The country’s complex sociopolitical fabric is shaped by intensifying religious polarization, entrenched patriarchy, and a rising digital connectivity. As one of the world’s largest and most active digital publics, India has recently experienced exponential growth in online political engagement, accompanied by new forms of gendered and religiously motivated harassment (Ahmed et al., 2025; Elliott, 2024; Gurumurthy and Dasarathy, 2022).
The contemporary Indian politics is characterized by the ascendance of Hindu nationalism and the reconfiguration of communal politics. Since 2014, the Bharatiya Janata Party (BJP) has consolidated its power through a political ideology that centers on Hindu identity as synonymous with the Indian national identity (Jaffrelot, 2021; Subramanian, 2019). This ideology represents not merely religious conservatism but a majoritarian political project that reimagines India’s secular, pluralistic constitutional framework through an exclusionary lens (Hansen, 1999; Jaffrelot, 2021; Subramanian, 2019). For example, Muslims, constituting approximately 14% of India’s population, have increasingly been positioned as the internal “other” in this political discourse, facing systematic marginalization and discrimination (Dhattiwala and Biggs, 2012; The Associated Press, 2024). This political climate has created what scholars describe as a “politics of fear and humiliation” directed at religious minorities, where questions of loyalty, belonging, and national identity are continuously contested (Dhattiwala and Biggs, 2012).
Moreover, gender norms in India are inextricably linked to religious identity, creating distinct forms of vulnerability for women across religious communities. Within Hindu nationalist frameworks, women’s roles are constructed through references to ancient texts and mythological figures, positioning them as guardians of cultural tradition and family honor while simultaneously mobilizing them as political actors in defense of the Hindu nation (Banerjee, 2006; Basu and Jeffery, 2011). However, this political mobilization coexists with deeply patriarchal structures that constrain women’s autonomy and public participation (Prillaman, 2023). For women from minority communities in India, the intersection of religious identity and gender creates compounded marginalization. They navigate not only the patriarchal norms within their own communities but also the external stigmatization and targeting rooted in gendered nationalist discourse (Parveen and Sanyal, 2024).
Digital platforms in India have increasingly become spaces for coordinated attacks on religious minorities, particularly Muslim women, who are targeted with hate speech, cyberbullying, and fabricated accusations (Barthwal and Shukla, 2022; BBC, 2022). These incidents are not isolated as they reflect a broader trend of misogynistic nationalism, where women from minority communities are rendered symbols of communal tension. For instance, a report by Amnesty International (2020) found that Muslim women politicians in India face disproportionately higher volumes of online abuse. Research confirms that online political expression in India is deeply entwined with gender and religious hierarchies, often reinforcing offline exclusions. As such, India presents a critical case for understanding how digitally mediated political engagement interacts with intersectional vulnerabilities, offering valuable insights into the structural dimensions of online harassment and democratic participation.
Political use of social media and online harassment
“Online harassment” refers to hostile actions committed against someone online, typically intended to cause harm (Sengupta and Chaudhuri, 2011). It encompasses a wide range of abusive behaviors, including threats of violence, intimidation, and public humiliation (Sengupta and Chaudhuri, 2011). In today’s politically charged digital environment, harassment directed at individuals for their beliefs has become increasingly prevalent (Walther, 2022). This form of belief-based harassment targets people’s opinions and attitudes (Nadim and Fladmoe, 2021), often singling out those who express controversial or dissenting viewpoints against a powerful majority. Such attacks are typically intended to silence, discredit, or discourage political expression (Walther, 2022).
As social media has become a significant space for political discourse, online harassment has been used to silence opposing views and reinforce ideological divides (Sharma et al., 2023). However, research on how political use of social media contributes to online harassment experiences remains limited, with existing studies primarily documenting the prevalence of harassment and the motivations behind harassment behaviors (e.g. Chen et al., 2024; Lu and Myrick, 2016; Sharma et al., 2023). Examining the experience of harassment from the perspectives of both observers and experiencers is crucial to participatory democracy in the digital age.
Social media has the potential to expose users to a broad and diverse range of political views (Chen et al., 2024; Christner et al., 2025; Lu and Myrick, 2016; Masood et al., 2024), even when they are not seeking them (Zhu et al., 2024). The way social media works, bringing together people with different political beliefs, can sometimes turn conversations into conflicts (Barnidge, 2017; Lu and Myrick, 2016; Masood et al., 2024). Thus, social media is an arena where political disagreements escalate more than in other political settings (Barnidge, 2017). What starts as a disagreement can quickly escalate into personal attacks, name-calling, and even organized harassment (Chen et al., 2024; Goyanes et al., 2021). Therefore, one’s use of political social media opens the floor to political disagreements and experiencing attacks targeting their beliefs (Barnidge, 2017; Goyanes et al., 2021; Sharma et al., 2023).
Moreover, the existence of partisan enclaves and echo chambers on social media facilitates collective harassment by reinforcing political identity and intensifying polarization (Walther, 2022). The extreme political identity motivates out-group hatred and in-group enthusiasm (Zhu et al., 2024), promoting identity-based attacks (Walther, 2022). Individuals are also motivated to harass ideological out-group others to seek approval and acceptance from their like-minded peers (Walther, 2022; Waruwu et al., 2021). Thus, the use of social media for political purposes may contribute to online harassment due to its role in political polarization and extreme identity formation. In addition, it is worth noting that interventions designed to encourage stepping outside social media echo chambers can paradoxically increase rather than reduce polarization (Bail, 2022). In Bail’s (2022) Social Media Prism, the author reveals that political extremists are drawn to social media because it provides them with status they cannot obtain elsewhere, and their hyperactivity dominates public discourse, making moderates less visible due to the anticipated fear of attacks on social media. Online hostility does not emerge simply from ideological isolation, but from the interaction between platform affordances, heightened visibility, and perceived identity threats in politically charged exchanges. Therefore, actively seeking political disagreements on social media, produced by extremist opposition actors, further fosters the polarized attitudes and attacks (Bail, 2022), which have been established in the social media environment (Chen et al., 2024; Goyanes et al., 2021).
Moreover, the anonymity of social media and decreased social accountability of digital networks further facilitate online harassment (Tandoc et al., 2023). Compared with offline interactions, there are fewer real-world repercussions for online harassment precursors. Many of the offenders may feel empowered to behave more aggressively than they would in real-world social situations, where they face direct consequences. The hostile political environment further amplifies this disinhibition effect (Rossini, 2021). The divisive political atmosphere online encourages individuals to engage in hostile behaviors without considering the immediate consequences.
Studies have found that the time spent on social media is not a direct driver of online harassment; rather, activities that disclose one’s demographics, beliefs, and identities make them more likely to be targeted by online harassment (Sengupta and Chaudhuri, 2011). Online attacks often target a specific group of people based on certain characteristics, such as political ideologies (Castaño-Pulgarín et al., 2021). Individuals who engage more politically on social media expose their political views and stance to the public, including strangers holding opposing opinions. This publicly salient political identity increases the chance of being targeted by others online. Studies demonstrate that active engagement in political activities on social media, such as expressing one’s political opinions, can lead to negative experiences due to representing differing political opinions (Chen et al., 2024), particularly for individuals with more central political identities (Masood et al., 2024). In sum, when individuals’ political identity is apparent on social media, they are more likely to experience online harassment from opposing groups, as many malicious online behaviors are driven by ideological hatred (Chen et al., 2024; Lu and Myrick, 2016; Sharma et al., 2023). Therefore, we propose as follows:
H1: Political social media use is positively associated with online harassment.
Religious minorities and online harassment
Online harassment also involves identity-based attacks, targeting individuals with specific group characteristics (Wintterlin et al., 2020). Social hierarchy and deep-rooted biases are the root causes of this type of harassment. Although belief-based harassment and identity-based harassment are directed at different traits, the two can overlap. When ideological disagreements and identity-related vulnerabilities intersect, some individuals may experience intensified attacks (Nadim and Fladmoe, 2021). This line of research lacks empirical operationalization. Therefore, one of the primary goals of this study is to theorize the mechanisms by which political use of social media triggers individuals to become targets of coordinated attacks, particularly for those at the intersection of marginalized identities.
Religion is one group characteristic that may be targeted (Wintterlin et al., 2020). As political discourse increasingly integrates identity-based appeals, religious affiliations become markers of political alignment (Perry, 2023), making religious minorities more vulnerable to harassment. Religion is not just a private belief system but also a significant element in political rhetoric (Nigam, 2016). Dominant religious groups depict themselves as aligned with specific political ideologies, such as Christian nationalism in the United States (Perry, 2022) and Hindu privilege in India (Nigam, 2016). They strategically exclude religious minorities by associating them with broader political or cultural threats (Perry, 2023). This kind of rhetoric fuels online harassment directed at non-dominant religions, especially when minority religious groups try to voice their political opinions on social media. Speaking up can be seen as challenging the dominant political or cultural narrative, making them more likely targets for online attacks.
Research shows that minorities and marginalized groups face more online harassment than those in the majority (Castaño-Pulgarín et al., 2021). For religious minorities, political social media use often leads to more identity-based harassment, targeting them simply for their faith (Nadim and Fladmoe, 2021; Wintterlin et al., 2020). Therefore, even though political social media use increases the chance of belief-based online harassment, religious minorities face one more layer of vulnerability that is directed toward their religious identity (Nigam, 2016). Therefore, we propose as follows:
H2: Religious status will moderate the association between political social media use and online harassment, such that religious minorities face more substantial online harassment as compared with the religious majority.
The intersection of gender and religion in online harassment
Furthermore, the intersection of religion with other marginalized identities, such as gender, can amplify the severity of online abuse, an area of inquiry that remains understudied in existing political communication research. Previous research has largely treated gender-based and religion-based harassment and discrimination separately (e.g. Ahmed et al., 2025; Castaño-Pulgarín et al., 2021), overlooking their unique intersection. This study operationalizes and examines this intersectionality, focusing specifically on religious minority women. Religious identity plays a crucial role in shaping individuals’ experiences of online harassment; however, when intertwined with gender, the effects become even more pronounced due to the compounded vulnerabilities faced by women (Nigam, 2016). This intersection is particularly salient in less democratic and developing societies, where social hierarchies are more rigid and entrenched. Women from religious minority groups often experience dual marginalization, being targeted both for their faith and their gender (Wintterlin et al., 2020). By contrast, women belonging to the religious majority may find that their religious affiliation mitigates some of the inequalities and discrimination they face (Schnabel, 2021). Consequently, it can be expected that gender moderates the moderating effect of religion on experiences of online harassment.
Gender has long been a central issue in discussions of social justice and equality. Despite progress, women continue to face numerous barriers, particularly in patriarchal societies. These challenges become especially evident in the political sphere, where women encounter heightened resistance, hostility, and even threats to their safety (Stevens et al., 2024), which may function as a reason for strategic political participation of women in less visible and less contentious behaviors than men (Bode, 2017). Therefore, anticipating potential harassment and threats, women may choose to limit their visible political expression and participation on social media when disclosing their minority identities is required. Evidence supports this phenomenon, showing that minorities who filter political opponents from their social media networks tend to increase their political expression online (Zhu and Skoric, 2022). Female identity may bring a similar social media experience. Female politicians are also disproportionately subjected to digital abuse (Ahmed et al., 2025). Compared with men, women are more frequently targeted by identity-based online harassment (Nadim and Fladmoe, 2021). As women’s participation in politics challenges entrenched gender norms within male-dominated systems, their visibility often increases their vulnerability to such attacks (Nigam, 2016). Consequently, many women hesitate to express their opinions online due to concerns about safety (Stevens et al., 2024). Their active engagement on social media not only amplifies their public presence but also exposes them to intensified harassment directed at both their personal identities and collective group affiliations.
Moreover, gender may moderate the moderating effect of religion on the relationship between political social media use and online harassment due to the intersecting influences of gender and religious identity. For women in particular, an already underprivileged group, religion can function both as a source of empowerment and as a marker of vulnerability. Women from religious majority groups may benefit from the protective and supportive aspects of their faith communities, as religion often provides them with social resources, legitimacy, and status in various domains, including politics (Schnabel, 2021; Williams, 1996). By contrast, women belonging to religious minorities face compounded forms of discrimination: they are targeted not only for their gender but also for their minority religious identity, leading to heightened levels of online harassment (Wintterlin et al., 2020). Because they are often perceived as outsiders or as threats to dominant sociopolitical or cultural norms, these women are more likely to encounter hostility online than their counterparts (Schnabel, 2021). Hence, religion can mitigate inequality for some groups while exacerbating it for others, depending on their social position within the religious hierarchy. Thus, the following hypothesis is proposed:
H3: Gender will further moderate the association between religious status, political social media use, and online harassment, such that women from religious minorities will experience the most substantial online harassment.
The moderated moderation model predicting online harassment is illustrated in Figure 1.
Moderated moderation model predicting online harassment.
The personal costs of political use of social media
Personal consequences often accompany political participation. Our study does not conceptualize political costs as barriers to formal political careers or electoral participation, although such participation also brings personal costs. Rather, we focus on the costs ordinary citizens incur when engaging in everyday political expression in digital spaces. While early research primarily emphasized the tangible costs of political engagement, such as time and financial resources (Isaksson, 2014), more recent scholarship has expanded this focus to encompass a broader range of effects (Smith et al., 2019). These encompass emotional costs, including heightened stress, irritation, and anxiety, as well as social costs, such as strained relationships and even physical tolls (Smith et al., 2019). This concept encompasses the perceptual and experiential dimensions of disengagement. We conceptualize these perceptions as proximal psychological costs that are theoretically linked to downstream behavioral outcomes, such as self-censorship and reduced participation.
Despite these advances, important gaps remain in the literature. Many scholars continue to assume that social media platforms lower structural barriers to political participation by reducing costs and expanding access, thereby fostering greater equality in political engagement (Boulianne, 2015; Tufekci, 2017). This optimistic view overlooks the possibility that digitally mediated political participation may generate new forms of personal costs, particularly in polarized online environments and experienced by specific group members. Empirical research has yet to systematically examine the influence of social media political engagement on individuals’ perceptions of political costs and the mechanisms underlying this influence. Addressing these gaps is crucial for understanding the implications of contemporary political participation in the digital sphere and for evaluating whether social media genuinely fosters a healthy, sustainable participatory democracy.
We argue that the use of political social media could intensify the costs of politics. Political social media use refers to activities through which individuals express or disclose their political opinions and preferences online. Such engagement, however, carries inherent risks and potential negative consequences (Dhir et al., 2019). Because social media networks typically include both close acquaintances and distant connections, revealing one’s political views can strain interpersonal relationships, particularly when political disagreements arise (Weeks et al., 2024). In today’s increasingly polarized political climate, these disagreements can easily escalate into hostile or emotionally charged exchanges. Moreover, individuals who express political opinions publicly on social media are subject to heightened visibility and scrutiny, exposing them to criticism, abuse, and reputational harm (Weeks et al., 2024). These perceived social and emotional costs can instill reluctance or fear in social media users when engaging in political expression online (Weeks et al., 2024).
Moreover, engagement with political content on social media can lead to emotional exhaustion, frustration, and heightened stress. Political discussions in online spaces often expose individuals to hostile or uncivil language (Masood et al., 2024). Frequent exposure to such hostile or emotionally charged discourse can harm mental well-being and diminish individuals’ perceived social support (Dhir et al., 2019). In addition, the algorithmic design of social media platforms often amplifies these negative experiences. In pursuit of engagement and attention, platforms often prioritize content that is controversial or emotionally provocative (Lim, 2020), which further intensifies the psychological strain associated with online political participation. Consequently, the increased use of political social media may increase the emotional and cognitive “costs” of politics, potentially discouraging individuals from engaging in active politics. Based on the above arguments, we believe as follows:
H4: Political social media use will be positively associated with the cost of politics.
Online harassment and the personal costs of political social media use
The costs of politics may also increase if the use of political social media exacerbates online harassment. Individuals are more vulnerable to belief-based digital harassment when they interact with political content and conversations on social media (Nadim and Fladmoe, 2021; Weeks et al., 2024). Hostility can swiftly rise in the open and unregulated environment that social media platforms provide. Online harassment motivated by ideological hate and identity threats regularly targets people who express political beliefs and attitudes on social media, particularly those who voice dissenting perspectives (Sharma et al., 2023).
These online traumatic experiences due to their political disclosure on social media reinforce the perception that politics is costly, not only in terms of time and skills but also in terms of bringing in emotional, psychological, and social consequences (Smith et al., 2019). A comprehensive review found that victims of online harassment suffer from a wide range of severe mental health consequences (Stevens et al., 2021). These individuals frequently experience heightened levels of depression, anxiety, and fear (Short et al., 2014). In more severe cases, prolonged exposure to online harassment has been linked to suicidal ideation, as victims struggle with overwhelming pyschological distress (Pashang et al., 2019; Thomas et al., 2023). These results show that the impact of online harassment on victims’ general mental health goes well beyond the virtual world.
Furthermore, political motivations frequently drive Internet harassment. Digital users may purposefully spread false information to undermine their political rivals (Sharma et al., 2023). When encountering disagreeing ideas on social media, individuals turn irrational and attack the person expressing opposing opinions with uncivil language (Masood et al., 2024). Political polarization and ideological divisions fuel hostile interactions on social media (Chen et al., 2024), where individuals who express political viewpoints risk facing attacks from opposing groups. Online harassment serves as a means of silencing or discrediting individuals holding opposing ideologies. Thus, the use of political social media becomes increasingly risky. It may contribute to the perception that participating in politics comes with high personal costs due to the enlarged possibility of experiencing belief-based online attacks. Studies have shown that online harassment deters individuals from participating in politics online due to the fear of potential costs (Nadim and Fladmoe, 2021; Weeks et al., 2024). Therefore, we propose as follows:
H5: The relationship between political social media use and the cost of politics will be positively mediated by online harassment.
Intersecting identities and unequal costs of politics
For individuals from underprivileged groups, the impact of political social media use on the perceived costs of politics through experiences of online harassment can be especially pronounced. Because harassment based on gender and religion remains prevalent, members of marginalized communities may be discouraged from engaging in politics due to fear of retaliation (Southern, 2024). While political social media use does not always lead to online attacks, individuals from historically disadvantaged groups, such as women and religious minorities, are disproportionately targeted (Southern, 2024; Wintterlin et al., 2020).
In online environments, gender and religion serve as visible social identities that often attract identity-based harassment, alongside political attitudes and ideologies (Nadim and Fladmoe, 2021; Southern, 2024; Wintterlin et al., 2020). Both gender (Krook and Sanín, 2020; Stevens et al., 2024; Tandoc et al., 2023) and religion (Nigam, 2016; Perry, 2023; Schnabel, 2021) are significant factors shaping discrimination in politics. Political rhetoric frequently intertwines with gender and religious stereotypes (Perry, 2023; Wald et al., 2005), reinforcing exclusionary narratives. Women and religious minorities, already marginalized in many political contexts, are often further silenced through online harassment (Faith, 2022; Lazarus et al., 2022; Nadim and Fladmoe, 2021; Southern, 2024). As a result, when these groups engage politically on social media, their group identities themselves become potential triggers for attack.
Empirical evidence indicates that women politicians and journalists experience disproportionate levels of online violence and harassment rooted in gender bias (Krook and Sanín, 2022; Tandoc et al., 2023). Compared with men, women face higher emotional and psychological costs when engaging in political discourse online, often resulting in fear of self-expression due to safety concerns (Stevens et al., 2024). While this body of research primarily focuses on elite actors in highly visible public roles, it nonetheless reveals the broader gendered structure of digital political spaces. These dynamics are not confined to formal political actors but also shape the everyday experiences of ordinary women who engage in political discussion online (Southern, 2024). Even without institutional power or public office, women who engage in politics online are subject to similar gendered scrutiny and hostility, which translates into heightened psychological costs and reluctance to participate in politics (Southern, 2024) compared with men. Similarly, religious minorities are frequently portrayed as threats to national or cultural identity (Perry, 2023). Such narratives, reinforced by ethno-nationalist rhetoric, render religious minorities particularly vulnerable to online abuse (Perry, 2023). Consequently, gender and religion may jointly moderate the mediating role of online harassment in the relationship between political social media use and the perceived costs of politics, as these marginalized identities amplify the likelihood of experiencing harassment.
Moreover, gender and religion may interactively influence this mediation process. Women, who often face systemic disadvantages in political participation, may find not only religious grounding but also social and emotional support (Williams, 1996). This can enhance resilience and a sense of belonging among marginalized groups (Nigam, 2016). In this sense, religious membership, particularly within majority faith communities, may buffer women against gender-based harassment (Schnabel, 2021). Conversely, women belonging to religious minorities face “double risk,” as both gender and religion become sources of identity-based targeting. However, given the limited empirical evidence in this context, we pose the following research question:
RQ1: How do gender and religious status moderate the mediated relationship between political use of social media, online harassment, and the cost of politics?
We included Figure 2 to demonstrate the moderated-mediation model predicting the costs of politics.
Moderated-mediation model predicting the costs of politics.
Method
Sample
We used the survey agency Dynata to recruit a diverse online panel of Indian citizens. To ensure representation, a quota sampling method was applied, targeting demographic benchmarks based on age and gender distributions. This strategy, commonly used in previous research to enhance generalizability, was implemented during survey administration in September 2023. The final sample consisted of 1533 respondents. The study received approval from the Institutional Review Board at the host University, and all participants provided informed consent before participating in the research.
Measures
Social media political news use was assessed using five items asking respondents how often they engaged in political news-related activities on social media. These included (a) posting on their timeline, (b) commenting on posts, (c) sharing posts, (d) reading their news feed, and (e) reading friends’ timelines, all in relation to political or public affairs news. Responses were recorded on a 5-point scale (1 = never to 5 = always) and averaged to form a composite index (M = 3.24, SD = 1.09, α = .90) (e.g., Wang et al., 2024).
Online harassment was measured by asking participants how frequently they experienced specific types of harassment over the past year as a result of their political expression on social media. The prompt read: “During the past year, how often have you faced the following situations on the Internet or social media because of your political expression (posts, likes, comments about political issues)?” Responses ranged from 1 (never) to 5 (daily). The measure included 20 items that reflected varying forms of hate and harassment, ranging from personal insults to severe threats. Sample items include: “You have been underestimated or criticized because of your age,” “You have been attacked because of your religion or ideology,” “False statements about you have been spread on social media,” and “You have been threatened with violence.” Responses were averaged to create a composite scale (M = 2.15, SD = 1.25, α = .96). The scale was adapted from existing validated instruments designed to assess hate and harassment on digital platforms (e.g., Oksanen et al., 2021).
Cost of politics was assessed using a 10-item battery from Smith et al. (2019), measuring the psychological and emotional toll of political engagement. Participants indicated their level of agreement on statements such as: “Politics has caused me to be stressed,” “I have become depressed when a preferred candidate lost,” and “Differences in political views have damaged a friendship I valued.” Responses were provided on a 5-point scale (1 = strongly disagree to 5 = strongly agree). Items were averaged to create a composite measure (M = 3.06, SD = 1.02, α = .94).
Religious status was captured by asking respondents to indicate their religious affiliation. Most participants identified as Hindu (78.7%), followed by Muslim (10.1%), Christian (5.6%), and others (5.6%). For analytic purposes, Hindu respondents were classified as the religious majority (coded = 1), while all others were categorized as religious minorities (coded = 0).
Gender was initially measured using three categories, male, female, and other, rather than a strictly binary framework. The “other” category was intended to capture respondents whose gender identities do not align with male or female classifications. However, only a negligible number of respondents selected this category (N = 6), which precluded a more detailed or fine-grained analysis of specific identities within the sample (e.g., moderations or moderated mediation). To test our hypotheses, we coded females as the reference group (coded = 0) as a comparison for males (coded = 1).
Control variables included both demographic and motivational factors. Demographics comprised age (M = 38.26, SD = 13.95), education (median = bachelor’s degree), and income (median = ₹60,000 to ₹79,000). Motivational factors were traditional media use, political interest, and partisanship (support for the ruling party, BJP). Traditional media use was measured by asking participants how frequently they consumed political or public affairs news via television, radio, and print newspapers. Political interest was measured using a single item that asked respondents how interested they were in politics (1 = not at all interested to 5 = extremely interested; M = 3.59, SD = 1.11). Partisanship was measured by asking respondents whether they supported the BJP, the ruling party, 63% responding affirmatively. These demographic and motivational variables were controlled for to isolate the effects of key predictors.
Results
We ran a hierarchical regression model to test our first set of hypotheses (H1–H3). The results are summarized in Table 1. First, the results support H1, indicating that the political use of social media is positively associated with online harassment (b = .40, SE = .04, p < .001). Individuals who engage more with political news on social media experience higher levels of online harassment. This finding aligns with past studies indicating that visible political expression on digital platforms can lead to increased exposure to hostile online interactions (e.g. Gorwa et al., 2020; Theocharis et al., 2020).
OLS regression results predicting online harassment.
p < 0.001, ** p < 0.01, * p < 0.05.
Next, H2 was supported; the two-way interaction between the political use of social media and religious status was significant (b = −.14, SE = .06, p < .05). The results (illustrated in Figure 3) indicate that religious status moderates the relationship. Specifically, those from religious minority groups who frequently engage in political use of social media experience higher levels of online harassment (b = .51, SE = .06, p < .001) compared with individuals from the religious majority (b = .36, SE = .04, p < .001). These findings underscore how religious identity can shape online vulnerability, especially in politicized contexts (Nadim and Fladmoe, 2021; Wintterlin et al., 2020). This pattern is particularly meaningful in the Indian context, where Hindu nationalism has become a dominant political ideology (Jaffrelot, 2021; Subramanian, 2019). Political discourse often draws on majoritarian narratives that frame religious minorities as outsiders or threats, thereby increasing their exposure to identity-based harassment when they participate in political discussions online (Jaffrelot, 2021).
Illustrating the moderation between political news on social media and religious status in predicting online harassment.
Finally, H3 was also supported by a statistically significant three-way interaction among political use of social media, religious status, and gender (b = .22, SE = .11, p < .05, Table 1). The results are illustrated in Figure 4 (for females) and Figure 5 (for males). Women from religious minority communities who are politically active on social media experience the highest levels of online harassment (b = .58, SE = .08, p < .001). These results highlight the intersectional vulnerabilities faced by individuals at the nexus of gender and minority religious identity in digital political spaces. This pattern is consistent with the broader Indian sociopolitical context, where religious identity and gender are both highly politicized and deeply embedded in public discourse (Parveen and Sanyal, 2024). For example, Muslim women often occupy a doubly marginalized position, as they are simultaneously subject to gendered norms and religious othering within a political climate shaped by Hindu nationalist narratives (Parveen and Sanyal, 2024).
Illustrating the moderation between political news use of social media, religious status, and gender (for females) in predicting online harassment.
Illustrating the moderation between political news use of social media, religious status, and gender (for males) in predicting online harassment.
Next, to test the next set of hypotheses focusing on the cost of politics (H4 and H5), we ran a mediation analysis using the PROCESS macro for SPSS (Model 4; Hayes, 2018). The results of the ordinary least squares (OLS) regression predicting the cost of politics are included in Table 2. The results of the mediation analysis pathways are illustrated in Figure 6.
OLS regression results predicting cost of politics.
p < 0.001, ** p < 0.01, * p < 0.05.
Illustrating the pathways in the mediation between political news use of social media and the cost of politics through online harassment.
The results suggest that the political use of social media is not directly associated with the cost of politics (b = .01, SE = .03, p = .91). As such, mere political use of social media does not necessarily translate into perceptions of political vulnerability. Support for H4 is denied. However, we find support for H5, as the indirect effect is statistically significant (b = .23, SE = .02, CI = [.188 to .273]).
Finally, we test the moderating effects of religious status and gender on the mediated path (social media political use → online harassment → cost of politics). We ran a moderated-mediation model using the PROCESS macro (Model 76, Hayes, 2018). The conditional indirect effects of social media political use on the cost of politics, through online harassment, are presented in Table 3. The patterns indicate that the pathway from social media political use to perceived political costs via online harassment is stronger for religious minorities than for the majority. The gender effects are subtle, with marginal differences between males and females. Overall, the results reflect the compounded disadvantage of minority identity in social media political use and expression, confirming the presence of structural vulnerabilities in mediated political participation.
Conditional indirect effects of social media political use on the cost of politics via online harassment.
Additional results on predicting online harassment and the cost of politics are included in Appendix 1 and Appendix 2.
Discussion
This study addresses critical gaps in social media political communication research by developing an intersectional theoretical framework that examines how religious and gender minorities interact with social media political use, which in turn shapes online harassment experiences and the costs of politics. Our findings demonstrate that political expression on social media carries distinct emotional, social, and psychological burdens, particularly for individuals with marginalized identities.
While prior research has often treated social media as a low-cost avenue for participation and examined its democratic consequences from an optimistic perspective (Boulianne, 2015; Tufekci, 2017), our study foregrounds the personal costs of digital political engagement. Consistent with our expectation (H1), we found that the political use of social media is positively associated with online harassment, reinforcing the idea that visibility in digital political spaces increases exposure to hostility. This finding aligns with previous research, which has shown that online political expression increases the likelihood of harassment and incivility (Gorwa et al., 2020; Theocharis et al., 2020), particularly in polarized digital environments. Because social media political use can publicly disclose individuals’ political ideologies, opinions, and positions, which can then attract attention and harassment from political opponents. Social media not only enables political participation but also facilitates silencing through attacks (Barnidge, 2017; Chen et al., 2024; Goyanes et al., 2021; Lu and Myrick, 2016; Sharma et al., 2023), a phenomenon identified in many democracies. Our findings advance a more critical perspective by conceptualizing political visibility on social media as a source of personal risk, where increased engagement simultaneously invites hostility. This shifts theoretical attention from social media participation as a largely beneficial outcome from a democratic perspective (Tufekci, 2017) to a potentially costly process in digital spaces.
Although not hypothesized, our data also show structural disparity in online harassment. Younger, less educated, and lower-income individuals experience more harassment, suggesting that structural disadvantage compounds exposure to online abuse. This pattern aligns with broader research on political inequality (Ahmed, 2020).
Furthermore, this research posits a dynamic relationship between political use of social media and harassment experiences, moderated by marginalized identities. Our findings supporting H2 confirm that religious status significantly moderates the link between political social media use and online harassment. Religious minorities who engage politically online report substantially higher levels of harassment compared with those from the majority faith. This aligns with past scholarship indicating that religion functions as a visible identity marker in politicized online contexts (Nadim and Fladmoe, 2021; Nigam, 2016; Perry, 2023; Wintterlin et al., 2020), making minority groups more vulnerable to attack. Research from diverse contexts globally demonstrates that when religion becomes intertwined with political ideologies and identity, minority religious groups are more frequently framed as cultural or political outsiders and are disproportionately targeted in online discourse (Nadim and Fladmoe, 2021; Nigam, 2016; Perry, 2023; Wintterlin et al., 2020). Such findings underscore how digital spaces replicate offline hierarchies, where dominant religious groups exercise symbolic power through exclusionary narratives (Gardiner, 2018).
The findings for H3 extend this pattern by revealing that gender further intensifies these risks. The three-way interaction indicates that women from religious minority communities who are frequent users of social media experience the highest levels of harassment. This intersectional pattern substantiates claims of intersectionality theory (Cho et al., 2013) by showing how gender and religion jointly shape vulnerability in digital political spaces. In practical terms, women from minority religious groups face “double threats,” where both their gender and religion attract identity-based abuse (Parveen and Sanyal, 2024; Stevens et al., 2024). These empirical patterns highlight that compounded marginalization, rather than political expression alone, drives much of the hostility in online political interactions.
By centering minority women’s experiences in India’s contemporary political landscape, the study moves beyond Western-centric models and offers insights into how digital political risks are shaped by overlapping structures of power and marginalization. The study contributes theoretically by reconceptualizing online harassment not as individual victimization but as structural exclusion, revealing how digital platforms simultaneously enable political voice and facilitate coordinated mechanisms of silencing. This framework advances the understanding of digital citizenship by illuminating how sociopolitical contexts shape the relationship between social media political use and harassment.
Turning to the perceived cost of politics, our findings reject H4, indicating no direct relationship between political social media use and the perceived cost of politics. This suggests that political use of social media online does not automatically increase perceptions of political vulnerability. However, support for H5 reveals that this relationship is indirect and mediated through online harassment. In other words, harassment acts as the psychological bridge linking digital political use to the emotional and cognitive costs of participation. This mediation confirms earlier theoretical claims that repeated exposure to online hostility translates into political fatigue, anxiety, and social withdrawal (Smith et al., 2019).
Furthermore, the moderated-mediation analysis provides a nuanced understanding of this pathway. The indirect effect of political social media use on perceived political costs through harassment is significantly stronger for religious minorities than for majorities, while gender differences are marginal. This suggests that religious identity remains a more salient vulnerability marker in online political spaces than gender alone, at least within the Indian context where religious tensions are heightened. This finding advances the digital intersectionality literature by illustrating that not all marginalized identities operate with equal weight; religious affiliation can sometimes overshadow gender as a determinant of online risk (Perry, 2023; Schnabel, 2021).
These patterns challenge the notion that social media democratizes participation. Rather than serving as politically neutral arenas, platforms are structured to reward visibility and engagement at the expense of safety (Rheault et al., 2019; Walther, 2022). The positive relationship between political use of social media and harassment empirically validates these critiques, demonstrating that algorithmic amplification and engagement-driven design may exacerbate exposure to abuse. For example, as past work on India’s digital sphere suggests, politically vocal women and minority users are often targeted through coordinated trolling and disinformation campaigns (Amnesty International, 2020; Banaji et al., 2019; Gurumurthy and Dasarathy, 2022).
Our findings also reveal broader inequalities in who bears the emotional and psychological costs of digital engagement more. While men in our sample reported slightly higher perceived political costs overall, this pattern may reflect different thresholds for reporting or variations in the types of political risk experienced. For marginalized users, particularly religious minorities, the sense of exposure and insecurity is compounded by structural exclusion. The stronger indirect effect for religious minorities suggests that harassment not only silences individuals but also reinforces systemic barriers to equitable participation (Ahmed and Madrid-Morales, 2021; Rossini, 2021).
Ultimately, these findings extend ongoing debates on digital participation inequality. The results illustrate that social media’s participatory affordances are not equally distributed; those from dominant social groups enjoy relative safety, while marginalized users incur disproportionate emotional and psychological costs. This asymmetry may contribute to a selective withdrawal from political discourse, consistent with both the spiral of silence (Gearhart and Zhang, 2015; Weeks et al., 2024) and minority stress frameworks (Meyer, 2003). When harassment drives users to self-censor, digital spaces become echo chambers dominated by the majority’s voices, thereby diminishing political diversity and the deliberative quality of discourse essential for the well-functioning of any democratic society.
Conclusion
In summary, this study provides empirical evidence of how identity-based vulnerabilities mediate and moderate the costs of digital political participation. By showing that online harassment mediates the link between political social media use and perceived political costs and that this pathway is intensified for minorities, our research underscores the compounded nature of digital inequality. Far from being inclusive public spheres, social media platforms often perpetuate societal hierarchies, where the cost of visibility is disproportionately borne by those who are already marginalized.
To address these challenges, social media companies must implement stronger protective measures. This includes investing in real-time content moderation and developing proactive tools to detect and prevent coordinated harassment. Such efforts are particularly critical in developing democracies like India, where institutional responses to digital harms remain limited. The Bulli Bai and Sulli Deals incidents exemplify the consequences of corporate inaction. Both cases involved fake auction apps circulating doctored images of Muslim women, resulting in public humiliation and emotional distress (Barthwal and Shukla, 2022; BBC, 2022). Despite widespread outrage, platform responses were delayed and insufficient, revealing stark inequities in moderation standards.
These failures underscore the urgent need for equity in global content governance. Social media companies must decentralize safety investments, establish regionally informed moderation teams, enhance cultural and linguistic sensitivity, and increase transparency in enforcement. Without these reforms, marginalized users, particularly in the Global South, will remain disproportionately exposed to digital harm. Beyond corporate accountability, civil society and public institutions also play a critical role. Digital literacy initiatives that integrate mental health awareness and legal education can equip users to navigate online risks more safely. Meanwhile, nongovernmental organizations can provide vital support networks, including reporting mechanisms on social media.
Finally, while this study is grounded in strong theoretical frameworks, its cross-sectional design limits the ability to make causal claims. We cannot definitively determine whether experiencing harassment increases perceived political cost or whether those who already anticipate harm are more sensitive to online hostility. Future research should investigate these dynamics through longitudinal studies, controlled experiments, or the analysis of digital trace data. Such approaches could offer more nuanced insights into how risk perceptions shape political outcomes over time and across diverse contexts. In addition, due to the self-reported nature of the data, such as respondents’ assessments of online harassment experiences and political costs, they may be influenced by recall bias or subjective interpretation, which could lead to measurement error. Furthermore, although the study adopts an intersectional perspective by focusing on religion and gender, other relevant dimensions of social stratification, such as caste, language, and geography, are not examined. These factors may further condition experiences of online harassment and political engagement and warrant attention in future research.
Footnotes
Appendix 1
Appendix 2
Conditional direct effects of social media political use on the cost of politics.
| Religious status | Gender | Effect | SE | 95% CI [LL, UL] |
|---|---|---|---|---|
| Religious minority | Female | 0.04 | 0.05 | [-0.07, 0.14] |
| Religious minority | Male | 0.04 | 0.05 | [-0.05, 0.14] |
| Religious majority | Female | -0.01 | 0.04 | [-0.08, 0.06] |
| Religious majority | Male | -0.00 | 0.03 | [-0.07, 0.06] |
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The project received funding from the Ministry of Education (MOE), Singapore.
