Dismissive Affect in Anti-Immigration TikTok Content: The Construction of Social Crisis Through Platform Affordances and Affective Persuasion

Abstract

Keywords: dismissive affect; TikTok; symbolic threat; immigration; framing; platform affordances; affective politics

1. Introduction

TikTok was created to be an app that encourages everyone to be a creator with its wide range of effects and challenges, and due to the nature of its algorithm, it is easier to go viral on TikTok compared to other social networking websites (Schellewald, 2025). It appeals to younger generations and becomes popular among them because its creators designed it with teens as the target audience from the start. As a result, the app matches the habits and preferences of younger people.

TikTok’s primary goal is user engagement, not information provision. Due to its growing use as an information resource, concerns about information quality, misinformation, surveillance, and general well-being may be only the tip of the iceberg (O’Brien, Davoudi, & Nelson, 2025). Lax platform security allows harmful content to circulate (Weimann & Masri, 2023).

Given the decreased security, instead of professional, enjoyable, and ephemeral consumption, nationalist, racist, and white supremacist messages have increased (Ozduzen, Ferenczi, & Holmes, 2023). Research by Albertazzi and Bonansinga (2024) shows that radical parties do not necessarily spread negative content. They state that positive or connecting messages also play an important role in content engagement. In this regard, some extremist groups and political parties employ strategies to evoke the deepest sentiments of voters using not only negative but also positive narratives, mobilising those who are uncertain, uninformed, or abandoned by traditional political parties or current systems (Cartes-Barroso, García-Estévez, & Méndez-Muros, 2025).

In their studies, Manik and Tarisayi (2025) found that in the aftermath of such incidents, accusations of criminal intent reveal strong anti-foreigner sentiment, expressed through criminal stereotypes, dehumanising attitudes, and the denial of immigrants’ right to fair treatment. According to them, there is also the scapegoating of immigrant groups that is evident in comments and content of the videos, revealing that foreigners are under constant threat, as they are often perceived as inherently criminal, dangerous, and a drain on local resources and accused of usurping jobs and economic opportunities from locals. Such perceptions fuel demands for mass deportation and restrictions on immigrants’ rights to business ownership, challenging foreigners’ desire to earn a livelihood and coexist with the local population. These types of criminalisation and resource competition align with the findings in Stephan and Stephan’s (2000) study. The findings echo their account of how host populations perceive economic and safety-related threats from newcomers. Their study was developed through survey and offline media research, whereas the present study examines short-form platform content.

Although all this research has touched on the issue from different perspectives (narrative, platform affordance and framing) most of it focuses on how far-right activists are using this affordance to spread misinformation, and, most importantly, these studies assume that anger and fear are the primary emotional drivers of this content, an assumption more often asserted than tested against what actually dominates highly engaged videos. There are relatively few studies focusing on how such content is framed and which emotions they are trying to provoke. This study does not necessarily focus on the specific groups and how they spread information but on how the most engaged and popular videos are framed and which emotions and formats are dominant. As the research findings show, the emotions that actually dominate highly engaged immigration content are not anger or fear but sarcasm, disappointment and irritation. This is a dismissive, ridiculing form of affect that has received comparatively little attention in either the platform studies or migration studies literature on online hostility towards immigrants.

Chouliaraki and Georgiou (2019) divide how the symbolic border is created between immigrants and the local community into two separate but interrelated processes: digital disinformation, where unreliable or malevolent online sources of knowledge may perpetuate the precarity of migrants’ lives, and linguistic misrepresentation, where the dominant vocabularies of migration ‘other’ and dehumanise migrants in western mediascapes. The aim of the study is not to identify misinformation but to show how malevolent online sources are distributed through specific techniques and platform affordances, allowing selective, though not necessarily false, information to attract more engagement and serve as proof without context. Ultimately, it aims to establish how immigrants specifically are framed within the TikTok environment, and how this relates to their being posed as a realistic or symbolic threat online. This realistic and symbolic distinction draws on a long tradition in intergroup relations research on perceived threat from outgroups (Stephan & Stephan, 2000), applied here to a platform-specific communicative environment.

Previous research examined anti-immigrant narratives, platform affordances, and misinformation on TikTok separately and focused on single actors or narratives. Less attention has been paid to how highly engaged videos combine framing, affective persuasion, and platform-specific affordances to construct immigration as an immediate social crisis. Considering this, the research focuses on recurring communication patterns through which immigration is constructed as a crisis within TikTok’s communication environment, combining framing, affective persuasion and platform-specific affordances. In doing so, the study identifies recurring patterns in the construction of immigration as a social crisis and contributes to broader literature on crisis construction in platform-mediated communication by demonstrating how framing, affective persuasion, and platform-specific features work together to produce persuasive crisis narratives. Based on the analysis, the study proposes the term dismissive affect. The term is used to describe a low-intensity affective register in which exclusionary claims about immigrants are presented not as alarming threats but as self-evident and mildly amusing, inviting the audience to share a faintly superior certainty rather than fear or outrage. Because it does not resemble open hostility, the study argues that this register is both harder for platforms to moderate under existing frameworks and harder for audiences to consciously resist.

Based on this aim, the research answers the following questions:

1. RQ1 How is immigration framed as a social crisis in highly engaged TikTok videos?
2. RQ2 What emotional appeals are most used in immigration-related TikTok videos, and how do they interact with specific frames?
3. RQ3 How is responsibility for immigration-related problems attributed in TikTok videos, and how do platform affordances increase the visibility and persuasion?

2. Theoretical Framework

The research combines three theoretical frameworks to address the research questions. The framing theory (Entman, 1993), affective politics (Papacharissi, 2015), and platform affordance theory (van Dijck, 2013) are used to analyse and offer insights into the content that was presented in TikTok. The analysis includes how content is framed to point out certain problems that are discursively attributed to immigrants, how content is emotionally charged to be persuasive for the audience, and how TikTok as a platform may reward emotionally charged short-form videos over deeper and more contextual explanations of the problem. As a result, audiences may be directed towards the emotionally charged short-form videos that are framing immigrants for certain social problems, shifting the attention from the authorities or policies to the immigrants themselves.

Platform affordance theory in this case shows that regardless of the individual creator’s personal intentions, these patterns can be understood as a platform reward system giving greater visibility to certain emotionally charged content rather than analytical content or deeper policy analysis. Together these frameworks help to analyse content in a systematic way, showing how immigration is framed as a social problem and how audiences are steered towards particular interpretations and emotional responses against immigration.

2.1 Framing theory: how immigrants are attributed in immigration-related content

Framing theory by Entman (1993) explains framing as a process in which content creators construct information about social problems in a way that highlights one aspect while hiding another. For this study, framing theory is one of the main theories used to understand how immigration-related videos are constructed, defined, and morally evaluated.

In platformed media, immigration is often constructed as the primary cause of various social problems. Especially in recent years, considering short attention spans (Hendrickx, 2025) and the rise of short-form media consumption, information tends to be simplified and presented in an emotionally dramatic or entertaining way to gain more visibility (Barta, Belanche, Fernandez, & Flavi, 2023). This encourages immigration to be simplified and framed as a threat to demographic values, the economy, and public order rather than as a complex problem. This type of framing directly puts immigrants in the limelight and problematises individuals rather than structural factors such as the complexity of the policies, their applications, or the government’s capacity.

The study extends framing theory into short-form videos on TikTok, highlighting how attribution and contextual understanding operate in rapid, affect-driven narrative structures. Using the variables “threat type” and “attribution”, the research analyses the main attribution of blame in each video, such as governments, policies, certain political parties or immigrants, and how the video claims that this attributed variable is the cause of real or symbolic threats.

2.2 Affective politics: why certain emotions are more popular than others

Affective politics, as explained by Papacharissi (2015), highlights the central role of affect in political communication, explaining how political events are expressed and understood by the public. Platform power is described as the inclusion of these affordances within economic, algorithmic, and governance structures that shape visibility and participation. This frame helps to understand why certain frames resonate and gain traction with mass audiences. Existing research on anti-immigrant content on social media platforms identifies anger, disgust, and fear as the dominant register. Survey and computational text analysis of anti-immigrant discourse on Facebook by Ahmed et al. (2024) finds such discourse strongly associated with anger and negative sentiment, fuelled by perceptions of symbolic and realistic threat. Another experimental work by Igartua and Ballesteros-Herencia (2026) on hate speech similarly finds that anti-immigrant hostility is closely tied to anger, disgust, and rage. One might expect the same emotions to dominate persuasion within TikTok’s content; however, this research shows that the most engaged content there departs from that pattern.

In this research, I analyse the dominant affective register elicited in the video and how it intensifies the attribution and blame, making the content more persuasive. Affective cues reinforce specific frames and can be analytically examined as a form of emotional influence, as they trigger visceral reactions over contextual understanding of social problems. Considering the content posted on TikTok, its algorithm privileges the content that triggers stronger emotions and attracts attention quickly (Gillespie, 2018). Within this algorithmically curated environment, such videos are more likely to attract engagement and visibility. In this platform environment, affective framing becomes the main platform-optimised mode of political expression and of understanding political and social problems. By identifying the dominant affective register, the research aims to examine whether these emotions appear incidentally or systematically.

2.3 How platform affordances and power shape user behaviours actively

An affordance, in general terms, refers to the possibilities and constraints a technology’s design offers for action: what a given interface makes easy, difficult, or invisible to do. In pursuit of understanding the environment in which this type of content circulates and is amplified, this study draws on the framework built on platform affordances theory (van Dijck, 2013; van Dijck, Poell, & de Waal, 2018; Zeng & Abidin, 2021). Platform affordances and platform power explain how a platform’s design, interface, and algorithm favour certain types of content, shaping user behaviour and visibility of content.

This study focuses on TikTok and its affordances, such as short-form content, audiovisual effects, simplified content, and engagement-based recommendations driven by likes and shares. These affordances suggest that TikTok favours short-form, simple yet emotionally engaging videos, encouraging people to watch the content that is visually striking, emotionally charged and easily understandable (Gillespie, 2018). Fang, Wang, and Hao (2019) define this consumption dynamic as the “anaesthetic effect”, in which users consume content for long continuous periods without being fully aware that they are doing so. A study by Siles et al. (2024) also reveals that TikTok’s algorithm offers users specific content based on their likes, dislikes, and engagement. TikTok’s machine learning-enabled recommendation system does not require users to follow creators or to explicitly choose types of content; the platform itself decides and suggests content to its users as they swipe through short videos (Gray, 2021).

Considering this, TikTok can be conceptualised as an active mediator in political and social discussions. As an active mediator, TikTok’s engagement-based algorithm tends to amplify narratives and incidental news consumption (Vázquez-Herrero & Negreira-Rey, 2022) that produce stronger affective reactions and clearer attribution of responsibility. As a result, instead of nuanced explanations, content blaming immigrants for social problems through affective cues circulates more widely.

This research also discusses platform power alongside individual creators and consumers. The approach aligns with Bucher’s (2018) view of algorithms not as static, independent forces but as emergent accomplishments produced through everyday human practices and platform interactions.

2.4 Threat types: symbolic and realistic threats

The attribution and threat-type variables coded in this study also connect to migration and intergroup-relations research by Stephan and Stephan (2000). Their research lies outside media studies, distinguishing realistic threats related to immigrants, involving tangible competition over resources, safety, or economic opportunity, from symbolic threats, involving perceived challenges to values, culture, and group identity. I add their distinction to this study to explain the dominant emotional response on anti-immigrant content on TikTok, which was earlier used to identify emotions on other platforms (Ahmed et al., 2024). By adding a “threat type” variable within the framework, the study’s analysis connects platform studies to the broader migration studies literature.

3. Methodology

3.1 Research design

The research is based on Mayring’s (2021) approach to qualitative content analysis, applied here to immigration-related TikTok videos. Theory-driven coding combined with systematic empirical observation is used to interpret the discourse through which immigration is framed. The research is interpretative and focuses on identifying recurring communication patterns in immigration-related videos and analysing how they are framed and emotionally charged to persuade audiences within the TikTok environment.

The study examines how attribution mechanisms, affective cues and platform-specific affordances are used to present immigration as a primary reason for social problems. Accordingly, the analysis extends beyond the identification of misinformation or explicit manipulation tactics to examine broader communicative patterns through which immigration is presented as a social crisis.

TikTok was chosen because of its growing popularity in political communication and opinion shaping through engagement-based algorithms and short audiovisual videos. As one of the world’s fastest growing platforms, TikTok’s engagement-based algorithm and recommendation system make the platform relevant for researching how emotionally charged political and social content is presented to users within a personalised recommendation environment.

3.2 Data collection

The sample of 120 videos maintains the interpretative depth appropriate for qualitative content analysis across a range of framing patterns, emotions and formats. The aim of this research is not to collect large-scale statistics and generate statistically representative findings but to identify recurring communication patterns within a manageable qualitative dataset (Mayring, 2021).

For the data collection, the researcher followed a similar approach to Fuchs (2017) and Chen et al. (2021), using TikTok’s recommendation environment as a sampling strategy. The researcher created a new TikTok account and began searching for and engaging with videos carrying immigration hashtags or keywords (see Appendix 2). Following engagement with immigration-related videos, the platform’s algorithm started suggesting similar content. Relying on the platform’s suggestions, the researcher identified highly engaged immigration-related content within one algorithmically curated recommendation environment. The procedure served as a pragmatic sampling strategy for identifying relevant videos in order to examine recurring framing patterns within a specific recommendation environment. The study does not claim to represent all immigration-related content available on the platform and is not an audit of TikTok’s recommendation system.

To maintain the validity of engagement and traction, the researcher only collected videos that had received more than 2,000 likes, were posted on TikTok, were related to immigration, and were in English. To ensure only highly visible and circulated content was included, the 2,000-like threshold was applied as a pragmatic operational indicator to identify highly engaged content while maintaining a manageable dataset for qualitative analysis. The researcher applied this measure based on van Dijck’s (2013) discussion of engagement metrics such as likes functioning as signals of circulation and recommendation potential. However, the researcher recognises that engagement alone cannot measure TikTok’s internal recommendation processes. It also aligns with Siles, Valerio-Alfaro and Meléndez-Moran’s (2024) research, which emphasises the algorithmic curation shaping users’ exposure to political content.

Although all 120 videos had received more than 2,000 likes and gained traction among users, the research acknowledges that it is not representative of all immigration-related videos on the platform, and this research does not generalise across populations, but it aims to provide deeper insights into recurring framing and affective patterns within a clearly defined recommendation environment (Tracy, 2010). While the results of the commentary analysis could be applied to understand users’ interpretations, this study focuses on the representation of the immigration problem within the videos themselves and on the communicative techniques used to construct those representations.

All the data were collected and listed with the date of collection, URL, postdate, format, content, caption text, onscreen text, attribution, threat type, evidence style, emotion, us-versus-them narrative, and whether a remedy was present or absent.

The research organised the following variables and linked them to research questions. Attribution of responsibility is aimed at answering RQ1, the threat type addresses RQ2, while evidence style and dominant affective register address RQ3. Other variables such as format, content, onscreen text, and caption text are used to understand platform-specific presentation patterns and communication techniques used in framing the attribution of immigration.

3.3 Data analysis

After archiving the data of 120 videos, each video was analysed to identify the subject of the attribution, threat type, evidence style, emotion, us-versus-them narrative, and whether a remedy was presented. Initial coding was conducted on the first 20 videos to establish the coding variables; the remaining videos were then coded using these pre-identified variables. To avoid ambiguity in unusual cases, the researcher documented additional exceptions during the analysis.

To keep consistency, each coding variable was given an operational definition, and exceptions defined in the analysis were documented. Operational definitions are provided in Appendix 1. In the study’s definition, the newly introduced dismissive affect can be identified through a deadpan or exaggeratedly calm delivery; rhetorical questions that imply their own answer (“I mean, obviously…”); an absence of urgent or fearful language; and a structural reliance on implied consensus rather than argument. Three criteria distinguish this register from similar categories. First, dismissive affect differs from humour in that it carries an implicit claim about a social group, whereas humour does not. Second, satire typically signals its own exaggeration or artifice to the viewer; dismissive affect presents itself as sincere, unremarkable observation. Third, earnest complaint carries emotional urgency or alarm triggering fear or anger; dismissive affect remains low-intensity and amused throughout.

This study acknowledges that sarcasm as a register can co-occur with emotions such as underlying anger or fear. In this analysis a video expressing anger through a sarcastic delivery was coded as sarcasm. The findings of this study focus on the dominant affective register rather than on how often discrete emotional states appear. Therefore, the study is not directly comparable to works that coded emotional content rather than rhetorical mode.

Each video was analysed based on affective cues, framing, and platform affordance features to answer the research questions. The coded data were analysed to identify recurring patterns in the attribution, the dominant affective cues, evidential strategies, the construction of immigration-related problems, and the platform-specific communication features accompanying these narratives.

Systematic coding was used to ensure the consistent application of the theories and coding variables across the research dataset. Coding was conducted by a single researcher rather than through multi-coder validation. This approach follows Tracy’s (2010) criteria for qualitative research quality, which do not depend on inter-coder reliability statistics and are better suited to this study’s interpretive design than a psychometric framework it was never built around. To further reduce the risk of disagreement at category boundaries (e.g., sarcasm versus disappointment), findings are reported mainly at the level of the broader affective register rather than individual emotions. In this way, the results remain robust even if some videos were classified differently by another coder.

The study aims to provide theoretical rather than quantitative or statistical generalisation. Consequently, the findings identify visible patterns while avoiding speculative interpretations beyond the research dataset.

3.4 Ethical considerations

In the collective archive, a link was recorded for each of the 120 videos as evidence of analysis, and the data were entered by the researcher. Videos were publicly available at the date of data collection. Although the analysed videos were publicly accessible, privacy is protected by not recording creators’ names, their personal profiles, or their appearance. In line with ethical judgement calls about qualitative social media research on sensitive issues (Harrington, 2025), the researcher did not contact any content creator, and no personal interaction occurred. No analysis of usernames, content creators, or individual creator motivations was conducted, and the research focuses solely on visible platform features, communicative patterns and narrative construction.

3.5 Limitations

As the study focuses on English-language TikTok, its findings do not capture attributions and affective registers present in other linguistic and cultural contexts. Additionally, it focuses on videos that have already gained traction and engagement; less engaged videos, and their framing, platform-based features and affective registers, could yield different results.

Data collection took place between 18 January 2026 and 2 February 2026. TikTok videos analysed here may become unavailable over time due to deletion, account removal, or platform moderation, meaning some content referenced in this study may not be independently verifiable by future readers. Furthermore, the study examines one algorithmically curated recommendation environment generated through a newly created TikTok account and therefore it does not claim to represent the full diversity of immigration-related content available on the platform.

Finally, as the research does not have direct access to TikTok’s internal metrics, the interpretations of platform influence are solely based on observable engagement patterns captured in the format, content, caption text, and onscreen text variables. Because the dataset is limited and was analysed by a single researcher, the study does not claim statistical representativeness; future research could strengthen these findings through multiple recommendation environments and multi-coder validation.

4. Findings

The dataset of 120 videos was coded for format (original, stitch, AI-generated, and minor additional types), attribution (who or what is positioned as responsible for the problem), and threat type (symbolic versus realistic). While symbolic threats refer to fears of cultural, demographic, and national erosion or identity loss, realistic threats concern tangible problems such as economic crises and safety issues arising from criminal activity or social disorder. How claims are supported was captured by the evidence style variable, which distinguished techniques such as visual anecdotes, textual overlays, or statements from influencers or authorities.

The results are presented in the following subsections, combining framing, affective persuasion, and platform-specific communication features. The subsections address the research questions while showing how these dimensions interact within highly engaged TikTok videos.

4.1 Attribution patterns: how immigrants become scapegoats

The analysis demonstrates that instead of emphasising government policy, institutional capacity, or broader socioeconomic conditions, the videos directly position immigrants individually as the explanation for social problems. Based on the analysis shown in Figure 1, 97 of the 120 videos (80.8%) directly blamed immigrants. This category included 20 videos targeting Muslim immigrants specifically (including two in which Muslims were represented by the political figure Zohran Mamdani), combining religious identity with immigration in the attribution of responsibility.

This shows that most content creators shift the blame solely onto immigrants rather than authorities, governments, and policies. The actions of individuals and social groups are blamed more than the complexity of policies or gaps in their application.

Additionally, seven videos (7/120; 5.8%) blamed pro-immigration actors, liberals, and left-wing actors for their ideology. Nine videos (7.5%) blamed race or ethnicity, two videos (1.7%) blamed authorities and structural factors, and two videos (1.7%) blamed pro-Palestine actors or supporters. A further two videos called for the protection of European, white, and Christian identity, blaming immigrants for violating national identity. Finally, one video (0.8%) attributed responsibility to immigration as a process rather than to a specific social group.

Geographically, the corpus was concentrated on the United Kingdom (28/120 videos, 23%), with smaller numbers focusing on the United States, France, Germany, Poland, Hungary, Norway, Belgium, Romania, Croatia, Ireland, Greenland, Thailand, Sweden, Italy, Australia, and South Africa. A further 41 videos did not mention a specific country.

Overall, the results show that the dominant pattern is the simplification of complex political and social issues. Most of the videos assigned responsibility to identifiable social groups on the basis of individual misconduct rather than institutional or policy-related factors. Such an approach frames certain social groups as the primary cause of misconduct or socio-economic problems, causing friction between the host nation and immigrants instead of attributing broader government or policy failure.

4.2 Threats constructed around immigrants

The findings indicate that immigration is predominantly constructed as a symbolic rather than a material or administrative threat. Based on the analysis, videos mostly address symbolic threats (89 videos; 74.2%), while realistic threats were found in only 31 videos (25.8%). Highly engaged videos are more likely to frame immigrants as a threat to demography, national identity and cultural values than as a threat to economic stability or a source of policy-related problems.

Videos concerning changes in lifestyle and demographics appear to provoke the most nostalgia and irritation, blaming immigrants for displacing a previous social and demographic order. Notably, the majority of videos do not address a single problem explicitly but touch on several symbolic issues and reinforce them within one clip.

Economic- and policy-related videos appeared less frequently in the dataset. This leads the study to conclude that immigration is generally presented as a social and cultural problem in short videos rather than as an administrative, economic, or policy-related problem that requires longer policy discussions.

Zhang’s (2025) research provides empirical evidence that political memes on TikTok measurably increased engagement among cynical viewers during the 2024 U.S. election, with effects moderated by humour and political cynicism. If humour and simplification through memes can increase political engagement independent of substantive content, then similar dynamics may increase engagement with the short-form, low-intensity, mocking videos that simplify complex social and economic issues into an immigrant-as-threat narrative. This research suggests that engagement with this type of content could plausibly contribute to friction between local people and immigrants, and to moral justification for hostility or violence towards immigrants. However, this suggestion remains a theoretical extension beyond Zhang’s (2025) findings and would benefit from direct empirical testing in future work.

4.3 Illustrative examples from the dataset

Exemplar 1 — Visual anecdote, textual overlay, realistic threat, dismissive affect through disappointment

The video shows a littered historical site in Frankfurt with a text overlay implying that unregulated Muslim immigration is responsible for it. The creator does not make an explicit claim on camera; instead, he expresses surprise at the state of the site while the footage does the persuasive work. This was coded as a realistic threat since the framing centres on visible harm, namely litter in a public space, rather than on cultural, demographic or identity-based concerns. It was coded as disappointment, given the resigned, mildly surprised tone rather than a mocking or ironic delivery. Notably, comments on the video suggested the litter was in fact left behind after a local football match rather than being connected to immigration, illustrating how a visual anecdote can imply causation that the footage itself does not establish.

Exemplar 2 — Visual anecdote, textual overlay, stitch, symbolic threat, dismissive affect through sarcasm

The video stitches two clips together, showing a crowd of visibly diverse ethnic backgrounds gathered for New Year’s Eve fireworks in Sydney. A text overlay frames the scene as a challenge to spot an “Aussie” in the crowd, implying that the visible ethnic diversity represents a loss of dominant white demography. In this exemplar as well, the creator does not state this claim directly; the combination of the footage and caption performs the persuasion. This was coded as a symbolic threat since the concern centres on perceived demographic change rather than concrete safety or economic harm. Given the actively mocking tone challenging the viewer to find a “real Aussie”, this video was categorised as sarcasm.

Exemplar 3 — Reused media clip, textual overlay, symbolic threat, dismissive affect through sarcasm

The video reuses a romantic scene from “Top Gun”, paired with a text overlay reframing the moment as a call to “outbreed” foreigners to protect national identity, invoking a demographic-replacement narrative. This was coded as a symbolic rather than a realistic threat, since the underlying concern is demographic change rather than concrete economic or safety harm. It was coded as sarcasm given the humorous, meme-like repurposing of an unrelated media clip, which frames an extreme claim as a knowing joke rather than an urgent warning. This exemplar illustrates how dismissive affect can carry genuinely extreme ideological content such as demographic-replacement rhetoric, precisely because the humorous packaging makes the claim register as a joke rather than as an explicit political argument, reinforcing the paper’s central point that this register’s danger lies in how it evades being read as hostile.

4.4 How threats are proven and persuaded

Analysis of the videos showed that 44 (36.7%) videos used visual anecdotes as the main form of evidence, followed by 24 (20.0%) textual overlay claims and 20 (16.7%) combined visual anecdote with text overlay. Influencer statements appeared in 12 videos (10.0%), and elite or authority statements in only six (5.0%). The remaining evidence types included interview or anecdotal testimony (four videos; 3.3%), comparison clips, videos providing no supporting evidence, visual anecdotes combined with elite or voiceover speeches (two videos each; 1.7%), and several isolated formats, each represented by a single video (0.8%).

The popularity of the visual anecdote and textual overlay claim shows that creators seek to persuade through selective framing that appears to prove their point, combined with their own statements as textual overlay claims. Captions do more to guide viewers towards particular understandings of the visual material than extended discussion or analysis of the problem. This encourages causal inference and biased interpretation based on captions.

One of the main findings concerned the dominant register across the videos in the dataset. Contrary to the expectation formed by the previous studies that anti-immigrant communication primarily relies on fear or anger, dismissive affect was the most frequently used affective register.

As shown in Figure 2, dismissive affect was the most frequently coded register, of which sarcasm accounted for 36 videos (30.0%), followed by disappointment 22 (18.3%) and irritation 19 (15.8%). Pride and defiance were expressed in only 15 (12.5%), fear in 11 (9.2%), nostalgia in eight (6.7%), anger in six (5.0%), and moral certainty in three (2.5%). Original videos were mainly expressed through the dismissive affective register: sarcasm (23/79; 29.1%), irritation (18/79; 22.8%) and disappointment (13/79; 16.5%).

Although the number is small, eight of 120 videos (6.7%) were AI-generated. Their dominant affective register was also sarcasm, in six of the eight AI-generated videos (75.0%). Use of AI on this topic may increase in future, given how easily such content can be produced without presenting real evidence.

Dismissive affect (see Appendix 1), comprising lower-intensity registers such as sarcasm, disappointment and irritation, accounted for 77 of 120 videos (64.2%) and was used to persuade rather than the sharper, fear- or anger-based persuasion found in 17 videos (14.2%). As explained by Zulli and Zulli (2022), sarcastic content strengthens in-group bonding through ridicule. Content that makes fun of situations, and dismissive “common sense” disappointment framing, subtly provokes certain emotions rather than stating something sharply enough to provoke disapproval in the audience.

Similarly, Matamoros-Fernández (2023) states that users who are not necessarily trying to spread hateful ideologies or threaten democratic values may nonetheless participate online in ways that can harm others. The findings of the study also support this interpretation by showing how microaggressions towards immigrants, expressed through humorous sarcasm, normalise anti-immigration narratives and subtle exclusion while remaining less explicit than open hate speech.

4.5 Stitch-to-frame mechanism

Beyond the primary research questions, the analysis found that video formats on TikTok are not merely descriptive but shape how evidence is presented. The dismissive register is substantially produced by the stitch feature. This suggests a platform-based influence on how symbolic threat narratives are communicated and reinforced.

Almost all stitched videos, 25 of 28 (89.3%), provide the audience with footage relevant to symbolic threats. In addition, 19 of 28 stitched videos (67.9%) rely heavily on visual anecdotes as proof alone.

These findings suggest that the stitch format enables creators to reinterpret existing videos by combining selective visual material with additional commentary or textual framing. Considering this, stitched videos frequently guide viewers towards particular interpretations of the original content by reinforcing symbolic threat narratives.

4.6 Remedy and polarisation in videos

Almost all of these videos include polarisation narratives: an us-versus-them framing appears in 114 of 120 videos (95.0%). From this, the study concludes that in these types of videos polarisation is conveyed regardless of emotion and persuasive format.

Remedies were rarely offered: only 24 of 120 videos (20.0%) proposed any kind of remedy. Out of the offered remedies, 9 of 24 (37.5%) proposed racialised boundaries based on racial or third-country exclusions, and 2 of 24 (8.3%) offered non-actionable or implausible remedies. The remaining 13 (54.2%) proposed restrictive policy measures, electoral solutions, or integration-oriented remedies.

The absence of remedies fuels perceptions of crisis rather than encouraging discussion of practical policy responses. These narratives primarily construct immigration as an ongoing social crisis by offering limited opportunities for contextualisation or solution-oriented conversation.

5. Discussion

To answer RQ1, the research found that attribution is directed towards individuals rather than policies and broader patterns. This demonstrates a preference for causal simplification rather than framing immigration-related issues as the result of governance decisions and institutional capacity. These frames are combined with demographic and cultural identity change, contributing to polarisation and to the moral justification of symbolic boundaries between groups without necessarily addressing policies or contextual reasons.

Another finding is that the scarcity of remedies contributes to polarisation in this form of short-video persuasion: only 24 videos offered any remedy, of which 13 proposed policy-based remedies. Almost all videos create morally justified us-versus-them narratives. This leads to the conclusion that anti-immigration videos are frame packages that encourage a particular negative ideology towards immigration based on causal attribution and emotionally charged moral judgement, lacking deeper contextual analysis of the problems.

In regard to RQ2, the study analysed videos from the perspective of the dominant register, looking at how these immigration-related issues are affectively presented. The dominance of sarcasm shows that narratives against immigrants often strengthen in-group bonding through ridicule (Zulli & Zulli, 2022) and limit openness to different interpretations of the same situation. Disappointment also creates affective agreement and ties common problems to a narrative of decline, blaming immigrants alone by simplifying the issue.

This finding runs counter to an assumption in the previous literature on anti-immigrant content online, where anger and fear are typically treated as the primary affective drivers of hostility (Cartes-Barroso, García-Estévez, & Méndez-Muros, 2025; Manik & Tarisayi, 2025). On the contrary, in this dataset, the dominant affective register is dismissive: sarcasm, disappointment and irritation. The prevalence of dismissive registers in TikTok’s most engaged immigration content suggests that these videos are not trying to provoke alarm about a concrete danger but to consolidate a shared, faintly amused certainty that immigration represents an obvious, almost self-evident decline. Because they do not present open and intense hostility, these affective registers may be harder for platforms to moderate and for audiences to consciously resist. This is consistent with the possibility raised in Section 2.4: rather than simply replicating the anger and fear-dominant pattern documented in the anti-immigrant social media literature, TikTok’s highly engaged immigration content offers a different, more dismissive affective register for the same attribution and threat structure.

To answer RQ3, based on van Dijck’s (2013) account of platform affordances, the research finds that TikTok’s short-form affordances and platform power favour and reward affectively charged content. Among the analysed videos, visual anecdotes and textual overlays show that such forms can serve as proof in this environment. A short clip with casual textual overlays can serve as proof and increase engagement. In this sense, platform-compatible evidence forms support the dominant patterns found by the research, namely attribution to immigrants and simplified explanation, aligning with causal narratives that involve minimal contextualisation.

Another notable finding was the relationship between stitches and symbolic threat framing. Stitched videos allow creators to combine several clips out of context as proof, to curate selectively, and sometimes to supply commentary related to identity- and culture-based social crisis narratives. Selective curation of videos tends towards symbolic framing of immigration-related problems. A further finding is the presence of highly engaged AI-generated videos. Although these appeared less frequently, they may signal more frequent future use of AI for effective messaging with a low production and contextual burden.

6. Conclusion

This research analysed how immigrants are framed as scapegoats in the TikTok environment, especially the short-form video environment, through framing, affective persuasion and platform affordance. The study’s main contribution is presenting sarcasm, disappointment and irritation as a dominant affective register rather than the anger and fear which was typically assumed to dominate anti-immigrant content online. This distinction matters for identifying and addressing anti-immigrant content. These patterns are drawn from one algorithmically curated recommendation environment and should be read as a case study rather than a general claim about TikTok content as a whole (see Section 3.5).

Beyond its theoretical contribution, the research findings raise questions for practical consideration in the field of platform governance, regulatory policy, digital manipulation and media literacy. The findings show that digital manipulation in media platforms may not necessarily spread through false information but through information that was taken out of context. This itself shows that there is a regulatory challenge in these types of selective digital manipulation, as this type of content does not violate misinformation standards but still manipulates audiences towards social polarisation.

For policymakers, this shows a need for definitional language targeting selective decontextualisation. It should explain (a) that the underlying material is genuine rather than fabricated, (b) that it has been presented in a way that materially misrepresents its original context, and (c) that this misrepresentation supports a claim the source material does not itself prove. The difficulty in drafting such a definition is preventing its use against legitimate satire, commentary, or journalism, which routinely recontextualise existing material for critical purposes. The definition would need to focus on the recontextualised material itself instead of the act of recontextualisation, to avoid encroaching on protected forms of freedom of speech.

For platforms, it indicates the need for contextualisation mechanisms, such as the ability to flag stitched videos in which the caption claims something the original footage does not show. This should be based on factual checking of whether the claim matches the source material, rather than on a casual judgement of hostility.

Finally, for the media literacy practitioners, it shows the importance of informing the audience about dismissive affect. A dismissive affect module can teach people to notice when a calm, joking tone is being used to make a radical claim seem obvious and less hostile, and to check whether a stitched video shows real evidence or merely repurposed footage presented as evidence.

The findings of this study are timely for research on affective narratives and platform affordances, building on earlier indications of systematic patterns. This could be especially useful for political communication research examining how platforms may contribute to the growth of certain patterns while limiting the others. The findings also offer early insights into how anti-immigration discourse is shaped on digital platforms where rapid emotional engagement is valued over conventional policy discussion.

Appendix 1. Operational definitions of coding variables

* Unlike the other emotions listed above, sarcasm is coded here as a dismissive affective register that can co-occur with other emotional content (e.g., anger, disappointment) rather than as a mutually exclusive discrete emotion. See Section 3.3 for full diagnostic criteria and boundary conditions.

* Besides the abovementioned categories, only two videos also called for the protection of European, white, and Christian identity together by attributing immigrants’ existence as a violation or oppression of these values.

Appendix 2. Seeding protocol

The newly opened account was seeded by using hashtags #immigration, #migrants, and #migration. During the seeding stage engagement maintained through only by watching 30 videos with these hashtags; no videos were liked, shared, commented on, or saved, and no creators were followed. After watching videos, formal sampling started as TikTok’s recommendation system began to suggest similar content through the For You feed. Videos were registered based on the study’s inclusion criteria (English-language content, immigration-related topic, and more than 2,000 likes).

Ethics Declaration

This study analysed publicly available TikTok content. No personal data were collected, no content creators were contacted, and no analysis of usernames, individual profiles, or creator identities was undertaken. Creator names, profile details, and appearance were not recorded, and no material identifying individual creators is reproduced in this article. On this basis the research did not require review by an institutional ethics board. The study follows established guidance on the ethical conduct of qualitative social media research on sensitive topics.

Funding Declaration

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of Competing Interest

The author has no conflicts of interest to declare that are relevant to the content of this article.

Declaration of Generative AI and AI-assisted technologies in the writing process

During the preparation of this work, the author used Claude to improve language and readability and to check for grammatical correctness. The author reviewed and edited the content as needed following these suggestions and takes full responsibility for the content of the publication.

References