First Monday

Exploring political topics that connect to antisemitism on Twitter: U.S. midterms Pennsylvania gubernatorial race 2022 by Gabrielle Beacken

Online antisemitism rears its head during times of political instability, especially during contentious elections. This paper uses the 2022 Pennsylvania gubernatorial governor race as a case study to examine what political topics spur antisemitic expression on Twitter. Both rich political and antisemitism discourses are offered by this election that featured a Jewish Democrat candidate Josh Shapiro and far-right Republican candidate Doug Mastriano. Over 800 tweets were qualitatively analyzed through critical discourse analysis to understand both Shapiro’s political tweets and the antisemitic reply tweets. This study addresses the gap of what online political contexts leads to instances of antisemitism on Twitter during political elections. Grounded in antisemitism and political propaganda literature, this study shows that particularly controversial topics of abortion rights and extremism in politics led to the highest amount of antisemitic expression. Dominantly, Jewish political control conspiracies were used as simple explanations for particularly turbulent political topics.


Literature review




Antisemitism is not a new prejudice. Thousands-years-old conspiracies and ideologies are consistently repackaged by actors across the political spectrum to fit the modern socio, economic and political tribulations of a particular society and time. Especially when contentious political contests are at play and a political controllability is unclear, for example the 2016 U.S. presidential election and the 2018 U.S. midterms election, antisemitism rears its familiar head through a significant online spike, especially on social media platforms (Woolley and Joseff, 2018; Zannettou, et al., 2020). Online antisemitism experiences a hefty increase during offline political events, and yet, the online political dialogues that lead to this increase are notably under-analyzed.

To fill this gap, this study explores the political and antisemitic discourse on the platform X (formerly Twitter) surrounding an electoral race to connect political tweets with their antisemitic replies. X activity within the lead up to Pennsylvania’s governor race in the 2022 U.S. midterm elections was used as a case study to understand and explore political linkages to antisemitism. This election and data collection processes preceded the limitations set on the X API and before new company ownership in October 2022 (Conger and Hirsch, 2022). Therefore, the time period analyzed is before investigated increases in hate speech — including antisemitic rhetoric — due to content moderation layoffs and laxed expression policies at X (Lima-Strong, 2023).

This election revealed an intriguing political landscape due to the nominated Democrat and Republican candidates: Democrat Josh Shapiro — a publicly religious Jewish candidate who incorporated his faith into his campaign program — and Republican Doug Mastriano — a candidate accused of espousing antisemitic ‘dog whistles’ and promoting Christian nationalism. Investigating the social media political conversations around this contest produced a vast political and antisemitism discourse. Furthermore, this study situates its results within the existing literature surrounding offline political elections generating online antisemitism.

A holistic framework is adopted to investigate, one, the antisemitic reply tweet and, two, the original crafted political message within the leadup to the 2022 gubernatorial governor race in Pennsylvania. Due to the overt religious tones and sharp partisan differences between the two candidates, this race produced a rich antisemitism and political discourse. Critical discourse analysis — informed by popular antisemitic political conspiracies and highly circulated propaganda (Allington, et al., 2021) — is used to identify, classify, and analyze both political and antisemitic content. Results of this analysis show that hotly contested political topics — including abortion rights and extremism — spurred the most antisemitic replies that primarily expressed political conspiracy, indicating a reaction to political uncontrollability. The results of this study reflect the analysis of 806 tweets. This number reflects reply and original tweets collected and evaluated.

Studies have investigated antisemitism within political parties or actors (Gaber, 2021), however, there is a gap in studying political discussion that possibly connects, or eventually leads to, antisemitic expression on social media platforms. Antisemitism has been lengthily studied within fringe or extreme groups (Zannettou, et al., 2020), on both the far left and right of the political spectrum (Staetsky, 2017), but this study aims to examine how antisemitism arises in daily political conversations on the mainstream social media app X.

By understanding political hotbeds for antisemitism, communities, policy-makers, and elected officials may sufficiently prescribe and implement solutions that effectively combat online antisemitism within political arenas. Furthermore, this study aims to extract overarching political topics and circumstances that can help identify and predict political derivations that spark platformed antisemitism. In a modern digital ecology where content moderation efforts are limited and fallacious, it is even more imperative for researchers and practitioners to understand the particularities of online hate speech. Hate speech and antisemitic rhetoric are symptoms of political and social turmoil (Kofta, et al., 2020). Qualitative analysis provides a critical lens to investigate these political and social topics that connect to antisemitic speech. In order to fully understand and challenge antisemitic rhetoric, understanding political and social trends of modern society is required.



2. Methodology

The Pennsylvania governor race & antisemitism

Religion was oftentimes at the forefront for both political candidates in this gubernatorial governor race. Democrat candidate Josh Shapiro keeps kosher as well as Shabbat, and discussed how Jewish values influenced his political ideologies (Glueck, 2022b). Republican candidate Doug Mastriano conveyed his support for Christian values, yet expressed his resentment towards a separation of church and state with a desire to create a Christian state of Pennsylvania (Dias, 2022).

While both candidates exhibited strong religious identities, they were notable instances of Mastriano and his close ties negatively invoking Shapiro’s Jewish identity. Mastriano sparked alarm in the Jewish community when he repeatedly criticized Shapiro for sending his children to Shapiro’s alma mater, a “privileged, exclusive, elite” school, claiming this shows “disdain for people like us” (Glueck, 2022b; Lapin, 2022). Jewish groups took Mastriano’s descriptions of the school as antisemitic ‘dog whistling’ since Shapiro sends his kids to a Jewish day school. A dog whistle describes phrases or words used in messages that appear to be neutral to an overall audience, but connect with a specific audience that understands its intended meaning (Langer, 2022). For example, instead of using the word ‘Jewish’, Mastriano replaces with ‘elite’ and ‘privileged.’ Mastriano also engaged in ‘othering’ — emphasizing a sharp divide between groups as a tool for debasement — by expressing that people from this ‘elite’ school are different from ‘people like us’.

Mastriano also circulated the conspiracy that famous American-Hungarian Holocaust survivor George Soros — a liberal billionaire who donates to progressive causes around the world — was a Nazi collaborator (Glueck, 2022b). Mastriano has also accused Shapiro of having a “real grudge” against the Catholic Church within discussions surrounding abortion (Smith, 2022). Conspiracies of Soros and fabrications about Shapiro play an important role in this political election, both on and offline. It is significant that common online antisemitic tropes — Soros conspiracy, ‘othering’ Jews, dog whistles — appeared within Mastriano’s governor campaign.

Mastriano’s connection to the far-right platform Gab was also was called into question, especially due to popular usage of the platform by the shooter who killed 11 congregants at the Tree of Life Synagogue in Pittsburgh, Pa. in October 2018 (Roose, 2018). Mastriano’s ties to Gab include pay-outs for campaign consulting and a joint interview with Gab’s founder Andrew Torba (who later told outlets that neither he or Mastriano would speak to non-Christian journalists). Torba defended Mastriano’s connection to Gab saying, “We’re not bending the knee to the 2 percent anymore,” a reference to the approximate percentage of the Jewish population in the United States (Gabriel, 2022; Homans, 2022). At that time, Mastriano refused to comment on accusations of antisemitism against Torba or the platform (Gabriel, 2022).

Antisemitism: Othering, dog whistles, and conspiracies

The conceptual underpinnings of antisemitism are key to the critical analysis of antisemitic expression in political settings. Understanding how antisemitism [1] is expressed through othering, conspiracies and dog whistling allows the antisemitic discourse towards political candidate Josh Shapiro to be effectively examined. The prejudice against a person simply because they are perceived as or identify as Jewish is expressed in numerous forms. This study primarily inspected antisemitism through 1) othering, 2) conspiracies, and 3) dog whistling.

Othering is a process “labelling and defining those who we perceive to be in a different group as somehow ‘deviant’” [2]. The goal of othering is to create stark divisions that allow the dominant group to look down on the subordinate group, particularly within ethnic, religious, and racial circles (Modood and Thompson, 2022). Othering appears in several forms within the discourse of antisemitism. For example, the concept ‘dual loyalty’ refers to a person’s two distinct interests that may conflict with one another (American Jewish Committee, 2021). This is common for American Jews who are perceived to have a disloyalty to the U.S. because of a perceived loyalty to Israel (World Jewish Congress, 2022).

‘Othering’ can also be used to demonize or dehumanize Jewish people. This dehumanization was showcased through Nazi propaganda and statements made by Nation of Islam’s leader Louis Farrakhan that labelled Jews as ‘vermin’ or ‘termites’ that need to be exterminated (American Jewish Committee, 2021; Smith, 2011). Today on social media, images of reptiles, tentacled-men and other ‘monsters’ are illustrated to ‘other’ Jews as subhuman, which allows antisemites and others to justify the prejudice against Jews (Antisemitism Policy Trust, 2020).

One effective way to spread antisemitic propaganda in support of or against an actor or movement is to fabricate consensus via conspiracies (DiMaggio, 2022). A political conspiracy is defined as a “claim, often without evidence, that powerful coalitions (for example, elites from an opposing party) coordinate in secret to deceive the rest of the population to its own benefit” [3] meaning, a public’s deception is blamed at the hands of a secret, disliked group. Therefore, a common topic throughout political conspiracies is ‘scapegoating’, which describes shifting the blame onto outcast groups to present a solution to a problem (Debney, 2020). Recently, the COVID-19 pandemic was the subject of several antisemitic conspiracies involving scapegoating, such as the claim that COVID-19 was a Jewish hoax (Community Security Trust, 2022).

Furthermore, antisemitic conspiracies have traditionally been embraced as a justification for “all the bad things happening in society” [4]. There are three key components, composed by Kofta, et al. (2020), that attempt to illuminate why antisemitic conspiracies are so buoyant throughout centuries. One, the ‘ego defense’, discusses how individuals handle an increasingly uncontrollable world by attempting to restore order and build self-esteem through identifying enemies to blame for the problems of the day (Kofta, et al., 2020). Secondly, ‘intentional collectives’, tells how these disparate groups who feel helpless band together to promote conspiracy theories (Kofta, et al., 2020). Thirdly, ‘cognitive exhaustion’, asserts that since people jump to the simplest answers to explain an ever-complicating world, they have a weakened intellectual capability to understand their surrounding realty (Kofta, et al., 2020). An understanding of the numerous political conspiracies of antisemitism aids the recognition and measurement of antisemitic content on social media.

Dog whistles, or political dog whistles, refers to the “use of messages embedded in speeches that seem innocent to a general audience but resonate with a specific public attuned to receive them” [5]. Meaning, the intended audience will understand the message — in this case antisemitism — but the general listenership or readership will not detect anything untoward. For example, on social media one may see the terms ‘Jewish cabal’ or ‘Jewish lobby.’ On the surface these words may not seem antisemitic, but when ‘Jewish cabal’ is used to imply ‘Jewish control’ or when ‘Jewish lobby’’is suggesting ‘Jews control politics’, then its antisemitism can be detected (American Jewish Committee, 2021). Furthermore, dog whistling can be even more obscure. For example, when Mastriano called Shapiro’s alma mater ‘elite’, he was evoking the antisemitic ‘clan’ conspiracy through the dog whistle of ‘elite.’ Due to its obscurity, an understanding of antisemitic dog whistles is vital to recognizing and understanding antisemitism on social media.

Social media and antisemitism: Platformed antisemitism

The concept platformed antisemitism is central to this study because it allows the governor race’s social media, political and antisemitic discourse to be investigated and understood in a technological framework. The presence of online antisemitism has increased with the rise of online networks and platform affordances allow for the rapid spread of information by all actors (Woolley and Joseff, 2018). Platformed antisemitism is defined as, “the particularities of antisemitic discourse, ridden with conspiracy theories, prejudice, stereotypes, Holocaust denial and more veiled forms of hatred aimed at Jews, amplified and moving within and through platforms, on the fringes and in the mainstream” [6]. Platformed antisemitism focuses on the unique characteristics of antisemitic discourse and how the affordances of social media platforms intensify its dissemination. Yet, platformed antisemitism emphasizes the strong existence of antisemitism within mainstream social and political discussions. Within this study, platformed antisemitism will be focused on to address research questions concerning antisemitic conspiracies and prejudice.

Propaganda, political events, and antisemitism on social media

Modern researchers have defined propaganda as a “deliberate, systematic attempt to shape perceptions, manipulate cognitions, and direct behavior to achieve a response that furthers the desired intent of the propagandist” [7]. Propaganda is not the exchange of ideas, but rather a sender attempting to persuade others effectively with a particular message, in this instance, hatred against Jews. Due to social media’s participatory environment, users are not only encouraged to consume information, but also to create new or promote existing information. An active audience’s ability to modify and share propagandists’ messages is a key element within a “participatory propaganda model” (Wanless and Berk, 2020). This model extends off a traditional definition of propaganda to include the propagandist’s desire to invite audiences to dynamically spread this strategic communication (Jowett and O’Donnell, 2019; Wanless and Berk, 2020).

While typically a propaganda model speaks to a one-way form of communication — where the flow of propaganda starts with an active sender and ends with a passive, targeted receiver — social media platforms have transformed this top-down communication to be multi-directional (Jowett and O’Donnell, 2019). Through the usage of social media features that modify, amplify, and disseminate messages, targeted audiences may unintentionally spread messages they relate to that were intentionally created by propagandists (Wanless and Berk, 2020). It is important to note that the existence of platforms themselves have enabled the quick speed and scale of propaganda campaigns. Users may take advantage of platform features, such as hashtags, sharing, commenting and more, to organically circulate propaganda and hate speech online (Woolley and Joseff, 2018). Platforms offer the chance for human actors to organically create and circulate content while also offering automated actors to take advantage of the virality of social media. Bots — or simple code scripts that dictate automation of iterative functions like liking, commenting, and more — is one example of a pervasive automated tool that is specifically deployed on social media platforms to achieve political objectives (Woolley and Howard, 2018).

As a technological infrastructure, social media platforms enable technological innovations — such as tools of artificial intelligence — to be custom applied to platform machinations. While these innovative applications may be used for normative positively values, such as efforts to remove harmful content targeting children on the Internet (Bulger, et al., 2017), they are also used for nefarious purposes by political actors, such as deploying propagandistic campaigns targeting marginalized communities (Freelon, et al., 2020; Riedl, et al., 2024).

Spreading false information — in this instance antisemitic propaganda — on social media is a tool used by propagandists and active participatory users alike to discredit and delegitimize political opponents (Dehghan and Glazunova, 2021). As political conspiracies have demonstrated, in times of uncertainty or lack of control, antisemitic conspiracies arise (Galliford and Furnham, 2017). Political elections offer opportunities for change — some desired, some feared. When emotions run high, politics turns divisive, and voters and politicians alike are searching for a group to blame for the pitfalls of life (Kofta, et al., 2020).

It is important to note that the repercussions of offline discourse onto online spheres has been well documented in media and communication literature, especially connecting offline political events to platformed political expression (Bucy, et al., 2020; Doroshenko and Lukito, 2021). Political online and offline spheres do not exist in isolated vacuums but rather are constantly evolving and shaping one another (Bossetta, et al., 2023; Woolley, 2023). There are computationally deduced direct lines from offline events to online fervor. Whether it be from the spoken ideology and language by public individuals or news reported of international conflict, an influx in engaged actors and user responses on social media platforms is likely to occur (Bucy, et al., 2020; Doroshenko and Lukito, 2021). Political expression on social media does not exist in a void but rather is an extension of offline political discourse.

This study attempts to bridge the online political discourse of X with the offline governor election because researchers have discovered a link between offline political events and online spikes in the spread of antisemitism (Zannettou, et al., 2020). According to Zannettou, et al.’s (2020) computational textual analysis, after the 2016 presidential election, Donald Trump’s inauguration, and the “Unite the Right” rally in Charlottesville, Va., the number of antisemitic messages on social media substantially increased, in some cases doubled, on the fringe Web community Gab (Zannettou, et al., 2020). Ozalp, et al.’s (2020) machine learning analysis around antisemitic tweets supported Zannettou, et al.’s (2020) finding of offline political events triggering spikes in online antisemitism on X, including the offline event of a far-right demonstration in the Jewish neighborhood Golders Green in London. This migration from offline events to online antisemitism is not restricted to a fringe social media app, like Gab, but extends to traditionally considered mainstream social media app X.

Studies leading up to and during the 2018 midterm elections also show spikes of antisemitism on X (Riedl, et al., 2024; Woolley and Joseff, 2018). Particularly for Jewish journalists on X, online antisemitic harassment in the lead up to the 2018 midterms was shown as a coordinated response: either by automated bots or trolls (Woolley and Joseff, 2018). During the 2018 midterm elections, X affordances of tweets and hashtags allowed for the spike of Jewish global conspiracy fantasies; antisemitic QAnon propaganda, antisemitic Trump/MAGA support, and demonization of famous Jewish figures George Soros and the Rothschild family (Riedl, et al., 2024). This study aims to extend explorations of antisemitism during political seasons with the Pennsylvania governor race case study.

Based on the analysis of platformed antisemitism, propaganda and connection between offline political events and online antisemitism, the following research questions, situated in the 2022 Pennsylvania governor race, aim to address:

RQ1: What political subjects are connected to different categories of antisemitism on the platform Twitter?

RQ2: How are the connections between political subjects and categories of antisemitism placed within the existing discourse antisemitism on social media during political elections?




Critical discourse analysis

Critical discourse analysis (CDA) was used in this study because of its ability to analyze a political discourse in relation to antisemitic discourse. In this study, the broader historical, social, and political contexts — antisemitism, politics, and social media — were examined to fully address the research questions. Additionally, CDA aims to tackle intricate social problems, such as antisemitism, which may be investigated through social, historical, and political perspectives (Fairclough, 2012; Wodak, 2011). CDA sheds light on social inequalities by examining the social contexts in which a discourse in situated and how these social elements directly contribute to the reproduction of a dominating force (van Dijk, 1993). Here, the social context of social media, the political context of the governor race in Pennsylvania, and the historical context of antisemitism were all considered to address how and which political topics connect to platformed antisemitism.

CDA investigates and reveals how tools of domination seek to sway political and societal ideologies (Mullet, 2018). Van Dijk (1993) defines dominance as a social power, which can be wielded by institutions, elites, and more, whose outcome is social inequality. One of the ways to solidify this dominance is through the manufacturing of consensus, which may lead to the acquiescence and credibility of this dominance (Herman and Chomsky, 1988). Critically, this study analyzes the tools of this domination through antisemitic attacks against a political actor. Furthermore, the tools to craft an accepted form of dominance were investigated through the categorization and analysis of antisemitic tropes of dog whistling, othering, and conspiracy.


Data was purposively pulled in the months leading up to election day: 1 September to 8 November 2022. This timeframe was determined to collect a sufficient and accurate data sample for critical analysis. Events concerning major political elections are not constrained to a singular election day. Rather, key campaign rhetoric, decisions, and newsworthy issues occur within the preceding months. September was chosen as the start date due to an intense increase of campaign coverage within mainstream news and social media platforms. Therefore, September onwards constituted the most salient period to examine heated campaign discussions on platform X. In December 2022, X’s API developer tool was utilized to extract relevant tweets within the months leading up to the Pennsylvania governor election day. A keyword list composed of 40 terms was used to filter the X sample (see Appendix A). This set of keywords was influenced by lists in previous studies of predictive antisemitic keywords (Ozalp, et al., 2020; Riedl, et al., 2024). The primary focus was on keywords relevant to antisemitic conspiracy theories (American Jewish Committee, 2021). For example, common markers of antisemitic conspiracies ‘globalist’, ‘shill’, and ‘Soros’ were adopted, as well as campaign-relevant keywords such as ‘Jewish school’ (discussed earlier). A second filter was also used to gather the data. All reply tweets included a mention of Democrat candidate Josh Shapiro. These filters created a data set that contained antisemitic messaging towards Democrat candidate Josh Shapiro.

In total,806 tweets (reply tweets + original tweets) were examined in this study. A simple breakdown of this tweet sample was required to fully understand tweet content and and assist in interpretation. The tweet sample of reply tweets included both antisemitic and non-antisemitic tweets (n=552). However, since this research study explored the linkages between political subjects and antisemitism, only antisemitic reply tweets were analyzed and interpreted within this sample (n=254). Furthermore, this research study examined antisemitic reply tweets’ original tweets (n=254). Yet, since this study is situated within an antisemitism and political discourse, only the explicitly political original tweets scribed by official Josh Shapiro accounts were examined (n=221). The investigation of antisemitic reply tweets and their original political tweets were reflected within this study’s results (see Appendix B for a breakdown of the data collection process).

Measuring antisemitism

Aligned with the literature review, tweets were deemed antisemitic based on the attributes of the three outlined categories: othering, dog whistling, or conspiracy. Othering was categorized as a term or phrase that expressed demonization, dehumanization, division, or deviance. Conspiracy was selected when a term or phrase connected Judaism to scapegoating, secret plots, global/political/financial dominance, and elitism. Lastly, dog whistling required a two-step operationalization. First, familiarization with the existing critical literature surrounding political dog whistling was required. Secondly, dog whistling was categorized as a specific term or phrase that evoked an antisemitic trope or conspiracy but was not further elaborated upon or clear from context. Due to this operationalization, dog whistling and conspiracy were often paired.




Emergent themes


Table 1: Categories, themes, and frequency in tweet sample.
Note: * More than one theme can be assigned to one Tweet; ** Political subjects with < 2.0 percent frequency in pop. are not shown in this table. See Appendix C for full list.*** Percentages reflect proportion of total antisemitic reply tweets (n=254); **** Percentages reflect proportion of total political original tweets (n=221).
Core categoryThemePercentage of theme frequency
Antisemitic expression***Conspiracy95.3
 Dog whistle87.0
Themes within antisemitic tweets***Intention to control & deceive62.6
 Reason for bad things happening44.1
 Ideological demonization15.3
 Political extremism8.6
 Suspicion of Jewish people6.3
 Antisemitic anti-Zionism1.2
 Holocaust imagery0.8
 The abortion Holocaust0.4
Conspiracy type***George Soros83.1
 Jewish global domination/power15.0
 Mark Zuckerberg2.0
Political subject****Attacking opponent: Mastriano40.3
 Abortion rights23.5
 Extremism in politics22.6
 Election day19.5
 Worker rights/unions11.3
 Crime and prosecution10.4
 LGBTQ+ rights5.9
 Pro/saving democracy5.4
 Voting rules4.5
 January 6 insurrection/election denying4.5
 Political accountability4.5
 Environment/climate change4.1
 Individual rights/freedom for all4.1
 Political endorsements3.6
 Attacking opponent: Trump3.6
 Small business2.7
 Police and community2.3


Antisemitic expression

RQ1, asking how political subjects were linked to categories of antisemitism, was first approached by understanding the frequency of both main categories of antisemitism and political subjects within the analyzed sample.

Overwhelmingly, the greatest category of an antisemitic reply tweet was conspiracy, appearing in 95.3 percent of reply tweets. The most popular conspiracy category was George Soros — a liberal Jewish billionaire positioned at the center of a global Jewish financial and political conspiracy — appearing in 83.1 percent of reply tweets (Glueck, 2022b). For X users unaware of the accusations made against Soros — for example, blame for rise in crime, acting as a global puppet master — this term could simply denote a person’s name. Soros is often used as a dog whistle to signal to fellow conspiracists a belief in a world domination conspiracy. Dog whistling appeared within 87 percent of reply tweets due to its frequent coupling with conspiracy. It is important to note that the mere mention of Soros was not deemed antisemitic. When Soros was used as a symbol to convey Jewish power, wealth, or control — especially the control of Jewish political figures — then this counted as an antisemitic trope. For example:

‘Soros’ Conspiracy of Control
@JoshShapiroPA Clean up Philly, DA. Your revolving door system of “justice” is pathetic. Being a pawn for @georgesoros & @AlexanderSoros MUST PAY OFF since you’re running for governor now to adhere to MORE of their wishes on how society is run. You are NOT going by the law, but NWO handlers

The notion of a global Jewish domination conspiracy made up 15 percent of antisemitic reply tweets primarily through keywords ’NWO’, ‘New World Order’, and ’Globalist’. Here in these NWO examples the theme of secret control and obedience to a concealed power with ulterior motives is prevalent and coded as such. For example:

‘New World Order (NWO)’ Jewish Conspiracy of Global Dominance
@JoshShapiroPA Look at this political hack, wanting to murder your babies for the NWO depopulation agenda

The third category, Othering, appeared the fewest times with appearances in 23.6 percent of reply tweets. A found ‘dual loyalty’ trope within Othering was Josh’s alleged obedience to Soros versus his duty to the citizens of Pennsylvania. Meaning, Shapiro was framed as a candidate who did not care for or understand the concerns of Pennsylvania constituents due to his supposed prioritization of non-Pennsylvanian political actors. For example:

@JoshShapiroPA Josh takes orders from George Soros! Josh works for Soros not the people of Pennsylvania!
@JoshShapiroPA Shapiro is already screwing Pennsylvanians over by refusing to debate Mastriano.
Little big man only cares about climbing the ladder to become the first Jewish President ... he’ll following orders from Soros and destroy PA to get there.
@JoshShapiroPA This is Philly, do something. You can’t even be tough on your NWO handler that makes you be SAWFT on crime. Where are these drugs these people are strung up on, coming from, Josh?

Here, the employment of othering was used to discredit Shapiro as a reliable political candidate. Specifically, the accusation that Shapiro is disloyal to Pennsylvanians because of his loyalty to Soros also engages themes of conspiracy and dog whistling.

Themes within antisemitic tweets

Efforts to answer RQ1 attempted to, first, understand extant antisemitic and political discourse around the election for governor of Pennsylvania and, secondly, draw connections between the two discourses. Identifying and interpreting related themes within antisemitic reply tweets aided efforts to address RQ1.

A popular related theme, appearing in 44.1 percent of reply tweets, was ‘Reason for bad things happening (violence, crime).’ Concern over rising crime in Philadelphia and Shapiro’s former position as Attorney General of Pennsylvania contributed to this category’s high popularity. However, what was distinct about many of these tweets was not simply the blame (passivity) on Shapiro for an increase in crime, but an accusation of his participation (active) in this trend. Here, Shapiro’s mere connection to Soros was posed as a viable explanation for a rising crime rate in Philadelphia. Alleging active participation in the destruction of Pennsylvanian safety extended beyond a mere critique or scepticalness of his connection to Soros. For example:

But people like Fetterman #Josh Shapiro would have to go after them, and they aren’t doing that. Because flooding Philly’s streets, all PA streets with crime, vioIent criminaIs is part of George Soros’ agenda for America. And both Fetterman #Shaprio are Soros-backed puppets.
@JoshShapiroPA Why is PA’s top law enforcement official trying to distract voters from the real issues such as violent crime in PA’s largest city, Philadelphia? This is what can be expected from a George Soros funded @PAAttorneyGen !

Here, Shapiro is accused of intentionally raising crime in Philadelphia because, in this instance, it is at the will of his globalist master Soros. This belief that bad things are occurring not only because Shapiro wills it, but his ‘master’ directs it, is showcased through the most popular related theme appearing in 62.6 percent of reply tweets, ‘Intention to control and deceive.’ This control and deception plays into aforementioned popular global Jewish domination conspiracies, i.e., Soros, globalist, New World Order. The third and fourth most popular categories, ‘Ideological demonization’ (15.3 percent) and ‘Political extremism’ (8.6 percent) also demonstrated the popularity of Othering through assigning an inherent evilness and differentiation to a Jewish candidate.

Political subjects

Attacking opponent. Efforts to address RQ1 involved identifying the original political tweet political subjects with the highest frequency among antisemitic reply tweets. Attacking opponent (Republican candidate Doug Mastriano) was the most popular political tweet categorization appearing in 40.3 percent of political original tweets. While each tweet may be categorized by more than one political subject (see ‘relationships’ below), this attack category appeared most frequently. For example:

“God said marriage is between man and woman. What’d Obama say? No, I have another idea.” — Doug Mastriano. These are Doug Mastriano’s words criticizing President Obama for finally allowing millions of Americans to marry who they love. Shameful.
Doug Mastriano just took the stage with Dr. Oz and Donald Trump to spread election lies, undermine our democracy, and shine a national spotlight on his dangerous extremism He is showing us who he is. Let’s show him who we are, Pennsylvania.

As this Pennsylvania governor race was particularly ideologically combative, it is unsurprising that Shapiro’s tweets to his followers exhibited condemnation of his opponent. This too follows a trend of hyper partisanship and antagonistic nature of the 2022 midterm political elections (Epstein, 2022).

Abortion rights and extremism in politics. The subsequent most popular political categories, Abortion rights (23.5 percent) and Extremism in politics (22.6 percent) shed further light on the high political combativeness of this election. Taking account for the sharp policy and ideological divide of abortion rights policy between Shapiro (for) and Mastriano (against) and the 2022 overturn of Roe v. Wade, it is unsurprising that the controversial topic of Abortion rights was frequently discussed by Shapiro. For example:

Doug Mastriano has made clear: the single most important issue of his campaign is banning abortion outright, with absolutely no exceptions. He’s not focused on our economy, workforce, or defending our freedoms. He’s only focused on forcing his extremist agenda on Pennsylvania.
I am the only candidate for Governor of Pennsylvania who will keep abortion legal.

Discussing his stance on abortion rights was a key tactic for Shapiro to separate himself from his opponent Mastriano; both in terms of legal and ethical matters. With popular reporting of Mastriano’s connections to far-right platform Gab, Christian Nationalism and election denying (Glueck, 2022b), Shapiro also emphasized the Extremism in politics phenomenon. For example:

I am Josh Shapiro, Democratic nominee for Governor of Pennsylvania. Tomorrow is Election Day. And I’m running against an extremist looking to take our state back decades. Retweet this tweet and give me a follow to help us get our message out this last day.

References to and reminders of Mastriano’s extremism was a continuous theme throughout Shapiro’s X campaign.

Emergent patterns

Links between political subjects

Over 59.5 percent of antisemitic reply rweets were spurred by political original tweets including more than one political topic (see popular political pairings in Appendix D). The popularity of linking multiple political subjects fleshes out the question posed in RQ1 asking which political subject combinations are most connected with antisemitic tweets. Attacking opponent was a consistent politically paired topic within this sample. Top associations included Attacking opponent & Abortion rights (33.9 percent) and Attacking opponent & extremism in politics (27.8 percent). Antisemitic reply tweets were sparked by Attacking opponent tweets that traversed the range of political issues.

Out of Abortion rights’ original political tweets containing more than one subject, Abortion rights & Attacking opponent appear the most frequently together (55.7 percent). A common political tactic for Shapiro was to one, criticize Mastriano’s rejection of abortion rights, and two, affirm his own support for abortion rights. For example:

I am pro-choice‚ which means I believe women deserve the right to make decisions about their own bodies. Doug Mastriano disagrees, he wants to ban abortion outright, no exceptions for rape, incest, or the life of the mother. These are the facts.
Doug Mastriano’s extreme abortion ban would force companies to move elsewhere because employees don’t want to work in states where their rights are limited. His anti-women, anti-union, and anti-workforce values will destroy our economy if we elect him Governor of Pennsylvania.

Shapiro disseminated his pro-abortion rights messaging by incorporating additional political topics he supported into the same tweets, including Abortion rights & Workers rights/unions (17.1 percent), Abortion rights & crime and prosecution (15.7 percent) and Abortion rights & Pro/saving democracy (11.4 percent). For example:

Abortion rights & Crime and prosecution
As Attorney General, I’ve fought to put predator priests, drug dealers, and scammers behind bars to protect the good people of Pennsylvania. Doug Mastriano wants to throw doctors behind bars for performing live-saving abortions for women in Pennsylvania.
Abortion rights, Workers rights/unions & Pro/saving democracy
I’m pro-choice, pro-democracy, and pro-union, and I’ve fought to protect the reproductive and workers’ rights of every Pennsylvanian. Doug Mastriano has pledged to ban abortion outright and undermine the union way of life. He doesn’t reflect our values.

In addition to the frequency and diversity of Abortion rights original political tweets, the category Extremism in politics was also repeatedly used and associated with varying political issues within multi-subject tweets: Extremism in politics & Attacking opponent (82.1 percent) and Extremism in politics & Attacking opponent: Trump (17.9 percent). Shapiro consistently emphasized Mastriano’s extremist ties and accused him of being too ‘dangerous’ or ‘extreme’ for Pennsylvania. For example:

You have to ask yourself: what makes Doug Mastriano so appealing to white supremacists, antisemites, and extremists? He speaks their language and empowers their hate and he’s recruiting them to join his campaign. He’s too dangerous to be governor of Pennsylvania.
In 2020, Donald Trump sued to throw out our votes, and I took him to court to ensure every ballot was counted. Now, he's backing an extreme candidate for Governor of Pennsylvania, hoping he’ll carry out his attack on our democracy. He won’t.

Antisemitic tweets prevalently stemmed from Shapiro’s condemnation of Mastriano; whether it be a condemnation of his stance on abortion rights (including criminality), an accusation of extremism, criticizing his relationship to Trump or the reiteration that Mastriano is anti-union.

Antisemitism and political linkages


Connections between main antisemitic categories and political categories
Figure 1: Connections between main antisemitic categories and political categories. Note. Numbers represent frequency count. Darker colors represent higher frequencies.


The objective of RQ1 was to understand the relationships between antisemitic categories and political subjects. As a result of this analysis, the highest correlations were Conspiracy & Attacking opponent (38.2 percent), Dog whistling & Attacking opponent (35.5 percent), Conspiracy & Abortion rights (22.7 percent), and Conspiracy & Extremism in politics (20.7 percent).

Conspiracy appeared most frequently within topics of extremism, abortion, and attacking opponents; either through naming George Soros and stating his control over Shapiro and the subsequent deceitfulness of Shapiro’s actions or through the accusations of Shapiro partaking in a globalist New World Order that seeks to ascertain worldwide power. Within hot-button issues of extremism, abortion and Mastriano’s contentious reputation, antisemitic conspiracies flourished. Since existing antisemitic discourses tell that conspiracies and dog whistling are activities closely linked, it is logical that Dog whistling & Attacking opponent were highly popular connections.

Related themes and political linkages


Relationship between antisemitic related themes and political categories
Figure 2: Relationship between antisemitic related themes and political categories. Note. Percentages reflects percent of related theme within political category tweets.


As discussed earlier, antisemitic tweets were prevalently linked with political topics of extremism in politics, abortion rights, and attacking Republican opponent Mastriano. The relevacy of ‘Intention to control and deceive’ and ‘Reason for bad things happening’ reflect the widespread usage of conspiracy in this sample. Conspiracies were used to both explain perils — ‘Reason for bad things happening’ — and blame perils on an individual, in this case Shapiro — ‘Intention to deceive and control.’

Additionally, antisemitic attacks in response to Abortion rights invoked the related theme ‘Ideological demonization’ (30.8 percent) (i.e., perversion, evilness, murder, body penetration). The comparison is stark when viewing this related theme’s overlap with other popular political subjects: Crime and prosecution (17.4 percent), Extremism in politics (16.0 percent), and Attacking opponent (11.2 percent). Surrounding the criticism of Mastriano’s dismissal of abortion rights and stating Shapiro’s own pro-choice stance was commonly met with accusations of a natural deviance. For example:

REPLY: All you have is your nonsense abortion rhetoric. You are a Soros puppet, pushing to sexualize grade school kids and abortion up to birth. You are human scum.
ORIGINAL: Doug Mastriano wants to charge women who have abortions with murder. He’s too extreme to be governor, and he’s too dangerous to represent the women of Pennsylvania.

REPLY: The difference is YOU support GUN CONFISCATION, Child sex change surgical mutilation, affirmative action, anti-White rhetoric, anti-Energy independence, you are an America Last Globalist Scumbag. Oh Michael Bloomberg, George Soros co-own you! #WhinyJoshMamasBoyWimp
ORIGINAL: I’m the pro-freedom, pro-choice, pro-democracy candidate for Governor of PA. Doug Mastriano is the candidate who would take away your freedoms to love and choose and vote, and surrounds himself with white supremacists and conspiracy theorists. That’s the difference.

Here, ideological demonization of sexualizing young children goes hand in hand with issues of abortion and the conspiracy of Soros control. Original political tweets of Abortion rights were met with a particular type of ideological ferocity that was not as present in other antisemitic reply tweets.

Interestingly, as stated earlier, Shapiro’s attacks of Doug Mastriano popularly surrounded his declaration that Mastriano was too ‘extreme’ and ‘dangerous’ to serve as governor of Pennsylvania. These accusations of extremism against Mastriano led to claims of political extremism against Shapiro himself. The related theme ‘political extremism’ displayed a high correlation with political subjects Extremism in politics (14.0 percent) and Attacking opponent (12.4 percent). For example:

REPLY: @JoshShapiroPA is a do-nothing, Soros-backed, Regime following, ultra-Left, Socialist AG who intends to destroy PA the same way Biden has destroyed America. Doug Mastriano is *extremely* different and is a *dangerous* challenger to Shapiro’s stupidity.
ORIGINAL: Doug Mastriano is a dangerous extremist. Spread the word.

REPLY: Not k*llin babies is extreme? How about the Left’s New World Order? The war on energy, censorship/arrests of the opposition, Stasi-DOJ, inflation, food shortages, shut/lockdowns, forced masks, forced vaxxes, racial/gender indoctrination, grooming and mutilating kids?
ORIGINAL: Fear and extremism have no home in Pennsylvania.

Accusations of political extremism were launched from both ends: Shapiro to Mastriano, and Mastriano supporters to Shapiro. The idea of being an extremist — outside the desired norm — was a tool used by both sides to discredit political opponents.




Connecting political subjects and antisemitism

Antisemitic replies to Democrat Josh Shapiro correlated most highly with political topics Attacking opponent, Extremism in politics, and Abortion rights. The midterm governor race in Pennsylvania was seeped in discussion of extremism (Glueck, 2022a) and the potential legal and policy repercussions of the U.S. Supreme Court’s June 2022 decision on Roe v. Wade (Epstein, 2022). Politicians were indebted to Pennsylvanians to describe their political path in a post-Roe era, which exacerbated an already highly partisan race. Typically, a Shapiro political original tweet condemning Mastriano for his rejection of abortion rights and accusations of extremism sparked the largest number of antisemitic replies.

Antisemitic categories of Conspiracy and Dog whistle were most prevalent, compared to the reply tweet makeup of Othering. This breakout of antisemitic categories may indicate, one, the ease of employing conspiracies or dog whistles in a tweet versus the description of ‘othering’ or, two, the difficulty in the gathering and subsequent analysis of antisemitic tweets that denote ‘othering.’ Antisemitic conspiracies appeared in over 95.3 percent of all reply tweets; within those reply tweets, the George Soros conspiracy was invoked over 83.1 percent of the time. Conspiracy was most highly correlated with Attacking opponent, Abortion rights, and Extremism in politics, but its presence was felt across 30 political categories: from LGBTQ+ rights and Environment/climate change to Voting rules and Small business. Conspiracies, specifically the Soros conspiracy invoking global Jewish domination, was particularly frequent amongst Extremism in politics, Abortion rights, and Attacking opponent, yet its usage was seemingly politically universal. This study demonstrated that despite the political topic at hand, conspiracies of Soros and Jewish global control were consistently asserted.

The categorization of related themes reveals that many antisemitic reply tweets were tailored to the political discussion at hand. Reply tweets invoking ‘ideological demonization’ — i.e., baby murderer, Satan Soros, pedophilia — were most commonly used in response to tweets about abortion. Abortion rights original political tweets typically condemned Mastriano and often included Shapiro’s supportive stance. The strong ideological and religious undercurrents of Abortion rights was reflected in these antisemitic replies. Additionally, accusations of Mastriano’s extremism were most popularly met with antisemitic reply tweets related to Extremism in politicsi.e., fascism, radical socialism, Marxism. Adopting the popular tool of extremist-calling was used to fight fire with fire.

Connections to political and antisemitism discourses

This study’s answer to RQ2, how thematic connections discovered fit into existing literature surrounding online antisemitism and offline political elections, revealed both affirmations and extensions. As past research has demonstrated, political contests — especially contentious ones — spark an increase in antisemitism on social media (Ozalp, et al., 2020; Riedl, et al., 2024; Woolley and Joseff, 2018; Zannettou, et al., 2020). This study reaffirmed the results of these earlier studies around antisemitism’s presence in certain elections. An openly Jewish candidate who incorporated his faith into his campaign and a far-right opposition candidate accused of dog whistling, offered X users a prime and convenient outlet for expressing antisemitism. This case study — a Jewish candidate and Far Right opponent — provided a rich political and social discourse for critical investigation.

This study corroborated the popular use of the dog whistle and conspiracy notion of George Soro’s global domination, yet, the results indicated a much higher frequency than previously explored. This study’s findings contrast with past research whose results reflected a Soros conspiracy presence, but not an overwhelming domination (Allington, et al., 2021; Riedl, et al., 2024; Woolley and Joseff, 2018). For decades, the conspiracy of the Jewish European Rothschild family was employed as an antisemitic dog whistle for global Jewish domination (Allington, et al., 2021). However, within this study’s sample, ‘Soros’ appeared within 84.6 percent of antisemitic tweets and ‘Rothschild’ appeared in one instance. This noted absence speaks to a transition from ‘Rothschild’ to ‘Soros’ as a key dog whistle for antisemitic conspiracies.

Investigations into antisemitism’s persistence throughout time suggest that people turn to this prejudice during times of uncontrollability in an ever-complicating world (Kofta, et al., 2020). The popular political categories that sparked the most antisemitic replies revolved around divisiveness and uncontrollability. Within the recent U.S. political arena, abortion rights has been a prominent and contested issue, particularly after the 2022 U.S. Supreme Court’s overturn of Roe v. Wade (Chemerinsky, 2022). Furthermore, discussions of fear and accusations surrounding extremism within the political sphere has sharply heightened since the 2016 presidential election of Donald Trump (Baker, 2022). Two paramount political issues of the day — abortion and extremism — played a vital part in this study’s investigation into antisemitism. As the literature denoted, antisemitism promptly appears during turbulent societal, economic, and political moments. This Pennsylvania governor race was situated in a turbulent political time, thus acting as an optimal host for antisemitic discourse.

Within the three most prominent political subjects — Abortion rights, Extremism, Attacking opponent — a common topic was revealed: control. Furthermore, the urge to return to simple, quick answers in times of political uncontrollability supported the popularity of Jewish conspiracies found in this study (Kofta, et al., 2020). Whether it be control over reproduction, control over defining parameters of mainstream politics, or control over the power of adversaries, users are employing repackaged antisemitic tropes towards an openly Jewish candidate in response to perceived threats over political control. Due to the heightened loss of control and cognitive exhaustion that derives from the flood of information online, social media provides an apt environment for the spread of conspiracy. With these factors in mind, it is notable that in this instance of political instability, antisemitic conspiracies flourished.




Guided by CDA, the political, historical, and social context of antisemitic and political discourses on X surrounding the 2022 Pennsylvania governor race were explored and interpreted. The popular political subjects of abortion rights and extremism in politics and the prolific use of antisemitic conspiracies reflected the turbulent nature of this political race and the ease to fall back on simple, quick explanations for an ever-complicating political world. The hot button political issues of the day — extremism, abortion, attacking adversaries — coupled with sharp political divisiveness between two candidates led to the highest linkage of antisemitic attacks against a Jewish political candidate in this study. It appears that when political contests heat up offline, antisemitic conspiracies are fueled online.

As an old and ever-present prejudice, the discourse of antisemitism is naturally entangled with a high variety of important issues, events, and phenomenon throughout political and social history. Antisemitic expression and tropes are constantly repackaged to reflect modern machinations but the underpinning themes remain constant: conspiracy of secret control, relegation to ‘otherness’, and ideological villainization. An examination of the antisemitic rhetoric appearing within the X discourse of Josh Shapiro’s election reflect all three of these categories. The names, places, and times may change but the underlying antisemitic message does not. It is imperative to pinpoint and fight against the adaptive phenomenon of antisemitism within the political realm, as it continuously rears it heads in both familiar — but critically temporally varied — ways.




This qualitative study focuses on a particular case study on the platform X, therefore the results of this paper cannot be generalized to all scenarios. Crucially, the results of this paper indicate that the mostly hotly debated, socio-relevant issues of the day are connected to increased bouts in antisemitism online, particularly in the adoption and evolution of Jewish conspiracy theories. Researchers and policy-makers monitoring these spaces may be better equipped to predict, identify, categorize, and understand online antisemitism within politically controversial spaces. This is key — as noted in this paper — because antisemitism often is detected through contextual means, i.e., the usage of conspiracies and dog whistling. Studying online antisemitism qualitatively digs deeper into online antisemitic spaces that draw on the historical and present socio-political contexts that continuously shape antisemitic expression. End of article


About the author

Gabrielle Dora Beacken (M.Sc., London School of Economics and Political Science) is a Ph.D. student in the School of Journalism and Media at The University of Texas at Austin (UT Austin) and a graduate research assistant at the Propaganda Research Lab at the Center for Media Engagement.
ORCID: 0009-0007-4987-877X
E-mail: Gabrielle [dot] Beacken [at] utexas [dot] edu



1. The International Holocaust Remembrance Alliance (2020) coined a working definition of antisemitism: “Antisemitism is a certain perception of Jews, which may be expressed as hatred toward Jews. Rhetorical and physical manifestations of antisemitism are directed toward Jewish or non-Jewish individuals and/or their property, toward Jewish community institutions and religious facilities.”

2. Van Der Zwet, et al., 2020, p. 517.

3. Marie and Petersen, 2022, p. 1.

4. Kofta, et al., 2020, p. 903.

5. Langer, 2022, p. 221.

6. Riedl, et al., 2024, p. 2,231.

7. Jowett and O’Donnell, 2019, p. 7.



D. Allington, B.L. Buarque, and D. Barker Flores, 2021. “Antisemitic conspiracy fantasy in the age of digital media: Three ‘conspiracy theorists’ and their YouTube audiences,” Language and Literature, volume 30, number 1, pp. 78–102.
doi:, accessed 25 April 2024.

American Jewish Committee, 2021. “Translating hate glossary: Stopping antisemitism starts with understanding it,” at, accessed 25 April 2024.

Antisemitism Policy Trust, 2020. “Antisemitic imagery and caricatures,” at, accessed 25 April 2024.

P. Baker, 2022, “Trump embraces extremism as he seeks to reclaim office,” New York Times (1 December), at, accessed 25 April 2024.

M. Bossetta, A. Dutceac Segesten, and D. Bonacci, 2023. “Reconceptualizing cross-cutting political expression on social media: A case study of Facebook comments during the 2016 Brexit referendum,” Political Communication, volume 40, number 6, pp. 719–741.
doi:, accessed 25 April 2024.

E.P. Bucy, J.M. Foley, J. Lukito, L. Doroshenko, D.V. Shah, J.C. Pevehouse, and C. Wells, 2020. “Performing populism: Trump’s transgressive debate style and the dynamics of Twitter response,” New Media & Society, volume 22, number 4, pp. 634–658.
doi:, accessed 25 April 2024.

M. Bulger, P. Burton, B. O’Neill, and E. Staksrud, 2017. “Where policy and practice collide: Comparing United States, South African and European Union approaches to protecting children online,” New Media & Society, volume 19, number 5, pp. 750—764.
doi:, accessed 25 April 2024.

E. Chemerinsky, 2022. “Abortion is about to dominate American politics like never before,” Time (27 June), at, accessed 25 April 2024.

Community Security Trust, 2022. “Coronavirus and the plague of antisemitism,” at, accessed 25 April 2024.

K. Conger and L. Hirsch, 2022. “Elon Musk completes $44 billion deal to own Twitter,” New York Times (27 October), at, accessed 25 April 2024.

B.M. Debney, 2020. The oldest trick in the book: Panic-driven scapegoating in history and recurring patterns of persecution. Singapore: Palgrave Macmillan.
doi:, accessed 25 April 2024.

E. Dehghan and S. Glazunova, 2021. “‘Fake news’ discourses: An exploration of Russian and Persian tweets,” Journal of Language and Politics, volume 20, number 5, pp. 741—760.
doi:, accessed 25 April 2024.

E. Dias, 2022. “The far-right Christian quest for power: ‘We are seeing them emboldened’,” New York Times (8 July), at, accessed 25 April 2024.

A.R. DiMaggio, 2022. “Conspiracy theories and the manufacture of dissent: QAnon, the ‘big lie’, Covid-19, and the rise of rightwing propaganda,” Critical Sociology, volume 48, number 6, pp. 1,025–1,048.
doi:, accessed 25 April 2024.

L. Doroshenko and J. Lukito, 2021. “Trollfare: Russia’s disinformation campaign during military conflict in Ukraine,” International Journal of Communication, volume 15, pp. 4,462–4,689, and at, accessed 25 April 2024.

R.J. Epstein, 2022. “Doug Mastriano, a far-right 2020 election denier, is Pennsylvania Republicans’ choice for governor,” New York Times (17 May), at, accessed 25 April 2024.

N. Fairclough, 2012. “Critical discourse analysis,” In: J.P. Gee and M. Handford (editors). Routledge handbook of discourse analysis. London: Routledge, pp. 9–20.

D. Freelon, M. Bossetta, C. Wells, J. Lukito, Y. Xia, and K. Adams, 2020. “Black trolls matter: Racial and ideological asymmetries in social media disinformation,” Social Science Computer Review, volume 40, number 3, pp. 560–578.
doi:, accessed 25 April 2024.

I. Gaber, 2021. “Why the EHRC Report into anti-semitism in the Labour Party is about so much more than anti-semitism in the Labour Party — A personal view,” Political Quarterly, volume 92, number 1, pp. 139–141.
doi:, accessed 25 April 2024.

T. Gabriel, 2022. “Doug Mastriano faces criticism over his backing from antisemitic ally,” New York Times (27 July), at, accessed 25 April 2024.

N. Galliford and A. Furnham, 2017. “Individual difference factors and beliefs in medical and political conspiracy theories,” Scandinavian Journal of Psychology, volume 58, number 5, pp. 422–428.
doi:, accessed 25 April 2024.

K. Glueck, 2022a. “Pennsylvania stakes its claim as center of the political universe,” New York Times (2 September), at, accessed 25 April 2024.

K. Glueck, 2022b. “Mastriano’s attacks on Jewish school set off outcry over antisemitic signaling,” New York Times (10 October), at, accessed 25 April 2024.

E.S. Herman and N. Chomsky, 1988. Manufacturing consent: The political economy of the mass media. New York: Pantheon.

C. Homans, 2022. “The extremely online rise of Doug Mastriano,” New York Times (31 August), at, accessed 25 April 2024.

International Holocaust Remembrance Alliance, 2020. “What is antisemitism?” at, accessed 25 April 2024.

G.S. Jowett and V. O’Donnell, 2019. Propaganda & persuasion. Seventh edition. Thousand Oaks, Calif.: Sage.

M. Kofta, W. Soral, and M. Bilewicz, 2020. “What breeds conspiracy antisemitism? The role of political uncontrollability and uncertainty in the belief in Jewish conspiracy,” Journal of Personality and Social Psyhcology, volume 118, number 5, pp. 900–918.
doi:, accessed 25 April 2024.

A. Langer, 2022. “Deep state, child sacrifices, and the ‘plandemic’: The historical background of antisemitic tropes within the QAnon movement,” In: M. Hübscher and S. von Mering (editors). Antisemitism on social media. London: Routledge.
doi:, accessed 25 April 2024.

A. Lapin, 2022. “GOP candidate makes opponent’s Jewish day school an issue in PA governor race,” Times of Israel (17 September), at, accessed 25 April 2024.

C. Lima-Strong, 2023. “Antisemitic tweets soared on Twitter after Musk took over, study finds,” Washington Post (20 March), at, accessed 25 April 2024.

A. Marie and M.B. Petersen, 2022. “Political conspiracy theories as tools for mobilization and signaling,” Current Opinion in Psychology, volume 48, 101440.
doi:, accessed 25 April 2024.

T. Modood and S. Thompson, 2022. “Othering, alienation and establishment,” Political Studies, volume 70, number 3, pp. 780–796.
doi:, accessed 25 April 2024.

D.R. Mullet, 2018. “A general critical discourse analysis framework for educational research,” Journal of Advanced Academics, volume 29, number 2, pp. 116–142.
doi:, accessed 25 April 2024.

S. Ozalp, M.L. Williams, P. Burnap, H. Liu, and M. Mostafa, 2020. “Antisemitism on Twitter: Collective efficacy and the role of community organisations in challenging online hate speech,” Social Media + Society (18 June).
doi:, accessed 25 April 2024.

M.J. Riedl, K. Joseff, S. Soorholtz, and S. Woolley, 2024. “Platformed antisemitism on Twitter: Anti-Jewish rhetoric in political discourse surrounding the 2018 US midterm election,” New Media & Society, volume 26, number 4, 2,213–2,233.
doi:, accessed 25 April 2024.

K. Roose, 2018. “On Gab, an extremist-friendly site, Pittsburgh shooting suspect aired his hatred in full,” New York Times (28 October), at, accessed 25 April 2024.

A. Smith, 2022. “Six days, 25 stops: Josh Shapiro sprints across Pennsylvania with Democrats’ hopes along for the ride,” NBC News (3 November), at, accessed 25 April 2024.

D.L. Smith, 2011. “‘Less than human’: The psychology of cruelty,” National Public Radio (20 March), at, accessed 25 April 2024.

D. Staetsky, 2017. “Antisemitism in contemporary Great Britain,” Institute for Jewish Policy Research (17 September), at, accessed 25 April 2024.

A. Van Der Zwet, M.S. Leith, D. Sim, and E. Boyle, 2020. “Brexit, Europe and othering,” Contemporary Social Science, volume 15, number 5, pp. 517–532.
doi:, accessed 25 April 2024.

T.A. van Dijk, 1993. “Principles of critical discourse analysis,” Discourse & Society, volume 4, number 2, pp. 249–283.
doi:, accessed 25 April 2024.

A. Wanless and M. Berk, 2020. “The audience is the amplifier: Participatory propaganda,” In: P. Baines, N. O’Shaughnessy, and N. Snow (editors). Sage handbook of propaganda. London: Sage.
doi:, accessed 25 April 2024.

R. Wodak, 2011. “Critical linguistics and critical discourse analysis,” In: J. Zienkowski, J.-O. Östman, and J. Verschueren (editors). Discursive pragmatics. Philadelphia: John Benjamins, pp. 50–70.
doi:, accessed 25 April 2024.

S. Woolley, 2023. Manufacturing consensus: Understanding propaganda in the era of automation and anonymity. New Haven, Conn.: Yale University Press.
doi:, accessed 25 April 2024.

S. Woolley and P.N. Howard (editors), 2018. Computational propaganda: Political parties, politicians, and political manipulation on social media. New York: Oxford University Press.
doi:, accessed 25 April 2024.

S. Woolley and K. Joseff, 2018. “Computational propaganda, Jewish-Americans and the 2018 midterms: The amplification of anti-semitic harassment online,” Anti-Defamation League (26 October), at, accessed 25 April 2024.

World Jewish Congress, 2022. “Antisemitism defined: Allegations of dual loyalty” (23 February), at, accessed 25 April 2024.

S. Zannettou, J. Finkelstein, B. Bradlyn, and J. Blackburn, 2020. “A quantitative approach to understanding online antisemitism,” Proceedings of the Fourteenth International AAAI Conference on Web and Social Media (ICWSM 2020), pp. 786–797.
doi:, accessed 25 April 2024.


Appendix A: Keyword list

Keywords in Table A1 show keywords that were used to pull data. Keywords that did not appear in the sample are not included in this table.


Table A1: Keyword list with frequency count and percent of antisemitic reply tweets. Percentage reflects percentage of all antisemitic reply tweets.
jewish school20.79
New World Order51.97



Appendix B: Data collection process

Figure B1 illustrates how many tweets were originally collected, sampled, and processed.


Antisemitic reply tweets and political original tweets examination
Figure B1: Antisemitic reply tweets and political original tweets examination reflected in this study’s Results section.



Appendix C: Full category, theme, and frequency table

Categories, themes, and frequency of themes within the X sample are shown in Table C1.


Table C1: Tweets analyzed and interpreted through categories and themes informed by existing literature of political, historical, and antisemitism discourse.
Note: * More than one theme can be assigned to one tweet; ** Political subjects with < 2.0 percent frequency in pop. are not shown in this table; *** Percentages reflect proportion of total antisemitic reply tweets (n=254); **** Percentages reflect proportion of total political original tweets (n=221).
Core categoryThemePercentage of theme frequency
Antisemitic expression***Conspiracy95.3
 Dog whistle87.0
Themes within antisemitic tweets***Intention to control & deceive62.6
 Reason for bad things happening 44.1
 Ideological demonization15.4
 Political extremism8.7
 Suspicion of Jewish people6.3
 Antisemitic anti-Zionism1.2
 Holocaust imagery0.8
 The abortion holocaust0.4
Conspiracy type***George Soros83.1
 Jewish global domination/power15.0
 Mark Zuckerberg2.0
Political subject***Attacking opponent: Mastriano40.3
 Abortion rights23.5
 Extremism in politics22.6
 Election day19.5
 Workers rights/unions11.3
 Crime and prosecution10.4
 LGBTQ rights5.9
 Pro/saving democracy5.4
 Voting rules4.5
 January 6 insurrection/election denying4.5
 Political accountability4.5
 Environment/climate change4.1
 Individual rights/freedom for all4.1
 Political endorsements3.6
 Attacking opponent: Trump3.6
 College /education3.2
 Small business2.7
 Police and community2.3
 COVID-19 /vaccinations/masking1.4
 Minimum wage1.4
 Gun rules/gun violence0.9
 Racial justice0.9
 Attacking opponent: GOP0.9
 Children/the future0.9
 Drug epidemic0.5



Appendix D: Multi-subject appearances in original political tweets

Political-subject pairings are illustrated in Figure D1, with an emphasis on the three most popular political categories of Attacking opponent, Abortion cover, and Extremism in politics.


The three most frequent political topics – Abortion rights, Attacking opponent, and Extremism in politics – were typically connected to additional political subjects
Figure D1: The three most frequent political topics — Abortion rights, Attacking opponent, and Extremism in politics — were typically connected to additional political subjects. Percentages represent the percent of total multi-subject tweets for each individual political category.



Editorial history

Received 21 November 2023; revised 2 February 2024; accepted 25 April 2024.

Creative Commons License
This paper is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Exploring political topics that connect to antisemitism on Twitter: U.S. midterms Pennsylvania gubernatorial race 2022
by Gabrielle Beacken.
First Monday, Volume 29, Number 5 - 6 May 2024