Click here to read our latest report “Transmisogyny, Colonialism and Online Anti‐Trans Activism Following Violent Extremist Attacks in the US and EU”

Examining Support for the 2023 Brazilian Coup Attempt: Gender Stereotypes and Occupational Trends in a Radicalised Digital Ecosystem

Examining Support for the 2023 Brazilian Coup Attempt: Gender Stereotypes and Occupational Trends in a Radicalised Digital Ecosystem
21st August 2023 Rosana Pinheiro Machado


On 8 January 2023, Brazil witnessed the most threatening coup attempt in its recent democratic history. Inspired by the January 6th Capitol attack in 2021, Jair Bolsonaro supporters invaded the headquarters of the Executive Branch, Supreme Court, and the National Congress in Brasilia to contest President Lula da Silva’s victory and claim for military intervention. The riots resulted in 1,138 detentions and caused more than $5 million of damages as the rioters destroyed public property and vandalised national historical treasures—the damage incurred by the latter is immeasurable. 

Data from the Brazilian Federal Prosecution Chief showed that 60% of the people who were detained by the police received the cash transfer welfare benefit Auxilio Emergencial, which was given during the pandemic. This indicates a strong presence of low-income people, informal workers, and ‘micro entrepreneurs’ within the rioting mob. Among the detained group, 60%  were men between the ages of 36 and 55. Data from Brasilia police suggested that 67% of the detained people were between 40 and  59 years old, and 3.5% were older than 60. A recent report by the Brazilian Intelligence Agency (ABIN) revealed that these groups planned a series of attacks, including blocking energy transmission towers and forming paramilitary cells in a bid to overthrow the elected President.

While much is known about the social strata of far-right supporters around the world, there remains a gap in our understanding of the vocations of radicalised individuals,  disaggregated by gender. As a research group, we have been exploring the connections between labour precarity, gender stereotypes, and far-right mobilisation. This Insight describes how the three elements played a role in the 8 January attack in Brazil, offering preliminary findings into 89,865 accounts that supported the coup by posting active information about the events and amplifying calls to action. We distinctly focus on Twitter bios and personal websites displayed on user profiles, specifically their self-identifications and occupations, disaggregating the results by gender. The primary objective of the research is to gain insight into how the attacker ecosystem operates and how ordinary people who are engaged in such a network display their online persona.

 A Massively Male Ecosystem 

The accounts we observed materialised from 2018 onwards, coinciding with Jair Bolsonaro’s election as president. Since then, we identified three major peaks in the emergence of new accounts, indicating online mobilisation between April 2022 and January 2023. The first was a day after Elon Musk announced the purchase of Twitter, which also happened in between two massive pro-Bolsonaro moto rallies, in March and May 2022). A second peak occurred during the Brazilian election in October 2022. The final peak was in November, after the defeat of Jair Bolsonaro. Many users proudly proclaimed that they were onto their second or third accounts because their primary ones had been removed by Twitter for violating community standards.

A total of 5,181 bios displayed links to additional information about the users. A qualitative analysis of the links to external websites indicated that several accounts used personal websites like Blogspot, where people share writings, poems, art creations, and worldviews. We also noted a proliferation of semi-professional or outdated websites of local news radios, town newspapers, or small businesses. Many self-employed workers had amateur websites or Instagram accounts offering their services as manicurists or electricians,  for instance. 

The overall semantic analysis confirmed what is already known about Bolsonarism. Our studies show that these accounts are all a part of a homogenous ecosystem that circulates content replete with conservative rhetoric and dis/misinformation, in their efforts for mutual identification and contagious effect that produces virality. The bios largely defined users as conservative, right-wing, faithful to Christian God, patriotic, pro-gun, and devoted to the traditional family. ‘God’ is the sole most recurring word in their bios (9.2 %). In the ecosystem, 4.5% of accounts define themselves as ‘anti-PT’ (Workers’ Party, the party of the current President Lula da Silva) and ‘anti-communist’. This does not include several bios that expressed the same view in other words, i.e., ‘PTphobic’. 

The compelling results, however, emerged through the breakdown based on occupation and gender. Adhering to the anti-gender stance of the group, we worked with a binary frame for gender. There was no evidence of gender diversity in the realm of bios and related websites. We identified the gender of 77% of accounts. Among them, we encountered a substantial male presence, with 85% of the users identifying as male. 

The same political figures and outlets impacted both male and female accounts. The most retweeted politicians, whom we call spreaders, were male politicians Jair Bolsonaro, his son Eduardo Bolsonaro, and the congressmen Nikolas Ferreira and Gustavo Gayer. The news channels that received the most retweets were Terra Brasil Notícias, Revista Oeste, and Jornal da Cidade Online, all of them on the edge of mainstream media and subjected to demonetisation and/or content removal by recent court rulings or online platform sanctions for disseminating disinformation. 


In the datasets organised by gender, 24.4% of male accounts and 19.4% of female accounts disclosed their professional occupations. They often mentioned more than one occupation, for instance, teacher and entrepreneur. As a result, the numbers below represent the total occupation mentions.

Men define themselves more by their public activities than women. These included a diverse range of occupations and interests, mainly football and music. Following a tradition of Bolsonaro’s guru, the far-right thinker Olavo de Carvalho, they generally wanted to show their artistic and athletic hobbies or secondary/amateur occupations, demonstrating intellectual power. Typical male bios would be ‘conservative, engineer, businessman, and a writer in my free time’ or ‘patriot, entrepreneur and amateur historian/philosopher/astrophysicist’. 

The most prevalent career sectors in the male dataset were: business, finance, and trade (18%), engineering (12%), law (10%), education (10%), administration and management (7%), and a wide range of other activities, including security forces (4%). The self-designated professor (meaning both teacher and professor in Portuguese) was vaguely defined as a secondary activity mentioned to signify authority in a field and political capital to influence others. Likewise, the words ‘entrepreneur’ and ‘businessman’ were used without precision and could mean a corporation CEO or a self-employed trader. Men did not shy from expressing their financial prowess by using words like ‘money’, ‘bitcoins’, and ‘investments’. 

The participation of the security forces – police and military, for example – in the attack is a question under investigation by the current government. Our data showed a small presence of these groups in the Twitter ecosystem, which can be explained by the institutional constraints related to political activities and the non-disclosure of their affiliations.  

Women, on the other hand, demonstrated a more cohesive professional identification, with fewer careers being displayed. They usually relied on general terms such as ‘entrepreneurs’ and ‘lawyers’, allowing more space for family belonging. For example, one of the most common female bios was “married, conservative/right-wing, businesswoman”. The most prevailing career sectors in the female data set were: law (18%), education (18%), business (12%), and management (12%). In the female dataset, the word ‘professora’ (teacher/professor) was usually precise, generally meaning a school teacher. The word ‘empresária’ (businesswoman) was vague and could mean anything in the trade or business world, such as an owner of an architecture office or a craftswoman.

Lawyers Against the Rule of Law

Two additional aspects can be highlighted about careers. First is the ambiguity in using the words ‘entrepreneur’ and ‘businessperson’. Users identified themselves through more socially prestigious terms to their ideological ecosystem. Qualitative analysis of a representative sample of the bios and related websites showed that many users were self-employed traders (clothing and craftwork traders) or freelancers and service providers (i.e., electricians, handymen, hairdressers, manicurists, make-up artists). However, they rarely named these activities in their bios. For example, hairdressers and manicurists are popular careers for Brazilian women, however, only 0.01% of female users defined themselves as such. The occupation ‘vendedor’ (seller), also a common occupation, was statistically non-existent in the entire database. Yet, by checking the users’ websites, we could see these types of services and activities being offered. 

Secondly, we found a prevalence of attorneys and law-related professionals supporting an unconstitutional attack against democracy. Whilst, objectively, the attacks were against the law, men and women were not discreet in posting content rallying against the election results and the Supreme Court, and in support of a coup. 

Mentions of law careers in the bios prevailed in the ecosystem and were the highest amongst women. In our understanding, it was a sign of prestige amongst the ecosystem of people who supported the attacks; users were proud to define themselves as ‘lawyers’. Although Law is socially perceived as an elite career, Brazil has the highest number of Bachelors in Law worldwide, most of them graduating from private colleges that have proliferated in recent years and have focused on attracting low-income students. Amongst the 1,500 degrees offered by academic institutions in the country, only 282 have a minimum satisfactory standard. 

A high prevalence of law professionals does not necessarily mean that it was an ecosystem of predominantly high-income people. Paradoxically, it is possible that a large part of these professionals – along with other users with degrees in Management and Pedagogy, for example – are the ‘formerly poor’, namely part of the masses of low-income people who benefited from the democratisation of higher education during the previous left-wing Worker’s Party administration.

The Families of the Coup

Women were prouder to say they were part of a family: 20.5% of women described themselves as mothers in their bios, 7.7% of men described themselves as fathers, 7.3% of women and 2.1% of men displayed they were married. There were 3.3% grandmothers and 1.6% grandfathers. 

Women appeared more devoted to Bolsonaro than men, while men were more against President Lula than women: 12.9% of women compared with 8.8% of men included the name of Bolsonaro (i.e., ‘supporter of Bolsonaro’) or ‘Bolsonarist’ in their bios. In turn, men were more anti-Lula (1.5% mentioned words against the President in their profile), while only 0.1% of women did the same. 

Women were also twice as ‘passionate’ about things than men. They mentioned loving ‘Jesus’, ‘the family’, and ‘Brazil’. Men loved ‘hobbies’, ‘amateur knowledge’ ‘male football teams’, ‘Brazil’, and ‘the family’. The analysis of female bios and websites showed a recurring tendency to publicly display positive emotions. They described themselves as ‘fulfilled’, ‘happy’ and ‘blessed’ in the family realm. Expressions like ‘positivity’, ‘hard worker’, ‘fighter’, ‘never give up’, and ‘overcoming obstacles’ were used to describe the professional sphere. The word ‘hope’ was frequently used to refer to the future of Brazil.  We did not find the same pattern in the male dataset.


In this Insight, we presented the preliminary results of our research examining 89,865 Twitter accounts that were engaged to some degree with  the Brazilian coup attempt. On the whole, this ecosystem is cohesive: people displayed similar self-definition and followed the same figures, demonstrating an intention to connect and the power of social media virality.

At the individual level, differentiation based on the users’ occupations was apparent. A disaggregated analysis by gender was able to reveal novel aspects of occupations and political emotions, such as: 

  1. Men displayed more of their public activities and personal passions. They were proud to show off knowledge in various areas as forms of power and influence. 
  2. Women were as professionalised as men, but this was only part of a broader identity that combined intense professional, political and religious engagement with family belonging. Being a mother is a strong sign of distinction and virtue in the ecosystem. 
  3. Women strongly relied on positive emotions to describe their families, occupations and hopes for their country’s future. This finding expands Euro-American theories on the far-right, which gives considerable attention to male negative emotions, such as resentment, anger, and nostalgia.
  4. The high prevalence of users designating themselves as businesspeople and entrepreneurs is a political and religious phenomenon that rejects the working-class identity. It derives from a combination of strong free-market ideology and prosperity ethics from neo-Pentecostal evangelicals.
  5. Comparing female and male datasets, law-related occupations were among the most popular careers. This is a paradoxical result, considering their positions and actions were unambiguously against the rule of law.

Finally, in the face of the worrying coup attempt in Brazil and the wider concern of a global democratic backslide, this piece aims to contribute to a deeper understanding of how the far-right has been organised by gender stereotypes and prestigious occupational ethos amongst various social strata.


This research results from a collaboration DeepLab (UCD, Ireland) and between Labic (UFES, Brazil). Co-funded by the European Union (ERC, WorkPoliticsBIP, 101045738). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible for them.

Rosana Pinheiro-Machado is a Professor in the School of Geography at the University College Dublin. She is the Director of the Digital Economy and Extreme Politics Lab and the Principal Investigator of the project Flexible Work, Rigid Politics in Brazil, India, and the Philippines, funded by the European Research Council, Consolidator Grant. Twitter: @_pinheira

Debora Diniz is a Brazilian university professor. In 2020, she won the prestigious Dan David Prize, a lifetime achievement recognition for her contributions to gender justice and, in 2023, the Global Health Ethics Leadership, Oxford University. Twitter: @Debora_D_Diniz

Athus Cavalini is an MSc student in the Postgraduate Program in Computer Science (PPGI-Ufes) at the Federal University of Espírito Santo (Brazil). He is a member of the Data Science Lab (DSL-Ufes) and focuses his research on online social networks and information disorders.

Fabio Malini is an Associate Professor in Communication at the Federal University of Espirito Santo (UFES, Brazil). He is the Director of Labic – Studies on Images and Cyberculture. Twitter: @fabiomalini

Wagner Silva Alves is a PhD student at University College Dublin (UCD). His research focuses on extreme politics and the digital economy. He is a member of DeepLab (Digital Economy and Extreme Politics) and has worked on misinformation, the internet, and public health. He holds a PhD in Social Anthropology (2023, National Museum, Brazil).