“How a Facebook Update Can Cost You Your Job”: News Coverage of Employment Terminations Following Social Media Disclosures,From Racist Cops to Queer Teachers

Abstract

Social media posts and profiles have become a key part of hiring and firing processes, producing a “hidden curriculum of surveillance.” When hiring, employers routinely engage in “cybervetting” job candidates, making judgments based on their social media presence (or absence), and so too can social media disclosures impact (positively and negatively) employment progression and even result in termination. Where is the line between personal social media use and professional identities? What is the difference between holding people in positions of power to account and invading the privacy of everyday people? What kinds of social media posts get people fired? In this article, we report on a study of 312 news media articles that document stories of people being fired because of a social media post. We divide the corpus into posts made by the individuals who are fired (“self-posts,” n = 264) and posts made by others that resulted in the subject of those posts losing their job (“third party,” n = 48). Racism was the most common reason people were fired in these news stories, followed by other forms of discriminatory behavior (such as queerphobia), offensive content, workplace conflict, political content, acts of violence, and abuse. We examine these narratives through the lens of what van Dijck describes as “professional value,” and ultimately seek to question how these stories normalize the “hidden curriculum of surveillance,” putting additional pressure on employees and young people who are called to act on social media through the prism of future employment.

Introduction

In 2013, hospitality workers in Japan were fired for posting pictures to social media involving them disregarding food safety standards (Ryall, 2013). In 2014, a basketball coach in the United States was fired after posting a photo of herself in a bikini on Facebook (“Confusion Over E. Idaho Basketball Coach’s Status,” 2014). In 2015, an Australian man was fired from his job at a hotel after abusing a woman on Facebook (“Clementine Ford Describes Deluge of Abuse Over Facebook Post Which Saw Hotel Worker Michael Nolan Sacked,” 2015). In 2017, a retail employee in South Africa was fired for posting racist apartheid-era signs on Facebook (O’Reilly, 2017). In 2018, a bisexual school teacher was fired after he came out on Instagram in the United States (Allen, 2018). In 2020, a man in Dubai was fired for insulting Islam on Facebook (Farooqui, 2020). Also in 2020, four police officers were fired after killing unarmed black man George Floyd in the United States, with a video of the incident going viral on social media, fueling a renewed and global #BlackLivesMatter movement and sustained protests against police violence and racism (Shammas et al., 2020).

Some of these stories are not like the others. Some of these stories are indicators of structural social problems like racism and misogyny, where people (especially people in positions of power and authority) are being held to account through the capacity for social media to reveal and circulate at scale (Mundt et al., 2018). Other stories raise important issues about the boundaries between work and private life (if such boundaries do exist), including employer–employee surveillance, the limits of professional identity, and the rights of employees (Batza, 2017; Brandstater & Chaparro-Domínguez, 2020).

For more than a decade now, stories of people being fired for posts on social media—posts that they themselves have made, or posts by others that resulted in a firing—have circulated widely. In this article, we draw on an analysis of 312 news stories of firings related to social media posts from around the world, to analyze the patterns in these stories. We seek to examine who these “firing stories” are about and what reasons are given for being fired. These stories evidence a public fascination with employment termination involving social media posts that we explore. Here we set out to contribute to a wider scholarship that considers how we can begin to reconcile the productive capacity for social media to help make visible structural problems like systemic racism, misogyny, and queerphobia, at the same time as these logics of social media and visibility are used to erode boundaries between personal and public lives. In doing so, we also seek to reflect critically on the amplification of pressures on workers and extensions to the boundaries of “professional identities.”

Context Collapse, Cybervetting, and Professional Value on Social Media

Over the past 10–15 years, both organizations (“employers”) and individuals (“employees” or potential employees) have taken to social media to find each other (Nikolaou, 2014). Employers use social media for general marketing, public relations, and online brand building while also using social media for job recruitment and employee screening (Carrillat et al., 2014). On the other side of the employer/employee dynamic, individuals are able to create curated profiles that can represent an idealized version of themselves to improve their chances of employment (Frantz et al., 2016). These might be profiles focused on a professional identity with employment histories, major projects, and work experience, such as on LinkedIn, or personal but public-facing profiles, like an Instagram or Facebook profile, or a blog that presents an individual’s everyday experiences such as travel, relationships, consumption, and politics. There can be significant overlaps between the processes of impression management employed across these kinds of profiles, but as van Dijck (2013) argues, users “have a need for multiple ‘stories’ about themselves, each story concerning different parts of their identities and addressing a limited audience” (p. 211) and these stories can be managed across platforms or within them. Ultimately, however, from an employment perspective, van Dijck (2013) suggests that these platforms are being used to assess “professional value” (p. 212). The incorporation of social media in professional environments and employment decision-making can benefit organizational and individual imperatives but raises important questions around the ethics of these methods of recruitment, and the introduction of biases and discrimination, along with a wider erosion of boundaries between personal and professional lives (Jeske & Shultz, 2016).

In this article, we take up van Dijck’s (2013) theorization of “professional value” as related to how employees conduct themselves (or are presented) on social media. Van Dijck was focused on the notion of professional value as a disciplining strategy, where employers might monitor the ways employees conduct themselves on social media. She suggests that “employers appear particularly interested in personal information even if users intend to keep this private, going as far as to force job seekers to let them access their Facebook passwords during interviews” (van Dijck, 2013, p. 212). Similarly, Berkelaar’s (2014, 2017) observations of “cybervetting” echoes Van Dijck’s theorization of professional value. Cybervetting is often used to ensure organizational fit, with employers using social media platforms to surveil potential and current employees across multiple social contexts, ostensibly to ensure alignment in terms of values, and assess various attributes and workplace “fit.” Employers justify the use of cybervetting as a form of “risk work,” with the information collected online used to mitigate risk by determining compatibility and exposing “red flags” (Berkelaar, 2017, p. 1126) that would suggest a misfit, compromise the image of the employer, or disrupt the workplace. Berkelaar (2014) observes that the normalization of this practice represents a new or extended “digital social contract” which raises important ethical questions about employee visibility, transparency, and availability (p. 502).

In this article, we focus on the other end of the “employment life cycle”—termination of employment. Here, we are interested in how “professional value” is compromised by certain actions, such as social media posts from an employee that might reveal (i.e., bodies, sexuality, politics) or posts by others that depict employees (i.e., being negligent, racist, violent). As we will show, how these acts compromise a sense of “professional value” vary based on (a) the values of the employers/company; (b) the kind of ‘transgression’; and (c) forms of public attention or “backlash” that are also facilitated by social media.

A key strategy for mitigating “context collapse” (boyd, 2011; Vitak, 2012) on social media and managing boundaries between different “audiences” in everyday life is to maintain multiple social media profiles or accounts, engaging in “audience segregation” in the language of Goffman’s (1959) dramaturgical framework. This kind of audience segregation might operate within or across platforms, and is vital for those looking to separate “work” from other aspects of life and other social worlds. Within platforms, audience segregation might involve a single profile that has different privacy settings and filters such that person A (a close friend) can see everything the user posts, whereas person B (a distant relative or co-worker) can see less or restricted content. Instagram currently allows users to post stories to either their “close friends” or all followers as well as hide stories from specific followers, for instance. Alternatively, audience segregation may involve maintaining multiple accounts or profiles on the one platform as is common on Instagram with “finstas” (Duffy & Chan, 2019, p. 121) or on reddit with “throwaway” accounts (Leavitt, 2015) where users post different kinds of content under different usernames for different imagined audiences. Audience segregation across different platforms is also common, where in an era of “polymedia” use (Madianou & Miller, 2012), individuals turn to different platforms to connect with different audiences (like “work colleagues” and “school friends” and “family”) and in different ways, as van Dijck (2013) points to in the different uses of Facebook and LinkedIn. van der Nagel (2017) explains that managing multiple accounts and profiles on different social media platforms is a strategy to “consciously compartmentalize different facets of identity” (p. 312). Not managing context to compartmentalize can compromise a sense of “professional value,” and for employees whose employers engage in cybervetting, this can jeopardize employment.

Duffy and Chan (2019) suggest that young people are particularly attuned to being surveilled across social media platforms and in turn this sense of “imagined surveillance” steers young people’s self-presentation practices on social media toward tactics including finely tuned privacy settings, deleting content, and maintaining multiple profiles for different imagined audiences. They suggest that the “hidden curriculum” young people navigate and are socialized into (punctuality, conforming to hierarchies, and internalizing drives toward productivity, for instance) now also includes an internalization of ubiquitous monitoring and surveillance that plays out in networked publics. In an employment context then—where people are being fired for what they post on social media, or because of the posts made by others that depict them on social media—there is a clear rupture in this hidden curriculum, resulting in an employee’s “professional value” being called into question.

As a result of the increased use of social media for screening and employee monitoring, employees and job seekers have expressed increased awareness around its use by employers. Hurrell et al. (2017), in their research on working students in the United Kingdom, found that 27% of participants experienced employer “disapproval” of their social media activities, and some even experienced attempted influence of their use of social media by employers. Their participants also reported concern over the misuse of their data and often managed their social media with employers or potential employers in mind, aligning neatly with Duffy and Chan’s (2019) theorization of the hidden curriculum of surveillance. The study by Hurrell et al. (2017) points to concern among young employees about the blurring of professional and personal life boundaries. Hurrell et al. (2017) also highlight the potential for employer misuse of personal information and at the same time point to the strategies young people are using to manage this intrusion such as being “discerning” with what is posted, “restricting” who could see certain posts, and not adding managers as contacts (Hurrell et al., 2017, p. 79).

In their study of 637 students in criminal justice courses in the United States, Rossler and Scheer (2020) found that—contrasting with the findings of Hurrell et al. (2017)—74% of students thought it was appropriate for police departments to conduct social media background checks on potential recruits. Two-thirds strongly agreed that looking at social media activity is a good indicator of character, which was important in determining whether or not that individual was suitable for a police career. On the other hand, Leott and Hodgson (2019), who also undertook a study with undergraduate students studying criminal justice in the United States (n = 231) found these students were “unequivocal in their expectation of privacy” (p. 1) when it came to social media. Leott’s participants thought employers should avoid using the information on social media to make employment decisions if those candidates or employees did not give consent to do so.

Drouin et al. (2015) also asked students in the United States (n = 442) about their opinions on the use of social media in surveilling the behavior of teachers, but this group was from a range of areas of study. In contrast to Rossler and Scheer’s participants, and aligning more with Leott’s findings, Drouin et al.’s participants mostly thought a person’s social media profiles should not be used to make hiring or firing decisions (only a third of their participants supported that practice) but when asked about specific cases related to school teachers, there was some variation. Most thought it was acceptable for a teacher to post a photo holding a beer (52% agree, 24% neutral, and 24% disagree) but fewer said it was acceptable to post a photo of a teacher engaging in “lewd behaviour” such as at a bachelorette party (33% said that teacher should lose their job, 28% were neutral, and 39% said the photo should not result in them losing their job). In other, more general research on the role of social media in hiring and firing decisions, researchers have also found generally negative perceptions of the use of nonprofessional social media in vetting and hiring processes (Aguado et al., 2016; Parker et al., 2019).

These differences in results—when comparing studies from Hurrell et al. (2017), Rossler and Scheer (2020), Drouin et al. (2015), and Leott and Hodgson (2019)—are presumably related to the kinds of work, career types, and corresponding understandings of power and authority different jobs are imbued with. Police officers and teachers, as discussed here via Rossler and Scheer (2020) and Drouin et al. (2015), for instance, are held to higher standards which require them to take on what Leott and Hodgson (2019, p. 13) describes as “guardian roles.” These career-contextual expectations are reflected in attitudes around what is and is not “appropriate” (linked to “professional value”) when it comes to visibility and thus accountability in networked publics. Some careers require higher levels of impression management and people in those careers experience higher levels of scrutiny around “professional value.” The costs for violating these expectations are also higher, as we will demonstrate. Across the research discussed here, it is also likely that different cultural contexts play a role, and the predominance of studies centered on perspectives of undergraduate students is a major concern in how this issue is understood. Compared to the research on perceptions of social media use in hiring and “cybervetting,” there is relatively little research on the other side of employment: terminations related to social media use.

Method

In this article, our goal is to answer the research question: What kinds of social media posts get people fired, according to news media stories? To answer this question, we draw on a content analysis of 312 news media articles, sourced through three news databases: Factiva, MediaCloud, and the Australia & New Zealand Newsstream. We used these databases because of their comprehensive and global collection of news articles. The Australia & New Zealand Newsstream database was selected to increase the sources from the Oceanic region as this research took place in Australia, and the broader project it is part of is concerned with national points of comparison. We used search terms “social media AND employment AND fired” and also replaced “social media” with specific popular platforms such as Facebook, Twitter, and Instagram. Searches were conducted in the first half of 2020, to include a ten year window, with all articles collected published between 1 January 2010 and 17 July 2020. This timeframe was set because social media platforms became widely adopted and ubiquitous during this time. The number of stories about people being fired as a result of a social media post increased over time, with just 16 stories in 2010 and 14 stories in 2011, compared with 71 stories from the first half of 2020 alone in the final corpus. This demonstrates a growing public interest in these kinds of firings and continued scrutiny of social media by employers.

The initial sample of articles included more than 7,000 entries, mostly from Factiva (n = 2,400) and NewsBank (n = 4,800). From this initial sample, duplications (including the same story covered in multiple outputs and/or outlets) were removed along with a range of entries that were transcripts, references to primary sources, and irrelevant entries (such as news articles on a range of topics that included embedded links to the news agencies own social media accounts to share stories). Only news stories in print or digital formats that included a story about an employee being fired following a post made on social media were included—either by the employee or a post that included them (which we coded as “third party” posts). This eventually led to a final corpus of 312 news stories for a more detailed analysis. These 312 articles were cataloged in an Excel spreadsheet including title, a summary, date published, and country of publication, along with a range of codes assigned to each entry such as whether the story involved a high profile or public figure, an element of public backlash, the social media platforms referred to, the occupation the person represented in the story was fired from, and—importantly for our analysis—the reason for termination. These codes were reached inductively and deductively based both on initial assumptions and what we discovered in reading and analyzing these stories. Articles were then organized into two broad categories and then divided into subgroups as detailed in the findings section.

Articles selected were predominantly sourced from Western, English-speaking countries with the majority of the articles sourced from the United States (n = 227), followed by Australia (n = 18) because of our purposive sampling, then Canada (n = 16), the United Kingdom (n = 13), and New Zealand (n = 12), with a smaller number (between 1 and 5 stories) from a long list of countries including the United Arab Emirates, India, Israel, Singapore, and The Philippines.

Within this corpus, 62 stories were coded as “meta.” That is, meta stories about the phenomenon of being fired following a social media post, that centered on opinion, advice or expert analysis of social media in the professional environment. For example: “#boredatwork means #yourefired” (Keen, 2014), and the piece that the title of this article refers to “How a Facebook update can cost you your job” (O’Shea, 2010). These kinds of meta stories were included in the sample because within them they did contain specific stories of firings we were interested in.

A key limitation with this method for answering the research question “what kinds of social media posts get people fired?” is that looking only at stories that make it into the news is a very narrow lens. Spectacular, controversial, and the most disturbing stories are most likely to be circulated, whereas more banal stories—transgressions that might be considered “obvious” or not newsworthy—could be buried. We are relying on judgments made by journalists, editors, and publishers on what kinds of stories are centered. We acknowledge their expertise and the work involved in writing these stories while also recognizing how this lens can limit and shape what stories are told. There were also undoubtedly more news stories out there on this topic over the past decade. We do not claim our corpus of stories is exhaustive or representative; however, our study has allowed us to point to emergent themes and patterns in these stories, which come to serve as the backdrop to dominant narratives about the impact of social media on employment and careers. This is especially the case for young people, precisely because the prevalence of these media narratives shapes that “hidden curriculum of surveillance” (Duffy & Chan, 2019).

We want to acknowledge from the outset that our own subject positions shape and inform our coding processes, how we interpreted the stories we were reading, and how we organized the results here. Robards identifies as a White queer cis man, and Graf identifies as a mixed-race straight cis man. While the stories were being coded, we met regularly to double-code, to discuss the organization of codes, and this article was written collaboratively.

Findings

Our corpus of 312 news stories can be divided into two categories: those centered on “self-posts” (n = 264) and those on “third-party posts” (n = 48). Self-posts include news stories covering posts made by individuals that resulted in them getting fired, such as when a paramedic was fired in New Zealand following an altercation she had with a colleague that partly played out on Facebook (“Facebook Spat Goes to Court,” 2010) or when a pilot in India was fired over an offensive tweet (“GoAir Fires Pilot Over Objectionable Tweets, Employee Says Being Falsely Accused,” 2020). These self-posts make up the bulk of the corpus, with 264 news stories being coded into this category. The second category we have coded as stories related to third-party posts—posts made by people other than the person fired, like bystanders or colleagues. For example, when a hospitality worker was fired when a colleague called him a “Nazi” on Facebook for attending a “Unite the Right” rally in the United States (“Mojo Burrito Fires Employee Who Attended Charlottesville, Va., Rally,” 2017) or when a car salesperson was fired in Canada when a video someone else recorded of her yelling at and threatening a group of men was widely circulated (Issawi, 2018). These stories on third-party posts were fewer in number in our corpus (n = 48) but also tended to be higher profile. The key difference here is that for the stories on self-posts, these are scenarios where individuals ostensibly got themselves into trouble, whereas the stories on third-party posts are about others (co-workers, customers, and strangers) recording and sharing the actions of the person who is eventually fired. Sometimes this is intentional, like holding a police officer to account for brutality (Elshahawy, 2019), but in other cases, holding the person who was eventually fired to account was not the initial intent of the post (Oppenheim, 2017).

Facebook was the social media platform most commonly referred to in our corpus with 170 stories referring to it as would be expected for this 10-year period from 2010 to 2020 and the nature of Facebook as a platform where content can be shared rapidly and widely through a broad set of networks, often without contextual information (Marwick & boyd, 2014). Other platforms less frequently connected to firings in our corpus of news stories included Twitter (n = 47), Instagram (n = 12), and Snapchat (n = 9). In all, 74 articles referred only to social media in a general sense, not naming the exact platform used. Some articles cited multiple platforms simultaneously, with terminations the result of social media “misuse” across multiple platforms or public backlash spreading across different platforms, highlighting the enmeshing of social media and the flow of content across them in this “polymedia” environment (Madianou & Miller, 2012). An example of this is a story about a pilot in India who was fired because of a Twitter post, with both the backlash against the airline and the abuse directed at the pilot playing out on Facebook, Twitter, and YouTube (“GoAir Fires Pilot Over Objectionable Tweets, Employee Says Being Falsely Accused,” 2020).

In the following sections, we take a closer look at our two broad categories of analysis (self-posts and third-party posts), before moving on to discussions of key themes in these data: the “social media backlash” and fame; and discrimination, hate, and violence.

Self-Posts

Figure 1 shows a breakdown of news articles centered on stories of job termination, where the ex-employee had themselves posted something on social media (self-post) that resulted in them getting fired, categorized here by type of transgression. The visual presentation of this breakdown was inspired by McCosker et al.’s coding and visual presentation approach taken in their research on how dating and hook-up apps are discussed in the news media (McCosker et al., 2019).

Figure 1.

Self-posts. Reasons for termination reported, categorized by kind of transgression.

For the 264 stories coded into the “self-post” category, the most common reason for being fired was “hate and discrimination” (n = 112, or 42% of this group of self-posts), followed by “offensive” (n = 72, 27%), “workplace conflict” (n = 53, 20%), and “other” (n = 27, 10%). There is clearly some overlap in conceptual categories here, as content that is discriminatory can also be read as offensive and could also be the subject of workplace conflict. So we further coded the reasons for termination into subcategories.

For our “hate and discrimination” category, these included “racism” (n = 68), lesbian, gay, bisexual, and trans “(LGBT)” (n = 18), “threats/violence” (n = 14), “abuse/hate” (n = 8), and “misogyny” (n = 4). That racist posts were the most common kind of post to get someone fired in our entire sample is clearly significant, and this is a finding we will return to in our discussion subsequently. Stories centered around posts related to LGBT content were more complicated, as these included stories where people were fired for being “out” on social media, or posting pro-LGBTQ+ content, and also stories of people being fired for posting LGBTQ+-phobic content. For instance, “Bourne Firefighter/Paramedic Is Fired Over His Facebook Posts” (2011) where the terminated employee was fired for being “disrespectful to gay individuals,” and “School chaplain sacked over Facebook post calling homosexuality ‘not normal’” (Shine, 2014) were coded into the same category as “Catholic Church Has School’s Back on Firing” (2014) where a teacher was fired for posting about his plans to marry his male partner. All three stories were coded under hate and discrimination in our assessment, but the “source” of the discrimination was different: The firefighter and the school chaplain (employees) were discriminatory in stories one and two, but it was the school (employer) who discriminated against the teacher in story three. We combined these stories for our analysis here to draw attention to enduring “controversy” around queerness in public and professional life, in the context of employment.

Under the category of “offensive” (n = 72 stories), the most common subcategories were “insensitive” content (n = 32), such as “Nurse Fired for Comment About Fatal Crash” (2012) where a nurse blamed the victim of a fatal collision between an ambulance and a motorcycle on social media; followed by “controversial” content (n = 19), such as “Teacher’s Criticism of Israel Leads to Dismissal” (2020) where a teacher tweeted what the school described as “hurtful and offensive” comments, and “Texas Man Fired, Under Police Investigation Due to Online Threat Over Whole Foods Mask Policy” (Nelson, 2020). The remaining stories in this category ranged from teachers acting inappropriately with students through to unprofessional behavior like the Air New Zealand cabin crew member recorded in a Snapchat video spitting water with the caption “wish I could spit on passengers like this” (Cole, 2016). This “offensive” category was a particularly slippery one, as “offense” can be interpreted so differently. Offensive to whom? To the employer, clearly, but that was often shaped by “public outcry” as posts were circulated and recontextualized in different social media spaces, such as when the man in Texas referred to in the San Antonio Current story earlier had his Facebook posts screenshot and reposted by another person on Twitter, thus compromising his “professional value” (van Dijck, 2013). It could also be argued that the racist and queerphobic posts covered in stories we categorized under our biggest category, “hate and discrimation,” are also “offensive,” but we separated these out here to highlight how dominant those stories were.

The third category of self-posts was “workplace conflict” which included “management criticism / dispute” (n = 11), “disparaging coworker / workplace” (n = 10), “privacy / policy violation” (n = 7) and an assortment of others that were more varied and related to absenteeism, social media usage during work hours, whistleblowing, promoting competitors, and unfair workplace terminations. These were more straightforward and less about the kind of “public outcry” that would often drive stories in the other categories (like racism and queerphobia from the first category; and offensive, inappropriate, and unprofessional content in category two) and were more about internal workplace issues. These are clearly more mundane stories and presumably underrepresented in media coverage of social media-related firings because of their banality.

Finally, we were left with 27 “other” stories about self-posts that did not clearly fall into a defined category or contained a small number of unique reasons for termination. These included “political content” (n = 17), “abuse of power” (n = 5), and individual cases such as being anti-vaccination (pre-COVID 19) and exposing sexual harassment.

Third Party

Our second broad group of news stories shown in Figure 2 are centered on “third-party” posts or posts made by social media users documenting actions, behaviors, and words by others (usually strangers) and sharing them on social media which results in the subject of those posts being fired. For the 48 stories coded into the third-party group, we further divided these terminations into four categories that were then broken down into subcategories. The four top-level categories were “racism” (n = 19), “violence” (n = 14), “abuse” (n = 8), and “other” (n = 7). While this broad group of stories on “third-party” posts is smaller, there are also clear overlaps with our first and bigger group of stories.

Figure 2.

Third-party posts. Reasons for termination reported, categorized by kind of transgression.

The racism category included “racist group involvement” (n = 9) such as with white nationalist groups (“Charlottesville White Nationalists Outed on Social Media; 1 Fired From Job,” 2017), “verbal” (n = 8) which was usually recordings of verbal racism, such as “Experts say woman who told men they were ‘not Canadian’ in viral video was rightfully fired” (Issawi, 2018), and finally one story about a video shared on social media showing police engaging in racial profiling and another story on a video showing a corrections officer mocking the death of George Floyd.

The second category was “violence” (n = 14), which centered on stories of physical violence recorded and shared on social media, with just two subcategories: “police violence” (n = 5) and “other” (n = 9), where acts of physical violence were recorded and perpetrated by journalists, TV hosts, hospitality workers, teachers, and a bus driver. The five stories centered on violence perpetrated by police were all based in the United States, including the story that has received global attention—the killing of George Floyd (Shammas et al., 2020). The social media and employment angle here is of course secondary to the way these stories reveal patterns of violence and brutality related to institutionalized racism and cultures of violence. We note them here as important and present in our corpus of news articles but also acknowledge a deeper discussion of institutionalized racism, police violence, and cultures of violence and oppression more generally is beyond the scope of this article and our expertise.

The third subcategory was “abuse” (n = 8), which included incidents of abuse that were mostly verbal and threatening, such as “Highway worker fired for making threats” (Wimmera Mail Times, 2019) where the worker was recorded threatening conservation protesters in Australia, and “Online Petition Calling for Firing of Ramesh Erramalli From JP Morgan Singapore Gathers More Than 26,000 Signatures” (2019) after a video of a JP Morgan employee verbally abusing a security guard was circulated on Facebook. Finally, the “other” subcategory included stories related to “sexual content” (n = 2), “sexual harassment” (n = 2), and single cases of “derelection of duties,” “political content,” and “unhygienic practice.”

In the following sections, we examine two key themes that emerged from our analysis: first, public backlash and fame for being fired; and second, the centrality of racism, discrimination, hate, and violence in these news stories.

Backlash and Fame for Being Fired

One recurring theme in our analysis of these stories was a cycle where an employee losing their job because of a social media post generated considerable public interest that itself played out on social media. Sometimes the interest preceded the news story in our corpus (with public response becoming part of the story), and in other cases, the news story itself worked to generate that interest. As we document subsequently, the public interest sometimes resulted in a “backlash” against the employer for firing the employee because of a social media post. Other times, the public interest resulted in far-reaching reputational damage for the person who was fired. These narratives represent a public conflict between employer and employee, playing out both in news and in social media channels. The backlash/fame phenomenon has two sides. On one side, the public sharing of these stories can garner sympathy and public goodwill toward the person who was fired, sometimes even resulting in the employee being reinstated or finding work elsewhere. The other side of this phenomenon is infamy, where the person being fired is publicly “canceled” to the point that their reputation is damaged significantly and enduringly.

One example of this phenomenon was the story of an employee at a stone business in Detroit in the United States who was fired for sharing a meme on Facebook. The meme was an image of television character Elmo on a potty captioned with “Boss makes a dollar, I make a dime. That’s why I poop on company time” (“Metro Detroit man says he lost his job for sharing a meme”, WXYZ Detroit, 2019). After sharing the image, the man’s employer sent him a profanity-laced message and what appeared to be a dismissal, which is how it was reported. However, another journalist (Chowdhry, 2019) investigated further and found that the employee continued to work several shifts before eventually leaving the job due to “bad blood” from the meme incident. A lawyer for the employer made a statement claiming the ex-employee “started a Gofundme page to attempt to capitalize on his newfound exposure” and that his client (the employer) had subsequently “been subjected to significant monetary damage, as well as threats, profanity and harassment by phone and email” (Chowdhry, 2019). This public response and backlash against the employer reflect Drouin et al. (2015) and Parker et al.’s (2019) findings that the public holds negative perceptions of companies who fire their employees as a result of personal social media use, although in this case the employer’s lawyer argued the employee was not terminated over the meme although that appeared to be implicated in the initial response to it.

Another example of public backlash to a firing was the case of a trainee pilot, who we referred to earlier, who was fired from an airline in India over what were described as “objectionable tweets” (“GoAir Fires Pilot Over Objectionable Tweets, Employee Says Being Falsely Accused,” 2020). The pilot claimed he was wrongfully terminated in a case of mistaken identity and that the tweets came from another person with the same name. The pilot was subsequently harassed and received death threats. A backlash campaign emerged against the airline (#BoycottGoAir) because of the unfair dismissal, and eventually, the airline reversed the termination to investigate further. For both the GoAir pilot and the meme sharer in Detroit, ultimately it could be argued that they benefited from the digitally mediated public backlash—with the Detroit man establishing a Gofundme page and the pilot eventually being reinstated—but at least in the case of the pilot, the mid-way “cancellation” would have no doubt been traumatic.

Writing from a legal perspective about the “scarring” reputational damage that can be done in the “social media backlash,” Batza (2017) argues that social media enables what he calls the “digital mob.” Batza suggests the social media backlash represents a “new defamatory phenomenon” (p. 438) and cautions that “each of us” (p. 439) is potentially vulnerable to it. And yet, Batza’s opening example of such a backlash is the story of a public relations worker who tweeted, before getting on a flight, “Going to Africa. Hope I don’t get AIDS. Just kidding. I’m white!.” Batza, sympathetic, lamented that after the tweet (which he conceded was “ill-advised”) went viral the tweeter lost her job. Batza found this to be an example of how a reaction to such a post can ruin a reputation. Was it the backlash or the racist tweet that did the damage here? And is everyone truly vulnerable to the backlash, or only those whose views and behaviors are deemed to be unacceptable? Certainly, had it been included in our corpus, we would have coded this story about a self-posted social media disclosure as racist. The reputational harm from the tweet being widely circulated (“going viral”) compromised this woman’s “professional value” (van Dijck, 2013) for her employer, resulting in her losing her job, but there is an important counterpoint here about people being held responsible for their views and behaviors. While we do not agree with Batza’s (2017) framing of the issue, in calling for defamation law to be applied in cases like this, he highlights an important question about where the line is drawn between public figures and private individuals when it comes to social media.

Racism, Discrimination, Hate, and Violence on Social Media

The most common reason for firings in our corpus was racism. In both the “self” posts and “third-party” posts, racism was the most frequently occurring focus in stories. This is unsurprising in some ways, given the increasing focus on specific instances of racism and more intense discussion of structural and institutional racism in the countries that were the most dominant and well-represented in our corpus: the United States, the United Kingdom, Canada, Australia, and New Zealand. These are countries with their own histories of colonialism and very current issues of continuing forms of discrimination, so stories of people losing their jobs for being racist (and being called out on it, often with clear documentation and evidence) are “news-worthy” is unsurprising. This is part of a wider cultural reckoning around structural racism.

Can the visibility of these cases—and placing instances of racism in the public eye—be productive in addressing these systems of oppression and marginalization? In reflecting on the “#Gamergate” phenomenon in 2014, where “online gaming communities erupted into vicious arguments—ostensibly about ethics in video game journalism, but more pointedly about gender, privilege, and gaming” (Braithwaite, 2016, p. 1), Braithwaite pointed out how #Gamergate might have been productive in revealing everyday experiences of harassment women face. Similar parallels could be drawn with the way racism is being made visible and persistent in digital spaces. For Braithwaite (2016), the coordinated efforts of #Gamergate across multiple platforms at huge scale, “made this kind of ‘ordinary’ harassment newsworthy, calling our collective attention to the sustained abuse many people endure in order to participate in online spaces” (p. 7). So too can we look to patterns of firings in news articles as we have done in this article, to draw attention to racism as pervasive and embedded, but also—perhaps increasingly, and promisingly?—risky. That is, as the patterns of firing evidenced in our corpus reveal, when racism and other forms of discrimination (like queerphobia, although this is still more contingent and varied) are revealed in a “public” way, through and on social media, there are repercussions. At least in terms of employment (which is, admittedly, a very narrow lens), employers appear to very quickly drop employees who are publicly outed as racist, precisely because it compromises their “professional value” (van Dijck, 2013) and risks the reputation of the employer.

Despite this pattern in our own data, it is also clear that speaking out on social media about racism and other issues of harassment and discrimination can also be detrimental in professional terms. Howard et al. (2020) found that posting about racism on social media can have professional consequences for Black Americans. They asked a diverse group of 154 participants in the United States—many of whom had some experience in hiring decisions—to rate candidates for an office supervisor job, based on their cover letter, resume, and a printout of their Facebook profiles. For the fictional people with profiles that included posts complaining about racism, Howard et al.’s (2020) participants rated them less favorably in terms of likability and likelihood to offer them a job interview. This resonates with Ahmed’s wider work on the complaint. She has observed that “the mere fact of making a complaint can be used to cast doubt upon the person who complained” (Ahmed, 2018). In drawing attention to racism, misogyny, queerphobia, and other forms of discrimination, bigotry, violence, and exclusion, people can themselves become framed as “complainers,” which can perversely also have a negative impact on their “professional value” (van Dijck, 2013).

While saying racist things on, and being recorded doing racist things on social media were clearly damaging when it came to the “professional value” of employees in our corpus, the reputational harm associated with “queerphobia” (including homophobia, biphobia, transphobia, and other forms of bigotry toward LGBTQ + people) was more complex and uneven. As we described earlier, while less prevalent than racism, queerphobia was the second most common form of discrimination in stories of self-posts leading to a firing. Importantly, this included stories where employees were fired for being queerphobic or bigoted toward queer people but also stories where employers (usually religious organizations) fired employees for being queer. In these cases, being out as “non-normative” (i.e., not straight and not cisgender) was itself a risk to “professional value” in organizations where employers are able to discriminate based on sexuality and gender. While it is beyond the scope of this article to explore the terrain of queerphobia globally, it is important to note here that when it comes to employment there are very different cultural standards and legal protections for LGBTQ + people in different parts of the world (McPhail et al., 2016). As our data suggest, being out as queer on social media is still an employment risk even in otherwise “liberal” or “progressive” countries (Carrasco & Kerne, 2018; Duguay, 2014).

Who Is Held to Account?

Who is held to account when it comes to social media and employment? Law enforcement was the most highly represented occupation in our corpus of stories on job terminations (n = 62), followed by education workers (n = 41) (including school teachers, principals, academics, coaches, and teaching support staff), hospitality (n = 27), media (n = 24), medical (n = 22), retail (n = 12), government (n = 10), and transport workers (n = 10). The remaining stories covered firings from a wide range of occupations, including finance, manufacturing, public service, technology, and real estate. It was unsurprising that the jobs most represented in our study were police, teachers, and health workers. These are “everyday people” in highly accountable and “public” roles, responsible for the safety and health of the people they serve. Indeed, this finding in our own research aligns neatly with Rossler and Scheer’s (2020) study that found most students in criminal justice courses in the United States thought it was appropriate for police departments to use social media as an indicator of character, and Drouin et al.’s (2015) research that indicated teachers were also held to a higher standard when it came to their social media conduct than people in other professions. As such, it is unsurprising that people in these kinds of jobs are subject to what Berkelaar (2014) describes as “cybervetting,” not only at the hiring point but also subsequently throughout their employment. It follows then that the dismissals of people in these “public” roles are more newsworthy too, given their responsibilities and the associated standards to which they are held. Similarly, however, hospitality workers—from restaurant managers and wait staff to cashiers and dishwashers—are on the front line of large customer service industries, representing the companies they work for, although they arguably have significantly less power and autonomy. Should everyone be held to account in the same way for their conduct on social media? Where is the line between accountability and an invasion of privacy? Power clearly plays an important role here, and so too does the exercise of “moral judgement.”

Limitations

This article is based on an analysis of 312 news media articles, predominantly from the United States, covering employment terminations linked to social media posts. While this corpus of stories does reveal the contours of public discourse around employment terminations linked to social media posts that are the subject of public fascination, the stories that are “newsworthy” are not necessarily representative. Stories that journalists pick up are likely to be more spectacular, controversial, and disturbing, whereas everyday or banal stories are less likely to attract the same interest. The patterns we observe here are not generalizable. Our analysis was also not able to properly explore variations from country to country or over time.

Future research could look to directly recruit individuals who have been fired because of a social media post, or employers who have terminated an employee because of a social media post, to more deeply explore experiences, perceptions, and impact, beyond what makes it to the news. Further research could also attend to cultural differences, both in terms of employer/employee boundaries and in local modulations of sociostructural problems like racism, misogyny, and queerphobia.

Conclusion

In their study on the perceptions of using social media to surveil employees, Drouin et al. (2015) conclude by suggesting that young people need to be warned that “short-term social media use could have a long-term effect on their future careers.” They suggest that, for the majority of their young respondents who were opposed to the use of social media in hiring and firing decisions, this was either because they “have more to hide” or that they were “embracing of the current culture, in which the sharing of all types of information (both appropriate and inappropriate) via social media is commonplace” (Drouin et al., 2015, p. 128). This reading of the rationale that people—young people in particular—might have for resisting what Duffy and Chan (2019) call the hidden curriculum of surveillance is extremely limited. This explanation also pushes the burden of responsibility here onto young people themselves, who are historically the target of moral panics, especially as related to new technologies (boyd & Hargittai, 2013). Instead, we would argue a more critical line of inquiry should be directed at employers, recruiters, HR managers, and others who engage in sustained “cybervetting” (Berkelaar, 2014), blurring the boundaries between personal lives and professional contexts (see also Smith & Kidder, 2010, who were similarly calling for more ethical organizational guidelines around cybervetting practices more than a decade ago). As Berkelaar (2014) observed, these pressures produce an extended “digital social contract” that prescribes “normative expectations for workers’ digital visibility” (p. 479). Under this paradigm of both a hidden curriculum of surveillance and intense pressure around employment futures, young people in particular then are being asked to navigate social media use (documenting their lives, hanging out with friends, and engaging in self-expression) with this threat of future reputational harm looming.

In this article, we have sought to map out some of the key reasons why, in news coverage of employment terminations, people are fired for social media posts—both posts they have made themselves and posts by others making their behavior and actions visible. Key themes have been racism, discrimination, hate, a range of different offensive content, and workplace conflict. These stories revealed patterns around who is held to account, and which stories become dominant when it comes to “professional value” being compromised on social media. Importantly, this corpus of stories allowed us to reflect on the line between personal social media use and professional identities, and to critically consider the tension between holding people in positions of power to account and invading the privacy of everyday people.

The continued perpetuation and socialization of the “hidden curriculum of surveillance” (Duffy & Chan, 2019) works to produce compliant, self-governing citizen-employees, pushed to curate and manage often highly sterile representations of their lives in networked publics, under threat of employment doom. At the same time, there is a clear productive role in exposing violations of power, bad behavior, misconduct, racism, misogyny, homophobia, transphobia, and other forms of bigotry, harassment, and violence in these very same digitally mediated networked publics. We suggest that visibility here is a double-edged sword: invasive and violent in some cases and for some people (especially for young people whose digitally mediated youths are managed through this prism of futurity) but also liberating and enabling justice, accountability, and transparency in other scenarios and for other actors.