Keywords

1 Introduction

Although state-sponsored homophobia is an alien legacy introduced by the former colonial power [1, 2], the Ugandan government has increasingly become an active actor in its own right in institutionalizing and normalizing violent discrimination of LGBT+ individuals [3,4,5]. The Ugandan government has, since the formation of a visible LGBT+ rights activism in the early 2000s, actively suppressed its rights claims [6]. The Ugandan government has introduced several laws that provide state and non-state actors with an expansive legal toolbox that can be used for systematic discrimination of LGBT+ individuals. Although it was the 2009 Anti-Homosexuality Bill that catapulted Ugandan state-sponsored discrimination into the international limelight [7], it is only one example of the state’s efforts to create a comprehensive legal framework of oppression. In 2005, the Ugandan constitution was changed to ban same-sex marriage in 2005. The Anti-pornography Act of 2014 and The Ugandan Computer Misuse Act of 2011 provide the Ugandan state with far-reaching discretionary power to control and restrict citizens’ expression and private communication [8]. It should be noted that several scholars have tried to nuance our understanding and move beyond simplistic explanations of Ugandan lawmakers as intrinsically homophobic, arguing that so-called legal reforms need to be situated in a post-colonial context [3, 9, 10]. State actors’ actions can be explained by the challenges with nation-building and post-colonial anxiety, whereby sexual policing allows the state to assert its autonomy by “defining a gendered and racialized African subject,” far removed from unwanted Western influences [11].

The 2009 bill, which sought to introduce life imprisonment for same-sex sexual activities and the death penalty for aggravated homosexuality, as well as criminalize support to LGBT+ activism [12], set several processes in motion. First, it forced the local LGBT+ community to become better organized in its resistance against state oppression. In collaboration with international development partners, the local human rights community successfully managed to stall the 2009 AHB for several years and eventually had it overturned ten months after being passed in late 2013. The 2009 bill was also a key contributor to securing the international human rights and development community’s attention to the plight of the LGBT+ community, resulting in moral, tactical, and financial support [6, 13]. The Ugandan LGBT+ community has enjoyed a privileged position over the past ten years in terms of receiving international funding. A review of the Global Philanthropy Project’s data, which tracks dedicated LGBT+ funding globally, shows that the Ugandan LGBT+ community has been among the six top receivers since 2012 [14]. This influx of resources supported the formation of the contemporary Ugandan LGBT+ activist community, characterized by a significant increase in organizations and intra-community diversification. The community has grown from 24 organizations in 2012 [15] to over a hundred in 2022, according to the leading umbrella organization, Sexual Minority Uganda (SMUG) [16]. The expansion has also resulted in specialization and geographical dispersion.

The 2009 Anti-homosexuality bill, unfortunately, set other processes in motion as well. Even if the 2014 Anti-Homosexuality Act was repealed, its proponents, consisting of a mix of norm-conservative and populist religious and political leaders backed by ultra-conservative American churches [17,18,19], normalized hate speech and discrimination against LGBT+ individuals [20, 21]. Hate crimes surged in the aftermath of the 2009 bill [22]. Furthermore, Parliament debates and efforts to address the ‘problem of homosexuality’ did not cease after the annulment of the 2014 Anti-homosexuality Act. Parliamentarians have continued to work to strengthen regulation against homosexuality based on notions of widespread so-called ‘promotion of homosexuality’ and ‘recruitment of Ugandan children into homosexuality’ [23]. Lawmakers and proponents of discrimination against LGBT+ individuals have been successful in creating a homo-hostile public climate. A recent public opinion survey indicates that Uganda is the most intolerant country of the 37 African countries surveyed when it comes to LGBT+ individuals and human rights for the community. Ugandan state-sponsored discrimination enjoys strong public support across all demographic groups [24]. It has, however, been argued that the emergence of homophobic public discourses and widespread discrimination is a relatively recent phenomenon [25].

1.1 Resistance Against Repression

Human rights advocacy consists of systematic documentation of human rights abuse, formulation, and dissemination of information that skillfully situates local abuses into a broader human rights framework. Human rights activists also need to create broad strategic alliances that provide moral, tactic, and financial support, which can be leveraged towards convincing the public and policymakers to accept the proposed change. Historically, mainstream media has been central in reaching groups beyond the directly afflicted community. Uganda LGBT+ activists have, however, been systematically denied access to public discourse-producing spaces, such as mainstream media, town halls, and political forums [6, 13, 26]. Traditional media spaces have displayed a range of exclusion and silencing tactics and more overt discriminatory practices [27,28,29,30]. Tabloid media actors have even actively encouraged violent hate crimes against Ugandan LGBT+ by calling for their public hanging and publishing home and work addresses of suspected homosexuals [31, 32].

The Ugandan LGBT+ community has, however, similar to LGBT+ activists across the world, attempted to bypass exclusion from discourse-producing spaces and also escape the risks associated with entering a hostile physical environment by adopting digital tools and media [26, 33,34,35,36]. In repressive contexts, digital tools and social media spaces are particularly useful for documenting abuse, as well as ensuring that remote publics become ‘distant witnesses’. By publishing reports on human rights abuse, activists hope to raise awareness, which potentially generates solidarity as well as international pressure to address human rights abuses [37, 38]. Despite the many limitations of digital activism, Ugandan LGBT+ organizations have developed an active presence on one or several online social platforms (OSP), as well as use various messaging services for coordination.

Although repression of LGBT+ Ugandans has gone in waves over the past 20 years, the situation deteriorated significantly in the spring of 2023 (Fig. 1). On the 29th of May, the Uganda government succeeded where it had failed in the past and passed a piece of legislation with unprecedented far-reaching suppression of rights for LGBT+ individuals and organizations working for their human rights [39,40,41]. The new Anti-homosexuality Act’s (AHA2023) clauses that criminalize the “promotion of homosexuality” and the vague offense of promoting or encouraging homosexuality or “the observance or normalisation of conduct prohibited under this Act” (unnumbered) [42]; all directly aim to silence voices calling for equal rights for Ugandan LGBT+.

Fig. 1.
figure 1

Passing the Ugandan Anti-Homosexuality Bill/Act timeline in the spring of 2023.

Full size image

1.2 Research Question

Against the backdrop that the Ugandan judiciary’s interpretation of the new offense ‘promotion of homosexuality’ has yet to be established through legal precedents, and the stakes are high for perpetrators, that is, Ugandan LGBT+ activists and their organizations; this paper seeks to explore if the Anti-homosexuality Act of 2023, had a chilling effect on Ugandan LGBT+ advocacy in their key online social platform (OSP), Twitter. The following section explores the concept of chilling effects in greater detail.

1.3 Unpacking Chilling Effects and the AHA2023

The AHA 2023 contains a number of far-reaching clauses that could be used to arbitrarily silence LGBT+ activism and cut activists off from essential resources, as well as introduce surveillance by mandating citizens to report crimes listed in the Act. The clause on promoting homosexuality directly targets LGBT+ rights advocacy and is committed when a person publishes, broadcasts, or distributes “any material promoting or encouraging homosexuality or the commission of an offense under this Act”; or “operates an organization which promotes or encourages homosexuality or the observance or normalization of conduct prohibited under this Act”. The crime of promotion also includes leasing property, or providing “financial support, whether in kind or cash, to facilitate activities that encourage homosexuality or the observance or normalisation of conduct prohibited under this Act.” The Act also includes a duty to report crimes listed in the act. “A person who knows or has a reasonable suspicion that a person has committed or intends to commit the offense of homosexuality or any other offense under this Act shall report the matter to police for appropriate action” [42].

The vagueness of the offenses, combined with the severe punishment if convictedFootnote 1, creates a high level of uncertainty amongst activists, which is likely to have a chilling effect on activism, similar to that reported during the attempts to pass the predecessor- the 2009 Anti-Homosexuality Bill [6]. A chilling effect exists when an individual is deterred by fear of legal punishment or other adverse consequence and “engages in self-censorship, that is, censors themselves and does not speak or engage in some activity, despite that activity being lawful or even desirable” [43]: 1454.

Research on the phenomena of chilling effect is still fairly limited outside the legal realm. Prior research focuses on the role of laws to curb certain activities, including specific speech acts, such as hate speech, false statements, and defamation. Penney notes the existence of two conventional theories of chilling effects, and both operate on an assumption of deterrence as a core mechanism to understand an individual’s behavior modification. Previous research has primarily been centered around two causal models to explain how chilling effects arise [43]. Chills may occur as a consequence of fear of legal consequences, and/or fear of privacy harm as a consequence of various actors’ increased capacity to engage in extensive surveillance. Chills due to surveillance are postulated to be significantly fueled by the unrestricted circulation of personal information online. However, these two deterrence-based theories assume that individuals conduct a rational assessment of legal risk and/or harm to privacy, and they fail to account for chilling effects when legal or privacy harms are not present. Penney, therefore, argues that chilling effects are “best understood as an act of compliance with, or conforming to, social norms in the context” (1488) and are thus less a consequence of sole avoidance of legal and privacy harms. It is argued that the desire to comply with social norms is central to human sociality and proceeds to introduce a theory of chilling effects as social conformity. Chilling effects due to social conformity are greatly facilitated by the context collapse and the peer-to-peer real-time surveillance digital spaces afford users [44]. Most users are aware of social surveillance, and multiple studies have documented significant chilling effects on dissident voices as a consequence of online social policing and in various sectors, such as academia [45], gaming [46], journalism [47, 48], and female journalist in particular [49].

However, in the case of LGBT+ activism across the world, activists are consciously and very deliberately challenging social norms they perceive as unjust or even unlawful. Indeed, a disregard for and desire to challenge existing social norms and policies lies at the very heart of sexual rights activism. It could even be argued that when it comes to LGBT+ activism, the desire to conform to dominant social norms on sexuality and gender identity ideals is non-existent and thus unlikely to produce a chilling effect on different forms of communication and behavior. The Ugandan LGBT+ community has, for example, continued to challenge state-sponsored homophobia for two decades despite pervasive negative public opinion and enduring both state and non-state violent policing and discrimination.

Besides understanding the mechanisms behind chilling effects, which, according to some, are “most likely unprovable” [50]: 270, it is also important to understand the chilling effects’ impact on various types of content and conduct. Penney (2023) argues that besides understanding the mechanism producing chilling effects, there is also a need for a better understanding of its impacts. That is, how chilling effects may produce a more compliant, docile, and conforming speech and/or conduct, or perhaps even silence. A recent study finds that in the context of content moderation of social media, a chilling effect was confirmed, but its effect was small and had only a minimal impact on content. At most, it subtly altered the specific style and/or tone used [50]. Even if the results are interesting, it should be noted that participants in that study were not threatened with lengthy jail time if they failed to comply with content moderation restrictions.

2 Methods and Sample

Bedi (2021) attributes the lack of empirical studies confirming the existence of chilling effects and how they manifest to the difficulties of measuring them in content, whether speech, visuals, or text. To successfully capture chills manifesting themselves as silence, that is, absent of speech and/or behavior, and/or at the discursive level, the empirical set needs to contain data before the suspected chill and after. Given the methodological difficulties in capturing chilling effects and that potential chilling effects can manifest in different ways, an explorative approach was deemed suitable, drawing on both quantitative and qualitative methods for analysis. Chilling effects were hypothesized to become manifest in either a significant decrease in activity on digital spaces and/or a discursive shift at the content level.

As Facebook remains blocked in Uganda since Meta de-platformed fake accounts connected to the country’s Ministry of Information during the elections in 2021, Twitter, now X, was the primary OSP for activism and public communication during the spring of 2023. An account‐driven data collection approach was used. Tweets from 46 active Ugandan LGBT+ activist accounts were downloaded from the Twitter Application Programming Interface v2 using the Academic Research product track, spanning 1st January 2022 to 11th June 2022 and the corresponding period for 2023. The 2023 sample thus covered the timeframe where a shift in policy became increasingly likely (Fig. 1). The somewhat arbitrary cutoff date, the 11th of June, is a result of the new owner of Twitter blocking capture after the 11th of June. Even if the lack of access to accounts beyond the 11th of June is a study limitation, the dataset still provides almost two weeks of data.

The decision to compare the critical period in 2023 (Fig. 1) with a corresponding time frame from 2022, is motivated by two rationales. Firstly, prior research has indicated that Ugandan LGBT+ advocacy on Twitter is event-triggered and, to a significant degree, connected to an international LGBT+ calendar, such as Zero Discrimination Day 1st of March, International Day Against Homophobia, Transphobia, and Biphobia-IDAHOBIT 17th May, and Pride month [51]. In 2020, the Ugandan advocacy and event calendar was supplemented with the Kuchu Memorial Day. The 26th of January was created to celebrate the legacy of the murdered Ugandan LGBT+ activist David Kato. Secondly, towards the end of the summer of 2022, there were looming signs of increased state repression. In early August, the National Bureau for Non-governmental Organizations banned SMUG, the most prominent organization in the Ugandan LGBT+ activist ecosystem. Since its inception in 2004, SMUG has functioned as an umbrella for smaller organizations and grown into an essential coordinating body for the community. Furthermore, SMUG is well connected internationally, and its director, Frank Mugisha, has won multiple prestigious human rights awards for his activism and is frequently invited as a keynote speaker and adviser. The decision to ban SMUG from operating signaled a new level of confrontation and disregard for international standards and relations with development partners.

The first analysis consisted of comparing the activity level from January 1st to June 11th, 2022, with that of the corresponding 2023 dates. The two sample periods generated 7089 tweets (2022) and 6603 tweets (2023), which indicated a minor decrease in activity (6.9%). The activity level in the immediate post-bill period, 29th of May- 11th June 2023, was, however, significantly higher than that of the corresponding timeframe in 2022. Furthermore, the two 2022 and 2023 samples displayed a similar pattern of activity being centered around external trigger events. For 2022, activity was centered around key activism events, such as the UN Human Rights Council’s period review of Uganda, Zero Discrimination Day, Transgender Day of Visibility, and International Day Against Homophobia, Transphobia, and Biphobia. The corresponding spikes in 2023 were found around the introduction of AHA to parliament in March, the parliament’s first passing of AHA2023 later in March, the parliament’s vote on the amended AHA2023, and finally, the President’s ascension of the AHA2023, the 29th of May. Based on similar activity levels in the two data sets, i.e., less than a 10% decrease between 2022 and 2023, and activity patterns centered around external events, it can be argued that the introduction and the legislative process leading to AHA2023 did not suppress the level of activity in any significant manner.

With no apparent chilling effect on activity or patterns of use of Twitter, the next step of analysis consisted of qualitative analysis of the post-AHA corpus, i.e., 29th of May- 11th of June, which consisted of 611 tweets. Given the uncharted terrain of capturing chilling effects on content, tweets were explored using a conventional content analysis [52]. Conventional content analysis is appropriate in cases where existing theory or prior research is limited. In more exploratory work, researchers should avoid using preconceived categories and instead allow content categories to emerge organically from the data. Furthermore, the open analysis of the post-AHA2023 corpus was used as a method for exploring and assessing the grounds for further analysis rather than a method for singularly capturing the existence and determining the characteristics of chilling effects on content. Determining the nature of and full extent of chilling effects on content requires a comparison with content from the same 2022 time period, which is arguably untouched by both the closure of SMUG in the late summer of 2022 and the ultimate chill, the policy arena’s introduction and passing of AHA2023.

Finally, the author’s in-depth familiarity with the Ugandan activist scene greatly facilitated the content analysis.

3 Results

The post-AHA2023 corpus is entirely dominated by content related to the Act. Less than 10% of the content covered unrelated topics, such as news on SOGIEFootnote 2 rights from around the world and educational and training opportunities offered to the community. The content analysis generated four overarching themes (Table 1): Condemnation of the 2023 AHA, Human rights advocacy, the legal battle to repeal AHA 2023, and finally, Unpacking the AHA.

Table 1. Overarching themes emerging from the content analysis. n = 611.
Full size table

In isolation, the analysis of the post-AHA2023 corpus does not provide sufficient grounds to determine the existence of a chilling effect on the content level. However, when situated against a recent study that analyzed Ugandan LGBT+ organizations’ digital activism [51], the results strongly suggest chilling effects at the content level. The previous study, which focused on Ugandan LGBT+ activists’ Twitter activity in the first half of 2022, that is, well before the Ugandan Government’s crackdown on civil society and closure of the largest LGBT+ organization, Sexual minorities Uganda (SMUG) in August 2022; found that Twitter appeared to have two primary functions for the organizations: 1) Self-representation and organizational promotion, and 2) Service to the community. Self-representation and organizational promotion primarily served to showcase the organizations’ activities and output, as well as spotlight the LGBT+ community’s accomplishments and contributions to the cause. The second category, Service to the Ugandan LGBT+ community, consisted of a range of services such as disseminating relevant news and information on SOGIE rights, advocating for equal rights in Uganda, exposing the Ugandan government’s failures, and calling for accountability, as well as engaging in community care by increasing awareness on the importance of self-care, including mental health and offering various off-line and online health services.

In the 13-day post AHA2023 corpus, self-representation and organizational promotion, disappears. That is, none of the organizations spotlighted their organization’s work, output, agency, and general achievements. The disappearance of this type of content is very likely an effect of that it may be interpreted as “promotion of homosexuality”. The area of Twitter use, previously categorized as Service to the community, remained in the 2023 material but was significantly altered, and in particular in respect to human rights advocacy.

The most conspicuous change is the local activists’ silence on equal rights for Ugandan LGBT+ in the material. With few exceptions, self-penned explicit LGBT+ rights advocacy disappears. The 13-day corpus still contains explicit calls for equal rights for Ugandan LGBT+ and efforts to hold the Ugandan government accountable to its human rights commitments. However, this type of rights advocacy is primarily advanced by international actors that have little to fear from the Ugandan government. Explicit calls for equal rights for LGBT+ and LGBT+ rights as part of a human rights framework are thus still very present and still actively promoted, but local organizations do so by re-tweeting international actors material, such as the UN, and other multilateral actors, bilateral development partners, including heads of state, as well as international human rights organizations. Ugandan LGBT+ organizations thus appear to have adopted a strategy of re-circulating international actors’ vocal rejection of AHA2023, rather than convey the same message through self-penned posts.

Service to the community was also provided by explaining and educating on the Act’s clauses as well as raising awareness of its potential impact on all spheres of life.

To sum up, there is a clear discursive shift in the content. The local activists’ 2022 content production was explicitly self-promotional and self-celebratory, as well as emphasized the community’s agency despite a hostile environment. In the 2023 material, Ugandan activists appear to have adopted a safety margin to what could constitute and be persecuted on grounds of promoting and normalization of homosexuality. They community also begins using the retweet affordance rather than creating original content to a higher degree. Another conspicuous change at the content level is the use of descriptive identity labels. The LGBT+ acronym has historically been preferred over the local term for queer, kuchu, and the LGBT+ is widely adopted as a self-referencing label in Uganda. In the post AHA 2023 landscape there are signs that both terms are viewed as a potential liability. LGBT+ is still used, but primarily in the post AHA2023 landscape to refer to community’s shock, anguish, and descriptions of the AHA’s negative impact. Instead a new, and safe(r) public health term, ‘key population’ emerge as term for referring to the LGBT+ community. The term ‘key population’ is widely used in the public health sector and typically refers to individuals who, due to societal pressures and/or social circumstances, are more vulnerable to exposure to infections, including HIV. The term thus refers to several other groups, besides LGBT+ individuals.

It should be noted that even if the LGBT+ activist and their organizations appear to have adopted a new type of strategic invisibility and toned down their explicit rights demands; local LGBT+ organizations still attempt to service their community through Twitter. A significant part of the corpus focuses on unpacking and explaining the implications of AHA’s draconian provisions. Service to the community is thus manifested through an educational effort, and awareness raising of how the different clauses may impact the community’s ability to access health care, education, employment, and housing. In short, the Ugandan LGBT+ organizations used Twitter to educate the community about the AHA’s heavy-handed reach into all spheres of life, in an effort to provide the community with new essential life skills.

4 Discussion

This study sought to contribute to the growing body of literature on chilling effects and to the methodological toolbox for studying chilling effects empirically. Furthermore, the article approached chilling effects in a repressive context that has increasingly become synonymous with political homophobia and state-sanctioned extra-judiciary violence against LGBT+ individuals. The Ugandan government has since 2009 explored different ways to equip itself with the legal tools to sanction the oppression of Ugandan sexual minorities. With the passing of the AHA2023, the Ugandan Government has granted itself unprecedented powers to persecute LGBT+ activists, human rights defenders, and any allies that would dare to challenge the repression.

Despite the Act’s vague legal language, which creates significant situational uncertainty in terms of what speech acts are henceforward illegal and can be arbitrarily prosecuted by the Ugandan Government, combined with harsh penalties for offenders, activists were not deterred from actively using their Twitter accounts to voice their shock, condemnation, and opposition to the AHA 2023A. The results do, however, clearly indicate that the Act had a noticeable chilling effect on the content level, which is conspicuous enough to warrant arguing that a discursive shift takes place in the wake of the AHA2023. The most conspicuous shift in content is the loss of local voices explicitly advocating for equal human rights for Ugandan LGBT+ individuals. Advocacy for equal rights for LGBT+ individuals does not disappear but rather appears to be primarily pushed by actors outside the Ugandan government’s jurisdiction. Although the Ugandan LGBT+ organizations’ editorial decisions are understandable, given the government’s appetite for harsh penalties for offenders, the shift may impact the community’s long-term ability to conduct human rights advocacy. In contexts where LGBT+ rights are dismissed on the grounds that sexual and gender diversity are Western creations and impositions pushed upon African nations by neo-imperialist forces; rights advocacy needs to be conducted by African voices to be perceives as legitimate [53]. That is, advocacy for equal rights must come from Ugandans to have any chance of countering the popular belief that sexual and gender diversity and LGBT+ rights are un-African and alien to Ugandan culture. There is thus a risk that even if the Ugandan organizations may have opted to circulate international criticism, rather than creating their own content, in an attempt to lessen the risk of being prosecuted for promoting homosexuality as well as show that their rights struggle enjoys strong international support; it undermines their local legitimacy. It simply fuels pre-existing discourse around uninvited and corrosive Western interferences.

Finally, this study reminds us that digital spaces can and are successfully attacked without applying sophisticated surveillance technology. Optimistic interpretations of digital spaces as separate and free in contexts where they are not yet heavily surveilled or locked up behind state-run firewalls, thus, need to be revisited once more. The AH2023’s chilling effects on digital content suggest that the Ugandan LGBT+ community, at least the part that resides inside Ugandan borders, may be losing a vital space. With limited access to other discourse-producing spaces, the community risks being starved of an essential communicative space.

With events still unfolding, it is arguably too early to fully assess the long-term impact of the identified chilling effects. The Ugandan case can, however, serve as a critical case study for understanding the chilling effects of draconian suppression of human rights advocacy in and outside digital spaces. Given the indication of chilling effects on the content level, and with several African countries exploring similar legislation, more research is urgently needed to deepen our understanding of the chilling effects of legal reforms copying the Ugandan AHA. Furthermore, in light of digital spaces’ ephemeral nature, sometimes caused by platform owners’ shifting interests, more research is needed on the methodological challenges connected with studying chilling effects in digital spaces. Besides ensuring access to data, we also need a better understanding of how to empirically capture the scope, facets, and duration of chilling effects.